Abstract
The use of piezoelectric materials to treat critical-sized bone defects typically requires additional stimulation to generate their piezoelectric properties and the implantation of scaffolds to promote bone repair. Here we present a self-reinforced piezoelectric chip and demonstrate its efficacy in the treatment of critical-sized bone defects. Specifically, the chip is comprised of the third-generation semiconductor aluminum nitride (AlN) as a piezoelectric layer, molybdenum (Mo) electrodes, and a silicon substrate with an optimized internal cavity structure. All these components are confirmed to be non-cytotoxic. This design enables the chip to provide self-sustained and long-term electrical signals in response to physiological vibrations. After being implanted into a rabbit critical-sized femoral defect model, the chip creates a localized bioelectric microenvironment, thereby promoting vascularized bone repair within 4 weeks without using any scaffolds and additional tools. Moreover, the chip can be fixed onto the clinically used orthopedic plate system, representing a universal plug-and-play strategy.
Similar content being viewed by others
Introduction
The treatment of critical-sized bone defects remains a significant challenge because bone loss of those defects exceeds 50% of a specific bone circumference and requires additional interventions for bone regeneration after fixation1. In the clinic, bone scaffolds such as autografts, allografts, and synthetic grafts are used to fill those defects2. These scaffolds are typically porous structures that mimic the native extracellular matrix (ECM) of bone and provide a three-dimensional microenvironment for new bone formation3. However, they face limitations, including donor site morbidity, immune rejection, inadequate integration, and extended surgery times4. Thus, new approaches to repair critical-sized bone defects are urgently needed. In recent years, piezoelectric materials, capable of generating electrical signals under mechanical stress, are shown to create a localized bioelectric microenvironment for promoting bone regeneration5. Inorganic piezoelectric materials (e.g., lead zirconate titanate, zinc oxide, and barium titanate) and organic piezoelectric materials (e.g., polyvinylidene fluoride, glycine, and poly-L-lactic acid) are developed. However, they require frequent stimulations such as stretching, electrical field6, and ultrasound7 for the generation and maintenance of piezoelectric properties. These materials are also integrated with bone scaffolds to treat bone defects. Namely, current piezoelectric materials for bone tissue engineering must rely on additional operations and tools8. Based on the fact that bone repair often takes more than three months, materials with spontaneous and long-term piezoelectric poverties are highly desired in clinical applications of critical-sized bone defects9,10.
Our experiences11,12,13,14,15 in semiconductor materials have driven us to choose aluminum nitride (AlN) as a candidate. As a representative third-generation semiconductor, AlN is extremely important in high-temperature, high-frequency, high-power electronic devices, and deep-ultraviolet optoelectronic devices due to its high breakdown electric field, high thermal conductivity, and high electron saturation drift velocity16,17. The piezoelectric property of AlN also gives it application potential in the biomedical fields. AlN-based energy harvesters, which transform mechanical energy within the human body into electrical energy, are of great significance because they can supply power to wearable, implantable, portable medical devices, and various medical sensors18,19. However, AlN-based biomedical devices are fraught with challenges. They exhibit a low current density and are constrained by limited energy-collection sources in terms of energy conversion efficiency. For instance, the efficiency of extracting electrical energy from the feeble energy sources inside the human body is meager, and their response to specific mechanical vibrations is inadequate20. In addition, they have been rarely explored in regenerative medicine.
Herein, we developed a self-reinforced smart chip composed of a piezoelectric layer of AlN, molybdenum (Mo) electrodes, and a silicon (Si) substrate (Fig. 1aI). To achieve self-reinforcement, an internal cavity structure within the Si substrate is designed for increasing the sensitivity of AlN in response to physiological vibrations such as muscle contraction and limb activity (Fig. 1aII). We optimized the cavity structure of the chip through simulation and electric signal collection tests. The introduction of the cavity structure into the AlN-based chip is applied in electrical stimulation-induced bone growth, and it showed a significant promoting effect on the proliferation and osteogenic differentiation of stem cells compared to the AlN-based film prepared with the same components but without the cavity structure. The underlying mechanism involved the activation of the PI3K/Akt pathways (Fig. 1aV). Meanwhile, the chip also presented angiogenic inductivity. When implanted the chip along with a clinically used Ti6Al4V plate in a critical-size femoral defect rabbit model (Fig. 1aIII), in vivo assays demonstrated that it could effectively accelerate vascularized bone repair without using any bone scaffolds, cell transfer, and additional tools (Fig. 1aIV).
a The smart chip is a multilayer structure, including a Si substrate, Mo electrodes, and an AlN piezoelectric layer (II). An internal cavity was etched into the Si substrate using a wet etching method to enhance sensitivity to physiological vibrations (I). The chip generates alternating electrical signals under vibration, converting mechanical energy into a localized bioelectric microenvironment. After being integrated with a clinically used plate and implanted at the critical-sized bone defect site (III), the chip synergistically activates the PI3K/Akt signaling pathway (V), thereby promoting the proliferation and osteogenic differentiation of stem cells (IV). Meanwhile, the chip enhances angiogenesis by upregulating VEGF-A and CD31 expression. Hence, the chip provides a promising strategy for critical-sized bone defect repair through robust electrical stimulation and vascularized bone regeneration, achieving an efficient scaffold-free bone repair. b Photos of the chip. c Cross-sectional SEM images of the chip with a 10 × 3 mm2 cavity. d Schematic diagram of the current testing system for the chips. A testing disk with a central cavity (10 × 5 mm2) is used to simulate a bone defect. The chips are adhered to a Ti6Al4V plate using Polydimethylsiloxane (PDMS), and then the plate is fixed across the defect with screws. A vertical cyclic force (0.5 N at 1−4 Hz) is applied perpendicular to the disk to mimic physiological vibration. e Electrical signal of the chips integrated with Ti6Al4V plates under vibration. f Statistical analysis of current density derived from panel (e). Data are presented as mean ± SD (n = 11). Two-sided ANOVA with a Tukey’s post hoc test for multiple comparisons. Source data are provided as a Source Data file.
Results
Simulation and characterization of chip
The photos of the as-prepared chip are depicted in Fig. 1b. Cross-sectional SEM images (Fig. 1c) show that the chip is the as-designed layer-by-layer (Mo/AlN/Mo/Si) structure. The first and third layers are Mo electrodes but do not completely cover the AlN intermediate layer, and Si with cavity is the bottom layer (Fig. 1aI and Supplementary Fig. 1a). In contrast to the Mo electrodes, the AlN surface exhibits distinct particulate features (Supplementary Fig. 2a). The root mean square (RMS) of Mo and AlN are tested by atomic force microscope (AFM) and the values are 1.2 and 3.5 nm, respectively (Supplementary Fig. 2b, c). Energy dispersive spectrum (EDS) analysis identifies the elemental composition and their content (Supplementary Fig. 1a). X-ray diffraction (XRD) results show the (110) lattice plane of Mo and the (002) lattice plane of AlN (Supplementary Fig. 1b).
To compare the piezoelectric performance and optimize the cavity structure, we performed a computational analysis. The results show that cavities significantly enhance the amplitude peak, with larger cavities further improving the chip’s amplitude range (Supplementary Fig. 3). To evaluate the piezoelectric performance of the chips under physiological mechanical stimulation, we constructed a bone-defect-mimicking testing platform with a central cavity (10 × 5 mm²). The chips are adhered to a Ti6Al4V plate system and fixed on the testing platform as illustrated in Fig. 1d. Upon exposure to vibration (1 − 4 Hz, 0.5 N), the chips generated a measurable and stable electrical output. When the cavity area was 6 × 3 mm², the average output current density increased significantly, reaching 2.31 ± 1.069 μA cm−². At 10 × 3 mm², the chip maintained better output stability with a maximum signal of 3.29 ± 0.966 μA cm−² (Fig. 1e, f). Further increase to 14 × 3 mm² led to incomplete etching and decreased output (1.34 ± 0.031 μA cm−²; Supplementary Fig. 4). Based on the signal performance and structural integrity, the 10 × 3 mm² cavity was selected for subsequent experiments.
