Carbon nanotube fiber electrodes modified with surface-arrayed nanowires for enhanced mass transport in electrochemical nanofabrication Authors: Lingchao Meng, Pengxiang Yi, Yabing Li, Zhicheng Ai, Yongchao Yang, Hao Li Published: 2026 ABSTRACT Carbon nanotube fiber (CNF) is a promising wire electrode material for wire electrochemical micromachining due to its high strength, electrical conductivity, and corrosion resistance. However, the smooth and hydrophobic surface of pristine CNF promotes bubble adhesion and interfacial velocity slip, which severely restricts mass transport and degrades machining stability. In this study, a surface-arrayed nanowire-modified carbon nanotube fiber electrode is proposed to enhance mass transport during micromachining. Vertically aligned NiCo₂S₄ nanowires, with an average height of approximately 10 μm, were grown in situ on the fiber surface, resulting in a surface that is both microtextured and functionalized. The nanowire array effectively penetrates the gas cushion, suppresses boundary velocity slip, and strengthens fluid-solid coupling. Flow field simulations and theoretical analysis demonstrate intensified near-wall convection and improved mass transport induced by the microtextured surface. Wire electrochemical micromachining experiments on Ni-based metallic glass show that the modified electrode significantly improves machining performance. The maximum stable feed rate was increased to 0.8 μm/s, the power fluctuation frequency was reduced from 2.2/min to 0.2/min, and the surface roughness Ra was decreased by approximately 60% to 25.5 nm. In addition, high-precision batch fabrication of 25 stepped cantilever micro-components was successfully achieved using the multiwire and multilayer electrochemical machining with a relative standard deviation below 0.1%, demonstrating the engineering applicability of the proposed electrode. This work provides a practical approach for overcoming mass transport limitations in wire electrochemical micromachining through electrode surface microstructure engineering. Introduction The ongoing push toward miniaturization in aerospace and biomedical fields, such as interventional surgical robots, implantable neural interfaces, and bio-absorbable vascular stents, has driven the development of high-precision micromanufacturing technologies1,2. In these applications, surface integrity is critical, as microcracks or thermal alteration can compromise biocompatibility and fatigue life. Among micromachining approaches, wire electrochemical micromachining (WECMM) has attracted considerable attention. Fundamentally, WECMM removes metallic materials through anodic dissolution at the atomic and ionic levels, which theoretically endows it with nanofabrication capabilities. Furthermore, because the process is conducted at room temperature, it generates no residual machining stress and avoids any microstructural deterioration, thereby fully preserving the excellent intrinsic properties of hard-to-machine conductive materials such as superalloys, titanium alloys, and metallic glasses3. In WECMM, machining accuracy critically depends on maintaining a narrow inter-electrode gap (IEG), typically below 20 μm, in which efficient mass transport and controlled electrolyte flow are required to stabilize the electrochemical environment and suppress accumulation of electrolytic products4. However, the confined geometry of the IEG inherently restricts mass transport. Hydrogen bubbles generated at the wire electrode surface tend to accumulate and form a transient gas cushion, distorting the current density distribution and inducing interfacial velocity slip5. Such a gas cushion weakens fluid-solid coupling and reduces the ability of the reciprocating wire to drive fresh electrolyte into the machining zone, leading to unstable dissolution, increased ohmic resistance, and frequent power fluctuation. In addition to mass-transport limitations, Ni-based metallic glass, which is increasingly used in orthopedic and orthodontic applications due to its high elastic limit and corrosion resistance, is prone to passivation in acidic electrolytes. The formation of a dense oxide layer hinders continuous dissolution6,7. While bipolar pulse currents have been shown to depassivate the electrode surface and improve machining stability, they cannot eliminate the severe gas cushion effect in the narrow IEG8,9. Therefore, combining a high-performance electrode structure with an optimized pulse waveform is necessary to address these limitations. Although axial reciprocating motion of the wire electrode can enhance flushing efficiency, the extent of improvement strongly depends on the surface drag characteristics of the electrode. Conventional metallic wires, such as tungsten, platinum, and molybdenum, possess inherently smooth surfaces that generate limited fluid-solid coupling, thereby constraining further enhancement of mass transport. To overcome the limitations of inherently smooth surfaces, recent studies have explored surface modification strategies for conventional metallic wires. For instance, geometric macroscopic modifications, such as the utilization of rotating helical electrodes, have been developed to mechanically stir the electrolyte and facilitate the evacuation of hydrogen bubbles and electrolytic products10. At the microscale, nanosecond laser ablation has been employed to fabricate spike-like microstructures on tungsten wires, which successfully improves surface hydrophilicity and significantly enhances mass transport efficiency during axial reciprocating motion11. However, these conventional modification strategies