Boron phosphide microwires based on-chip electrocatalytic oxygen evolution microdevice

Summary

Developing low-cost, high-performance metal-free electrocatalysts is crucial for sustainable oxygen evolution reactions (OERs) in alkaline media. Here, we synthesized ultra-long boron phosphide (BP) microwires along the [111] crystal axis with high yield. By controlling temperature and adding Ni, we identified Ni’s role as a flux and transport agent, optimizing BP microwires at 1050°C. Electrical property and band structure analysis revealed BP as a one-dimensional p-type semiconductor with a wide band gap. We constructed an on-chip electrocatalytic microdevice using individual BP microwires to evaluate OER performance. In 1M NaOH, the BP electrode achieved 50 mA cm⁻² at an overpotential of 320 mV, outperforming other boron-based catalysts. Additionally, the nanocatalyst exhibited a low Tafel value (88 mV·dec⁻¹) and excellent stability. This study offers valuable insights for developing future electrocatalysts and electrochemical reaction models.

Graphical abstract

Highlights

• •Prepared ultra-long boron phosphide microwires quickly via vapor-solid process
• •Microscopic study of OER on a single BP microwire in an on-chip electrochemical device
• •Non-precious metal oxygen evolution reaction catalyst comparable to IrO₂

Catalysis; Electrochemical materials science; Engineering

Introduction

To alleviate the dilemma of fossil energy depletion, hydrogen energy is widely pursued as a clean renewable energy source. One of the strategies to achieve mass production of hydrogen energy is the electrocatalytic decomposition of water; oxygen evolution reaction (OER) is a key half-reaction involved in the decomposition.^1,2,3 However, the industrial scale OER under alkaline conditions results in sluggish kinetics and expensive precious metal electrocatalysts. In contrast, metal-free catalysts exhibit a significantly lower susceptibility to corrosion and do not release metal ions that could contaminate the surrounding environment. Moreover, these catalysts can be conveniently reused. Consequently, there is a strong motivation to advance the development of metal-free catalysts for the purpose of enhancing the OER process. Drawing inspiration from semiconductor nanodevices, the on-chip electrocatalytic microdevice, employing an individual microwire as the working electrode, has emerged as a robust platform for extracting intrinsic catalytic performance at the nanoscale.^4,5,6,7 Moreover, the individual microwire located on the on-chip electrocatalytic microdevice can be meticulously crafted, which is vital in removing disturbances caused by various factors and facilitates quantitative or semi-quantitative analysis of fundamental principles.^8,9,10,11

Boron phosphide (BP), composed solely of boron and phosphorus, is a promising metal-free catalyst for OER in alkaline solutions^12,13,14,15 due to its low cost, environmental friendliness, and excellent chemical stability. Single-crystalline BP possesses a cubic-bonded structure characterized by a small lattice constant and robust covalent bonding with minimal ionicity. Additionally, it showcases exceptional chemical and thermal stabilities due to its elevated thermal conductivity and Debye temperature. However, the implementation of BP is constrained by the intricacy, low yield, insufficient purity, and elevated expenses associated with the current preparation methods. Various techniques have been employed in the synthesis of BP, including the reduction of PBI₂ using hydrogen gas,¹⁶ the decomposition of BBr₃PH₃,¹⁷ chemical vapor deposition,^17,18 gas phase reactions,¹⁹ reactions involving molten nickel or nickel phosphide,²⁰ and methods utilizing high pressure.^21,22,23 The product obtained through mechanochemical and self-propagating high-temperature synthesis methods exhibited impurities due to challenges in managing the reaction process effectively. Meanwhile, the flux methods and solvothermal synthesis methods were hindered by low yields and extended synthesis durations. Additionally, the toxic raw materials, significant amounts of impurities, and defect densities in the final products for CVD synthesis is well known. Therefore, the study of the synthesis and applications of pure BP is of great significance.

Herein, to address this issue, we effectively produced an ultralong pure BP microwire through a vapor-solid (VS) growth mechanism, thereby resolving the thermodynamic conflict posed by the high melting point of boron and the low sublimation temperature of phosphorus. These microwires exhibited high crystallinity and grew along the [111] crystal axis. By adjusting the synthesis conditions of BP microwires, we found that at temperatures around 1050°C, with the addition of Ni as a flux and transport agent, high-quality BP microwires with diameters ranging from 250 to 350 nm and lengths exceeding 250 μm could be prepared. Furthermore, we have confirmed that these microwires exhibited p-type semiconductor behavior. Subsequently, we fabricated individual BP microwire based on on-chip electrocatalytic microdevices to characterize the intrinsic oxidative behavior during the OER. When we tested the micro-OER in 1M NaOH, BP microwire achieved a low OER catalytic overpotential (320 mV) and fast kinetics (Tafel slope 88 mV·dec⁻¹). Our study provides new insights into the synthesis of one-dimensional boron-based materials and offers a pathway for investigating the synthesis mechanisms of other boron-based or phosphorus-based one-dimensional materials. The developed on-chip electrochemical microdevice platform facilitates future observations and understanding of the intrinsic catalytic behaviors of electrocatalysts at the microscale.

