Nanocone-Array-Based Platinum-Iridium Oxide Neural Microelectrodes: Structure, Electrochemistry, Durability and Biocompatibility Study

Neural interfaces provide a window for bio-signal modulation and recording with the assistance of neural microelectrodes. However, shrinking the size of electrodes results in high electrochemical impedance and low capacitance, thus limiting the stimulation/recording efficiency. In order to achieve critical stability and low power consumption, here, nanocone-shaped platinum (Pt) with an extensive surface area is proposed as an adhesive layer on a bare Pt substrate, followed by the deposition of a thin layer of iridium oxide (IrOx) to fabricate high-performance nanocone-array-based Pt-IrOx neural microelectrodes (200 μm in diameter). A uniform nanocone-shaped Pt with significant roughness is created via controlling the ratio of NH4+ and Pt4+ ions in the electrolyte, which can be widely applicable for batch production on multichannel flexible microelectrode arrays (fMEAs) and various substrates with different dimensions. The Pt-IrOx nanocomposite-coated microelectrode presents a significantly low impedance down to 0.72 ± 0.04 Ω cm2 at 1 kHz (reduction of ~92.95%). The cathodic charge storage capacity (CSCc) and charge injection capacity (CIC) reaches up to 52.44 ± 2.53 mC cm−2 and 4.39 ± 0.36 mC cm−2, respectively. Moreover, superior chronic stability and biocompatibility are also observed. The modified microelectrodes significantly enhance the adhesion of microglia, the major immune cells in the central nervous system. Therefore, such a coating strategy presents great potential for biomedical and other practical applications.


Introduction
Worldwide, there is a large unmet need for the detection, diagnosis, and treatment of neurological disorders. Our understanding of neuronal functions and dynamics is limited, largely due to our inability to monitor neural activities in both healthy and pathological disorders. Neural microelectrodes, acting as the key bridge of the neural interfaces to connect the inner tissues (biological) and external devices (technical), play a critical role in monitoring, diagnosing, and treating neurological diseases. They have been widely applied on various neural prostheses or interfaces, including retinal prosthesis for eliciting and reconstructing vision [1], cochlear implants for restoring hearing [2], deep brain stimulation for alleviating Parkinson's disease and severe psychiatric disorders [3], peripheral nerve electrodes for acquiring and understanding chronic or acute disease states [4], and other invasive or implantable devices [5][6][7]. In recent years, neural microelectrodes have developed towards miniaturization and integration on the micro/nano scale in order to provide higher electrical stimulation efficiency for clinical practice. However, the large reduction in electrode size leads to a sharp increase in electrode impedance and inevitably the decrease of electrode capacitance, which magnifies the energy consumption and seriously limits its application [8][9][10]. Therefore, it is critical to effectively solve the contradiction between electrode size reduction and performance. and consistent modified layers on the electrode surface with desirable impedance, CSC c and CIC, stability, and biocompatibility. Because ammonium salts have been reported to promote the formation of nickel (Ni) nanocones with excellent consistency [35], in this study we combined the advantages of Pt and IrO x to fabricate nanocone-array-based Pt-IrO x neural microelectrodes. Firstly, two kinds of ammonium salts, including organic and inorganic, were chosen in order to obtain a nanocone-array-based Pt coating as the adhesive layer on bare substrate. Subsequently, we utilized a template-free and electrodeposition method to prepare the functional nanocone-array-based Pt neural microelectrode, which was achieved by controlling the proportion of NH 4 + and Pt 4+ ions in the electrolyte. An IrO x layer with an extensive surface area was then deposited on the nanostructured Pt to achieve a high-performance neural interface. The morphology, electrochemistry, durability, and biocompatibility of the as-prepared nanomaterials were evaluated in detail before in vivo study.

Microelectrode Fabrication
Flexible polyimide (PI) was chosen as the packaging layer and insulating substrate to carry Pt microelectrodes (200 µm in diameter), which were fabricated as described in previous work [26,30]. As shown in Figure 1a, the process of creating 24-channel fMEAs began with spin-coating a thick PI layer (5 µm) as the substrate onto a silicon wafer, which was then cured at approximately 350 • C. A metal layer (Pt/Ti,~100 nm) was then sputtered and patterned as the conductive layer of the fMEA using standard photo-lithography (EVG 610, St. Florian, Austria), followed by spin-coating another PI layer (~5 µm). The fMEA was etched to expose the electrode/active sites and the interconnection pads by oxygen reactive ion etching using Al as the hard mask. Finally, the fMEA was released from the silicon wafer for electrodeposition and testing in subsequent experiments. Figure 1b shows a photographic image of the fMEA; the interconnection pads are welded onto the customized printed circuit board (PCB, 24-channel) for electrodeposition and further characterization. The magnified view shown in Figure 1c represents the microelectrodes without modification (bare Pt) and with modified coatings (Pt, IrO x , or IrO x /Pt nanostructure). Ten fMEAs were fabricated simultaneously on the same silicon wafer, and three were selected randomly for testing. More than three microelectrodes were modified in each of the different electrodeposition processes in parallel.
the Pt nanoflower grows rapidly in all directions, and the deposition potential and duration time must be controlled precisely.
Considering that a well-controlled microstructure has a great influence on the high stimulation efficiency of neural interfaces, we have here focused on preparing uniform and consistent modified layers on the electrode surface with desirable impedance, CSCc and CIC, stability, and biocompatibility. Because ammonium salts have been reported to promote the formation of nickel (Ni) nanocones with excellent consistency [35], in this study we combined the advantages of Pt and IrOx to fabricate nanocone-array-based Pt-IrOx neural microelectrodes. Firstly, two kinds of ammonium salts, including organic and inorganic, were chosen in order to obtain a nanocone-array-based Pt coating as the adhesive layer on bare substrate. Subsequently, we utilized a template-free and electrodeposition method to prepare the functional nanocone-array-based Pt neural microelectrode, which was achieved by controlling the proportion of NH4 + and Pt 4+ ions in the electrolyte. An IrOx layer with an extensive surface area was then deposited on the nanostructured Pt to achieve a high-performance neural interface. The morphology, electrochemistry, durability, and biocompatibility of the as-prepared nanomaterials were evaluated in detail before in vivo study.

