Advancing renewable energy technologies is essential for addressing the dual challenges of climate change and energy sustainability.1^–3 Among the wide range of renewable energy sources, solar energy, as the most abundant resource on earth, offers tremendous potential for scalable solutions and global deployment, making it a cornerstone of future energy strategies.4^–8 Silicon plays a vital role in solar energy applications,9^–12 where its semiconductor properties enable efficient light absorption and conversion capacity. However, its indirect bandgap nature and high reflectance usually around 40% from visible light to mid-infrared wavelength, greatly restrict its applicability in photon-sensitive devices.13^–16 Addressing these challenges to improve the efficiency and cost-effectiveness of silicon-based solar technologies remains a critical area of research and development.17

Black silicon (b-Si) is a unique silicon material characterized by its micro- and nanoscale structures, which have developed rapidly in recent years.18^–23 This material is notable for its ultra-low reflectivity, absorbing nearly all visible light and appearing pitch black to the naked eye, hence the designation “black silicon.”24 In contrast to conventional silicon which reflects light and appears gray, b-Si can absorb up to 99% in a broad wavelength range between 350 nm and 2250 nm, reflecting only about 1%.25^,26 It is usually produced through various fabrication techniques, such as electrochemical etching, stain etching, metal-assisted chemical etching (MACE), reactive ion etching (RIE), and laser treatment, which create a textured surface with high aspect ratios.22^,23^,27 This unique structure not only improves optical absorption over a wide range of wavelengths but also enhances charge carrier dynamics, making it a promising candidate for solar energy conversion.28 In addition, when doped with specific impurities, such as sulfur, b-Si can extend its absorption capabilities into the near-mid infrared (NIR-MIR) range. These remarkable properties have spurred extensive research into various solar energy conversion applications in recent years.29^–34 Beyond energy conversion, b-Si nanostructures have emerged as versatile materials for photonic applications, in which the tunable optical properties enable advancements in photodetectors,35^,36 surface-enhanced Raman scattering (SERS),37^–40 and imaging optoelectronic devices.41^–44 The incorporation of b-Si into these areas is paving the way for innovative technologies that can operate effectively across various wavelengths.

Although several prior review articles have examined individual nanostructures of b-Si,45^,46 and specific applications such as solar cells,19^,30 or photodetectors,22 they often overlook the diverse nanostructures of b-Si, including holes, pyramids, and pillars, as well as comprehensive applications in solar energy conversion (such as, solar to electricity, chemical and thermal energy) and photonics devices (such as photodetector, SERS, imaging). This review aims to offer a comprehensive and up-to-date exploration of nanostructured b-Si and its applications in solar conversion and photonic applications (Fig. 1). First, it will discuss the various fabrication techniques and unique properties of b-Si, which helps to understand the underlying mechanisms responsible for its optical characteristics. Next, the recent advancements in integrating b-Si into solar energy conversion and high-performance photonics devices will be elaborated across four key areas: (i) solar cells, (ii) photoelectrochemical solar energy conversion, (iii) solar thermal energy conversion, and (iv) photonics-related applications, including photodetectors, SERS, and imaging applications. Finally, the challenges and future perspectives in the continued development and optimization of b-Si for practical applications will be explored.

B-Si exhibits unique properties that make it highly suitable for solar energy conversion and photonic applications, distinguishing it from bulk silicon. In solar cells with thick wafers, optical management focuses on minimizing front-surface reflection by applying anti-reflection (AR) coatings and creating textured surfaces. The textured surface not only reduces reflection but also acts to trap photons, enhancing light capture and absorption, particularly at longer wavelengths.20

As is shown in Fig. 2(a), b-Si significantly reduces reflectance through various mechanisms, which are influenced by the dimensions and geometry of its nanoscale surface features. First, multiple light interactions with the textured surface decrease reflection. Second, when the features are considerably larger compared with the wavelength of sunlight, surface scattering extends the light path, increasing absorption. Third, extremely small texture features create a gradual refractive index transition between air and silicon, facilitated by the nanostructured surfaces, as illustrated in Fig. 2(b).48 In this graded-index AR effect, the porous layer of b-Si acts as a medium with a reduced refractive index. With uniform porosity, the b-Si layer presents a refractive index that is intermediate between that of air and bulk silicon, leading to a step change at the interfaces as illustrated in Fig. 2(b-ii). A porosity gradient in Fig. 2(b-iii) further smoothens the refractive index transition, reducing reflectance even more effectively. The greatest reduction occurs when the b-Si layer thickness increases and the nanostructure size decreases [Fig. 2(b-iv)]. As illustrated in Figs. 2(c) and 2(d), the nano-textured surface comprises sub-wavelength structures with two rough interfaces: one at the transition between air and the nanostructured (nT) layer (top), and the other between the nT and bulk silicon (bottom). These subwavelength structures, combined with boundary roughness, enable the nT layer to act as a transitional medium where the refractive index smoothly changes between air and silicon. As incident light is reflected at various depths within the nT layer, destructive interference reduces reflection losses across a broad wavelength range.50 Moreover, the bandgap of b-Si can be modified through doping with specific impurities, which may make it optically ideal for advanced photonic devices. Figure 2(e) illustrates the development of enhanced b-Si absorption across broad spectral ranges (from visible-NIR26^,52^,53 to MIR regions51^,54^,55) through optimizing the etching and doping parameters. Remarkably, by combining heavy phosphorus doping with a high-aspect-ratio nanostructure, b-Si can function as a broadband perfect absorber with exceptional spectral coverage extending to $25\mu \mathrm{m}$.