Biocompatibility and osteoinductivity of the chip
To assess the impact of the piezoelectric chip on biocompatibility, osteogenesis, and angiogenesis, in vitro experiments were designed with four experimental groups: (1) the control group, in which cells were cultured on conventional cell culture plates without any external stimulation, serving as a baseline for comparison; (2) the control + vibration group, where cells were subjected to mechanical vibration to mimic the physiological vibrations associated with limb movement in vivo, enabling the evaluation of the isolated effects of mechanical stimulation on cellular behavior; (3) the film + vibration group, in which cells were cultured on a AlN film under vibrational stimulation, allowing for the examination of the influence of material surface topography and composition on cellular responses; and (4) the chip + vibration group, where cells were cultured on the piezoelectric chip under vibrational stimulation, facilitating the investigation of the additional effects conferred by the piezoelectric properties of the chip on cellular behavior. Note that the AlN-based film has the same components as the AlN-based chip, which are also composed of the AlN piezoelectric layer, Mo electrodes, and Si substrate, but without the internal cavity structure. This experimental design ensures a systematic evaluation of the individual and combined contributions of mechanical stimulation, material properties, and piezoelectric effects on cellular responses.
We first incubated bone marrow mesenchymal stem cells (BMSCs) with AlN, Mo, Si, and PDMS. All materials exhibited negligible cytotoxicity. AlN and Mo slightly increased BMSC proliferation compared to the control group, indicating the biosafety of these components (Supplementary Fig. 5a). Then, we investigated the effects of the chip or film under vibration on cell proliferation. As shown in Supplementary Fig. 5b, an increase in cell numbers was observed in all groups along with the extended culture period. Specifically, the number of BMSCs in the chip + vibration group was significantly higher than other groups at all time intervals, indicating a marked enhancement in cell proliferation due to the electrical stimulation generated by the chip. Live/dead staining showed that negligible dead cells (red fluorescence) were observed in all groups and cell density was notably higher in the chip + vibration group after 3 days of culture (Supplementary Fig. 5d). The mean fluorescent intensity of Calcein-AM (green fluorescence; live cells) were highest in the chip + vibration group (Supplementary Fig. 5c), which was in accordance with the CCK-8 results. The cytotoxicity of the chip was further assessed through phalloidin staining. After seeding for 24 h, cells were well-spread on film and chip under vibration, and presented extended cellular actin fibrils and branched filopodia after seeding for 24 h (Supplementary Fig. 6).
To investigate the osteoinductivity of the chip under vibration, we performed alkaline phosphatase (ALP) staining and Alizarin Red (AR) staining. The ALP staining (Fig. 2a) and ALP activity (Fig. 2c) results revealed that BMSCs in the chip + vibration group exhibited the highest ALP activity among all groups after 7 days of incubation. The control + vibration and film + vibration groups slightly enhanced ALP activity of BMSCs. After 21 days of incubation, AR staining showed that much more calcium nodules (red color) were formed within BMSCs in the chip + vibration group relative to other groups (Fig. 2b). The film + vibration and control + vibration groups also showed enhanced mineralization compared to the control group. Quantitative analysis confirmed that the piezoelectric chip was best in terms of mineralization of BMSCs (Fig. 2d). EDS analysis confirmed that the calcium-to-phosphorus (Ca:P) ratio of the deposited minerals on the chip and the film was approximately 1.46 ± 0.21 and 1.00 ± 0.13, respectively (Supplementary Fig. 7).
ALP staining (a) and ALP activity (c) of BMSCs cultured in the control, control + vibration, film + vibration, and chip + vibration groups for 7 days. Alizarin Red staining (b) and its qualitative analysis (d) of the extracellular matrix of BMSCs cultured in the control, control + vibration, film + vibration, and chip + vibration groups for 21 days. e Expression levels of osteogenic-related genes, including Col-1a, BMP-2, and Runx2 of BMSCs, were measured and normalized to the GAPDH. f Representative images of immunofluorescence staining of BMP-2 (Red) and Runx2 (Green). g Semi-qualitative analysis of BMP-2 and Runx2 expression derived from panel (f). h Western blot results of the expression of Runx2, Col-1a, and BMP-2 after 7 days of culture. i Quantitative study of the grayscale of protein expression derived from panel (h). Data are presented as mean ± SD (n = 3 biological replicates). Two-sided ANOVA with a Tukey’s post hoc test for multiple comparisons. Source data are provided as a Source Data file.
To further validate these findings, the expression of osteogenesis-related genes was examined. As shown in Fig. 2e, the key osteogenic gene expressions in BMSCs, including RUNX family transcription factor 2 (Runx2), collagen type I alpha (Col-1a), and bone morphogenetic protein 2 (BMP-2), were significantly up-regulated in vibration groups. Notably, the chip + vibration group demonstrated the highest expression levels of these genes among all groups. Immunofluorescence staining of Runx2 and BMP-2 further confirmed enhanced osteogenic protein expression in the chip + vibration group, as evidenced by more intense fluorescence signals compared to other groups (Fig. 2f, g). The western blotting results of BMP-2 and Runx2 presented a consistent trend (Fig. 2h, i). Enzyme-linked immunosorbent (ELISA) assay indicated that the content of BMP-2 secreted by BMSCs was highest in the chip + vibration group (Supplementary Fig. 8a).
These above-mentioned results demonstrated that the piezoelectric chip with a 2D electrical surface not only created a favorable microenvironment for BMSCs proliferation but also enhanced osteogenic differentiation through the self-reinforced electrical stimulation under vibration.
Mechanism of the chip in promoting osteogenesis
To further investigate the mechanism underlying the osteoinductive effect of the chip, we analyzed transcriptomic changes in BMSCs cultured with the chip or film under vibration. Volcano plot analysis identified 740 differentially expressed genes (DEGs) in the chip + vibration group compared to the film + vibration group, with 427 upregulated and 313 downregulated DEGs (Fig. 3a). Gene Ontology (GO) enrichment analysis revealed significant enrichments in biological processes, cellular components, and molecular functions related to ECM organization, ECM binding, collagen-containing ECM, and focal adhesion (Fig. 3f). Heatmaps indicated the upregulated ECM-related genes, including Tgfbr3, Olfml2b, Dmp1, Postn, Mepe, Nav2, Col18a1, Col11a2, Col8a1, and Ltbp1 (Fig. 3d). Osteogenesis-related genes were also upregulated in the chip + vibration group, including Runx2, Spp1, Bmp8a, Fbn2, Gdpd2, Suco, Fgfr2, and Omd (Fig. 3b). Moreover, BMSCs in the chip + vibration group also showed upregulated genes involved in angiogenesis, such as Nrp1, Angpt1, Vash2, Ccn3, Ereg, Esm1, and Ecscr (Fig. 3c).
a Volcano plots of the chip + vibration versus the film + vibration. Heatmap of osteogenesis (b), angiogenesis (c), extracellular matrix (d), and PI3K/Akt signaling pathway (e) related genes. GO (f) and KEGG (g) pathway enrichment analysis. h GSEA for the PI3K/Akt signaling pathway. i Western blotting analysis for the expression of p-PI3K and p-Akt of BMSCs in the control, control + vibration, film + vibration, and chip + vibration groups for 7 days of incubation. j Quantitative analysis of the grayscale of protein expression derived from panel (f). Data are presented as mean ± SD (n = 3 biological replicates). Two-sided ANOVA with a Tukey’s post hoc test for multiple comparisons. Source data are provided as a Source Data file.