face inherent limitations. Directly applying physical texturing to wire electrodes often compromises their tensile strength and structural integrity, making them vulnerable to breakage during reciprocating motion. Furthermore, simple physical texturing struggles to achieve the ideal superhydrophilic state required to thoroughly eliminate the gas cushion effect. Therefore, there is an urgent need to develop a new type of wire electrode that can simultaneously support hierarchical surface microtextures and robust mechanical stability. Carbon nanotube fiber (CNF) has recently emerged as an alternative wire electrode material due to its high electrical conductivity, corrosion resistance, and hierarchical bundle structure12–14. Advances in wet spinning and dry spinning techniques have enabled the fabrication of CNF with mechanical and electrical properties suitable for microelectrochemical applications15. However, pristine CNF typically exhibits a relatively smooth and hydrophobic surface16,17, which promotes bubble adhesion and exacerbates interfacial slip, an undesirable characteristic for WECMM where strong fluid-solid coupling is required. Although surface treatments such as acid oxidation or electrochemical activation can improve CNF hydrophilicity, they primarily modify surface chemistry and introduce limited surface microtopography, which is insufficient to disrupt the gas cushion responsible for interfacial slip18. Consequently, existing CNF modification strategies do not simultaneously enhance surface roughness and superhydrophilicity, which is essential for suppressing interfacial slip and strengthening mass transport in the narrow IEG. To address this limitation, we developed a surface-arrayed nanowire-modified carbon nanotube fiber (SA-CNF) to fundamentally enhance mass transport during WECMM. A vertically aligned NiCo₂S₄ nanowire array was synthesized in situ on the CNF surface via a controllable hydrothermal method19,20. NiCo₂S₄ provided a conductive, stable, and highly wettable surface that improved fluid-solid coupling and mass transport. Simulations and experiments showed that SA-CNF enhanced flow dynamics, doubled the feed rate, and improved machining stability and surface finish compared with pristine CNF. These results demonstrated the potential of the proposed strategy for high-precision electrochemical nanofabrication in MEMS and biomedical manufacturing. Mechanistic analysis Influence of wall slip on mass transport in reciprocating WECMM In WECMM, mass transport within the narrow IEG critically affects machining stability and surface quality. Natural convection alone cannot sufficiently refresh the diffusion layer, so axial reciprocating motion of the wire electrode is introduced to enhance local flow and ion transport. As shown in Fig. 1, this motion generates periodic shear flows and transient pressure differences, which continuously disturb and renew the diffusion layer, improving ion replenishment and product removal. This forced convection reduces concentration polarization and suppresses bubble accumulation, especially in high-viscosity or low-diffusivity electrolytes where stable conductivity and uniform anodic dissolution are required.Fig. 1Schematic of mass transport enhancement via wire electrode axial reciprocating motion.a Schematic of the reciprocating WECMM process. b Comparison of bubble and electrolytic product transport around CNF and SA-CNF Schematic of mass transport enhancement via wire electrode axial reciprocating motion. a Schematic of the reciprocating WECMM process. b Comparison of bubble and electrolytic product transport around CNF and SA-CNF To analyze this mechanism, a flow field model of the reciprocating wire electrode in the IEG was established. While conventional fluid mechanics assumes a no-slip boundary, microscale studies show that boundary slip becomes significant and depends on surface wettability and roughness. Since the IEG in WECMM typically ranges from several to tens of micrometers and the flow is laminar, wall slip must be considered for smooth electrodes. As illustrated in Fig. 2, a model incorporating slip boundary conditions was developed. The axial motion of the wire relative to the stationary workpiece is analogous to Couette flow. The model assumes a fully filled gap, a stationary workpiece wall, a moving wire electrode with velocity U, and steady, incompressible laminar flow with velocity varying only in the x direction.Fig. 2Flow field model of the reciprocating wire electrode with wall slip boundary conditions.The left side illustrates the reciprocating motion of the wire electrode relative to the fixed workpiece, while the right side shows the simplified Couette flow model. Here, U denotes the wire velocity, H is the inter-electrode gap, and b1 and b2 represent the slip lengths of the workpiece and wire electrode Flow field model of the reciprocating wire electrode with wall slip boundary conditions. The left side illustrates the reciprocating motion of the wire electrode relative to the fixed workpiece, while the right side shows the simplified Couette flow model. Here, U denotes the wire velocity, H is the inter-electrode gap, and b1 and b2 represent the slip lengths of the workpiece and wire electrode According to the Navier slip condition, the velocity difference between the fluid and the solid wall is proportional to the local shear rate. In this model, b₁ and b₂ represent the slip lengths of the workpiece and wire electrode, respectively. The slip length corresponds to the distance at which the extrapolated velocity profile reaches zero. A larger value indicates lower wall friction. Accordingly, the slip boundary conditions are:1\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$u\left(0\right)={b}_{1}\frac{{d}_{u}}{{d}_{y}}{/}_{y=0}$$\end{document}u0=b1dudy/y=02\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$U-u\left(H\right)={b}_{2}\frac{{d}_{u}}{{d}_{y}}{/}_{y=H}$$\end{document}U−uH=b2dudy/y=H The simplified momentum equation can be written as:3\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\frac{{d}^{2}u}{{{dy}}^{2}}=0$$\end{document}d2udy2=0 Double integration of Eq. (3) yields:4\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$u\left(y\right)={Ay}+B$$\end{document}uy=Ay+B Substituting the boundary conditions into Eq. (4) yields:5\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\left\{\begin{array}{c}u\left(0\right)={b}_{1}A\\ U-\left({AH}+B\right)={b}_{2}A\end{array}\right.