Microelectrode Fabrication
Flexible polyimide (PI) was chosen as the packaging layer and insulating substrate to carry Pt microelectrodes (200 μm in diameter), which were fabricated as described in previous work [26,30]. As shown in Figure 1a, the process of creating 24-channel fMEAs began with spin-coating a thick PI layer (5 μm) as the substrate onto a silicon wafer, which was then cured at approximately 350 °C. A metal layer (Pt/Ti, ~100 nm) was then sputtered and patterned as the conductive layer of the fMEA using standard photo-lithography (EVG 610, St. Florian, Austria), followed by spin-coating another PI layer (~5 μm). The fMEA was etched to expose the electrode/active sites and the interconnection pads by oxygen reactive ion etching using Al as the hard mask. Finally, the fMEA was released from the silicon wafer for electrodeposition and testing in subsequent experiments. Figure 1b shows a photographic image of the fMEA; the interconnection pads are welded onto the customized printed circuit board (PCB, 24-channel) for electrodeposition and further characterization. The magnified view shown in Figure 1c represents the microelectrodes without modification (bare Pt) and with modified coatings (Pt, IrOx, or IrOx/Pt nanostructure). Ten fMEAs were fabricated simultaneously on the same silicon wafer, and three were selected randomly for testing. More than three microelectrodes were modified in each of the different electrodeposition processes in parallel.

Bare Pt electrode
Electrode/Active site

Electrodeposition of Pt and IrO x
The electrodeposition process was conducted on a three-electrode electrochemical workstation (Gamry Reference 600, Warminster, PA, USA), where the microelectrode (bare Pt and modified coating) was connected to the working electrode (WE) and an Ag/AgCl Nanomaterials 2022, 12, 3445 4 of 14 electrode, as well as a large-area Pt sheet. These were then employed as the reference electrode (RE) and counter electrode (CE). All the samples were carefully cleaned in acetone solution and electrochemically conditioned in 0.5 M H 2 SO 4 .
All the Pt nanodendrites were formed in different Pt electrolytes; 10 mM PtCl 4 and 20 mM (NH 4 ) 2 PtCl 6 were selected as the main sources of the Pt 4+ ion, with 10 mM (NH 2 CH 2 ) 2 ·2HCl and 10 mM NH 4 Cl as the crystal promoters (purchased from Sigma, St. Louis, MO, USA). Different Pt morphologies could be formed by varying the ratio of components. The pH of the electrolytes was stabilized at~7.4. The Pt layers were then electrodeposited at −0.6 V (vs. Ag/AgCl) for approximately 15 min by chronoamperometry.
The IrO x layer was electrodeposited in an Ir electrolyte containing 4.25 mM IrCl 4 (purchased from Sigma), with a pH of~10.5. A slow scanning rate between 0.05 and 0.55 V was applied to obtain a stable IrO x coating.
All the samples were carefully rinsed with deionized water after each modification.

Electrochemical Characterization
The electrochemical characteristics of different coated microelectrodes and bare Pt were assessed by a series of electrochemical methods, including electrochemical impedance spectroscopy (EIS), cyclic voltammetry (CV), and electrical pulse measurements. All measurements were performed in a 1× phosphate-buffered saline solution (PBS) at room temperature using the same three-electrode setup as described above.
EIS measurements were conducted with a 10 mV amplitude signal over a frequency range of 100 kHz-1 Hz. The impedance at 1 kHz was used here to evaluate the resistance of the coated microelectrodes for neural application. CV sweeps were applied between −0.6 V and 0.8 V at a scan rate of 50 mV s −1 . The cathodic charge storage capacity (CSC c ) was defined as the area of the cathode region enclosed by one CV cycle. Prior to the tests, all the samples were soaked in PBS for at least 30 min to stabilize the electrode sites in solution. The CSC c can be calculated by the following equation [26,34]: where E stands for the potential (V vs. Ag/AgCl), E a and E c are the corresponding anodic and cathodic potential limits, respectively, i is the current (A) responding to the potential, S represents the geometrical surface area of the microelectrode, and υ is the scanning rate of CV. In addition, the charge injection capacity (CIC) of the microelectrode was estimated by pulse-testing, which was determined by potential transient responses to a pulse current stimulation. It was realized with a neural stimulator (PENS0410, Greentek, Suzhou, China) to provide a symmetric biphasic current pulse (1 ms pulse width, 50 Hz) between a modified microelectrode and a Pt counter electrode in PBS. The potential excursion between the modified microelectrode and an Ag/AgCl reference electrode was monitored using a digital oscilloscope (TPS2012B, Tektronix, Suzhou, China). The maximum CIC was obtained by continuously increasing the pulse current density until the measured potential dropped at the boundary of "water window" (−0.6-0.8 V), after subtracting the ohmic voltage drop ("IR drop") [34,36]. Consequently, CIC can be calculated with the following equation [26,34]: where I max is the maximum of charge injection current and t is the pulse width of the given signal.