In addition, numerical modeling has emerged as a critical tool for both understanding the fundamental properties of b-Si and optimizing its nanostructures for photonic applications.30^,56^–58 State-of-the-art computational approaches, including finite-difference time-domain (FDTD) methods and rigorous coupled-wave analysis (RCWA), have quantitatively demonstrated that the unique nanostructured surface of b-Si can attribute to exceptional broadband antireflection capabilities. Zhu et al. systematically investigated the impact of variations in height ($h$) and bottom radius ($r$) on the absorptance of b-Si microstructures, demonstrating that $h$ is directly proportional to absorptance, whereas $r$ shows an inverse relationship.59 By simulating periodic pyramidal-like structures, Sai et al. further demonstrated that a sub-microstructured surface with a period exceeding $\sim 500\mathrm{nm}$ can simultaneously achieve minimal reflection and efficient light trapping.60 These modeling studies have proven instrumental in unraveling the intricate structure-property relationships between various nanostructure morphologies (e.g., nanowires, nanocones, and random textures) and optical performance, thereby facilitating the rational design of optimized light-trapping architectures. Most significantly, advanced multiphysics simulations integrating optical and electrical phenomena have elucidated crucial trade-offs between light absorption enhancement and surface recombination losses, providing critical guidance for developing optimized passivation strategies.

Furthermore, due to its relatively low mechanical strength, b-Si facilitates the separation of ultra-thin wafers from silicon ingots. The large surface area and high activity of b-Si enhance its ability to act as a getter, effectively reducing impurities within the wafer.20

Over the past several decades, a variety of manufacturing techniques for b-Si have been developed and tested, including electrochemical etching, stain etching, MACE, RIE, and laser processing (Fig. 3). Each of these etching methods produces unique micro- and nanostructures, resulting in distinct optical and electrical properties.

The electrochemical etching process utilizes silicon as the anode in a solution mainly containing hydrofluoric acid (HF). By fine-tuning parameters such as current density, HF concentration, etching duration, and illumination, this method enables precise control over the morphology of b-Si. Originally introduced by Memming and Schwandt,61 this technique gained prominence following the discovery of the luminescence phenomenon in porous silicon, signifying the advent of a novel silicon-based nanomaterial. As illustrated in Fig. 4(a), a silicon wafer (either n-type or p-type) is attached to the anode of an electrochemical cell and immersed in a mixture of HF, water, and ethanol.20 HF is essential for dissolving the silicon oxides produced during the etching, forming complexes that are soluble in water. Meanwhile, water participates in the oxidation reactions and helps regulate HF concentration to control reaction rates. Ethanol reduces silicon’s surface tension, improving wettability, facilitating the release of hydrogen (${\mathrm{H}}_2$) gas, and enhancing HF penetration into silicon pores.

During the process, charge transfer occurs between the silicon and the electrolyte until equilibrium is reached, aligning the silicon’s Fermi level (${\mathit{E}}_{\mathrm{F}}$) with the electrolyte’s electrochemical potential (${\mathit{E}}_{\mathrm{R}}$). For n-type silicon (n-Si), where ${\mathit{E}}_{\mathrm{F}}>{\mathit{E}}_{\mathrm{R}}$, electrons transfer from the silicon to the electrolyte, which creates a space-charge region (SCR) beneath the surface and induces an upward bending of the silicon energy bands. Conversely, in p-type silicon (p-Si), where ${\mathit{E}}_{\mathrm{F}}<{\mathit{E}}_{\mathrm{R}}$, electrons flow from the electrolyte to the silicon, resulting in the energy bands to bend downward [Fig. 4(b)]. The distinct behaviors of n- and p-Si are evident in their current density–voltage (J–V) characteristics. As illustrated in Fig. 4(c), the p-Si/HF junction is forward-biased for $V>0$, whereas the n-Si/HF junction is reverse-biased. Notably, the J–V curves of n-Si/HF resemble those of p-Si when exposed to light, indicating that silicon’s electronic properties minimally impact its electrochemical behavior. Both p-type and n-type silicon share similar dissolution mechanisms and chemistry during etching, and the peak current density (${\mathit{J}}_{\mathrm{PS}}$) signifies the transition between the two states.62^,64 Experimental studies reveal that ${\mathit{J}}_{\mathrm{PS}}$ follows an Arrhenius relationship with temperature and depends exponentially on the HF concentration. This behavior is captured by the semi-empirical equation: $$J_{\mathrm{PS}}=C_{\mathrm{PS}}·C_{\mathrm{HF}}^{\xi }·\exp (-\frac{E_{\text{a}}}{\mathrm{KT}}),$$(1)where ${\mathit{C}}_{\mathrm{PS}}$ ($\mathrm{A}{\mathrm{m}}^{-2}$) is an empirical constant, ${\mathit{C}}_{\mathrm{HF}}$ (%) is the volumetric HF fraction, $\xi$ is an empirical exponent, ${\mathit{E}}_{\mathrm{a}}$ is the activation energy, $K=1.38\times {10}^{-23}\mathrm{J}{\mathrm{K}}^{-1}$ is the Boltzmann constant, and $T$ is the thermodynamic temperature. This equation is critical for controlling electrochemical etching to form well-ordered silicon micro- and nanostructures. Simulated etching kinetics, corresponding morphologies, and SEM images of porous silicon are shown in Figs. 4(d) and 4(e).62^,63