$$\end{document}u0=b1AU−AH+B=b2A Solving for constants A and B yields:6\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$A=\frac{U}{H+{b}_{1}+{b}_{2}}$$\end{document}A=UH+b1+b27\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$B=\frac{{b}_{1}U}{H+{b}_{1}+{b}_{2}}$$\end{document}B=b1UH+b1+b2 The resulting velocity profile is expressed as:8\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$u\left(y\right)=\frac{U}{H+{b}_{1}+{b}_{2}}y+\frac{{b}_{1}U}{H+{b}_{1}+{b}_{2}}$$\end{document}uy=UH+b1+b2y+b1UH+b1+b2 In the absence of slip (b1 = b2 = 0), the velocity profile reduces to the classical Couette solution:9\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${u}_{{no}}\left(y\right)=U\frac{y}{H}$$\end{document}unoy=UyH Comparing Eqs. (8) and (9), the velocity difference can be expressed as:10\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\Delta u\left(y\right)={u}_{{slip}}\left(y\right)-{u}_{{no}}\left(y\right)=U\left[\frac{{b}_{1}H-y\left({b}_{1}+{b}_{2}\right)}{H\left(H+{b}_{1}+{b}_{2}\right)}\right]$$\end{document}Δuy=uslipy−unoy=Ub1H−yb1+b2HH+b1+b2 As shown in Fig. 3, a smaller wire electrode slip length b₂ shifts the intersection point α closer to the electrode wall, thereby reducing the velocity loss zone Δu. Furthermore, as b2 decreases, the fluid velocity at the wire electrode wall u(H) approaches the electrode’s motion speed U, and the velocity at the workpiece wall u(0) also increases. Therefore, reducing the wire electrode slip length b₂ is critical for improving mass transport in the IEG.Fig. 3Influence of wall slip on flow velocity distribution in the IEG.U denotes the wire velocity, b1 and b2 are the slip lengths of the workpiece and wire electrode, α is the intersection point of the two velocity profiles, and Δu represents the velocity loss caused by wall slip Influence of wall slip on flow velocity distribution in the IEG. U denotes the wire velocity, b1 and b2 are the slip lengths of the workpiece and wire electrode, α is the intersection point of the two velocity profiles, and Δu represents the velocity loss caused by wall slip Research indicates that slip behavior at the solid-liquid interface depends on microscopic wall morphology21,22. Surface macrotextures modify roughness and local flow fields, forming stagnation zones that enhance near-wall adhesion and reduce the slip length b2. Surface protrusions also disrupt streamlines and generate micro-vortices, promoting momentum exchange and suppressing interfacial slip. Debnath demonstrated that periodic macrotextures on tungsten wires enhance electrolyte flushing and reduce product accumulation within the IEG23. Therefore, modifying electrode surface morphology is an effective way to improve mass transport. Based on this, the proposed SA-CNF electrode features a surface-arrayed nanowire structure, and a fluid dynamics simulation was established to compare velocity distributions around SA-CNF and CNF electrodes, clarifying its effect on flow and slip. Flow field simulation and verification To quantitatively verify the hypothesis of enhanced fluid-solid coupling and visualize mass transport behavior, a CFD simulation was performed. Based on the actual WECMM, a physical model was established in Fig. 4a, including a wire electrode, a workpiece, and a fluid zone. The workpiece thickness Th, wire diameter D, and forward gap Δb were set to 50 μm, 30 μm, and 5 μm. To ensure mechanical stability, the wire electrode motion was defined as sinusoidal.Fig. 4Flow field simulation of smooth and SA-CNF wire electrodes.a Physical model and cross-section definitions. b 2D simulation model, surface mesh, and velocity comparison. c Velocity contours of the smooth electrode. d Velocity contours of the SA-CNF electrode Flow field simulation of smooth and SA-CNF wire electrodes. a Physical model and cross-section definitions. b 2D simulation model, surface mesh, and velocity comparison. c Velocity contours of the smooth electrode. d Velocity contours of the SA-CNF electrode To investigate how the electrode microtopography perturbs the IEG flow field, Section A in Fig. 4a was selected, and a detailed model was built in Fig. 4b. The electrode boundary conditions are also shown in Fig. 4b, with simulation parameters listed in Table 1. The fluid was assumed incompressible, and both the electrode and workpiece were treated as rigid bodies. The front and lateral sparking gaps were identical, and gap dimensions remained unchanged for different electrodes. Effects of electrolytic products and heat were neglected, and viscosity was set to 1.00 × 10⁻³ Pa s. ANSYS Fluent 14.5 was used to simulate the IEG flow field. Flow characteristics were analyzed on the XOZ plane along Section A, and the velocity distribution was further evaluated along Line 1.Table 1Flow field simulation