Stability Evaluation
The mechanical stability of the coated microelectrodes was estimated in PBS by applying an intensive ultrasonic bath (500 W, 1 kHz) for 1 h. The corresponding variation rate of impedance and CSC c of the microelectrode were then observed after it was stabilized in PBS for 30 min. In addition, a homemade bending/twist fatigue tester was also employed to evaluate the mechanical stability. The two ends of the microelectrode were clamped onto the equipment, followed by ±15 degrees of each twist cycle. The performance was then tested after 2000 cycles.
The electrochemical stability was then examined by applying long-time pulsing with continuous biphasic rectangular (cathodic-first) current pulses (500 µA, 200 µs) at 100 Hz for 0-30 days. After different pulsing times, the changes of CSC c were characterized.

Morphology Characterization
Optical images of samples were taken with a microscope (Nikon Ni-U, Tokyo, Japan). The surface morphology and microstructure of the as-fabricated coating was observed by field emission scanning electron microscopy (FESEM, Zeiss SIGMA 300, Oberkochen, Germany). The root-mean-square (RMS) roughness of the samples was measured by atomic force microscopy (AFM, NanoManVS, Veeco, Plainview, TX, USA).

Biocompatibility Study
Primary mouse microglia isolation and culture were performed as described below. Briefly, mouse brains were collected from P1~P3 neonatal C57BL/6J mice. After removing meninges and blood vessels, the mouse cortex was minced and dissociated by gentle mechanical disruption in 0.025% trypsin for 5 min. The cell suspension was then neutralized by 100% fetal bovine serum (FBS) and filtered through a 70 µm strainer. The culture was shaken at 180 rpm for 30 min on an orbital shaker. After centrifuging and removing supernatant, the culture was resuspended and cultured in Dulbecco's Modified Eagle Medium (DMEM) with 10% FBS, 1% penicillin/streptomycin, and 20% LADMAC-conditioned media (produced from the LADMAC cell line CRL-2420). Following these steps, the resulting cells were purified microglia, as previously established, and they were comprehensively evaluated in our laboratory [37].
For the microglia adhesion assay, the purified microglia were cultured on the surface of different microelectrodes in DMEM medium for 24 h; the microglia that remained adhered on different surfaces were fixed with 4% paraformaldehyde (PFA). The fixed microglia were permeabilized with 0.1% Triton-X100 and blocked with 5% bovine serum albumin (BSA), followed by immunostaining with Iba1 (Wako, Cat#: 019-19741) antibody to label microglia and 4 ,6-diamidino-2-phenylindole (DAPI).
A Cell Counting Kit-8 (CCK8) viability assay was performed to assay the effect of microelectrode extracts on microglia viability. Microelectrodes were sterilized by soaking them in 75% ethanol overnight, followed by UV radiation for 3 h. The sterile microelectrodes were immersed in microglia culture medium for 48 h for extract collection. The collected extracts were co-cultured with microglia for 72 h, followed by CCK8 viability test (MEDChemExpress, Cat#: HY-K0301). The CCK8 kit uses a water-soluble tetrazolium salt that produces an orange formazan dye upon bio-reduction in the presence of an electron carrier. The formazan product is soluble in culture medium, and the amount of formazan produced is directly proportional to the number of living cells. The CCK8 assay was performed as instructed by the manufacturer.