Results

Characterization of BP microwire

In the B-P-Ni system, nickel (Ni) acted as a flux, facilitating the transition of bulk boron to gaseous Ni-B. Ultra-long BP microwires were synthesized via VS process. The growth process principle and procedural temperatures were illustrated in Figures 1A and 1B, respectively. Ni-B reacted with gaseous phosphorus to form nanocrystals, which then continuously grown into BP microwires. The synthesized BP microwires exhibited high crystallinity and uniform diameter (Figure 1C). The X-ray diffraction (XRD) was conducted to comprehensively analyze the structure of BP microwires. The XRD diffraction peak of BP microwires was successfully characterized as cubic F43m structure (JCPDS NO.11-0119) (Figures 1D and 1E), and the intensity of the diffraction peak strength and the narrow half-peak width indicated that BP microwires had high crystallinity and no detectable impurities.²⁴ Additionally, BP microwires were exposed to a normal temperature air environment for a duration of 6 months. A comparative XRD analysis between the BP microwires from six months prior revealed no structural alterations (Figure S1). This suggested that the microwires maintained stability in atmospheric atmosphere. The Raman spectrum of BP microwires was displayed in Figure 1F, revealing the presence of two peaks. These peaks are situated at 798.2 cm⁻¹ (TO) and 832.7 cm⁻¹ (mixture of TO and LO) in the Raman spectrum of BP microwires. This was because in the case of the F43m structure without an inversion center, electrostatic interaction between ions presented in cubic BP caused the unique Lyddane–Sachs–Teller optical mode at the center of the Brillouin region TO split into a transverse optical phonons (TO) and a longitudinal optical phonons (LO) (TO-LO splitting), which resulted in the existence of two modes: TO (Г) and LO (Г).^25,26 However, due to the weak iconicity of BP, it’s TO-LO splitting was less than 3 cm⁻¹. The peak (TO) observed at 798.2 cm⁻¹ corresponding to TO isotopic disordered scattering of LO phonons near X or K in the Brillouin region. Additionally, the peak at 832.7 cm⁻¹ was attributed to mixed mode of TO (Γ) and LO (Γ).^12,27,28 According to atomic force microscopy (AFM) analysis, the BP microwires exhibited a consistently smooth surface and a uniform distribution of tube diameters around 230 nm. (Figure S2).

Figure 1.

Fabrication and characterization of BP microwires

(A) Schematic illustration of the growth process for BP microwires.

(B) Programmed temperature curve for BP microwires growth.

(C) SEM image of as-grown BP microwires.

(D) Unicellular structure of cubic BP.

(E) XRD of BP material, including comparison with PDF card of cubic BP.

(F) Raman spectrum of BP.

To further investigate the structural details at the atomic scale, the BP microwires were transferred onto a copper grid and investigated by Low-resolution transmission electron microscopy (TEM). TEM image of representative BP microwires (Figure 2A) showed clean surfaces and uniform diameters. The high-resolution transmission electron microscopy (HRTEM) image (Figure 2B) reflected the structural properties of monocrystalline BP microwires and the growth direction of the microwires. The crystal face spacing of 1.6 Å and 2.60 Å were respectively corresponded to the crystal plane of cubic structure BP (111) and ($2\overline{2}0$), which indicated that the BP microwires were grown along the crystallographic axis [111] direction. Through XRD and TEM analysis, the synthesized BP microwires exhibited crystal structure and lattice constant that were closely matched with those of previously reported isotope BP microwires.²⁹ The images produced by the energy dispersion system mapping demonstrate a uniform spread of B and P across the spatial area (Figure 2C). To further reveal the presence of B and P elements more extensively, we analyzed the X-ray photoelectron spectra (XPS) of BP in the B 1s, P 2P region, as well as the entire region. (Figures 2D, 2E, and S3). XPS clearly showed that the characteristic peaks of species B and species P in BP appeared at 187.9 eV and 128.4 eV, respectively, which indicated that B and P were bound by strong covalent bonds. Therefore, BP microwires generally had good mechanical properties, which was basically consistent with that had been reported in the literature.²⁵ Furthermore, the BP microwires were synthesized at an exceedingly elevated temperature of 1050°C. Considering the thermodynamics of the reaction, this high synthesis temperature typically signifies the strong binding energy of the atomic bonds, leading to the theoretical attainment of stable physical and chemical properties, which might contribute to the exceptional stability observed in BP microwires.