parametersParameterValueWire electrode diameter30 μmWorkpiece thickness50 μmSlit width50 μmFront gap5 μmWire reciprocating amplitude50 μmWire reciprocating frequency6 Hz Table 1Flow field simulation parametersParameterValueWire electrode diameter30 μmWorkpiece thickness50 μmSlit width50 μmFront gap5 μmWire reciprocating amplitude50 μmWire reciprocating frequency6 Hz Flow field simulation parameters Simulations were first conducted to establish a baseline by examining the y-direction velocity distribution of CNF over a quarter motion cycle (t = 0–0.05 s). As shown in Fig. 4c, the flow exhibits typical Couette characteristics. At t = 0, due to fluid viscosity and no-slip conditions, the fluid near the electrode reached a peak velocity, with momentum transferred mainly via viscous effects. However, the smooth surface led to rapid radial velocity decay and a stable laminar structure, limiting disturbance far from the wall and reducing mass transport efficiency. In contrast, the SA-CNF results in Fig. 4d show boundary layer thickening and fluid trapping. High-velocity regions extend toward the gap center, and fluid between nanowires moves synchronously with the electrode. As further confirmed in Fig. 4b, the SA-CNF velocity profile deviates from linear decay, showing a nonlinear convex distribution. Within 0–2 μm, velocity remains near peak, forming a stable locked zone, while the overall curve indicates slower kinetic energy decay. This behavior results from stronger form resistance induced by microstructures, enhancing momentum transfer and increasing flow rate. Consequently, SA-CNF improves fluid transport, facilitating the removal of bubbles and electrolytic products in the IEG. Experimental System and Instrumentation As shown in Fig. 5, the experimental system consisted of a numerically controlled XYZ motion stage, a piezoelectric transducer, a programmable bipolar nanosecond pulse generator, an oscilloscope, a computer-controlled charge-coupled device camera, a multifunction electrolyte tank, and associated fixtures. The multifunction electrolyte tank comprised two separate sections to accommodate different experimental stages. The first section was used for conductivity testing and connection to the fixture. A metallic coating was electroplated at the junction between the CNF and the wire electrode fixture to ensure a stable electrical interface. The second section served as the main processing zone. The CNF was mounted on the wire electrode fixture. The Z-axis motion stage was employed to generate axial reciprocating motion of the fiber, thereby assisting mass transport in the IEG during both electrochemical surface modification and WECMM.Fig. 5Schematic diagram of the WECMM experimental system.The system consists of the motion stages, nanosecond pulse power supply, oscilloscope, CCD camera, multifunction electrolyte tank, and control units Schematic diagram of the WECMM experimental system. The system consists of the motion stages, nanosecond pulse power supply, oscilloscope, CCD camera, multifunction electrolyte tank, and control units Materials In the WECMM experiments, a CNF and a Ni-based metallic glass sample (Ni72Cr19Si7B2) were used as the wire electrode and the anodic workpiece, respectively. Owing to its high corrosion resistance, the metallic glass exhibits typical passivation behavior in dilute acidic electrolytes. Previous studies have reported that metallic glass tends to generate a larger number of electrolytic products than crystalline alloys during WECMM24. To ensure the stability and uniformity of the wire electrode during machining, the preparation process of the CNF is described as follows. First, continuous and self-supporting carbon nanotube sheets were drawn from a spinnable forest via a dry-spinning process, as shown in Fig. 6a. Through this continuous drawing and subsequent twisting process, as shown in Fig. 6b, uniform carbon fibers were prepared. Multiple straight CNF bundles were then twisted into a single yarn and subsequently passed through an ethanol bath to induce densification, resulting in a helical fiber with periodic surface textures. The twisting process maintained dimensional consistency, ensuring the structural uniformity of the fiber. After being rinsed repeatedly with deionized water and dried at 60 °C, the pristine CNF was obtained. The uniformity of this fabricated fiber was confirmed by scanning electron microscopy, as shown in Fig. 6c, which exhibits a continuous and uniform helical morphology. A high magnification view of the surface texture shown in Fig. 6d demonstrates that the constituent carbon nanotube bundles are densely packed and tightly aligned along the twisting axis. Furthermore, the energy dispersive X-ray spectroscopy (EDS) spectrum shown in Fig. 6e confirms its chemical purity, which is dominated by a carbon peak at 99.14 atomic percent with negligible trace amounts of silicon and iron. In practice, the effective working length is selected according to the thickness of the workpiece. For instance, an electrode length of about 100 μm is utilized for machining a 50 μm thick single-layer metallic glass. Depending on the thickness of different workpieces, the length of the electrodes can be prepared to around or above 500 μm.Fig. 6Fabrication and characterization of the CNF.a, b Schematic illustration of the CNT drawing and the helical twisting for CNF fabrication. c, d SEM images showing the overall morphology and surface texture of the CNF (including 3D/2D profiles). e EDX spectrum of the CNF surface Fabrication and characterization of the CNF. a, b Schematic illustration of the CNT drawing and