Electrodeposition Process and Evaluation of Pt Nanocone
The micro-/nanostructure and morphology of microelectrode surfaces are quite important for the electrochemical performance of neural electrodes, and these can be realized by controlling the electrodeposition condition [34]. Figure 2 shows the representative morphologies and microstructures of different coated microelectrodes. Figure 2a represents the bare Pt microelectrode obtained using standard photo-lithography and reactive ion etching, with a flat surface and almost no microstructure. in the single PtCl 4 electrolyte (10 mM) is covered by large and heterogeneous angular grains with sizes in the order of 50-200 nm. Such structure increases the contact area between the electrode and the working medium to some extent. It has been reported that ammonium salts can promote the directional growth of Ni nanocones [35]; we here have also chosen two types of crystal accelerator (organic and inorganic) in order to promote the directional growth of the Pt nanocrystal. When an amount of 10 mM (NH 2 CH 2 ) 2 ·2HCl was added onto the PtCl 4 electrolyte, the microstructure did not improve in the expected direction ( Figure 2c). While an appropriate amount of 10 mM NH 4 Cl was added, small nanocone-shaped crystals could be observed obviously attached to the surface, shown in Figure 2d. This may indicate that the inorganic ammonium salt has a better promotion effect on the crystal formation. Considering that (NH 4 ) 2 PtCl 6 contains NH 4 + ion, we tested the effect of electrolyte containing 20 mM (NH 4 ) 2 PtCl 6 as the main Pt salt, and found that cluster-shaped Pt nanodendrites grew rapidly in all directions. Thus, the surface area could be increased by orders of magnitude without difficulty ( Figure 2e). In order to control the Pt nanocone growing perpendicular to the substrate, we further controlled the amount of PtCl 4 and (NH 4 ) 2 PtCl 6 in the same electrolyte, adjusting the ratio of n(NH 4 + ) to n(Pt 4+ ) to 4:5. Finally, a homogeneous and controllable nanocone-array-based Pt structure was achieved. A large nanocone-shaped structure can be easily formed, approximately 300 nm in width and 1 µm in height (Figure 2f). Such a nanocone array is directional electro-crystallization, and the aspect ratio of the nanocone structure can also be controlled. Therefore, both Pt nanocluster and nanocone structure are of particular interest due to their extensive surface area, which could further increase the charge capacity and reduce the impedance for improving neural stimulation/recording efficiency. portant for the electrochemical performance of neural electrodes, and these can be realized by controlling the electrodeposition condition [34]. Figure 2 shows the representative morphologies and microstructures of different coated microelectrodes. Figure 2a represents the bare Pt microelectrode obtained using standard photo-lithography and reactive ion etching, with a flat surface and almost no microstructure. Figure 2b-f demonstrates different micrographs of the Pt nanodendrites achieved at the same constant potential of −0.6 V in different Pt electrolytes. It can be seen from Figure 2b that the coating deposited in the single PtCl4 electrolyte (10 mM) is covered by large and heterogeneous angular grains with sizes in the order of 50-200 nm. Such structure increases the contact area between the electrode and the working medium to some extent. It has been reported that ammonium salts can promote the directional growth of Ni nanocones [35]; we here have also chosen two types of crystal accelerator (organic and inorganic) in order to promote the directional growth of the Pt nanocrystal. When an amount of 10 mM (NH2CH2)2·2HCl was added onto the PtCl4 electrolyte, the microstructure did not improve in the expected direction (Figure 2c). While an appropriate amount of 10 mM NH4Cl was added, small nanocone-shaped crystals could be observed obviously attached to the surface, shown in Figure 2d. This may indicate that the inorganic ammonium salt has a better promotion effect on the crystal formation. Considering that (NH4)2PtCl6 contains NH4 + ion, we tested the effect of electrolyte containing 20 mM (NH4)2PtCl6 as the main Pt salt, and found that cluster-shaped Pt nanodendrites grew rapidly in all directions. Thus, the surface area could be increased by orders of magnitude without difficulty (Figure 2e). In order to control the Pt nanocone growing perpendicular to the substrate, we further controlled the amount of PtCl4 and (NH4)2PtCl6 in the same electrolyte, adjusting the ratio of n(NH4 + ) to n(Pt 4+ ) to 4:5. Finally, a homogeneous and controllable nanocone-array-based Pt structure was achieved. A large nanocone-shaped structure can be easily formed, approximately 300 nm in width and 1 μm in height (Figure 2f). Such a nanocone array is directional electro-crystallization, and the aspect ratio of the nanocone structure can also be controlled. Therefore, both Pt nanocluster and nanocone structure are of particular interest due to their extensive surface area, which could further increase the charge capacity and reduce the impedance for improving neural stimulation/recording efficiency. The electrochemical performance is of great importance for neural microelectrodes. CV and EIS measurements were performed before and after modification in different Pt electrolytes in order to evaluate and confirm the proper electrolyte. The electrochemical performance is of great importance for neural microelectrodes. CV and EIS measurements were performed before and after modification in different Pt electrolytes in order to evaluate and confirm the proper electrolyte.
The impedance at 1 kHz is one of the critical parameters for neural electrodes, because the typical electrophysiological signals largely occur at that frequency, which corresponds to the power consumption of neural stimulation and the noise of neural recording applications [26,36]. Figure 3a shows the electrochemical impedance in PBS of those modified microelectrodes in different Pt electrolytes at their open circuit potential. It can be seen that the impedance of bare Pt microelectrode at 1 kHz was approximately 9.63 Ω cm 2 , which decreased significantly following the nano-Pt modification. After deposition under the same potential and duration time, the impedance values were reduced to 0.91 Ω cm 2 (PtCl 4 , reduction of 90.55%), 1.70 Ω cm 2 (PtCl 4 + (NH 2 CH 2 ) 2 ·2HCl, reduction of 82.35%), 0.75 Ω cm 2 (PtCl 4 + NH 4 Cl, reduction of 92.21%), 0.72 Ω cm 2 ((NH 4 ) 2 PtCl 6 , reduction of 92.52%), and 0.71 Ω cm 2 (PtCl 4 + (NH 4 ) 2 PtCl 6 , reduction of 92.63%), respectively. It seems that the organic ammonium salt has no significant effect on the decrement of impedance after mixing with PtCl 4 , which is consistent with the SEM results. Meanwhile, other Pt electrolytes lead to a dramatic decrease of the impedance of almost more than 1 order of magnitude. In addition, more significant reduction can be observed for lower frequencies, e.g., 10 Hz and 100 Hz. This may be useful for neural electrodes in other operation modes, such as neuromuscular electrical stimulation (NMES), transcutaneous electrical nerve stimulation (TENS), and deep brain stimulation (DBS), etc. Lower impedance means higher signal-to-noise ratio (SNR) when recording weak neural signals, as well as lower energy consumption when stimulating the target tissue. More importantly, the safety of electrical stimulation can be improved with low impedance.
cause the typical electrophysiological signals largely occur at that frequency, which corresponds to the power consumption of neural stimulation and the noise of neural recording applications [26,36]. Figure 3a shows the electrochemical impedance in PBS of those modified microelectrodes in different Pt electrolytes at their open circuit potential. It can be seen that the impedance of bare Pt microelectrode at 1 kHz was approximately 9.63 Ω cm 2 , which decreased significantly following the nano-Pt modification. After deposition under the same potential and duration time, the impedance values were reduced to 0.91 Ω cm 2 (PtCl4, reduction of 90.55%), 1.70 Ω cm 2 (PtCl4 + (NH2CH2)2·2HCl, reduction of 82.35%), 0.75 Ω cm 2 (PtCl4 + NH4Cl, reduction of 92.21%), 0.72 Ω cm 2 ((NH4)2PtCl6, reduction of 92.52%), and 0.71 Ω cm 2 (PtCl4 + (NH4)2PtCl6, reduction of 92.63%), respectively. It seems that the organic ammonium salt has no significant effect on the decrement of impedance after mixing with PtCl4, which is consistent with the SEM results. Meanwhile, other Pt electrolytes lead to a dramatic decrease of the impedance of almost more than 1 order of magnitude. In addition, more significant reduction can be observed for lower frequencies, e.g., 10 Hz and 100 Hz. This may be useful for neural electrodes in other operation modes, such as neuromuscular electrical stimulation (NMES), transcutaneous electrical nerve stimulation (TENS), and deep brain stimulation (DBS), etc. Lower impedance means higher signal-to-noise ratio (SNR) when recording weak neural signals, as well as lower energy consumption when stimulating the target tissue. More importantly, the safety of electrical stimulation can be improved with low impedance. In accordance with the low impedance, the CV measurements given in Figure 3b showed a distinct increase in CSCc for the modified microelectrodes. The loop area of the unmodified Pt microelectrode was the smallest, with a minimum CSCc of approximately 3.18 mC cm −2 . Meanwhile, the CSCc of nano-Pt-modified microelectrodes increased to ~25.48 mC cm −2 (PtCl4), ~25.02 mC cm −2 (PtCl4 + (NH2CH2)2·2HCl), and ~25.82 mC cm −2 (PtCl4 + NH4Cl), which were approximately 8.01, 7.87, and 8.12 times higher than that of bare Pt, respectively. The values of these three groups are relatively close. Nevertheless, the loop area of CV curves under (NH4)2PtCl6 and the mixture of PtCl4 + (NH4)2PtCl6 are similar and more extensively significant. The corresponding CSCc are approximately 37.25 mC cm −2 and 38.32 mC cm −2 , which are approximately 11.86 and 12.05 times higher than that of bare Pt, respectively. The insets in Figure 3b represent SEM images of microelectrodes deposited on (i) (NH4)2PtCl6 and (ii) the mixture of PtCl4 + (NH4)2PtCl6 electrolytes, respectively. It can be seen that the rapid growth in the (NH4)2PtCl6 electrolyte can easily result in the coating spilling over the edge of the microelectrode, due to the higher molarity of the NH4 + ion compared with the Pt 4+ ion. This is undesirable in the implantation of In accordance with the low impedance, the CV measurements given in Figure 3b showed a distinct increase in CSC c for the modified microelectrodes. The loop area of the unmodified Pt microelectrode was the smallest, with a minimum CSC c of approximately 3.18 mC cm −2 . Meanwhile, the CSC c of nano-Pt-modified microelectrodes increased tõ 25.48 mC cm −2 (PtCl 4 ),~25.02 mC cm −2 (PtCl 4 + (NH 2 CH 2 ) 2 ·2HCl), and~25.82 mC cm −2 (PtCl 4 + NH 4 Cl), which were approximately 8.01, 7.87, and 8.12 times higher than that of bare Pt, respectively. The values of these three groups are relatively close. Nevertheless, the loop area of CV curves under (NH 4 ) 2 PtCl 6 and the mixture of PtCl 4 + (NH 4 ) 2 PtCl 6 are similar and more extensively significant. The corresponding CSC c are approximately 37.25 mC cm −2 and 38.32 mC cm −2 , which are approximately 11.86 and 12.05 times higher than that of bare Pt, respectively. The insets in Figure 3b represent SEM images of microelectrodes deposited on (i) (NH 4 ) 2 PtCl 6 and (ii) the mixture of PtCl 4 + (NH 4 ) 2 PtCl 6 electrolytes, respectively. It can be seen that the rapid growth in the (NH 4 ) 2 PtCl 6 electrolyte can easily result in the coating spilling over the edge of the microelectrode, due to the higher molarity of the NH 4 + ion compared with the Pt 4+ ion. This is undesirable in the implantation of neural electrodes. After properly controlling the ratio of both ions in the mixture of PtCl 4 + (NH 4 ) 2 PtCl 6 , the microelectrode surface was covered by a uniform and consistent nanocone-shaped array, completely deposited in the groove of the electrode site. This will facilitate the tissue contact and long-term service of the neural electrode during implantation.