Using photolithographically defined masks, selective polishing of regions can be achieved through electrochemical etching, enabling the fabrication of three-dimensional micro- and nanostructures.65^–67 At lower current levels, factors such as current strength, HF concentration, etching time, and illumination influence the etching process. Increasing the current or reducing the HF concentration leads to larger pore diameters, while extended etching times result in deeper pores. The type and concentration of doping in the silicon wafer also impact its morphology: in p-type silicon, higher doping levels lead to an increase in pore diameter, with values spanning from 1 to 100 nm; conversely, in n-type silicon, higher doping concentrations result in a reduction of pore diameter, ranging from $10\mu \mathrm{m}$ to 10 nm.20^,68 These dependencies provide precise control for the fabrication of silicon nanostructures to meet various application needs. This method is unsuitable for large-scale production and causes fluoride wastewater in the process.

Stain etching usually utilizes a mixture of HF and ${\mathrm{HNO}}_3$ to chemically etch silicon, creating porous silicon structures [Fig. 5(a)]. This method is typically considered a localized electrochemical etching process, similar to electrochemical HF etching.20 To control the etching process, ${\mathrm{H}}_2\mathrm{O}$ or acetic acid is commonly used to reduce the concentrations of HF and ${\mathrm{HNO}}_3$, with acetic acid additionally acting as a wetting agent. The etching process follows a stepwise oxidation-dissolution mechanism: ${\mathrm{HNO}}_3$ enters the pores to oxidize the silicon, while HF eliminates the resulting silicon oxide. The porous layer becomes thicker as the etching duration increases, with Fig. 5(b) illustrating the reflectance spectra of porous silicon etched on a multicrystalline silicon (mc-Si) wafer for varying durations.70

In addition, the silicon surface morphology can be tailored by adjusting the HF-to-${\mathrm{HNO}}_3$ concentration ratio ($\mathit{C}=[\mathrm{HF}]/[{\mathrm{HNO}}_3]$). When [HF] exceeds $[{\mathrm{HNO}}_3]$, the etching rate is governed by the oxidation rate, resulting in porous silicon. Conversely, when $[{\mathrm{HNO}}_3]$ is much higher than [HF], the reaction is constrained by HF’s ability to remove silicon oxide. In this high-$[{\mathrm{HNO}}_3]$ regime, the etching process is typically employed for surface cleaning, damage removal, or polishing. By fine-tuning $C$, stain etching can texturize mc-Si, which is challenging for conventional alkaline etching due to the material’s diverse crystal orientations. Figures 5(c)–5(e) demonstrate how different etching parameters—such as silicon doping, solution pH, and etching time—yield varied nanostructures starting from the same geometry of pillars.71

It is worth noting that stain etching rapidly removes the silicon surface, creating a gradual transition from solid silicon to full porosity.72 However, this method has a limit on the maximum thickness of the porous layer20 and may also lead to fluoride wastewater and ${\mathrm{NO}}_x$ emissions. As the layer thickens, pore growth slows because the diffusion of reactants and products decreases in the deeper pores. The maximum thickness is reached when the pore growth rate matches the etching rate at the surface.

MACE is a widely employed technique for fabricating b-Si with high aspect ratios and tunable dimensions and morphologies.62^,73 In this approach, a metal catalyst layer is applied to a silicon substrate, guiding the etching process when exposed to HF and an oxidizing agent. The metal catalyst, commonly gold (Au),74^–77 silver (Ag),78^,79 copper (Cu),80^–82 or Ni,83^,84 facilitates the formation of micro- and nano-structured surfaces. Metals can be deposited using vacuum-based techniques, such as thermal evaporation, sputtering, or electron beam evaporation, or through solution-based methods such as electroless or electrodeposition. Vacuum deposition provides precise control over the morphology of the metal film, whereas electroless deposition provides a simpler approach, better suited for applications with less demanding surface requirements. Acting as an etching mask, the metal layer selectively protects parts of the silicon surface, allowing etching in exposed regions and controlling the direction and depth of the etching. Figure 6(a) illustrates the MACE process: (1) reduction of the oxidizer (e.g., ${\mathrm{H}}_2{\mathrm{O}}_2$) under the catalysis of noble metal particles; (2) hole injection into the silicon substrate, with the highest concentration beneath the metal particle; (3) hole migration to silicon sidewalls and surfaces; and (4) HF removes the silicon oxide. Equations (2) and (3) describe the chemical reactions involved, using gold nanoparticles as an example. $${\mathrm{H}}_2{\mathrm{O}}_2+2{\text{H}}^{+}\stackrel{\mathrm{Au}}{\to }2{\mathrm{H}}_2\mathrm{O}+2{\mathrm{h}}^{+},$$(2)$$\mathrm{Si}+4{\mathrm{HF}}_2^{-}+2{\mathrm{h}}^{+}\to {\mathrm{SiF}}_6^{2-}+2\mathrm{HF}+{\mathrm{H}}_2(\text{gas}).$$(3)