the helical twisting for CNF fabrication. c, d SEM images showing the overall morphology and surface texture of the CNF (including 3D/2D profiles). e EDX spectrum of the CNF surface Surface modification process The hydrothermal method is a widely used approach for nanomaterial synthesis, offering high crystallinity, good dispersion, high purity, controllable morphology, and low fabrication cost. In this work, a hydrothermal growth strategy was employed to construct surface-arrayed nanowires on CNF, thereby enhancing surface microtexture and improving electrode hydrophilicity. The resulting surface-arrayed nanostructure consisted of densely packed and vertically aligned NiCo₂S₄ nanowires with high uniformity. Compared with the pristine CNF surface, the SA-CNF exhibited a larger specific surface area and a well-developed porous network, as shown in Fig. 7. The nanowire synthesis process mainly involved two steps. First, Ni²⁺ and Co³⁺ ions in the solution reacted to form a NiCo₂O₄ precursor. Subsequently, a sulfidation reaction occurred to generate the NiCo₂S₄ nanostructure25. In the raw materials used in this study, graphene quantum dots are quasi-zero-dimensional nanomaterials with all three dimensions smaller than 100 nm. Their primary role is to introduce additional nanoscale textures onto the NiCo₂S₄ nanowires, thereby increasing the effective surface area and enhancing the electrical conductivity of the surface-arrayed structure. Other experimental conditions are listed in Table 2.Fig. 7Schematic diagram of the growth mechanism of NiCo2S4 nano arrays on the surface of CNF.Hydrothermal treatment and subsequent sulfidation lead to the formation of surface-arrayed NiCo2S4 nanowires on the CNF surfaceTable 2Hydrothermal growth conditions for NiCo₂S₄ParameterValueParameterValueCoCl3·H2O450 mgDeionized water30 mLNiCl2·6H2O250 mgHydrothermal reaction temperature120 °Curea180 mgHydrothermal reaction time6 hNa2S200 mgSulfidation reaction temperature150 °CGQD0.1 mgSulfidation reaction time3 h Schematic diagram of the growth mechanism of NiCo2S4 nano arrays on the surface of CNF. Hydrothermal treatment and subsequent sulfidation lead to the formation of surface-arrayed NiCo2S4 nanowires on the CNF surface Table 2Hydrothermal growth conditions for NiCo₂S₄ParameterValueParameterValueCoCl3·H2O450 mgDeionized water30 mLNiCl2·6H2O250 mgHydrothermal reaction temperature120 °Curea180 mgHydrothermal reaction time6 hNa2S200 mgSulfidation reaction temperature150 °CGQD0.1 mgSulfidation reaction time3 h Hydrothermal growth conditions for NiCo₂S₄ The experimental parameters in Table 2 were optimized through a series of comparative experiments. The solution composition and temperature dictated the initial nucleation and the feasibility of nanowire growth and were therefore fixed. The growth time governed the morphology of the nanowire arrays and was selected as the primary variable. Regarding the specific chemical kinetics, the nickel and cobalt salts were maintained at a 1–2 molar ratio to meet the stoichiometric requirement for the NiCo2S4 phase. The combination of solvent and urea ensured a controlled release of hydroxyl ions to maintain a supersaturation level that promotes uniform heterogeneous nucleation on the fiber surface while avoiding bulk precipitation. Graphene quantum dots were incorporated at a mass of 0.1 mg as a structure-directing agent to provide active sites and regulate nucleation kinetics. A hydrothermal temperature of 120 °C provided the thermodynamic driving force for growth while preventing thermal degradation of the CNF substrate. The sulfidation process utilizing 200 mg of Na2S at 150 °C for 3 h ensured anion exchange into crystalline NiCo2S4 without causing structural collapse of the nanowire skeleton. WECMM process To systematically investigate the effect of SA-CNF on mass transport efficiency and machining performance in WECMM, comparative experiments were performed using pristine CNF and SA-CNF as wire electrodes. To highlight the difference in mass transport effectiveness between the two fibers during machining, the voltage amplitude was set to 10 V. The specific experimental parameters employed in this study are summarized in Table 3.Table 3Experimental parameters for WECMMParameterValueParameterValueWire electrode diameter30 μmWire electrode feed rate0.1–1.0 μm/sWorkpiece materialsNi72Cr19Si7B2Workpiece thickness50 μmElectrolyte temperature25 °CElectrolyte0.1 M H2SO4Pulse width80 nsPulse period2 μsPositive applied voltage10 VNegative applied voltage−2 to 0 VWire reciprocating amplitude50 μmWire reciprocating frequency6 Hz Table 3Experimental parameters for WECMMParameterValueParameterValueWire electrode diameter30 μmWire electrode feed rate0.1–1.0 μm/sWorkpiece materialsNi72Cr19Si7B2Workpiece thickness50 μmElectrolyte temperature25 °CElectrolyte0.1 M H2SO4Pulse width80 nsPulse period2 μsPositive applied voltage10 VNegative applied voltage−2 to 0 VWire reciprocating amplitude50 μmWire reciprocating frequency6 Hz Experimental parameters for WECMM Characterization techniques The surface morphology and hierarchical nanostructures of pristine and SA-CNF were systematically characterized using field emission scanning electron microscopy equipped with EDS for elemental analysis. The wettability of the fiber electrodes was evaluated by measuring the meniscus height at the solid-liquid-vapor interface using the Wilhelmy method, serving as a direct indicator of surface hydrophilicity. Furthermore, the surface quality and average roughness of the machined metallic glass slit sidewalls were analyzed via atomic force microscopy (AFM) and high-resolution