Optimization and Batch Production of Pt Nanocone
In order to verify the universality of Pt nanocone for various geometries and substrates, we further fine-tuned the ratio of the solution and deposition conditions (n(NH 4 + )/n(Pt 4+ ) = 4/5, −0.6 V, 10~15 min), and successfully completed the modification on electrophysiological fMEA, platinized silicon wafers, Pt wire, and cochlear implants. In Figure 4a, Point 1 is the bare Pt without modification, as a reference. Points (2)(3)(4)(5), marked in Figure 4a-c, were chosen in order to observe the morphology. It was found that similar nanocone-array structures grew perpendicular to the surface and covered evenly on all samples with different channels and dimensions, including electrophysiological fMEA with different diameters (Figure 4a, 20 µm or 150 µm), platinized silicon wafers (Figure 4b), and a Pt wire and ring from a cochlear implant (Figure 4c). Subsequently, a 1 kg load was applied on a platinized silicon wafer for 72 h. The SEM result demonstrated almost no difference from the initial status (Figure 4d,e). The morphology illustrated that the coating consisted of billions of Pt nanocones with sizes of approximately 50-100 nm in width and 150-200 nm in height. Moreover, the AFM result also verified the homogeneity of nanocone structure.
cilitate the tissue contact and long-term service of the neural electrode during implantation.