In the MACE process, silicon oxide forms beneath metal particles and is etched away by HF, allowing the metal to penetrate into the silicon and develop a porous structure, in which the pore depth is controlled by the etching duration. After achieving the desired structure, the metal nanoparticles are dissolved with an etchant such as ${\mathrm{HNO}}_3$, followed by a cleaning procedure. Figure 6(b) illustrates the energy band diagrams of the Au/Si interface and the morphologies of n-type and p-type silicon after MACE. SEM images of the etched morphologies for various n- and p-Si electrodes are shown in Fig. 6(c).85 Toor et al. demonstrated nanoporous b-Si fabricated via cost-effective copper-assisted etching, achieving an average solar spectrum reflection as low as 3.1%.60 As shown in Fig. 6(d), Wang et al. demonstrated the formation of inverted pyramid structures on the silicon surface through a similar MACE process using Cu nanoparticles. Figure 6(e) illustrates the FDTD simulation results, showing the electric field intensity distribution in both inverted pyramid and upright pyramid structures.80 These findings highlight the potential of Cu nanoparticle-assisted etching for fabricating structures with distinct optical and electronic properties. In addition, by adopting masking materials, such as polystyrene (PS) spheres, anodic aluminum oxide (AAO) templates, and photoresists, this method can ensure consistent and uniform nanostructures on b-Si surfaces.23 Huang et al. fabricated b-Si nanowire arrays using PS sphere masks, achieving precise control over nanowire diameter, height, and spacing.87 They also employed AAO templates fixed on silicon via spin-coated polystyrene, producing silicon nanowires (SiNWs) with diameters below 10 nm.74

To conclude, MACE is a simple, fast, cost-effective, and adaptable technique for producing various nanostructures, requiring only basic equipment. Wet etching methods using Ag and Cu catalysts are especially promising for the affordable and large-scale production of b-Si in commercial silicon photovoltaic (PV) applications.88 The resulting surface morphology of b-Si depends on factors such as the type, size, shape, and coverage of metal nanoparticles, as well as the etching time, etchant composition, and temperature.88^–90 Over the past few decades, MACE has garnered widespread attention in research and remains a leading method for etching. However, metal contamination presents a major challenge for solar cell applications; thus, it is crucial to employ effective metal removal and cleaning procedures to guarantee optimal performance and reliability. In addition, the resulting Ag/heavy metal-contaminated fluoride wastewater may pose toxicity risks; therefore, further recovery of nanoparticles or the use of Fe/Cu catalysts could help mitigate heavy metal pollution.

RIE is a dry etching technique characterized by its high anisotropy and excellent selectivity, achieved through ion-induced chemical reactions [Fig. 7(a)].91 First introduced by Gittleman et al. at RCA Laboratories in 1979, RIE was used to create non-reflective b-Si surfaces for applications such as solar thermal energy conversion.94 In 1995, Jansen et al. demonstrated the application of RIE to produce grass-like b-Si surfaces.95 As illustrated in Fig. 7(b), this approach uses ${\mathrm{SF}}_6$ and ${\mathrm{O}}_2$ gases to produce ${\mathrm{F}}^{*}$ and ${\mathrm{O}}^{*}$ radicals. ${\mathrm{F}}^{*}$ etches silicon, forming volatile products such as ${\mathrm{SiF}}_x$, whereas ${\mathrm{SiF}}_4$ reacts with ${\mathrm{O}}^{*}$ to create a passivation layer of ${\mathrm{SiO}}_x{\mathrm{F}}_y$ on the silicon surface.91 Ion bombardment selectively removes this passivation layer on exposed areas, allowing further etching by ${\mathrm{F}}^{*}$. On the sidewalls of silicon columns, where ion bombardment is minimal, the passivation layer remains intact, shielding these areas from additional etching. This interplay between etching and passivation results in the self-masked formation of random, high-aspect-ratio silicon microstructures. Huang et al. reported a simple approach for creating aperiodic silicon nanotip (SiNT) arrays on 6-in. wafers with sub-wavelength structures. These structures significantly reduce light reflection across a wide spectral range, spanning from ultraviolet to terahertz. Figures 7(d)–7(f) display SEM images of the SiNTs and their corresponding absorption spectra.93

The morphology of b-Si fabricated through RIE can be controlled by adjusting factors such as gas composition, flow rate, system temperature, substrate bias, and RF power. A higher ${\mathrm{O}}_2$ flow rate promotes the formation of the passivation layer, whereas elevated temperatures accelerate its removal. By tuning the passivation layer coverage, the nanostructure density can be precisely controlled. In addition, increasing the substrate bias modifies ion energy, affecting ion bombardment and controlling the etching rate. However, RIE can impact the silicon surface, leading to damage that shortens the minority carrier lifetime near the surface [Fig. 7(c)].92 To address this, a post-etching step is commonly used to eliminate the uppermost 20 to 50 nm of the silicon layer. Alternative approaches include reducing RF power and substrate bias, minimizing ${\mathrm{O}}_2$ usage, shortening the etching time, or annealing the wafer at elevated temperatures (e.g., 400°C) to restore minority carrier lifetime. However, these measures may reduce etching efficiency, throughput, and increase operational costs.92 In addition, the process may produce greenhouse gases and toxic plasma byproducts, highlighting the critical need for future research into alternative gases.