Figure 2.

TEM and XPS characterization of BP microwires

(A) Low power TEM image of as-grown BP microwire.

(B) HRTEM image of as-grown BP microwire. Inset: The corresponding selective electron diffraction (SAED) image.

(C) EDS mapping of individual BP microwires.

(D) High-resolution XPS spectra of B 1s in BP microwires.

(E) High-resolution XPS spectra of P 2s in BP microwires.

Modulation synthesis of BP microwires

The optimal growth temperature of BP microwires was investigated, revealing significant morphological changes at various temperatures while maintaining identical growth parameters. Figures 3A–3D depicted the variations in BP microwires growth at different temperatures. Under the same growth conditions, at 750°C, the reaction products consisted of uniformly sized nanoparticles. As the temperature increased to 900°C, a small amount of short, irregular one-dimensional rod-like structures appeared in the products. At 1050°C, long and regularly shaped microwires emerged as predominant products, primarily because at this temperature, a significant portion of Ni-B undergoes a transformation into gas. The supersaturated vapor source frequently leads to an increase in nucleation and growth rates, ultimately resulting in the formation of denser and longer BP microwires. Figure 3E illustrated the lengths of BP microwires synthesized at different temperatures, with the longest BP microwires synthesized at 1050°C. Further temperature increased led to quartz tube recrystallization, which might block the growth endpoints of BP microwires and exacerbate the risk of fracture (Figure S4). Figures 3F–3I presented high-magnification SEM images of the synthesized products at different temperatures. Figure 3J provided statistical data on the diameters of BP microwires at different temperatures. At 900°C, due to low nucleation rates and slow molecular diffusion at lower temperatures, BP microwires exhibited a transition from coarse to fine. At 1200°C, crystallization of quartz tube impeded further axial growth of BP microwires, and supersaturated vapor sources in the atmosphere continue to deposit on the sidewalls of microwires, leading to an increase in wire diameter.

Figure 3.

The morphology of BP microwires prepared at different temperatures

(A–D) Low magnification SEM images of as-grown BP microwires, scale bars: 20 μm.

(E) Length of BP microwires synthesized at different temperatures (Data are represented as mean ± SEM).

(F–I) High magnification SEM images of as-grown BP microwires, scale bars: 5 μm.

(J) Diameter of BP microwires synthesized at different temperatures (Data are represented as mean ± SEM).

Band structure and electrical properties of BP microwires

To predict the electronic structure of BP microwires, we calculated the band structure and density of states (DOS) diagram (Figure S5B) of cubic phase BP by using the improved plane elastic band (PEB) method of hybrid functional theory (HSE 06). The cubic phase BP exhibited a distinct indirect band gap, which was easily observable. Because of the limitation of density functional theory, the calculated result was 1.93 eV. Therefore, we speculated that BP microwires might be an indirect band gap semiconductor with a band gap of about 1.93 eV. The absorption spectrum as shown in Figure S6A was measured by UV-vis spectrophotometer. By using the Kubelka–Munk equation and the Lambert–Beer law,³⁰ the calculated optical band gap of BP microwires was 2.19 eV (Figure S6B), which was significantly larger than the theoretical band gap calculated by density functional.

Additionally, we tested the PL of BP microwires under an excitation wavelength of 532 nm laser at room temperature, as shown in Figure S5A. Five luminescence peaks of BP microwires were obtained after the Gaussian fitting, which were located at 563.7, 791.2, 1096.9 and 1582.5 nm, respectively. Combined with reflection spectra, the multipeak photoluminescence of BP microwires was interpreted as the result of radiation recombination predominantly centered on defects.³¹

The electrical characteristics of the synthesized BP microwires were assessed by fabricating field-effect transistors. Figure 4A depicted the schematic representation of the device, while Figure 4B presents the optical image of the device. The I-V curves of BP microwires (Figure 4C) showed that the nonlinearity in the central region of each curve was derived from non-ohmic electrode contact, which was characteristic of semiconductor microwires. When the applied gate voltage V_gs = +80 V, the I-V curve became linear, which indicated the Schottky barrier disappeared and ohmic contact formed. As-grown BP microwires exhibited a characteristic behavior of a p-type semiconductor, wherein an increase in gate bias led to a decrease in current (Figure 4D). This behavior was due to the presence of holes as majority charge carriers in the BP microwires. When a positive gate bias was applied, it attracted these positively charged holes toward the surface of the microwires, effectively reducing the number of available charge carriers and therefore decreasing the current flow. This p-type behavior is essential for the functionality of many electronic devices, such as transistors and diodes, where the control of current flow is crucial for their operation.