Optimization and Batch Production of Pt Nanocone
In order to verify the universality of Pt nanocone for various geometries and substrates, we further fine-tuned the ratio of the solution and deposition conditions (n(NH4 + )/n(Pt 4+ ) = 4/5, −0.6 V, 10~15min), and successfully completed the modification on electrophysiological fMEA, platinized silicon wafers, Pt wire, and cochlear implants. In Figure 4a, Point 1 is the bare Pt without modification, as a reference. Points (2)(3)(4)(5), marked in Figure 4a-c, were chosen in order to observe the morphology. It was found that similar nanocone-array structures grew perpendicular to the surface and covered evenly on all samples with different channels and dimensions, including electrophysiological fMEA with different diameters (Figure 4a, 20 μm or 150 μm), platinized silicon wafers (Figure  4b), and a Pt wire and ring from a cochlear implant (Figure 4c). Subsequently, a 1 kg load was applied on a platinized silicon wafer for 72 h. The SEM result demonstrated almost no difference from the initial status (Figure 4d,e). The morphology illustrated that the coating consisted of billions of Pt nanocones with sizes of approximately 50-100 nm in width and 150-200 nm in height. Moreover, the AFM result also verified the homogeneity of nanocone structure. A high-intensity ultrasound test (500 W, 1 kHz) was performed in PBS for 1 h to further study the mechanical stability of coated microelectrodes, because an ultrasonic bath could easily remove the loosely attached substances on the surface. The corresponding variation rate is shown in Figure 4f. It can be seen that the impedance only increased by approximately 10.83%, and the CSCc only reduced by approximately 10.03%, reflecting the strong adhesion between the Pt coating and the substrate. Assuming that the electrode is implanted in the eyeball, the eye will blink approximately 40,000 times if it works for 24 h a day (the eye blinks at most 30 times per minute). Meanwhile, 1 h of ultrasound at 1 kHz corresponds to 3.6 × 10 6 times the amount of stimulation, which is equivalent to 90d of continuous usage. Hence, it presents excellent mechanical stability and shows superior potential for long-term applications.

IrOx/Pt nanocone for Better Neural Interface
Considering that low power consumption and chronic stability are critical requirements for neural implants, such as retinal prosthesis, we chose IrOx as a second layer on  A high-intensity ultrasound test (500 W, 1 kHz) was performed in PBS for 1 h to further study the mechanical stability of coated microelectrodes, because an ultrasonic bath could easily remove the loosely attached substances on the surface. The corresponding variation rate is shown in Figure 4f. It can be seen that the impedance only increased by approximately 10.83%, and the CSC c only reduced by approximately 10.03%, reflecting the strong adhesion between the Pt coating and the substrate. Assuming that the electrode is implanted in the eyeball, the eye will blink approximately 40,000 times if it works for 24 h a day (the eye blinks at most 30 times per minute). Meanwhile, 1 h of ultrasound at 1 kHz corresponds to 3.6 × 10 6 times the amount of stimulation, which is equivalent to 90d of continuous usage. Hence, it presents excellent mechanical stability and shows superior potential for long-term applications.