Femtosecond laser processing offers a flexible method for texturing the surface of silicon substrates.96^–98 The setup for b-Si preparation using femtosecond laser pulses is shown in Fig. 8(a).99 Various settings can be adjusted to enhance the properties and structure of b-Si, including laser polarization, spot size, intensity, pulse number, scanning parameters, and environmental conditions. The laser intensity controls the material removal rate and silicon evaporation, whereas the number of pulses determines the interaction time, with longer exposure allowing deeper energy penetration into the silicon wafer. By optimizing these parameters, very high spikes can be produced. Huang’s studies highlight the importance of laser fluence in shaping surface morphology. As fluence increases, the size of micro- and nanostructures on the surface expands, leading to increased surface roughness. The frequency of these features follows a distinct sequence (e.g., $2f$, $f$, $f/2$, $f/4$, and $f/8$), with higher-order grating couplings producing finer features.102 Sharp conical spikes created in ${\mathrm{SF}}_6$ at 500 Torr are shown in Fig. 8(b), viewed both at a 45-deg angle and parallel to the cleaved surface.100 Vorobyev et al. demonstrated the production of equally spaced parallel microgrooves with nanostructured textures on silicon using direct femtosecond laser treatment.101 The SEM image in Fig. 8(c) shows the treated surface, and the 3D optical image in Fig. 8(d) illustrates microgrooves with an average depth of $\sim 55\mu \mathrm{m}$. The optical spectrum [Fig. 8(e)] reveals significantly reduced reflectance for the b-Si sample compared with polished silicon over wavelengths spanning 0.25 to $16\mu \mathrm{m}$.

In addition, prior studies demonstrate that laser treatment in water or oil can generate much finer structures, reaching sub-100 nm—up to 100 times smaller compared with those formed in gas environments or vacuum conditions.103^,104 In addition, combining laser irradiation with a periodic mask allows for the formation of well-defined, regularly spaced structures on silicon wafers. Unlike many techniques, laser processing is not limited by grain orientation, making it suitable for both crystalline silicon (c-Si) and mc-Si.20 B-Si produced through laser treatment, especially with sulfur doping, has demonstrated versatility across a wide range of applications.

Although higher laser repetition rates, power levels, or larger spot sizes can improve the processing rate, laser manufacturing remains relatively slow compared with alternative etching methods, particularly for industrial applications. Furthermore, the material damage caused by lasers may be substantial, requiring thorough defect-removal etching to achieve high material quality, which is essential for PV applications. In addition, silicon dust or nanoparticles are generated during the fabrication process, necessitating the implementation of a further filtration system.

Various techniques have been developed to create b-Si surfaces with micro- and nanostructures, enabling highly controlled surface morphologies.20 A summary and comparison of different b-Si fabrication methods are presented in Table 1. For instance, porosity and thickness can be tailored during electrochemical etching in HF by adjusting the current density and etching time. Wet etching methods, such as MACE, are particularly promising for large-scale industrial applications due to their simplicity and high etching speed. The RIE process can produce silicon nanoneedles, enabling smooth refractive index transitions and achieving reflectance below 1% across a broad wavelength range and different polarizations. Thanks to the isotropic characteristics of fluorine etching and the physical removal of the passivation layer, this method is compatible with mc-Si, microcrystalline silicon ($\mu \mathrm{c}\text{-}\mathrm{Si}$), and amorphous silicon (a-Si).19

Due to its micro- and nanoscale surface structures that promote broad-spectrum light absorption, b-Si devices are anticipated to deliver improved solar energy conversion efficiency and perform exceptionally well in a range of photonic applications. Researchers and engineers worldwide continue to advance b-Si technologies, aiming to enhance its performance and economic viability.

Becquerel’s observation of the PV effect in 1839 inspired the development of solar cells capable of converting sunlight into electrical energy. 118 Nowadays, silicon wafer-based solar cells lead the PV market, accounting for $\sim 95\%$ of global production and serving as a cornerstone of renewable energy technologies. Over the past four decades, the costs of solar PV systems have significantly decreased due to lower manufacturing expenses and enhanced efficiencies. As silicon wafers account for a substantial portion of production expenses, the quality of the material plays a key role in determining device performance.119 At present, silicon solar cells are divided into three primary types according to their underlying device technology: (i) homojunction solar cells,120 (ii) passivating contacts, and (iii) silicon-based tandem cells [Fig. 9(a)].121

Among these, homojunction c-Si solar cells are the most prevalent. Compared with aluminum back surface field (Al-BSF) cells shown in Fig. 9(b), passivated emitter and rear cell (PERC) technology incorporates a layer of dielectric material to isolate the silicon backside from the aluminum contacts [Fig. 9(c)], thus minimizing rear surface recombination.122 Further advancements in silicon solar cells leveraging advanced passivating contact technology,32^,125 such as metal–insulator–semiconductor (MIS) contacts, tunnel oxide passivated contact (TOPCon), dopant free asymmetric heterocontact (DASH), and silicon heterojunctions (SHJs), are expected to push efficiencies to 25% or higher. The carrier transport process in solar cells with passivating contacts is illustrated in Fig. 9(d).123 In contrast to single-junction solar cells, which are constrained by their optical bandgap, tandem solar cells [Fig. 9(e)] utilize materials with varying bandgaps to harness a broader range of sunlight. The upper cell, characterized by a large bandgap, targets high-energy photons, whereas the lower cell, featuring a smaller bandgap, effectively collects lower-energy photons, including those in the NIR range.124 The applications of b-Si in solar cells will be discussed in relation to these three categories.