Figure 4.

Electrical performance of BP microwire based field-effect transistor

(A) Schematic view of the BP microwire based field-effect transistor.

(B) Optical image of single BP microwire device.

(C) Output curves (I_d-V_ds) of as-synthesized BP microwire.

(D) Transfer curve (I_d-V_gs) of as-synthesized BP microwire.

Single BP microwire micro-area OER test

To investigate the OER activity and oxygen evolution mechanism of as-grown BP microwire, we designed a typical three-electrode device to perform electrochemical tests in a micro-electrochemical test platform with 1.0 M NaOH as the electrolyte. As depicted in Figure S7, we first eliminated the interference of Ni element on the OER catalytic testing. Figure 5A illustrated the on-chip electrocatalytic microdevice concept for the electrochemical measurement. The measurements were conducted using a standard three-electrode system in an NaOH solution (1 mol L⁻¹) droplet, utilizing platinum wire and Ag/AgCl electrode as the counter and reference electrodes, respectively. Notably, the ends of individual BP microwire were firmly secured onto Ti/Au working electrode. The exposed catalytic regions in this experiment measured 5 × 12 μm, as determined by E-beam lithography (Figure 5B). The linear sweep voltammetry (LSV) curves of BP microwire showed that the OER performance of BP microwire under white light irradiation was significantly better than that measured in dark environment (Figure 5C). This might be due to that BP microwires were one-dimensional semiconductors with wide bandgaps, and their intrinsic exciton effects, together with traditional carriers, constitute the photoexcitation process, thus making important contributions to the photocatalytic behavior.³² In addition, the OER performance of BP microwire was better than that of BP nanoparticles (Figure S8), which could be attributed to the considerable high electron mobility of BP microwire and the exposed large area (111) and ($2\overline{2}0$) crystal planes. After compensation and calculation of the OER data, the OER overpotential of a single BP microwire was 320 mV at a current density of 50 mA cm⁻² (Figure 5D), which was equal to the performance of the excellent OER catalyst reported so far (Table 1). The Tafel slope of a single BP microwire was 88 mV·dec⁻¹ (Figure 5E), indicating that the OER reaction had a faster kinetic process, which was beneficial to accelerate the oxygen generation rate.

Figure 5.

OER activity of individual BP microwire on the electrocatalytic microdevice

(A) Schematic depiction of the on-chip electrocatalytic microdevice.

(B) Optical microscope image of the individual BP microwire microdevice.

(C) LSV curves of BP microwire under white light and dark conditions in 1 M NaOH.

(D) Overpotential histogram of BP microwire at different current densities.

(E) Tafel curve of BP microwire.

(F) V-t curve of BP microwire.

Table 1.

Comparison of characteristic parameters between BP microwire and high-performance OER electrocatalysts reported in the literature

Materials	Measurement conditions	Potential at 50 mA cm−2	Tafel slope [mV dec−1]	References
NiO/Co3O4	1.0 M KOH	1.57 V vs..RHE	78	Zhang et al.33
RuO2	1.0 M KOH	1.54 V vs..RHE	64.5	Zhang et al.34
IrO2	0.1 M KOH	1.63 V vs..RHE	215	Li et al.35
P-Co3O4	1.0 M KOH	1.76 V vs..RHE	51.6	Xiao et al.36
Pt/C	1.0 M KOH	2.2 V vs..RHE	138.5	Zhao et al.37
BP microwire	1.0 M NaOH	1.55 V vs..RHE	88	This work

To gain a deeper understanding of catalyst behavior, elucidating the transition processes of intermediates is a crucial aspect of studying catalyst performance. In our investigation of the OER mechanism involving BP microwires, the four-electron OER process in alkaline media is considered an effective approach for exploring the activity of BP compounds.^38,39,40 The steps of this process are as follows:

$${}^{\ast }+4{\text{OH}}^{-}\to \text{OH}{}^{\ast }+3{\text{OH}}^{-}+{\mathrm{e}}^{-}$$ (Equation 1)

$$\text{OH}{}^{\ast }+3{\text{OH}}^{-}+{\mathrm{e}}^{-}\to \mathrm{O}{}^{\ast }+{\mathrm{H}}_2\mathrm{O}+2{\text{OH}}^{-}+2{\mathrm{e}}^{-}$$ (Equation 2)

$$\mathrm{O}{}^{\ast }+{\mathrm{H}}_2\mathrm{O}+2{\mathrm{e}}^{-}\to \text{OOH}{}^{\ast }+{\mathrm{H}}_2\mathrm{O}+{\text{OH}}^{-}+3{\mathrm{e}}^{-}$$ (Equation 3)

$$\text{OOH}{}^{\ast }+{\mathrm{H}}_2\mathrm{O}+{\text{OH}}^{-}+3{\mathrm{e}}^{-}\to {\mathrm{O}}_2+2{\mathrm{H}}_2\mathrm{O}+4{\mathrm{e}}^{-}$$ (Equation 4)