SEM. To guarantee data reproducibility, all reported values represented the average of at least three individual samples. Results and discussion Growth evolution of NiCo₂S₄ nanowire arrays on the CNF surface Figure 8 shows the time-dependent morphological evolution of NiCo₂S₄ nanostructures on the CNF surface during the hydrothermal process. A clear growth sequence can be observed, evolving from a loose film layer to a dense film layer, followed by the formation of random nanowires, a vertical nanowire array, radial nanostructures, and finally flower-like nanowires. At 1 h, as shown in Fig. 8a, a loose film layer was formed on the CNF surface. When the reaction time increased to 2 h, Fig. 8b shows that the surface was covered by a dense film layer. At 3 h, as shown in Fig. 8c, disordered nanowires began to grow on the surface. When the reaction time reached 4 h, Fig. 8d shows the formation of a vertical nanowire array with a uniform and well-aligned morphology. Upon further increasing the reaction time to 5 h, Fig. 8e shows the appearance of a radiating nanostructure at the top region of the nanowire array. At 6 h, as shown in Fig. 8f, the nanostructures further evolved into radiating nanowire clusters with a flower-like morphology.Fig. 8Morphological evolution of NiCo2S4 nanostructures.a–f Surface morphologies at reaction times of 1–6 h. g Cross-sectional view of the optimal array (4 h). h Tilted view showing array alignment. i High magnification view of a single nanowire Morphological evolution of NiCo2S4 nanostructures. a–f Surface morphologies at reaction times of 1–6 h. g Cross-sectional view of the optimal array (4 h). h Tilted view showing array alignment. i High magnification view of a single nanowire Among all growth stages, the sample obtained at 4 h exhibits the most favorable architecture. Detailed characterization of this optimal structure is shown in Fig. 8g–i. Figure 8g shows that the nanowires are firmly anchored to the fiber surface and possess a uniform height of approximately 10 μm. Higher magnification images in Fig. 8h, i further reveal that the individual nanowires have relatively smooth surfaces and sharp tips, with an average midsection diameter of approximately 300 nm. Figure 9 shows a comparative analysis of the surface morphology and hydrophilicity of pristine CNF and SA-CNF. Figure 9a shows the overall morphology of CNF, and Fig. 9b shows that the pristine CNF exhibits a smooth and relatively flat surface microstructure. The nanowires have an average height of approximately 10 μm, which increases the overall diameter of the CNF, as shown in Fig. 9d. In contrast, the microscopic morphology is substantially modified after hydrothermal growth of the nanowire array, as shown in Fig. 9e.Fig. 9Comparison of surface morphology and wettability.a Overall morphology of pristine CNF. b Surface microstructure of pristine CNF. c Meniscus height of pristine CNF. d Overall morphology of SA-CNF. e Surface microstructure of SA-CNF. f Meniscus height of SA-CNF Comparison of surface morphology and wettability. a Overall morphology of pristine CNF. b Surface microstructure of pristine CNF. c Meniscus height of pristine CNF. d Overall morphology of SA-CNF. e Surface microstructure of SA-CNF. f Meniscus height of SA-CNF To evaluate the wettability of the wire electrodes, the Wilhelmy method was employed. As demonstrated in previous studies on WECMM26, this method is used to characterize curved microwires. In the Wilhelmy model, a meniscus spontaneously forms at the three-phase contact line driven by capillary or surface tension forces, and an increase in meniscus height indicates improved hydrophilicity of the wire. As shown in Fig. 9c, the meniscus height for the pristine CNF was measured to be 48.9 μm. As shown in Fig. 9f, the meniscus height for the SA-CNF increased to 70.6 μm. This increase indicates that the surface-arrayed nanowires enhance the hydrophilicity of the fiber electrode, which increases the adhesion energy between the electrolyte and the electrode, thereby improving the mass transport efficiency in the IEG. Enhanced performance of WECMM In the experiments, the thickness of the electrolytic products layer increased progressively with machining time, accompanied by a corresponding rise in power fluctuation frequency. Figure 10a shows the time-dependent evolution of both the adhered electrolytic products layer thickness and the power fluctuation frequency on the wire electrode surface. These quantitative variations are consistent with the corresponding morphological observations shown in Fig. 10b. During the initial 30 min of machining, accumulation of electrolytic products remained minimal and the power fluctuation frequency stayed low, ranging from 0 to 2/min. After approximately 60 min, the electrolytic products layer with a thickness of about 1.1 μm was formed, accompanied by an increase in fluctuation frequency to 7/min. At 90 min, the thickness further increased to 2.3 μm with a corresponding frequency of 13/min. Subsequently, the fluctuation frequency increased to 20/min at 120 min. As machining proceeded to 150 min, the electrolytic products layer thickness increased to approximately 6.2 μm, completely obscuring the original surface morphology, while the fluctuation frequency sharply increased to 33/min. By 180 min, the electrolytic products layer reached a thickness of about 9.1 μm, and the fluctuation frequency peaked at 48/min. Notably, the simultaneous acceleration of both electrolytic products layer growth and power fluctuation frequency during the 120–180 min interval indicates that the progressive accumulation of electrolytic