IrO x /Pt Nanocone for Better Neural Interface
Considering that low power consumption and chronic stability are critical requirements for neural implants, such as retinal prosthesis, we chose IrO x as a second layer on the Pt nanocone array. Due to its high pseudo-capacitance characteristics, IrO x can further improve the interface of neural microelectrodes equipped with lower impedance, higher CSC c , and higher CIC, as well as excellent chronic stability. Figure 5a shows an SEM image of IrO x modified on a bare Pt microelectrode via CV cycles. The microstructure of the IrO x coating was uniform and compact, with the size of nanoparticles (NPs) approximately 50-120 nm in diameter. Figure 5b demonstrates the morphology of an IrO x /Pt nanocone composite-coated microelectrode. IrO x adhered tightly to the surface of the nanocone array, with good homogeneity and a rough contact area. The top view of the IrO x /Pt nanocone structure displayed in Figure 5c reveals a good adhesion between the nanocones and the IrO x layer, which may be attributed to the nanocones having sharp angle (Figures 2f and 4d). The Pt nanocone array structure cannot only provide a large surface area for accommodating more IrO x NPs, but can also improve the adhesion and stability of these coating layers on bare Pt microelectrodes.  The electrochemical properties of bare Pt microelectrodes, including impedance at 1 kHz (Figure 5d), CSC c (Figure 5e), and CIC (Figure 5f), were compared to the herein proposed Pt nanocone, IrO x , and their composite coating in detail. The impedance of the IrO x coating was reduced down to 0.99 ± 0.03 Ω cm 2 , which was more than one order of magnitude lower than that of bare Pt. The Pt nanocone-modified microelectrode exhibited a much lower impedance of 0.74 ± 0.03 Ω cm 2 due to the extensive surface area of the nanocone array. Meanwhile, the introduction of IrO x coating slightly reduced the impedance of Pt-IrO x composites to 0.72 ± 0.04 Ω cm 2 , because the IrO x coating was thin and did not obviously change the contact area of the nanocone. Such low impedance would help in reducing the energy consumption, and less pulse current would be needed to achieve the same stimulation intensity during practical applications.
A clear increment of the CSC c of the coated microelectrodes can be observed in Figure 5e. The Pt nanocone increased CSC c sharply up to 38.28 ± 3.12 mC cm −2 , compared with that of bare Pt (3.12 ± 0.41 mC cm −2 ), while the addition of IrO x further increased the CSC c of Pt-IrO x composites by approximately 1.37 times (52.44 ± 2.53 mC cm −2 ). Notably, a single IrO x coating without particularly significant contact surface area could enlarge the corresponding CSC c by approximately 3.7 times (11.55 ± 1.61 mC cm −2 ). This could be attributed to the high pseudo-capacitance of IrO x . Moreover, the IrO x /Pt nanocone-coated microelectrode exhibited the highest CIC (4.39 ± 0.36 mC cm −2 ), compared with that of bare Pt (0.08 ± 0.02 mC cm −2 ), Pt nanocone (2.15 ± 0.19 mC cm −2 ), and IrO x (2.25 ± 0.26 mC cm −2 ). The nanocone-coated microelectrode here referred to the variant shown in Figure 2f in Section 3.1. Because the corresponding REDOX reactions occurred when the charge was injected through the reversible Faraday process, more charge could be accommodated between the interface of the electrode and tissue, at once combining the advantages of Pt and IrO x [30,36]. The morphology and the results with identical deposition conditions are highly reproducible and controllable. In particular, these superior CSC and CIC properties would allow the neural electrode to store more energy and improve the stimulation/recording efficiency. A comparison of the main features with similar materials reported in the literature, including electrochemical impedance, CSC c , and CIC, are listed in Table 1.

Chronic Stability Evaluation
The stability of neural electrodes is of great importance for chronic implantation applications, because the electrode surface will be subjected to hundreds of millions of electrical pulse stimulations. In addition, in the actual process, unavoidable friction occurs between the interface of the electrode and the tissue. In the above process, the mechanical stability of Pt nanocone has been verified. In this section we focused on the electrochemical stability of IrO x /Pt nanocone nanocomposite-coated microelectrodes, which was conducted in the same three-electrode electrochemical cell. It was carried out with an electronic stimulator to give the current pulse signal and a digital oscilloscope to collect the potential response. The CV measurements were performed after different pulse times, as shown in Figure 5g.
From the data in Figure 5g, it is evident that a superior electrochemical stability of the IrO x /Pt nanocone nanocomposites can be observed during the first 5 days (4.32 × 10 7 pulse cycles) of electrical stimulation, with less than 5% loss of CSC c (~50 mC cm −2 ). Meanwhile, the CSC c varied slightly to~44.91 mC cm −2 (reduction of~14.47%) when stimula-tion reached 15 d (1.30 × 10 8 pulse cycles), which was still approximately 14.39 and 1.17 times higher than that of bare Pt and Pt nanocone-coated microelectrodes, respectively. Subsequently, we continued the stimulation process for up to one month (2.60 × 10 8 pulse cycles), finding that the impedance of the nanocomposites only increased by approximately 4.29% and CSC c reduced by approximately 23.07%. The inset of Figure 5g shows that the nanocone-array structure was almost unchanged, confirming its robust nature. Table 2 compares the CSC c values of an IrO x /Pt nanocone-coated microelectrode with different electrostimulation duration times. Additionally, ultrasonic bath and bending/twist tests were also carried out; stable impedance and CSC levels can also be verified. Notably, there was no obvious degradation of the electrochemical performance after approximately 1 h of ultrasonic bath or 2000 torsional fatigue cycles. The stable and well-organized nanocone array is expected to promote excellent mechanical and electrochemical stability for long-term biomedical applications. Such structure is more conductive to uniform and close contact between the electrode and the nerve tissue/cells, which may further enhance the stimulation/recording efficiency. In future, we will continue to study the biotic stability in depth. Moreover, this simple strategy of nanocone-array-based structures can further be implemented on different conductive substrates with proper performance.