Here, the ∗ indicates the active sites on the surface, while molecules marked with ∗ represent their adsorbed states. Due to differences in electronegativity, oxygen atoms can more readily extract electrons from phosphorus (P),⁴¹ which suggests that the phosphorus sites on the surface of BP microwires may serve as the active sites for the OER. The electronegativity difference between phosphorus and boron allows adjacent boron atoms to induce charge separation within the lattice, further enhancing the adsorption of hydroxide ions (OH⁻) on the phosphorus atoms, thereby increasing OER activity. Under alkaline conditions, hydroxyl groups in the solution oxidize the P-OH species on the surface of the BP microwires, converting them into water (H₂O). Simultaneously, two O species couple directly to form oxygen gas. This process establishes a cycle that continuously produces oxygen.

Subsequently, to study the stability of BP microwire OER, the V-t curve was used to test the durability of a single BP microwire at 402 mV overpotential (under a current density of 100 mA cm⁻²), and the results showed that the catalyst maintained a current density of 100 mA cm⁻² after 2500 s, which confirmed that BP microwires had good OER stability (Figure 5F). Due to limitations such as electrolyte loss and the stability of products (O₂) affecting the testing window, on-chip electrocatalytic microdevices are unsuitable for long-term stability testing. As an alternative approach, we evaluated the V-t curves of BP microwires catalysts at a current density of 50 mA cm⁻² using a conventional electrochemical workstation. The results indicated that under testing conditions of 1 M KOH, the BP microwires exhibited exceptional long-term stability for the OER at 50 mA cm⁻² for over 48 h (Figure S9).

Discussion

Existing literature reports that the synthesis mechanism of BP microwires is the vapor-liquid-solid (VLS) mechanism.⁴² We had synthesized ultra-long BP microwires using the same procedure (Figure S10). However, the BP microwires we synthesized did not exhibit the conventional droplet-shaped catalyst tips (Figures S11 and S12), which clearly contradicts the VLS mechanism. To investigate the role of nickel in the synthesis process, we designed a nickel-free synthesis system. Under the same conditions, we successfully synthesized short rod-like products. XRD analysis indicated that the products contained cubic BP and boron phosphate (Figure S13), which is highly consistent with the expected products from our proposed VS synthesis mechanism.⁴³

Additionally, characterization of the ends of the BP microwires synthesized in the nickel-containing system revealed that the growth initiation sites contained only carbon (C), oxygen (O), and phosphorus (P), with no detectable nickel present (Figure S12). The energy dispersive spectroscopy (EDS) peaks for boron were very close to those for carbon, making it difficult to distinguish between them, while the carbon and oxygen elements originated from the carbon conductive tape used during sample preparation. Based on these results, we speculate that the VS mechanism may represent the actual synthesis mechanism for the microwires. Furthermore, we hypothesize that the role of nickel is primarily to form a lower melting point solid solution with boron, thereby accelerating the formation of gaseous boron sources and promoting the growth of the microwires. This also explains why the products of the nickel-free synthesis system predominantly consist of short rod-like BP structures.

The presence of oxygen in this process may be linked to the oxygen (O₂ and H₂O) found on the inner lining of the quartz tube, as well as the oxygen that is enclosed between the layers of two-dimensional black phosphorus. The evaporation of the latter might result in a rise in oxygen content. Consequently, the remaining oxygen underwent reactions with boron powder at elevated temperatures, producing B₂O₂ and B₂O₃ vapors as outlined in the subsequent reactions:

$$2\mathrm{B}\left(\mathrm{s}\right)+{\mathrm{O}}_2\left(\mathrm{g}\right)\to {\mathrm{B}}_2{\mathrm{O}}_2\left(\mathrm{g}\right)$$ (Equation 5)

$$4\mathrm{B}\left(\mathrm{s}\right)+{3\mathrm{O}}_2\left(\mathrm{g}\right)\to {2\mathrm{B}}_2{\mathrm{O}}_3\left(\mathrm{g}\right)$$ (Equation 6)