To enhance efficiency and reduce costs, b-Si eliminates the need for AR coatings while achieving comparable light-trapping, making it a cost-effective option to achieve $\sim 20\%$ cost reduction.126 Even though the nanostructures on the silicon surface significantly reduce reflectivity and improve light absorption, challenges arise from increased photo-carrier recombination. Oh and colleagues revealed that the high surface area often assumed to cause surface recombination is not the primary limiting factor as previously assumed.127 Instead, the primary limitation arises from excessive doping-induced Auger recombination linked to the diffusion of the p–n junction into the nanostructures, which hinders carrier collection and efficiency in the majority of nanostructured silicon solar cells [Fig. 10(a)]. The effective carrier lifetime ($t_{\mathrm{eff}}$) of nanostructured silicon featuring passivated $n^{+}$–$p$ junctions with four different surface areas is shown in Fig. 10(b). By mitigating the recombination through light and shallow doping, combined with minimizing surface area via tetramethylammonium hydroxide (TMAH) etching, b-Si solar cells achieved a validated efficiency of 18.2% under standard AM 1.5G simulated sunlight conditions [Fig. 10(c)]. For large-area b-Si cells, the structure resembles traditional solar cell designs but integrates a b-Si layer on the topmost surface.30 These cells generally consist of a base made from p- or n-type silicon, with heavily doped regions at the front and back, and are equipped with metal contacts on both sides. As shown in Fig. 10(d), Davidsen et al. developed a p-type solar cell featuring a laser-patterned selective emitter design, using copper and nickel for the front contacts and aluminum for the rear contacts.128 The emitter layer of b-Si was strongly doped with phosphorus and coated with a ${\mathrm{SiN}}_x$ passivation layer, achieving a conversion efficiency of 18.1%. By combining MACE with a post-etching process, inverted pyramidal nanostructure-based mc-Si solar cells were fabricated. Shen et al. and Su et al. reported this approach in their research.133^–135 By proposing nanostructure rebuilding (NSR) solution treatment, the obtained solar cell based on an inverted pyramid structure showed a high conversion efficiency of 18.62% in large size ($156\mathrm{mm}\times 156\mathrm{mm}$), which was 0.45% higher than the traditional acid textured devices.133

With the advantages of reduced rear surface recombination and enhanced rear surface reflectivity, PERC solar cells based on b-Si have been further explored in some works.136^–139 Huang et al. conducted a simulation-based analysis of PERC silicon solar cells with nano-inverted-pyramid (NIP) textures at both the cell and module levels.140 Their findings revealed that NIP-PERC solar cells could achieve an efficiency of 22.1%, representing a 1.4% relative gain over conventional micro-pyramid PERCs. Simulation results further indicated that 60-cell modules incorporating NIP-PERC technology could deliver a peak power output of 310 W—an 8 W increase compared with modules using standard textured PERCs. This improvement is attributed to better reflectivity performance across both short and long wavelength ranges.

Using a dielectric passivation layer is a widely recognized approach to minimize surface recombination rates in solar cells. This reduction occurs via two primary mechanisms: chemical passivation, which decreases the density of interfacial states, and field-effect passivation, where an internal electric field suppresses minority-carrier density near the interface. For b-Si solar cells, the main obstacle lies in reducing recombination losses resulting from the larger surface area and its associated defects. To address these issues, Wang et al. employed thermal atomic layer deposition (ALD) to fabricate passivation layers, such as conformal ${\mathrm{Al}}_2{\mathrm{O}}_3$ or ${\mathrm{Al}}_2{\mathrm{O}}_3/{\mathrm{TiO}}_2$ dual-layer stacks, effectively reducing surface recombination and total reflectance [Figs. 10(e) and 10(f)].129^,130 By employing these passivation strategies in an ${\text{n}}^{+}$-emitter/p-base configuration, nanotextured black silicon (NBSi) solar cells reached a maximum efficiency of 18.5%. Savin et al., in 2015, proposed an interdigitated back contact (IBC) architecture to fully passivate the front surface, reducing front surface recombination rates.131 Their approach demonstrated that a conformal ${\mathrm{Al}}_2{\mathrm{O}}_3$ film (20 nm) provided excellent chemical and electrical passivation [Fig. 10(g)], achieving a record efficiency of 22.1% in b-Si solar cells. The b-Si-based device demonstrates significantly reduced reflectance across both short (300 to 500 nm) and long ($>700\mathrm{nm}$) wavelength ranges compared with reference cells featuring conventional random micron pyramids with a thicker ${\mathrm{Al}}_2{\mathrm{O}}_3$ AR coating (90 nm). Due to the improved angular acceptance of b-Si shown in Fig. 10(h), the device based on b-Si can enhance daily energy production by 3%.

Currently, conventional micron-pyramid textures are approaching their performance limits in minimizing optical reflection losses and boosting power conversion efficiency (PCE) for TOPCon solar cells. Addressing this challenge, Li et al. developed an innovative micron-pyramid/nanopore pyramid (NPP) silicon architecture, achieving remarkable average efficiencies exceeding 23%.132^,141^,142Figure 10(i) shows the cross-sectional device architecture of TOPCon solar cells based on NPP structure. Compared with conventional micron-pyramid textures, the TOPCon solar cell based on NPP structure exhibited higher average PCE and exceptional wide-angle absorption capability [from 0 deg to 70 deg, Fig. 10(j)]—critical advantages for real-world outdoor photovoltaic applications.