Biocompatibility Study
Implantation of foreign devices causes tissue stress, and the inserted microelectrode can introduce potential cytotoxicity and foreign body response to brain cells. Microglia is the major resident immune cell type in the central nervous system, playing key roles in scavenging cell debris, maintaining neuronal networks, and orchestrating inflammatory response among others [43]. The effectiveness of the inserted microelectrode array could be limited by its biocompatibility to brain cells, including microglia. To assess the biocompatibility of the designed Pt/Ir nanocone-based microelectrode array in the central nervous system, we tested the effect of bare Pt microelectrodes and Pt nanocones (with or without IrO x coating) on the adhesion and viability of in vitro cultured primary microglia that had been isolated from mouse brain.
For microglia adhesion, microglia were plated on the surface of different microelectrodes, and microglia adhesion was determined by measuring the microglia-specific marker Iba1. Our results clearly showed that the Pt nanocone-based microelectrode, with or without IrO x coating, significantly improved microglia adhesion compared to the bare Pt microelectrode (Figure 6a,b). After culturing with different microelectrode extracts, we performed a CCK8 viability assay to determine the effects of different microelectrode extracts on microglia viability. It can be seen that bare Pt and nanocone Pt extracts showed no difference in microglia viability (Figure 6c). These results indicate that our designed nanocone Pt and Ir acts mainly on microglia adhesion rather than its viability, suggesting an interesting specificity on microglia biocompatibility. nanocone Pt and Ir acts mainly on microglia adhesion rather than its viability, suggesting an interesting specificity on microglia biocompatibility.

Conclusions
In this study, we fabricated nanocone-array-based Pt-IrOx neural microelectrodes, which were demonstrated to have high performance. NH4 + and Pt 4+ ions were controlled to fabricate nanocone-shaped Pt, with a uniform and extensive surface area. It formed an adhesive layer on the bare Pt substrate, which was followed by the deposition of a thin layer of IrOx. The Pt-IrOx nanocomposite-coated microelectrode presents a low impedance of 0.72 ± 0.04 Ω cm 2 and an enhanced CSCc and CIC of up to 52.44 ± 2.53 mC cm −2 and 4.39 ± 0.36 mC cm −2 , respectively. The impedance of the nanocomposites only increases by approximately 4.29%, and CSCc reduces by approximately 23.07% after a continued stimulation process of 2.60 × 10 8 pulse cycles, showing superior chronic stability. In addition, excellent biocompatibility is also obtained. These nanocone-array-based Pt-IrOx neural microelectrodes thus hold promise for alleviating neurological complications. In particular, the inspiring performance is attractive for various implantable/wearable devices, biosensing, energy storage, and other practical applications. Future studies should be directed at investigating the electrophysiological performance of these microelectrodes, both in vitro and in vivo. The capacity to measure neural activity is critical for the development and application of these microelectrodes. Microelectrodes embedded in Multi-Electrode Arrays (MEAs) have been routinely interfaced with isolated brain slices and implanted in the brains of freely moving animals. The MEA method enables the long-term recording of extracellular action potentials from a population of neurons at millisecond time scales and could be applied to the electrical stimulation of neuronal activity.
Author Contributions: Q.Z. and T.Z. conceived the study, designed the experiments, and supervised the study; Q.Z. carried out the fabrication and characterization of the materials; T.Z. performed the biocompatibility study; S.Y., Z.F., and Y.H. assisted in analyzing the results; Q.Z. and T.Z. analyzed the results and wrote the manuscript. T.Z. and B.S. acquired the funding and coordinated the project. All authors have read and agreed to the published version of the manuscript.

Conclusions
In this study, we fabricated nanocone-array-based Pt-IrO x neural microelectrodes, which were demonstrated to have high performance. NH 4 + and Pt 4+ ions were controlled to fabricate nanocone-shaped Pt, with a uniform and extensive surface area. It formed an adhesive layer on the bare Pt substrate, which was followed by the deposition of a thin layer of IrO x . The Pt-IrO x nanocomposite-coated microelectrode presents a low impedance of 0.72 ± 0.04 Ω cm 2 and an enhanced CSC c and CIC of up to 52.44 ± 2.53 mC cm −2 and 4.39 ± 0.36 mC cm −2 , respectively. The impedance of the nanocomposites only increases by approximately 4.29%, and CSC c reduces by approximately 23.07% after a continued stimulation process of 2.60 × 10 8 pulse cycles, showing superior chronic stability. In addition, excellent biocompatibility is also obtained. These nanocone-array-based Pt-IrO x neural microelectrodes thus hold promise for alleviating neurological complications. In particular, the inspiring performance is attractive for various implantable/wearable devices, biosensing, energy storage, and other practical applications. Future studies should be directed at investigating the electrophysiological performance of these microelectrodes, both in vitro and in vivo. The capacity to measure neural activity is critical for the development and application of these microelectrodes. Microelectrodes embedded in Multi-Electrode Arrays (MEAs) have been routinely interfaced with isolated brain slices and implanted in the brains of freely moving animals. The MEA method enables the long-term recording of extracellular action potentials from a population of neurons at millisecond time scales and could be applied to the electrical stimulation of neuronal activity.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.