products severely destabilizes the machining process.Fig. 10Analysis of the thickness and morphology of the electrolytic products layer.a Thickness of the adhesive electrolytic products layer and frequency of power fluctuation. b The morphology of the electrolytic products layer changes over time Analysis of the thickness and morphology of the electrolytic products layer. a Thickness of the adhesive electrolytic products layer and frequency of power fluctuation. b The morphology of the electrolytic products layer changes over time When electrolytic products adhere to the wire electrode, the adhered layer with a thickness of Δd can be simplified as a flat-contact model. The dynamic equilibrium of the adhered particle is governed by the following force balance:11\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${F}_{{\rm{g}}}+{F}_{{\rm{\tau }}}+{F}_{{\rm{b}}}+{F}_{{\rm{e}}} > {F}_{{\rm{v}}}+{F}_{{\rm{c}}}$$\end{document}Fg+Fτ+Fb+Fe>Fv+Fcwhere Fg, Fτ, Fb, and Fe represent gravity, fluid shear force, buoyancy, and electrophoretic force, which promote detachment. Conversely, Fv and Fc denote the wall viscous force and surface tension that hinder detachment. The fluid shear force Fτ is a function of the viscosity coefficient μ and reciprocating speed v. Theoretical calculations indicate that a critical speed vcrit of 33.6 mm/s is required for detachment. In these experiments, the axial speed vactual was approximately 1.0 mm/s, which is insufficient to overcome the adhesion forces Fv and Fc. This kinematic limitation indicates that vibration alone is insufficient to suppress electrolytic products adhesion on smooth electrodes, thereby necessitating functional surface modification of the electrode. To evaluate whether SA-CNF overcomes this bottleneck, comparative experiments were conducted to examine mass transport differences. As shown in Fig. 11a, pristine CNF exhibited insufficient transport capacity: although most bubbles detached, residual bubbles accumulated and coalesced around the electrode, while little debris was expelled. After 60 min, the fiber surface was covered with black electrolytic products, forming a thick, irregular layer in Fig. 11b, indicating poor transport. This is attributed to low surface hydrophilicity, which induces boundary slip, creates a stagnant layer, and leads to product accumulation with power fluctuations.Fig. 11Comparison of mass transport and electrolytic products.a Optical observation of IEG with pristine CNF. b SEM morphology of pristine CNF after machining. c Optical observation of IEG with SA-CNF. d SEM morphology of SA-CNF after machining Comparison of mass transport and electrolytic products. a Optical observation of IEG with pristine CNF. b SEM morphology of pristine CNF after machining. c Optical observation of IEG with SA-CNF. d SEM morphology of SA-CNF after machining In contrast, Fig. 11c shows that SA-CNF achieved more effective mass transport. Bubbles detached at smaller sizes, and fine electrolytic particles were continuously expelled from the IEG. After 60 min, almost no products adhered to the fiber surface, as confirmed in Fig. 11d. This demonstrates that enhanced hydrophilicity suppresses boundary slip and product adhesion, thereby reducing power fluctuations and improving machining stability. Figure 12 shows a comparative analysis of WECMM performance using pristine CNF and SA-CNF. The overall morphologies of the machined components are compared in Fig. 12a, b. Compared to the notable edge deterioration produced by the pristine CNF, the SA-CNF successfully preserved high geometric accuracy and sharp structural profiles. Further localized magnification reveals that the slit sidewalls machined by the pristine fiber were uneven, with conspicuous defects on the end face, as shown in Fig. 12c. In contrast, the SA-CNF yielded uniform sidewalls and a smooth, reflective end face as depicted in Fig. 12d. Figure 12e, f shows the significant improvement in surface integrity. The average surface roughness Ra for the pristine CNF was approximately 62.8 nm, reflecting unstable etching caused by stagnant electrolyte flushing in the IEG. The SA-CNF electrode reduced the Ra to 25.5 nm, representing an improvement of approximately 60%. This result is consistent with the topographical homogeneity observed in SEM micrographs. As shown in Fig. 12g, the feed rate of the SA-CNF increased to 0.8 μm/s, doubling that of the pristine CNF (0.4 μm/s), while the power fluctuation frequency dropped from 2.2/min to 0.2/min. Meanwhile, the average slit width decreased slightly from 53 μm for the CNF to 51 μm for the SA-CNF. This performance surge was driven by a synergistic mechanism of enhanced mass transport and interfacial activation.Fig. 12Enhanced WECMM performance and engineering applicability enabled by SA-CNF.a Overall SEM morphology of pristine CNF. b Overall SEM morphology of SA-CNF. c Magnified slit and end face of pristine CNF. d Magnified slit and end face of SA-CNF. e 3D AFM topography of pristine CNF. f 3D AFM topography of SA-CNF. g Frequency of power fluctuation, wire electrode feed rate, and average surface roughness of the workpiece. h Multiwire and multilayer machining configuration. i Dimensional deviations across five layers Enhanced WECMM performance and engineering applicability enabled by SA-CNF. a Overall SEM morphology of pristine CNF. b Overall SEM morphology of SA-CNF. c Magnified slit and end face of pristine CNF. d Magnified slit and end face of SA-CNF. e 3D AFM topography of pristine CNF. f 3D AFM topography of SA-CNF. g Frequency of power fluctuation, wire electrode feed rate, and average surface