In addition, Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis revealed that the DEGs were significantly enriched in the PI3K/Akt signaling pathway (Fig. 3g). GeneESt enrichment analysis (GESA) confirmed that the PI3K/Akt signaling pathway was activated in the chip + vibration group (Fig. 3h). The PI3K/Akt pathway-related genes such as Flt1, Pdgfd, Lama2, Pik3cd, Pik3r1, Itga2, Pdgfra, Igf2, Sgk1, and Tnr were upregulated in the chip + vibration group as identified by the heatmap (Fig. 3e). To validate these findings, western blotting (Fig. 3i) was conducted to evaluate protein expression levels of phosphorylated PI3K (p-PI3K) and Akt (p-Akt). Quantitative analysis indicated that BMSCs in the chip + vibration group presented significantly higher p-PI3K/β-actin and p-Akt/β-actin ratios relative to the film + vibration group (Fig. 3j). Hence, the chip and its piezoelectric effects could enhance osteogenic differentiation through activating the PI3K/Akt signaling pathway.
The chip enhances angiogenesis in vitro
The initial step in vascularization involves endothelial cell recruitment. To evaluate the response of human umbilical vein endothelial cells (HUVECs) to the chip, wound healing assay was performed. As shown in Fig. 4a and Supplementary Fig. 9a, the chip significantly enhanced HUVEC migration under vibration. HUVECs in the chip + vibration group exhibited enhanced sprouting, characterized by a higher number of branching nodes and elongated tube networks after 6-, 12- and 24 h post-seeding (Fig. 4b and Supplementary Fig. 9b). We further assessed the angiogenic gene expression in HUVECs. After 3 days of co-culture, both the film and chip groups exhibited higher gene expression levels of VEGF-A and CD31 compared to the control group (Fig. 4c). Specifically, HUVECs in the chip + vibration group displayed the highest gene expression levels of VEGF-A and CD31, followed by the film + vibration group. In addition, the control + vibration group also showed a statistically significant upregulation of these genes compared to the control group.
a Quantification of wound healing assay. b Quantitative evaluation of total branches length in tube formation assay. c Expression levels of angiogenic genes CD31 and VEGF-A of HUVECs cultured for 3 days. Semi-qualitative analysis (d) and representative immunofluorescence staining images (e) of CD31 (Red) and VEGF-A (Green) of HUVECs in different groups. Western blot results (f) and quantitative analysis (g) of protein expression of CD31 and VEGF-A. Data are presented as mean ± SD (n = 3 biological replicates). Two-sided ANOVA with a Tukey’s post hoc test for multiple comparisons. Source data are provided as a Source Data file.
Immunofluorescence staining, western blotting and ELISA of CD31 and VEGF-A were also performed after 3 days of incubation. As shown in Fig. 4d, e, the expression levels of CD31 and VEGF-A in the film + vibration and chip + vibration groups were significantly higher than those in the other groups. Western blotting indicated the highest expression of these angiogenic markers in HUVECs in the chip + vibration group (Fig. 4f, g). However, there were no statistical differences of these proteins among the film + vibration, control + vibration, and control groups. ELISA results further indicated that HUVECs in the chip + vibration group secreted significantly higher levels of VEGF-A and CD31 compared to the other groups (Supplementary Fig. 8b, c). Therefore, the electrical stimulation by the chip could promote angiogenesis in vitro.
The chip promotes critical-size bone repair in vivo
To evaluate the in vivo effects of the chip under physiological conditions, we established a critical-size bone defect rabbit model by creating a 10 mm segmental defect in the midshaft of the femur (Fig. 5c). Following the principles of fracture treatment, a clinically used 8-hole Ti6Al4V plate was applied to fix and secure to rabbit femurs with φ2.4 mm screws (Fig. 5a, b). Animals were randomly divided into three experimental groups: (1) the control group, in which only Ti6Al4V plates were implanted, serving as a baseline to evaluate the natural healing process of critical-size bone defects; (2) the film group, in which the film was affixed to the Ti6Al4V plates, allowing for the assessment of the impact of material composition, surface morphology, and physiological vibration (in the absence of piezoelectric effects) on bone healing; and (3) the chip group, in which the chip was attached to the Ti6Al4V plate, facilitating the evaluation of the combined effects of material properties, mechanical stimulation, and bioelectric activity on the repair of critical-size bone defects. The chip or the film was adhered at the center of the plate by using polydimethylsiloxane (PDMS) as a biosafe adhesive (Fig. 5a, b). It should be noted that no external vibration was applied in the in vivo study because the physiological movement and skeletal muscle activity of rabbits can induce vibration and mechanical stimulation at the defect site.
Photos of the chip on the Ti6Al4V plate (a, b) and the implantation of the chip (c). The chip is highlighted by the red arrow. Representative 2D X-ray images (d) and 3D reconstruction images (e) of bone defect in the control, film, and chip groups at 4 weeks post-operation. f Quantitative analysis of BV, BV/TV, Tb. N and Tb. Sp derived for panel (e). Data are presented as mean ± SD (n = 6 biological replicates). Two-sided ANOVA with a Tukey’s post hoc test for multiple comparisons. Source data are provided as a Source Data file.
At 4 weeks post-surgery, bone regeneration at the defect site was analyzed using X-ray and micro-CT. As presented in Fig. 5d, X-ray images suggested that the femurs were stably fixed with plates and screws in all groups. However, bone defect showed negligible healing in the control group. Obvious new bone formation was found in the film group, and the bone defects were almost completely healed in the chip group. 3D reconstructed micro-CT images (Fig. 5e) showed that negligible mineralized tissue was formed in the control, indicating that critical-size bone defects are difficult to achieve self-healing. The film + vibration slightly promoted bone repair while most defects remained unhealed. Notably, a large amount of new bone tissue formation was observed in the chip group. Quantitative analysis of bone morphological parameters further supported these findings (Fig. 5f). The chip group exhibited the highest bone volume (BV), bone volume/total volume (BV/TV), and trabeculae number (Tb.N). Trabecular separation (Tb.Sp) refers to the average width of the medullary cavity between the bone trabeculae. An increase in Tb. Sp indicates an increase in bone resorption and osteoporosis, and thus, the chip group had a significantly lower Tb.Sp value relative to the film and control groups.
H&E staining (Fig. 6a) revealed that the control and vibration groups formed only thin layers of fibrous tissue and small amounts of calcified tissue on the surface of plates, leaving the defect region largely unhealed. In contrast, the chip group showed extensive new bone tissue formation, with a structure and mineralization pattern closely resembling native bone. Remarkably, newly formed bone tissues nearly bridged the defect. Masson’s trichrome staining (Fig. 6b) confirmed the mature neo-bone (blue) and collagen fibrils in the chip group, while other groups showed fibrous tissue and sporadically distributed immature bone (red). Goldner’s trichrome staining (Fig. 6c) revealed that larger areas of mineralized bone (green) and osteoid-like tissue (orange) were formed in the chip group.
Representative images of H&E staining (a), Masson’s trichrome staining (b) and Goldner’s trichrome staining (c) in different groups. Immunohistochemistry staining of BMP-2 (d) and Runx2 (e) at 4 weeks post-operation. The newly formed bone tissues were highlight by black arrows. F represent fibrous tissues, O represents old bone tissues, Ti represents Ti6Al4V plate, and S represents screws. All experiments were repeated independently at least three times with similar results.
Immunohistochemical analyses of Runx2 and BMP-2 were performed at 4 weeks post-surgery. As shown in Fig. 6d, e, the chip group exhibited significantly higher expression levels of Runx2 and BMP-2 compared to the film and control groups. Extensive patches of brown-stained, newly formed bone tissue were found in the chip group, while the film and control groups displayed only fibrous connective tissues with pale staining.
Finally, H&E staining of major organs (Supplementary Fig. 10) showed that all organs had normal morphological structures, and no abnormalities or intergroup differences were found among all groups after treatment. Taken together, these results demonstrated that the chip provided a robust and biocompatible platform for accelerating bone regeneration through the self-reinforced piezoelectric property, leading to an efficient treatment of critical-sized bone defect models without causing any undesired side effects.