Recently, tandem solar cells combining perovskite and silicon have emerged as a promising innovation, offering the potential to exceed the efficiency limits of traditional silicon-based solar cells. This advancement is achieved through the strategic combination of perovskite top cells, which efficiently capture high-energy photons, and crystalline silicon bottom cells, which effectively harvest lower-energy photons. This tandem configuration enables a broader spectral response and significantly enhances overall energy conversion efficiency. In 2022, Ying et al. reported the first monolithic perovskite/silicon tandem solar cell that incorporates industrially applicable front-side nanostructured b-Si integrated with TOPCon technology [Fig. 10(k)].33 The device architecture of the nanotextured perovskite/silicon tandem solar cell is shown in Fig. 10(l). As the external quantum efficiency (EQE) and spectral response analysis shown in Fig. 10(m), the nanotexture significantly suppresses the reflectance and the nanotextured tandem achieves near-constant low reflectance due to the gradient of refractive index effect, resulting in the enhanced EQE responses for both perovskite and silicon. This design achieves effective surface passivation without compromising broadband light trapping. In addition, the reconstructed nanotexture significantly enhances perovskite wetting. As shown in Figs. 10(n) and 10(o), the reconstructed nanotexture acts as a nanoconfining scaffold, guiding vertical perovskite growth, whereas planar surfaces result in randomly oriented structures. These advances lead to a notable increase in both the current density and fill factor (FF) of tandems, ultimately achieving a certified PCE of 28.2%.

Furthermore, they conducted in-depth studies on perovskite/TOPCon tandem solar cells based on b-Si. By integrating a poly-$\mathrm{Si}({\mathrm{p}}^{+})/\text{poly}\text{-}\mathrm{Si}({\text{n}}^{+})$ tunneling recombination layer (poly-Si TRL), the device achieved a remarkable efficiency of 29.2%, significantly surpassing conventional indium zinc oxide (IZO)-based tandem solar cells (28%).143 By employing a facile surface reconstruction strategy to mitigate residual strain in perovskite films—post-treatment with an isopropanol solution containing $N,N$-dimethylformamide (DMF) and $n$-butylammonium iodide (BAI), an efficient and stable perovskite/b-Si tandem solar cell with a certified stabilized PCE of 29.0% was achieved.144 In addition, they developed an oblique angle evaporation method to deposit a conformal, ultra-thin ${\mathrm{SiO}}_x$ passivation layer, which can alleviate the recombination losses at the defective $\text{perovskite}/{\mathrm{C}}_{60}$ interface without damaging the underlying perovskite.145 When applied to 4-T and 2-T perovskite/b-Si tandem solar cells, this strategy achieved stable PCEs of 29.2% and 30.2%, respectively—far exceeding those of perovskite/TOPCon silicon tandem based on conventional textured silicon (e.g., 25.1%,146 25.84%147).

It is worth mentioning that the b-Si exhibits no crystallinity-related limitations during the fabrication,148 making it suitable for mc-Si with anisotropic texturing that is ineffective in alkaline solutions, as well as for diamond-wire cut mc-Si that is challenging to texture with conventional acidic solutions.149^,150 Moreover, b-Si demonstrates superior metal impurity-gettering efficiency compared with planar surfaces in PV applications.151 These advantages further enhance its potential for high-performance solar cells.

In PEC solar energy conversion, enhancing multiple properties through surface and interface engineering approaches is essential, as depicted schematically in Fig. 11. By implementing appropriate interface enhancements, such as optimized junction designs and protective/passivation coatings, it can attain strong mechanical durability and efficient charge transport. In addition, cocatalysts applied to the surface can significantly improve the reaction kinetics at the interface.10

A key factor limiting the performance of planar and b-Si photocathodes is the limited ${\mathit{V}}_{\mathrm{oc}}$ of the $\text{p}\text{-}\mathrm{Si}/{\mathrm{H}}_2\mathrm{O}$ interface, which constrains energy conversion efficiency. To overcome this limitation, the ${\mathrm{n}}^{+}/\mathrm{p}$ junction at the surface of b-Si, acting as a typical buried structure, is frequently employed to generate an internal electric field that facilitates the outward movement of photogenerated carriers. In addition, a buried metallurgical ${\mathrm{n}}^{+}/\mathrm{p}$ junction has also been integrated into Si wires [Fig. 12(b)].153 The buried junction separates the band bending and photovoltage properties of the electrode from the energy dynamics at the semiconductor/liquid interface. This design achieves higher ${\mathit{V}}_{\mathrm{oc}}$ values as a result of the enhanced band bending at the ${\text{n}}^{+}/\text{p}$ interface compared with that at the p-Si/solution interface. These findings align with previous studies on $\text{p}\text{-}\mathrm{Si}/{\mathrm{H}}_2\mathrm{O}$ systems, which demonstrate that the valence-band edge of Si is insufficiently positive relative to $\mathit{E}({\mathrm{H}}^{+}/{\mathrm{H}}_2)$ to achieve a high photovoltage.153

Besides buried p–n junctions, metal interfacial layers, including Ag 158 and Pt159 on the surface of b-Si, are effective in enhancing the separation and movement of photoexcited electrons. Furthermore, Lewis et al. developed an ${\mathrm{np}}^{+}$-Si microwire array featuring a coating of indium tin oxide (ITO) and $\mathrm{n}\text{-}{\mathrm{WO}}_3$ [Fig. 12(a)].152 The ITO layer provides an ohmic contact between the Si microwire PV unit and the ${\mathrm{WO}}_3$ layer, enabling the demonstration of a self-sustained, integrated solar-powered water-splitting system. This innovative design leverages the unique properties of the microwire array structure to achieve efficient solar water splitting.