When the temperature gradually increased, the black phosphorus vaporized, forming P vapor. Subsequently, B₂O₂ and B₂O₃ reacted with the P vapor to produce BP vapor through the following reaction:

$${2\mathrm{B}}_2{\mathrm{O}}_2\left(\mathrm{g}\right)+4\mathrm{P}\left(\mathrm{g}\right)\to \text{BP}\left(\mathrm{s}\right)+\mathrm{B}\left(\mathrm{P}{\mathrm{O}}_4\right)\left(\mathrm{s}\right)$$ (Equation 7)

$${4\mathrm{B}}_2{\mathrm{O}}_3\left(\mathrm{g}\right)+8\mathrm{P}\left(\mathrm{g}\right)\to 5\text{BP}\left(\mathrm{s}\right)+3\mathrm{B}\left(\mathrm{P}{\mathrm{O}}_4\right)\left(\mathrm{s}\right)$$ (Equation 8)

The above-mentioned gas-solid reaction was used to form BP crystal nuclei. Subsequently, due to the principle of lowest energy, the formed BP crystal nuclei acted as seed crystals and were connected to each other, forming BP microwires along the preferred crystallographic axis [111]. After the cessation of single crystal growth, a swift annealing technique was employed to minimize the imperfections.

Additionally, the influence of temperature on the growth status of BP microwires was investigated by controlling the same growth parameters. Initially, at a lower temperature of 750°C, the reaction products were primarily regular particles. As the temperature increased to 900°C, a small number of one-dimensional short rod-like structures developed on the basis of particles. No droplet structures were found at the ends of these short rods, providing further evidence that BP microwires grow via the VS mechanism at different temperatures. The optimum growth temperature for BP microwires was determined to be around 1050°C. As the temperature further increased, crystallization of the quartz tube onto the microwires increased their risk of fracture, and higher temperatures accelerated the deposition rate of the gas source on the microwires, thereby increasing the diameter of the BP microwires. Despite using sealed quartz tubes for the production of BP microwires, the current output still falls short of meeting practical application demands. To enhance the feasibility of scaling up production, we are considering the design of a device that has a large volume, high-temperature resistance and excellent airtightness. Additionally, controlling the temperature gradient within the device is a critical issue that needs to be addressed. Furthermore, improving the yield and stability of the nanowire quality during the scaling-up process presents a significant challenge that we must overcome.

Finally, we tested the OER catalytic performance of BP microwires, which exhibited a significant improvement in performance compared to traditional BP particles. This may be attributed to a dimensional transformation in BP material, where one-dimensional BP microwires may expose more catalytic active surfaces compared to bulk BP. Moreover, existing literature indicates that both p-type and n-type BP exhibit highly stable, metal-free photocatalytic performance for hydrogen evolution under visible light irradiation.^44,45 However, the hydrogen evolution reaction is only one half-reaction of the water-splitting process, with the OER representing another critical half-reaction. Inspired by the efficient photocatalytic water splitting of hole-rich Ni₂P-based photocatalysts under visible light,^46,47 we conducted photocatalytic OER tests on the synthesized hole-rich microwires samples. Due to the energy level of water oxidation being higher than the valence band maximum of BP microwires, it is theoretically permissible to transfer holes from the BP microwires, thereby facilitating the release of oxygen.⁴⁶ Additionally, under Xenon lump, BP microwires generate photo-induced carriers, which rapidly transport within the one-dimensional structure, enabling effective separation of the photo-generated carriers. This may also contribute to the enhancement of their photocatalytic water splitting efficiency.⁴⁸ To simulate natural sunlight, we employed a xenon lamp light source with a power of 100 mW/cm² (>420 nm) for testing. Under vertical illumination with visible light, the catalytic performance of the samples shows a slight improvement. This enhancement can be attributed to the ability of light to penetrate into the interior of the catalyst, generating electron-hole pairs. However, this also results in the generated electrons and holes having difficulty in effectively transferring to the catalyst surface to react with adsorbed atoms or ions, thereby hindering the progression of the OER. Consequently, utilizing metal-free BP catalysts for photocatalytic water splitting to release oxygen remains a challenge. One promising solution is the confined synthesis of few-layer BP, and we are currently actively pursuing research in this direction.