The chip promotes angiogenesis in vivo
To evaluate the angiogenic potential of the chip in vivo, a subcutaneous implantation model was established (Supplementary Fig. 11a). After 4 weeks of implantation, more vessels were observed in the chip group compared to the film group (Supplementary Fig. 11b). H&E staining of tissues beneath the materials indicated that there were abundant newly formed microvascular structures in the chip group, while vascular structures were negligible in the film group (Supplementary Fig. 12a). Immunohistochemical staining showed that tissues in the chip group highly expressed CD31 and VEGF-A. In contrast, CD31 and VEGF-A expressions were low in the film group (Supplementary Fig. 12, b c).
Discussion
Piezoelectric materials have gained growing attention for their potential to provide bioelectric signals to facilitate bone defect repair9,21. Traditional inorganic piezoelectric materials, such as piezoelectric lead zirconium titanate22 and BaTiO323, exhibit high piezoelectric coefficients but face significant limitations, including toxicity, chemical instability, and inadequate mechanical properties for biomedical applications. Organic piezoelectric materials, such as polyvinylidene fluoride24 and poly-L-lactic acid25, offer improved biocompatibility and flexibility. However, their piezoelectric performances heavily rely on external physical stimulation and tend to degrade under environmental factors such as humidity and temperature8. Given these facts, novel materials with stable piezoelectric performance, biocompatibility, and mechanical robustness are critical for advancing bone defect repair strategies26,27,28.
To address these unmet needs, we design and develop the smart self-reinforced chip based on the third-generation semiconductor AlN. Integrating the unique internal cavity structure with the inherent chemical stability and piezoelectric properties of AlN29, our chip is capable of generating sustained bioelectric signals in response to physiological-level vibrations without the need for external power sources. This feature distinguishes our chip from conventional organic piezoelectric materials, which often require pre-polarization or continuous mechanical stimulation to maintain functionality8. By optimizing its crystal orientation (e.g., c-axis alignment) and incorporating an internal cavity structure, the piezoelectric output of the chip was significantly enhanced30,31. Furthermore, the chip can be easily adhered to the Ti6Al4V plate to generate a localized bioelectric microenvironment in the bone defect region. It should be noted that the Si substrate of the chip or the film was fixed on the bottom of the culture plate or Ti6Al4V plate with PDMS. Hence, cells will only interact with the AlN and Mo surface. Interestingly, BMSCs in the Mo and AlN groups exhibited increased cell proliferation relative to the control group, suggesting AlN and Mo on the surface of the chip could promote cell proliferation32. Moreover, the Ti6Al4V plate and screws used in our study served as a fixation system according to the clinical principle of fracture management rather than as a scaffold. Unlike conventional scaffolds that are designed to provide a three-dimensional environment for cell proliferation and differentiation33,34, our chip-based strategy is a scaffold-free strategy to avoid scaffold-related complications. This strategy can also be extended to many other clinically used orthopedic plate systems, such as magnesium alloy, titanium alloy, stainless steel, and CoCrMo alloy (Supplementary Fig. 13).
Compared to previously reported piezoelectric biomaterials for bone repair, our study presents several advantages in both material design and application potential. As summarized in Supplementary Table 1, most existing strategies rely on scaffold-based materials with external physical stimulation (e.g., ultrasound, direct current) to activate piezoelectric effects. These approaches utilized non-load-bearing calvarial or mandibular bone defect models in small animals, where bone healing is relatively fast. In contrast, we achieved scaffold-free bone regeneration using a cavity-engineered and self-reinforced AlN-based chip in a critical-sized femoral defect model of a load-bearing long bone in rabbits. By optimizing the internal cavity structure, the piezoelectric performance of AlN was significantly enhanced, enabling sustained bioelectric signaling under physiological motion. This strategy led to substantial bone regeneration within four weeks, with improved micro-CT outcomes compared to previously reports35,36,37,38,39,40,41. Moreover, the entire regenerative process occurred without the need for any bone grafts, scaffolds, or external physical stimulation.
The in vitro experiments demonstrated the remarkable efficacy of the chip in bone repair. Previous studies have found that vibration can serve as a positive physical stimulus that can enhance cell migration42 and osteogenic differentiation43. Vibration can also upregulate BMP-244 and Runx245 expression. Our results were consistent with these reports. We also observed significantly increased BMP-2 secretion in the chip group, as validated by ELISA. Calcium deposition is the key marker of late-stage osteogenesis46,47. The Ca: P ratio of deposited minerals on the chip (~1.46) is very close to that of natural bone tissue (~1.71)48, indicating the chip could indeed promote osteogenic mineralization. As compared to the film + vibration group, we showed that the vibration accompanied with the piezoelectric effect by the chip significantly promoted proliferation and osteogenic differentiation of BMSCs by activating the PI3K/Akt signaling pathway in vitro. The PI3K/Akt pathway is extensively involved in osteoblast proliferation49, migration50, and differentiation51. This pathway also plays a central role in regulating bone formation, maintenance, and reconstruction52,53. Upon upstream stimulation, PI3K products activate Akt, leading to Akt phosphorylation (p-Akt), which in turn activates downstream target osteogenic proteins such as Runx2, ALP, and osteocalcin (OCN) to promote bone repair54. In vivo radiographic evaluation and histological study revealed that new bone tissue mainly formed from both ends of the defect (the native bone edges), and then bridged toward the center of the defect. Although some mineralized tissue was observed on the chip surface, the primary mode of new bone regeneration was bilateral end-to-center ingrowth, rather than island-like ossification. That is to say, the bone healing process follows the mechanism of “distance osteogenesis” rather than “contact osteogenesis”. Angiogenesis is closely related to osteogenesis during bone repair55. The formation of a functional vascular network is critical for successful bone regeneration, as it facilitates cell recruitment, cytokine secretion, and the transport of oxygen, nutrients, and metabolic waste56. The chip promoted endothelial cell migration and tube formation, with more extensive vascular networks. Specifically, HUVECs in the chip + vibration group displayed the highest gene expression levels of VEGF-A and CD31, followed by the film + vibration group, suggesting that electrical stimulation significantly enhanced the angiogenesis57. In addition, the control + vibration group also showed a statistically significant upregulation of these genes compared to the control group, suggesting the positive role of vibration in angiogenesis58. The expression levels of CD31 and VEGF-A were also significantly upregulated and further confirmed by ELISA, western blotting, and immunofluorescent staining. These results were coupled with the transcriptome sequencing results that the chip could activate the PI3K/Akt signaling pathway and upregulate the angiogenic genes, which play a role in vascular formation59,60. The in vivo angiogenic assay further demonstrated that the chip could promote angiogenesis (as determined by H&E staining and the elevated expression of CD31 and VEGF-A), consisting with the in vitro results. Hence, the chip significantly enhances mineralization and vascular density of the defect region in a critical-size bone defect model without using any bone scaffolds and assistant stimulation, demonstrating its strong osteoinductivity and good biocompatibility for long-term implantation.
There are also some limitations in this work. We only adopted a single vibration modality in vitro. Since different intensity and frequency of vibration modality will affect cell proliferation and differentiation, future work should optimize vibration parameter to mimic physiological mechanical stimuli. Moreover, future studies should focus on optimizing the chip design for large-animal models, evaluating long-term biocompatibility and stability. In addition, placing two chips on the two edges of the bone defect may further accelerate the bone healing process, and thus a dual-chip or multi-chip strategy placed at both ends of the bone defect can be explored.
In sum, this study develops a smart self-reinforced piezoelectric chip by using the third-generation semiconductor AlN with an optimal internal cavity. This chip provides an enhanced and sustained bioelectric environment that promotes osteogenesis and vascularization, ultimately accelerating critical-sized bone regeneration in vivo. Our strategy is universal in bone repair and shows great potential for clinical translation. First, the chip is biocompatible and will not cause side effects after long-term implantation. Second, the chip does not require any additional external stimulation to produce the piezoelectric effect, nor does it require any bone scaffolds or cell transfer to achieve bone repair. Finally, the chip might be applied to many clinically used plate systems.