roughness of the workpiece. h Multiwire and multilayer machining configuration. i Dimensional deviations across five layers To evaluate the engineering application potential of the proposed method, a high-efficiency batch machining experiment was performed. Figure 12h illustrates a simultaneous machining configuration with five wire electrodes and five stacked layers, enabling the fabrication of 25 micro-components in a single run and thus improving the processing efficiency by 25 times. This configuration places stringent demands on mass transport within the deep IEG. Compared with conventional CNF electrodes, the superhydrophilic nanowire surface of the SA-CNF electrode effectively suppresses wall slip, enhances fluid-solid interaction, and facilitates the replenishment of fresh electrolyte in deep channels. The dimensional consistency across the five layers was assessed, and the results in Fig. 12i show highly overlapping measurement curves. The relative standard deviation is below 0.1%, while the maximum dimensional deviation between the innermost and outermost layers is only 0.41 µm. No obvious electrode wear was observed after prolonged machining. Conclusions A surface-arrayed nanowire-modified CNF wire electrode was developed and systematically investigated for high-performance WECMM. The main conclusions are summarized as follows.NiCo₂S₄ nanowire arrays were synthesized in situ on the surface of CNF via a controllable hydrothermal method. Vertically aligned nanowires with a typical height of approximately 10 μm created a superhydrophilic interface and achieved geometric compatibility with the boundary layer thickness, enabling effective penetration into the gas cushion.Theoretical analysis combined with flow field simulations demonstrated that the hierarchical surface microtexture significantly suppresses boundary velocity slip by reducing slip length and disrupting the insulating gas cushion. Enhanced fluid-solid coupling allowed the reciprocating wire electrode to function as an efficient pump for mass transport and bubble detachment.WECMM experiments on Ni-based metallic glass confirmed superior machining stability and efficiency compared with pristine CNF. The feed rate increased to 0.8 μm/s, while power fluctuation frequency decreased from 2.2/min to 0.2/min. The surface roughness Ra was further reduced by approximately 60% to 25.5 nm, resulting in improved surface integrity.Engineering applicability was validated using a multiwire and multilayer batch machining configuration. Twenty-five stepped cantilever micro-components were simultaneously fabricated with high dimensional consistency, exhibiting a relative standard deviation below 0.1%. Uniform electrochemical dissolution was maintained even within high aspect ratio channels. NiCo₂S₄ nanowire arrays were synthesized in situ on the surface of CNF via a controllable hydrothermal method. Vertically aligned nanowires with a typical height of approximately 10 μm created a superhydrophilic interface and achieved geometric compatibility with the boundary layer thickness, enabling effective penetration into the gas cushion. NiCo₂S₄ nanowire arrays were synthesized in situ on the surface of CNF via a controllable hydrothermal method. Vertically aligned nanowires with a typical height of approximately 10 μm created a superhydrophilic interface and achieved geometric compatibility with the boundary layer thickness, enabling effective penetration into the gas cushion. Theoretical analysis combined with flow field simulations demonstrated that the hierarchical surface microtexture significantly suppresses boundary velocity slip by reducing slip length and disrupting the insulating gas cushion. Enhanced fluid-solid coupling allowed the reciprocating wire electrode to function as an efficient pump for mass transport and bubble detachment. Theoretical analysis combined with flow field simulations demonstrated that the hierarchical surface microtexture significantly suppresses boundary velocity slip by reducing slip length and disrupting the insulating gas cushion. Enhanced fluid-solid coupling allowed the reciprocating wire electrode to function as an efficient pump for mass transport and bubble detachment. WECMM experiments on Ni-based metallic glass confirmed superior machining stability and efficiency compared with pristine CNF. The feed rate increased to 0.8 μm/s, while power fluctuation frequency decreased from 2.2/min to 0.2/min. The surface roughness Ra was further reduced by approximately 60% to 25.5 nm, resulting in improved surface integrity. WECMM experiments on Ni-based metallic glass confirmed superior machining stability and efficiency compared with pristine CNF. The feed rate increased to 0.8 μm/s, while power fluctuation frequency decreased from 2.2/min to 0.2/min. The surface roughness Ra was further reduced by approximately 60% to 25.5 nm, resulting in improved surface integrity. Engineering applicability was validated using a multiwire and multilayer batch machining configuration. Twenty-five stepped cantilever micro-components were simultaneously fabricated with high dimensional consistency, exhibiting a relative standard deviation below 0.1%. Uniform electrochemical dissolution was maintained even within high aspect ratio channels. Engineering applicability was validated using a multiwire and multilayer batch machining configuration. Twenty-five stepped cantilever micro-components were simultaneously fabricated with high dimensional consistency, exhibiting a relative standard deviation below 0.1%. Uniform electrochemical dissolution was maintained even within high aspect ratio channels.