Despite their low overpotential compared to other high-performance composite materials, BP microwires exhibited inferior catalytic performance, which may be attributed to their limited conductivity and active site density. However, the OER performance of BP microwires was significantly improved compared to BP particles. This is due to the significant impact that the design and fabrication of nanostructured electrocatalysts have on catalytic efficiency. One-dimensional (1D) nanomaterials have been established as a unique platform for excellent electrocatalyst candidates. First, 1D BP microwires exhibit a high surface area, large roughness factor, and high density of active sites, which collectively enhance their catalytic activity for surface electrochemical reactions. Second, 1D BP microwires facilitate rapid charge transport by providing channels while minimizing the number of crystal boundaries, thereby reducing scattering losses. Third, the abundant open space and porosity between adjacent 1D nanostructures allow electrolyte molecules to rapidly diffuse and enhance chemical accessibility deep into the electrode/catalyst surface.^39,49,50

In summary, nickel (Ni) played a dual role as a flux in our growth system, significantly facilitating the reaction between boron and black phosphorus. High-density BP microwires were successfully synthesized using the chemical vapor transport (CVT) method at 1050°C for a relatively short duration. It is proposed that the synthesis of BP microwires may occur via a VS mechanism. These microwires exhibit uniform diameters and exceptionally high aspect ratios. Additionally, the prepared BP microwires exhibit evident one-dimensional wide bandgap P-type semiconductor characteristics. Furthermore, an on-chip electrocatalytic microdevice was proposed and fabricated to assess electrocatalytic OER based on individual BP microwire. When tested micro-OER in 1M NaOH, a single BP microwire exhibited excellent electrocatalytic performance (overpotential 320 mV) and fast kinetic process, and demonstrated long-term catalytic stability in a strong alkaline electrolyte. The catalytic performance of the synthesized BP microwire OER catalyst rivals that of recently reported advanced OER catalysts and surpasses some commercially available precious metal OER catalysts. Our study confirms the crucial role of Ni in the synthesis of BP microwires. Additionally, the unique on-chip electrocatalytic microdevice platform we developed can be applied to other relevant fields for in situ monitoring and understanding of the dynamic behaviors of energy materials at the nanoscale.

Limitations of the study

A limitation of on-chip studies of the OER performance of individual BP microwire lies in the fact that this method can only closely approximate the OER performance of a single BP microwire. However, in actual testing, various factors may affect the testing accuracy of the on-chip electrocatalytic microdevice.

The contact between a single microwire and the on-chip electrocatalytic microdevice substrate introduces significant interference in the calculation of the microwire’s reactive surface area. Additionally, since the current through the on-chip electrocatalytic microdevice samples is generally very small, a few test runs may result in slight errors in the accuracy of the current monitoring device.

To address this limitation, one approach is to etch channels on the substrate, allowing the sample to be suspended in the middle and thus reducing interference from the substrate contact. Additionally, increasing the sample size and employing higher-precision current monitoring devices can enhance the accuracy of the on-chip electrocatalytic microdevice testing.

Resource availability

Lead contact

Requests for further information and resources should be directed to and will be fulfilled by the lead contact, Lixing Kang (lxkang2013@sinano.ac.cn).

Materials availability

All unique/stable reagents generated in this study are available from the lead contact with a completed materials transfer agreement.

Data and code availability

Instructions for section 1: Data.

Raw data reported in this paper will be shared by the lead contact upon request.

Instructions for section 2: Code.

This paper does not report original code.

Instructions for section 3: Additional information.

Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.

Acknowledgments

The authors are grateful for the technical support for Nano-X from Suzhou Institute of Nano-Tech and Nano-Bionics, Chinese Academy of Sciences (SINANO). This work was financially supported by the National Natural Science Foundation of China (52372055), Jiangsu Funding Program for Excellent Postdoctoral Talent (2023ZB167, 2023ZB065, 2023ZB293), Basic Research Development Program of SuZhou (SJC2023004), the Natural Science Foundation of Jiangsu Province (BK20240484) and the China Postdoctoral Science Foundation (2022M722331, 2023M742561, 2023M742557).

Author contributions

Conceptualization: L.K. and L.W.; data curation: L.W. and H.S.; formal analysis: H.S.; investigation: L.W., H.S., X.P., Q.G., H.L., A.L., Y.Z., X.Z., and S.L.; resources: L.W. and H.S.; writing – original draft: H.S.; writing – review and editing: L.W. and Q.G.

Declaration of interests

The authors declare no competing interests.

STAR★Methods

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Software and algorithms
Excel	Microsoft Excel	Oprogramowanie arkuszy kalkulacyjnych Microsoft Excel | Microsoft 365
Macros in Excel environment	Microsoft Excel	Oprogramowanie arkuszy kalkulacyjnych Microsoft Excel | Microsoft 365
Origin	Open source	OriginLab - Origin and OriginPro - Data Analysis and Graphing Software

Experimental model and study participant details

There are no experimental models (animals, human subjects, plants, microbe strains, cell lines, primary cell cultures) used in the study.