Methods
Preparation of the chip
The shallow Si cavity structures with different areas (2 × 3, 6 × 3, 10 × 3 and 14 × 3 mm2) are etched on the surface of the Si substrate through optical lithography and chemical etching. On this patterned surface, epitaxial phosphorus-doped silicate glass (PSG) is grown using chemical vapor deposition (CVD). Subsequently, chemical mechanical polishing is performed to ensure that the retained PSG fills the shallow Si cavities with a certain thickness and the removal of excess PSG. Next, the preparation of the Mo/AlN/Mo layer is carried out. A 200 nm Mo bottom electrode is deposited onto the Si/PSG substrate via sputtering. Using lithography techniques, surface patterning is performed to further achieve a 1 µm AlN piezoelectric layer and a 200 nm Mo top electrode. The AlN layer is fabricated through sputtering using a pure Al target (99.9995%) under controlled working pressure in a direct current pulse power system. Finally, the Si cavity is revealed by using wet etching to remove the PSG. The thickness of the Si substrate is further polished and thinned, and the piezoelectric performance of the chip would be further enhanced. After that, the appropriate amount of PDMS is used to effectively stick the chip to the Ti6Al4V plate and prevent it from being squeezed and broken.
Characterization of the chip
The surface morphology of the chip was characterized by scanning electron microscope (SEM; Hitachi, SU-70) and atomic force microscope (AFM; Brucker Edge). The X-ray diffraction (XRD; RIGAKU, SmartLab 9 kW) was used to determine the quality of the AlN film. The electrical signals of the prepared chip with different active areas of Si cavity were measured using a source table, including an SR570 low-noise current preamplifier and a DS345 function generator.
Electrical performance simulation of the chip
A 2D simulation model for the film and chips was constructed using COMSOL Multiphysics 5.4 software. Relevant material parameters and resonator dimensions were referenced from settings in the ADS electrical equivalent model and from COMSOL’s built-in material properties. The thickness of the Mo top electrode, AlN, and Mo bottom electrode are 200 nm, 1 μm, and 200 nm, respectively, while the cavity depth is 3.3 μm, with variable lengths of 2, 6, 10, and 14 mm. To address the contradiction between the infinite propagation of acoustic waves in the real 2D model and the limited computational domain, a perfectly matched layer (PML) was introduced to eliminate boundary reflections, simplifying the computation. Consequently, PML was applied to truncate the boundaries in the 2D simulation model of chips, thereby enhancing computational efficiency. The materials from the software’s material library were assigned to different geometric regions, categorizing the device into the piezoelectric layer and the substrate layer based on structure. Furthermore, multiple physics settings were applied in the simulation software: within the solid mechanics field, the piezoelectric layer was selected, with fixed constraints applied to the longitudinal boundary of functional layers except for the top electrode, and free boundaries were defined for the transverse boundary. In the electrostatic field, the bottom electrode was grounded, and 1 V was applied to the top electrode. By adjusting the cavity size within the 2D simulation model of chips, the amplitude and potential distribution across the cross-section were analyzed. Displacement expressions and potential plots were then generated through the Model Developer window.
To evaluate the electrical performance of the AlN chip under physiological conditions, we designed a custom testing platform simulating a bone defect environment. A circular vibration stage (diameter: 3 cm) was fabricated with a central rectangular gap (10 × 5 mm²) to mimic a load-bearing bone defect. The AlN chip was adhered to a 4-hole Ti6Al4V plate using PDMS and fixed over the gap, ensuring that the central cavity region was suspended and could undergo mechanical deformation. The assembled structure was subjected to vibrational stimulation using a vibration bioreactor delivering a vertical compressive load of 0.5 N at 1 − 4 Hz61,62,63. This parameter was chosen to mimic physiological micro-movements transmitted to orthopedic implants during locomotion64. Copper conductors were connected to both electrodes of the chip by silver paste to record current output. The electrical signals of the prepared AlN chips were measured using a source table, including an SR570 low-noise current preamplifier and a DS345 function generator. For quantitative analysis, current values of chips with different cavities were recorded under vibration, and 11 randomly selected data points from each group were further analyzed to determine the current density on the chip surface.
Cell culture and vibration stimulation
The in vitro experiments aimed to investigate the piezoelectric effects of the chip on osteogenic and angiogenic activities. Both bone marrow-derived mesenchymal stem cells (BMSCs) and human umbilical vein endothelial cells (HUVECs) were divided into four experimental groups: (1) the control group, in which cells were cultured under standard conditions without additional materials or mechanical stimulation; (2) the control + vibration group, in which cells were seeded on the culture plate and subjected to daily vibrational stimulation; (3) the film + vibration group, in which the film was adhered to the culture plate, and cells were cultured under vibrational stimulation; and (4) the chip + vibration group, in which the chip was adhered to the culture plate, and cells were cultured under vibrational stimulation.
BMSCs derived from rats (RASMX-01001, Cyagen Biosciences (Guangzhou) Inc., China) were cultured in α-MEM with 10% fetal bovine serum and 1% penicillin-streptomycin. The cells were maintained at 37 °C in a 5% CO2 humidified incubator and switched to osteogenic medium (RAXMX-90021, Cyagen Biosciences (Guangzhou) Inc., China) upon reaching 90% confluency. Cells from passages 3–5 were used for subsequent in vitro experiments. HUVECs (HUVEC-20001, Cyagen Biosciences (Guangzhou) Inc., China) were cultured in endothelial cell medium (HUVEC-90011, Cyagen Biosciences (Guangzhou) Inc., China), and cells from passages 3–5 were used for angiogenesis assays.
Vibrational stimulation was applied by fixing the cell culture plate onto the stage of a custom-built vibration bioreactor. A vertical compressive load of 0.5 N at 1 Hz was delivered once daily for 10 min. These parameters were selected to mimic physiological micro-movements transmitted to the bone surface during locomotion, as previously reported in murine lower-limb models61,62,63,64.
Cell viability
For live/dead staining, BMSCs were seeded on the film or chip and incubated in 12-well plates at 2 × 104 cells/cm2. After 24, 48, and 72 h of incubation. Cell viability was assessed using a live/dead staining kit (C2015M, Beyotime, Shanghai, China), which utilizes Calcein-AM to stain live cells and propidium iodide (PI) to stain dead cells (red). Fluorescence microscopy was used to capture images. The mean fluorescent intensity of Calcein-AM (n = 3 biological replicates per group) was analyzed using ImageJ software.
As to the cell counting kit-8 (CCK-8) assay, BMSCs (2 × 104 cells/well) were seeded on the film or chip and cultured in 6-well plates (n = 3 biological replicates per group). Cell proliferation was assessed using a CCK-8 assay (CK04, Dojindo, Japan) following the manufacturer’s protocol. After 1, 2, and 3 days, the medium was replaced with culture medium containing 10% CCK-8 and incubated at 37 °C for 1.5 h. Subsequently, 100 μL of supernatant was transferred to a 96-well plate, and absorbance was measured at 450 nm using a microplate reader (SYNERGY H1 BioTek).
ALP activity and Alizarin red staining
BMSCs in different groups were cultured with osteogenic medium for 7 days, and stained using a BCIP/NBT alkaline phosphatase kit (C3206, Beyotime, Shanghai, China). ALP activity was measured using a kit (BC6085, Solarbio, Beijing, China) following the manufacturer’s instructions, and absorbance was recorded at 510 nm (n = 3 biological replicates per group).
For Alizarin red staining, BMSCs were cultured with osteogenic medium for 21 days, and mineralized matrix nodules were stained and quantified using an Alizarin Red S Staining Quantitative Detection Kit (G3283, Solarbio, Beijing, China) following the manufacturer’s instructions, and absorbance was recorded at 560 nm (n = 3 biological replicates per group).
Ca: P ratio measurement
To further evaluate the mineralization, the Ca:P ratio of the deposited minerals on the chip or the film was assessed using EDS after 21 days of osteogenic induction.
Wound healing assay and tube formation assay
To assess the angiogenic potential, HUVECs were seeded onto a 6-well plate and cultured until they reached 100 % confluence. A straight-line scratch was made in the middle of the well using a pipette tip. After 24 h, the wells were washed with PBS to remove the detached cells. HUVECs were stained with Calcein-AM and imaged using a fluorescence microscope (DM4000, Leica, Germany). The wound healing rate was calculated (n = 3 biological replicates per group).
To observe tube formation, HUVECs were seeded at a density of 2 × 104 cells/well onto 24-well plates pre-coated with Matrigel (C0372, Beyotime, China) and cultured in ECM. The plates were incubated at 37 °C with 5% CO2 for 6, 12, and 24 h. Tubule-like structures were stained with Calcein-AM and imaged using a fluorescence microscope. Total branch length was quantified using ImageJ software (n = 3 biological replicates per group).
Transcriptome sequencing
BMSCs were seeded on the film or chip under vibration and cultured in osteoconductive medium for 7 days. Total RNA was extracted using the Universal RNA Extraction Kit (R1200, Solarbio, China). RNA sequencing (n = 3 biological replicates per group) was performed by AiJi Co., Ltd using next-generation sequencing. Libraries were prepared with the NEBNext® Ultra™ RNA Library Prep Kit for Illumina® (NEB, USA), following the manufacturer’s instructions. mRNA was isolated with oligo d(T)25 magnetic beads, fragmented, and reverse-transcribed into cDNA, followed by second-strand synthesis, end-repair, adapter ligation, and size selection. Libraries were amplified by PCR and quality-checked on an Agilent Bioanalyzer 2100 system. Clean reads were aligned to the reference genome using Hisat2 v2.0.5, and gene expression was quantified with RSEM. DEGs were identified using DESeq2, with adjusted P-values < 0.05 and fold changes > 2. GO and KEGG pathway enrichment analyses were performed using the clusterProfiler R package, with P-values < 0.05 considered significant.
RT-qPCR
The RNA concentration was determined by a spectrophotometer (Nanodrop one, ThermoFisher Scientific). Complementary DNA was synthesized using PrimeScript RT Reagent kit (RR037A, Takara, Japan). The RT-qPCR was performed using the SYBR Green polymerase chain reaction (PCR) reagent on the LightCycler480 system (n = 3 biological replicates per group). The primer sequences are shown in Supplementary Table 2. GAPDH and β-actin were served as the house-keeping genes.
Immunofluorescence staining
Immunofluorescence staining was conducted to evaluate the effects of the chip on osteogenesis and angiogenesis. Cells were fixed with 4% paraformaldehyde and blocked with 3% BSA for 30 min each. Primary antibodies were applied overnight at 4 °C. Samples were then incubated with secondary antibody for 1 h. The nucleus was stained with 4,6-diamidino- 2-phenylindole (DAPI, Solarbio, China) for 5 min, and the composites were observed using a fluorescent microscope (FV1000, Olympus, Japan) in a glass-bottom plate (NEST, WuXi, China). The images were processed with the FV10-ASW (version 3.0, Olympus, Japan), and the fluorescence intensities were semi-quantified by the ImageJ software (n = 3 biological replicates per group).
Western blotting
Total protein was extracted from cells using high-efficiency RIPA lysis buffer containing 1% protease inhibitor and 1% phosphatase inhibitor. The protein concentration of each sample was measured using a BCA assay kit (P0010S, Beyotime, Shanghai, China). The sample was quantified by adding 5 times of loading buffer at different concentrations. The protein was separated by polyacrylamide gel electrophoresis and transferred to a PVDF membrane. Then the membrane was blocked with protein-free rapid blocking buffer at room temperature for 15 min, and then incubated with the primary antibodies at 4 °C overnight. On the second day, the membrane was incubated with a diluted solution of secondary antibodies at room temperature for 2 h. The membranes were visualized using a fluorescence gel documentation system (UVItec Ltd., Cambridge) and the protein gray values were quantified by ImageJ (n = 3 biological replicates per group).
ELISA
After osteogenic induction for 7 days or angiogenic induction for 3 days following the above-mentioned methods, the culture supernatants of BMSCs and HUVECs were collected (n = 3 biological replicates per group). The content of BMP-2 (MU30099, Bioswamp, China), CD31 (HM10909, Bioswamp, China), and VEGF-A (HM10978, Bioswamp, China) were measured using ELISA kits following the manufacturer’s instructions, respectively.
Surgical procedure and treatment
To evaluate the osteogenic effect of the chip in vivo, eighteen healthy New Zealand White rabbits (weighing 2.5–3.0 kg, average 2.71 kg; sex ratio 1:1) were randomly divided into three groups (n = 6 biological replicates per group): control, film, and chip. This study adhered to the ARRIVE 2.0 guidelines65. Anesthesia was induced via intramuscular injection of 3% pentobarbital (1 mL/kg) and xylazine hydrochloride (0.1 mL/kg). After shaving the posterior limbs, rabbits were positioned laterally on the operating table. The surgical area was disinfected with iodine, covered with a sterile drape, and a 10 cm incision was made to expose the femur. A 10 mm mid-femoral section was removed to create a bone defect model, which was thoroughly rinsed with saline to eliminate bone debris. Then, the rabbits were randomly divided into three experimental groups: (1) the control group, in which only Ti6Al4V plates were implanted; (2) the film group, in which the film was affixed to the Ti6Al4V plates; and (3) the chip group, in which the chip was adhered to the Ti6Al4V plates. Incisions were closed in layers using silk sutures, and the skin was sterilized with iodine. Postoperative care included intramuscular administration of penicillin sodium (800,000 U/d) for 3 days to prevent infection. The rabbits were not subjected to any immobilization post-surgery and were allowed full physiological weight-bearing activity throughout the recovery period. At 4 weeks post-surgery, the rabbits were euthanized. The implants and surrounding tissues, as well as the heart, liver, spleen, lungs, and kidneys, were collected and fixed in 4% paraformaldehyde for further analysis.
Subcutaneous implantation mice model
To evaluate the angiogenic potential of the chip in vivo, a subcutaneous implantation model was established in male 6-week-old BALB/c nude mice (n = 3 per group). Mice were anesthetized, a 1 cm incision was made in the dorsal skin, and the chip or the film were implanted with the Si substrate facing upwards. Thus, the AlN/Mo surface of the chip or the film could contact the dorsal tissue. After 28 days of implantation, the materials were taken out, and photos of the tissue beneath the materials were taken, and histological staining was performed to observe the formation of vessels.
Micro-CT analysis
New bone formation was analyzed using a micro-computed tomography (micro-CT) imaging system (ZKKS-MCT-Sharp). Specimens were scanned with a resolution of ~ 13 μm, 360° rotation, 70 kV source voltage, and 100 μA beam current. Two-dimensional images were generated using Medproject 4.1 software, while 3D images were reconstructed with Recon 1.0. The region of interest (ROI), corresponding to the implant area (length: 10 mm), was marked for analysis. The beam hardening correction was applied to reduce metal artifacts. The 3D microarchitecture within the ROI was reconstructed, and BV, BV/TV, Tb. N and Tb. Sp were calculated (n = 6 biological replicates per group).
Histologic staining
After micro-CT analysis, the samples were dehydrated, polymerized, and sectioned into 500 μm longitudinal slices. Sections were stained with H&E (G1120, Solarbio, Beijing, China), Masson’s trichrome (G1340, Solarbio, Beijing, China) and Goldner’s trichrome staining (G3550, Solarbio, Beijing, China) following standard protocols. All sections were examined, and images were captured for analysis. Critical organs, including the heart, lungs, liver, spleen, and kidneys, were also subjected to H&E staining.
For immunohistochemistry staining, sections underwent endogenous peroxidase reduction, antigen retrieval, and nonspecific background elimination before being incubated overnight at 4 °C with primary antibodies. Positive staining was visualized the following day using the secondary antibody tool kit (KIT-9707, MXB, China). Counterstaining with eosin was performed, followed by rinsing with distilled water. Finally, sections were dehydrated, cleared with xylene, and mounted with neutral resin for microscopic examination.