Method details

A mixture of red phosphorus (AR, 98.5%, Aladdin), tin (99.85%, Alfa) and tin iodide (99.999%, Macklin) was evacuated quartz tube as the precursor for the growth, then that was put into a muffle furnace (BEQ, MF-1600C) and heated to 650°C. After holding at the maximum temperature for 1 h, the furnace was cooled to 300°C at a constant cooling rate of 0.22°C per minute for 26.5 h, then the black phosphorus single crystals with high quality were successfully prepared.⁵¹ BP microwires produced in different batches at various time intervals were purified, and their optical images indicate that the yield across batches remains stable (Figure S14A). SEM images corresponding to the respective batches of BP microwires are shown in Figures S14B–S14D, further demonstrating that the quality of the microwires across different batches is also consistent.

The raw B powder (≥95%, Sigma-aldrich), Ni (99.7%, Alfa), and black phosphorus (synthesized in the previous step) were mixed and ground with mole ratio of 1:2.5:1, which was sealed in a vacuum. Samples were first heated to 1050°C within 480 min and incubated at this temperature for 1 h, then slowly cooled to 500°C at a rate of 16.5°C/h, and then cooled to room temperature at a rate of 120°C/h.⁴²

The Optical image was obtained by using Nikon Eclipse LV100ND microscope. The SEM images were observed by HITACHI S-4800. AFM images were captured with a Bruker Dimension ICON microscope in tapping mode in air. Raman and PL characterizations were performed using LabRAM HR Evolution Raman spectrometer (Horiba Jobin Yvon) with the excitation laser of 532 nm. XRD patterns were obtained from a Bruker D8 phase diffractometer with Cu Kα radiation (40 kV, 30 mA). TEM measurements were carried out by TEM (Talos F200STEM) operated at an accelerating voltage of 200 kV. XPS measurements were performed on an a ULVAC-PHI (PHI 5000 Versaprobe II) using monochromatic Al Kα X-ray radiation as the exciting source. The optical properties were studied using a UV-vis spectrophotometer (Hitachi U-3900H).

The bottom electrode pattern and overlay alignment marks were prepared by laser direct writing on Si/SiO₂ substrates, and 5 nm Ti and 80 nm Au were deposited by electron beam evaporation as working electrodes, and BP microwires were transferred to the working electrodes at fixed points. Take 3mg BP microwires dispersed in 1ml ethanol solution. A 20ul solution was dropped on the substrate with an electrode, and the microwires at the appropriate position were selected under an optical microscope. Subsequently, 600 nm thick PMMA photoresist was spun on the sample, and a window was prepared on the BP microwires by electron beam lithography overlaying. The exposed lateral surface area of BP microwires contributed to the electrochemical oxygen evolution reaction. The current measured by the on-chip electrocatalytic microdevice is prone to interference from current noise in the measurement system. The noise level cannot be easily improved by simply adjusting the settings of the source/measurement unit, as it is primarily limited by the measurement system of the electrolyte, which is already optimized with the best parameters and settings. It is worth noting that, aside from current noise, leakage current is a key factor contributing to measurement errors.⁵² Moreover, due to the extremely small size of most individual nanomaterials, measurements in the on-chip electrocatalytic microdevice typically provide qualitative and semi-quantitative analysis. Since the Tafel slope mainly reflects charge transfer kinetics and is not affected by the diffusion process, the exposed surface area of individual nanomaterials has little effect on the Tafel slope.⁸ It provides a more accurate measure of the catalytic performance of individual nanomaterials. Based on this principle, a lower Tafel slope indicates faster catalytic kinetics, reflecting superior performance of the catalyst.

NaOH solution with concentrations of 1 M was configured as the oxygen evolution reaction electrolyte. The reference electrode employed was an Ag/AgCl electrode. The platinum wire was served as the counter electrode. The keithley 2450 source meter was used as the test instrument. A voltage was applied to the counter electrode to collect the voltage at both ends of the reference electrode and the working electrode. The measured potentials vs. Ag/AgCl were translated onto a reversible hydrogen electrode (RHE) scale through application of the Nernst equation:

$${\mathrm{E}}_{\text{RHE}}\left(\mathrm{V}\right)={\mathrm{E}}_{\text{Ag}/\text{AgCl}}+{\mathrm{E}}_{\text{Ag}/\text{AgCl}}^0+0.059\times \text{PH}$$ (Equation 9)

Where E⁰_Ag/AgCl = 0.1976 V at 25°C, the experimentally measured potential against the Ag/AgCl reference electrode is denoted as E_Ag/AgCl, while E_RHE represents the converted potential relative to the RHE scale.

Quantification and statistical analysis

No methods were used to determine whether the data met the assumptions of the statistical approach.

Additional resources

Our study has not generated or contributed to a new website/forum or has not been part of a clinical trial.

Published: January 21, 2025