In addition, for the evaluation of angiogenesis in vivo, tissues beneath the materials were harvested and subjected to H&E staining (G1120, Solarbio, Beijing, China). Immunohistochemical staining of CD31 and VEGF-A were also performed following the same protocol as mentioned above.
Antibodies
Western blotting and immunofluorescence staining employ the following antibodies: anti-Runx2 (YM8347, Immunoway, USA), anti-BMP-2 (YM8215, Immunoway, USA), anti-Col-1a (ab255809, Abcam, UK), anti-CD31 (YM8079, Immunoway, USA), anti-VEGF-A (YT4870, Immunoway, USA), anti-p-PI3k (YP0765, Immunoway, USA), anti-p-Akt (YM8304, Immunoway, USA), and anti-β-actin (20536-1-AP, Proteintech, China). The dilutions of each antibody for western blotting and immunofluorescence staining were 1:1000 and 1:200, respectively. Immunohistochemistry staining employs the following antibodies: anti-Runx2 (1:1000, ab23981, Abcam, UK) and anti-BMP-2 (1:1000, ab6285, Abcam, UK); CD31 (1:1000, ab182981, Abcam, UK); VEGF-A (1:1000, ab46154, Abcam, UK).
Statistical analysis
All data were analyzed by the SPSS 26.0 (IBM Corp., Armonk, NY) software and presented as the mean ± standard deviation (SD). The two-tailed unpaired Student’s t test was used to compare two groups, and a one-way ANOVA with Tukey’s test for multiple comparisons. A P-value less than 0.05 was considered statistically significant.
Ethics statement
Every experiment involving animals have been carried out following the animal care and experimental protocols approved by the Animal Care and Use Committee of Zhujiang Hospital (Approved No.: LAEC-2022-209).
Reporting summary
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.
Data availability
All data supporting the findings of this study are available within the article and its supplementary files. Any additional requests for information can be directed to and will be fulfilled by the corresponding authors. Source data are provided in this paper.
References
Schemitsch, E. H. Size matters: Defining critical in bone defect size!. J. Orthop. Trauma 31, S20–S22 (2017).
Dimitriou, R., Jones, E., McGonagle, D. & Giannoudis, P. V. Bone regeneration: current concepts and future directions. BMC Med. 9, 66 (2011).
Yuan, X. et al. Recent advances in 3D printing of smart scaffolds for bone tissue engineering and regeneration. Adv. Mater. 36, e2403641 (2024).
Lu, J., Gao, Z., He, W. & Lu, Y. Harnessing the potential of hyaluronic acid methacrylate (HAMA) hydrogel for clinical applications in orthopaedic diseases. J. Orthop. Transl. 50, 111–128 (2025).
Nain, A. et al. Progress in the development of piezoelectric biomaterials for tissue remodeling. Biomaterials 307, 122528 (2024).
Zhang, C. et al. Modulating surface potential by controlling the β phase content in poly(vinylidene fluoridetrifluoroethylene) membranes enhances bone regeneration. Adv. Health. Mater. 7, e1701466 (2018).
Liu, C. et al. LIPUS activated piezoelectric pPLLA/SrSiO(3) composite scaffold promotes osteochondral regeneration through P2RX1 mediated Ca(2+) signaling pathway. Biomaterials 317, 123084 (2025).
Zhang, J. et al. Advanced piezoelectric materials, devices, and systems for orthopedic medicine. Adv. Sci.12, e2410400 (2024).
Yuan, X. et al. Piezoelectricity, pyroelectricity, and ferroelectricity in biomaterials and biomedical applications. Adv. Mater. 36, e2308726 (2024).
Bai, L., Song, P. & Su, J. Bioactive elements manipulate bone regeneration. Biomater. Transl. 4, 248–269 (2023).
Tang, X. et al. Wafer-scale vertical 1D GaN nanorods/2D MoS(2)/PEDOT:PSS for piezophototronic effect-enhanced self-powered flexible photodetectors. Nanomicro Lett. 17, 56 (2024).
Zheng, Y. et al. Vertical 1D/2D heterojunction architectures for self-powered photodetection application: GaN nanorods grown on transition metal dichalcogenides. ACS Nano 16, 2798–2810 (2022).
Zhang, C. et al. A novel approach to enhance bone regeneration by controlling the polarity of GaN/AlGaN heterostructures. Adv. Funct. Mater. 31, 2007487 (2021).
Wang, W. et al. 2D AlN Layers sandwiched between graphene and Si substrates. Adv. Mater. 31, e1803448 (2019).
Wang, H., Lin, Z., Wang, W., Li, G. & Luo, J. Growth mechanisms of GaN epitaxial films grown on ex situ low-temperature AlN templates on Si substrates by the combination methods of PLD and MOCVD. J. Alloy Compd. 718, 28–35 (2017).
Ben, J. et al. 2D III-Nitride materials: Properties, growth, and applications. Adv. Mater. 33, e2006761 (2021).
Wang, J. et al. Group-III nitride heteroepitaxial films approaching bulk-class quality. Nat. Mater. 22, 853–859 (2023).
Iqbal, A. & Mohd-Yasin, F. Reactive sputtering of aluminum nitride (002) thin films for piezoelectric applications: A review. Sensors 18, 1797 (2018).
Cinquino, M. et al. Enhancing cardiovascular health monitoring: Simultaneous multi-artery cardiac markers recording with flexible and bio-compatible AlN piezoelectric sensors. Biosens. Bioelectron. 267, 116790 (2024).
Desideri, D., Bernardo, E., Corso, A. J., Moro, F. & Pelizzo, M. G. Electrical properties of aluminum nitride thick films magnetron sputtered on aluminum substrates. Materials 15, 2090 (2022).
Luo, S. et al. Advances in electroactive biomaterials: Through the lens of electrical stimulation promoting bone regeneration strategy. J. Orthop. Transl. 47, 191–206 (2024).
Li, J. et al. Lead zirconate titanate ceramics with aligned crystallite grains. Science 380, 87–93 (2023).
Lee, M. et al. Ultrathin piezoelectric resonators based on graphene and free-standing single-crystal BaTiO3. Adv. Mater. 34, e2204630 (2022).
Fernandez-Yague, M. A. et al. A tympanic piezo-bioreactor modulates ion channel-associated mechanosignaling to stabilize phenotype and promote tenogenesis in human tendon-derived cells. Adv. Sci. 11, e2405711 (2024).
Liu, Y. et al. Exercise-induced piezoelectric stimulation for cartilage regeneration in rabbits. Sci. Transl. Med. 14, eabi7282 (2022).
Li, J. et al. Bioelectret materials and their bioelectric effects for tissue repair: A review. ACS Appl. Mater. Interfaces 16, 38852–38879 (2024).
Chen, H. et al. A sequential stimuli-responsive hydrogel promotes structural and functional recovery of severe spinal cord injury. Biomaterials 316, 122995 (2025).
Wu, C. et al. Efficiently directing differentiation and homing of mesenchymal stem cells to boost cartilage repair in osteoarthritis via a nanoparticle and peptide dual-engineering strategy. Biomaterials 312, 122720 (2024).
Lin, T. et al. Two-dimensional material/group-III nitride hetero-structures and devices. Rep. Prog. Phys. 88, 046501 (2025).
Chen, C. et al. Electric field stiffening effect in c-oriented aluminum nitride piezoelectric thin films. ACS Appl Mater. Interfaces 10, 1819–1827 (2018).
Shishkin, R. A., Elagin, A. A., Mayorova, E. S. & Beketov, A. R. The advanced aluminum nitride synthesis methods and its applications: Patent review. Recent Pat. Nanotechnol. 10, 146–156 (2016).
Luo, Y. et al. 3D Molybdenum disulfide nanospheres loaded with metformin to enhance SCPP scaffolds for bone regeneration. ACS Appl Mater. Interfaces 16, 201–216 (2024).
Bahraminasab, M. et al. Bone scaffolds: An incorporation of biomaterials, cells, and biofactors. ACS Biomater. Sci. Eng. 7, 5397–5431 (2021).
Koushik, T. M., Miller, C. M. & Antunes, E. Bone tissue engineering scaffolds: Function of multi-material hierarchically structured scaffolds. Adv. Health. Mater. 12, e2202766 (2023).
Bai, Y. et al. Biomimetic piezoelectric nanocomposite membranes synergistically enhance osteogenesis of deproteinized bovine bone grafts. Int. J. Nanomed. 14, 3015–3026 (2019).
Fan, Y. et al. Piezoelectric heterojunctions as bacteria-killing bone-regenerative implants. Adv. Mater. 37, e2413171 (2025).
Wang, L. et al. A biomimetic piezoelectric scaffold with sustained Mg(2+) release promotes neurogenic and angiogenic differentiation for enhanced bone regeneration. Bioact. Mater. 25, 399–414 (2023).
Fan, B. et al. Electroactive barium titanate coated titanium scaffold improves osteogenesis and osseointegration with low-intensity pulsed ultrasound for large segmental bone defects. Bioact. Mater. 5, 1087–1101 (2020).
Xu, Y. et al. Biomimetic bone-periosteum scaffold for spatiotemporal regulated innervated bone regeneration and therapy of osteosarcoma. J. Nanobiotechnol. 22, 250 (2024).
Li, Y., Fan, Y., Zhao, S. & Cheng, B. Ultrasound-triggered piezoelectric polyetheretherketone with boosted osteogenesis via regulating Akt/GSK3beta/beta-catenin pathway. J. Nanobiotechnol. 22, 539 (2024).
Wang, T. et al. Rehabilitation exercise-driven symbiotic electrical stimulation system accelerating bone regeneration. Sci. Adv. 10, eadi6799 (2024).
Mohammed, T., Murphy, M. F., Lilley, F., Burton, D. R. & Bezombes, F. The effects of acoustic vibration on fibroblast cell migration. Mater. Sci. Eng. C. Mater. Biol. Appl. 69, 1256–1262 (2016).
Shi, H. et al. Integrating physicomechanical and biological strategies for BTE: biomaterials-induced osteogenic differentiation of MSCs. Theranostics 13, 3245–3275 (2023).
Jing, D. et al. Mechanical vibration mitigates the decrease of bone quantity and bone quality of Leptin receptor-deficient Db/Db mice by promoting bone formation and inhibiting bone resorption. J. Bone Min. Res. 31, 1713–1724 (2016).
Yu, X. et al. Low-magnitude vibration induces osteogenic differentiation of bone marrow mesenchymal stem cells via miR-378a-3p/Grb2 pathway to promote bone formation in a rat model of age-related bone loss. FASEB J. 34, 11754–11771 (2020).
Li, Y. et al. The application of extracellular vesicles in orthopedic diseases. Wires Syst. Biol. Med. 2, e20230055 (2024).
Liu, H. & Su, J. Organoid and organoid extracellular vesicles for osteoporotic fractures therapy: Current status and future perspectives. Wires Syst. Biol. Med. 1, e2023011 (2023).
Shah, F. A. Revisiting the physical and chemical nature of the mineral component of bone. Acta Biomater. 196, 1–16 (2025).
Wu, S. S. et al. Omentin-1 stimulates human osteoblast proliferation through PI3K/Akt signal pathway. Int J. Endocrinol. 2013, 368970 (2013).
Katsianou, M. A. et al. Polycystin-1 regulates cell proliferation and migration through AKT/mTORC2 pathway in a human craniosynostosis cell model. J. Cell Mol. Med. 26, 2428–2437 (2022).
Tong, W. et al. Kaiso regulates osteoblast differentiation and mineralization via the Itga10/PI3K/AKT signaling pathway. Int. J. Mol. Med. 47, 41 (2021).
Zhao, S. J. et al. Macrophage MSR1 promotes BMSC osteogenic differentiation and M2-like polarization by activating PI3K/AKT/GSK3β/β-catenin pathway. Theranostics 10, 17–35 (2020).
E, L. et al. Insulin promotes the bone formation capability of human dental pulp stem cells through attenuating the IIS/PI3K/AKT/mTOR pathway axis. Stem Cell Res Ther. 15, 227 (2024).
Sun, K. et al. The PI3K/AKT/mTOR signaling pathway in osteoarthritis: a narrative review. Osteoarthr. Cartil. 28, 400–409 (2020).
Bixel, M. G. et al. Angiogenesis is uncoupled from osteogenesis during calvarial bone regeneration. Nat. Commun. 15, 4575 (2024).
Bosch-Rue, E., Diez-Tercero, L., Buitrago, J. O., Castro, E. & Perez, R. A. Angiogenic and immunomodulation role of ions for initial stages of bone tissue regeneration. Acta Biomater. 166, 14–41 (2023).
Jeong, G. J. et al. Therapeutic angiogenesis via solar cell-facilitated electrical stimulation. ACS Appl. Mater. Interfaces 9, 38344–38355(2017).
Wang, Z. et al. Nano-vibration exciter: Hypoxia-inducible factor 1 signaling pathway-mediated extracellular vesicles as bioactive glass substitutes for bone regeneration. Bioact. Mater. 40, 460–473 (2024).
Long, J. et al. Regulation of osteoimmune microenvironment and osteogenesis by 3D-printed PLAG/black phosphorus scaffolds for bone regeneration. Adv. Sci. 10, e2302539 (2023).
Han, L. et al. Dapagliflozin improves angiogenesis after hindlimb Ischemia through the PI3K-Akt-eNOS pathway. Biomolecules 14, 592 (2024).
Wang, A. et al. Piezoelectric nanofibrous scaffolds as in vivo energy harvesters for modifying fibroblast alignment and proliferation in wound healing. Nano Energy 43, 63–71 (2018).
Eng, J. J. & Winter, D. A. Kinetic analysis of the lower limbs during walking: what information can be gained from a three-dimensional model?. J. Biomech. 28, 753–758 (1995).
Schroeder, R. T., Croft, J. L. & Bertram, J. E. A. Amplitude and frequency of human gait synchronization with a machine oscillator system. Sci. Rep. 15, 1629 (2025).
Priyadarshini, B. S., Mitra, R. & Manju, U. Titania nanoparticle-stimulated ultralow frequency detection and high-pass filter behavior of a flexible piezoelectric nanogenerator: A self-sustaining energy harvester for active motion tracking. ACS Appl. Mater. Interfaces 15, 45812–45822 (2023).
Percie du Sert, N. et al. The ARRIVE guidelines 2.0: Updated guidelines for reporting animal research. PLoS Biol. 18, e3000410 (2020).
Acknowledgements
This study was supported by the Guangdong Basic and Applied Basic Research Foundation (No.: 2024B1515020015 and 2024A04J5347) and the National Key Research and Development Program of China (No.: 2022YFB3604500 and 2022YFB3604501).
Author information
Authors and Affiliations
Contributions
W.L. and Y.M. conceived the study, performed the majority of the experiments, analyzed the data, and wrote the manuscript. S.X. provided support for the characterization of the chip. L.F., Y.Y., Y.Z., and J.L. assisted with the in vivo surgical procedures and sample collection. W.W. and G.L. supervised the chip design and provided guidance for its fabrication. Y.L. and J.T. supervised both the in vitro and in vivo experiments. Y.L. also reviewed, edited, and revised the manuscript.
Corresponding authors
Ethics declarations
Competing interests
The authors declare no competing interests.
Peer review
Peer review information
Nature Communications thanks Gianluca Cidonio, and the other anonymous reviewer(s) for their contribution to the peer review of this work. A peer review file is available.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary information
Source data
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
About this article
Cite this article
Li, W., Mo, Y., Xie, S. et al. Self-reinforced piezoelectric chip for scaffold-free repair of critical-sized bone defects. Nat Commun 16, 5800 (2025). https://doi.org/10.1038/s41467-025-61243-w
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41467-025-61243-w