Many metal oxides are known for their optical transparency and chemical stability, making them effective for protecting Si-based photocathodes. Among these, ${\mathrm{TiO}}_2$ is particularly notable for its durability in a wide range of pH environments (acidic to alkaline) and its high transmittance in the visible spectrum, owing to its wide bandgap. In 2017, Yu et al. demonstrated the deposition of a conformal, ultrathin amorphous ${\mathrm{TiO}}_2$ film (8 nm) on b-Si using low-temperature ALD.154 This coating extended the functional lifespan of b-Si photocathodes from under 30 min to 4 h. Figure 12(c) illustrates the structural layout of the electrode along with the oxygen evolution mechanism. As illustrated in Fig. 12(d), b-Si demonstrated much lower reflectance ($\sim 10\%$) than planar Si (over 30%) across the UV-vis-NIR spectra. The combined reflectance and scattering data for b-Si and $\mathrm{b}\text{-}\mathrm{Si}/{\mathrm{TiO}}_2$ revealed minimal variation, confirming that the ${\mathrm{TiO}}_2$ layer did not adversely affect light absorption. By comparison, as depicted in Fig. 12(e), $\mathrm{b}\text{-}\mathrm{Si}/\mathrm{Co}(\mathrm{OH}{)}_2$ lacking a ${\mathrm{TiO}}_2$ layer showed a rapid decline in photocurrent density (${\mathit{J}}_{\mathrm{ph}}$) within 30 min, with deterioration becoming more pronounced after 2.5 h, resulting in a 77.3% loss of ${\mathit{J}}_{\mathrm{ph}}$ after 3 h, likely due to structural damage. Beyond ${\mathrm{TiO}}_2$, other materials such as ${\mathrm{Cu}}_2\mathrm{O}$,160^,161${\mathrm{Fe}}_2{\mathrm{O}}_3$,162^–165${\mathrm{SnO}}_2$,163${\mathrm{NiRuO}}_x$,166${\mathrm{MoS}}_2$, ${\mathrm{C}}_3{\mathrm{N}}_4$,167 carbon,168 and PEDOT155 have also been explored as protection layers. Figures 12(f)–12(h) show SEM and TEM images of a typical SiNW modified with PEDOT and Ag nanoparticles (AgNPs). The incident photon-to-current efficiency (IPCE) spectra in Fig. 12(i) demonstrate that modified SiNW arrays exhibit significantly enhanced IPCE compared to H-SiNW arrays, with the AgNPs/PEDOT/SiNW arrays reaching a peak efficiency of 42% at 510 nm.

Some previous studies have shown that decorating Si photocathodes with Pt can enhance their photoelectrochemical (PEC) properties.112^,169 However, the scarcity and high cost of Pt pose challenges for its use in widespread applications. As an alternative, earth-abundant transition metals (e.g., Mo, Co, and Ni) and their compounds (e.g., sulfides, phosphides, selenides) have been extensively explored due to their comparable catalytic activity.10 For example, the PEC performance of ${\mathrm{n}}^{+}\mathrm{p}$ b-Si photocathodes can be significantly improved with the addition of a cobalt (Co) layer, leading to a 560-mV positive shift in onset potential to 0.33 V (versus RHE) and an increase in the saturation photocurrent density from 22.7 to $32.6\mathrm{mA}/\mathrm{c}{\mathrm{m}}^2$.170 Similarly, M-B ($\text{M}=\mathrm{Ni}$, Co) compounds loaded onto SiNW arrays via electroless plating show enhanced PEC performance with a positively shifted onset potential [Fig. 12(j)].156 The optimized configurations of SiNWs/Co-B and SiNWs/Ni-B exhibit half-cell photo power conversion efficiencies of 2.53% and 2.45%, respectively, comparable to the 2.46% efficiency of SiNWs/Pt.

Among transition metal compounds, Mo-based sulfides are particularly effective cocatalysts for hydrogen generation on black Si, such as ${\mathrm{MoS}}_x$ quantum dot,171$[{\mathrm{Mo}}_3{\mathrm{S}}_{13}]$-2(MS) clusters,172${\mathrm{MoS}}_3$,173^,174 and cubane-like clusters (${\mathrm{Mo}}_3{\mathrm{S}}_4$).175 In 2022, Kim et al. fabricated a uniform amorphous ${\mathrm{MoS}}_x$ ($\mathrm{a}\text{-}{\mathrm{MoS}}_x$) coating on nanostructured b-Si using ALD [Fig. 12(k)].157 The $\mathrm{a}\text{-}{\mathrm{MoS}}_x$ layer, characterized by a reduced work function of 4.0 eV compared with the 4.5 eV of crystalline $2\mathrm{H}\text{-}{\mathrm{MoS}}_2$, improves the electronic alignment at the p-Si interface, promoting efficient charge transfer. It achieves an impressive overpotential of $\sim 0.2\mathrm{V}$ at $10\mathrm{mA}/\mathrm{c}{\mathrm{m}}^2$ [Fig. 12(l)] and provides robust protection against corrosion, enabling PEC operation for over 50 h while retaining 87% or more of its performance. In addition, transition metal bimetallic compounds have also been explored as cocatalysts on b-Si. For instance, heterometal-doped amorphous ${\mathrm{MMoS}}_x$ ($\text{M}=\mathrm{Fe}$, Co, Ni) materials improve water photoelectrolysis by exposing more S-terminated edges.176 Similarly, amorphous ${\mathrm{NiCoSe}}_x$ has emerged as a competitive HER electrocatalyst due to its high activity and optical transparency. The $\text{p}\text{-}\mathrm{Si}/{\mathrm{NiCoSe}}_x$ photocathode exhibits improved PEC activity, achieving a photocurrent density of $-37.5\mathrm{mA}/\mathrm{c}{\mathrm{m}}^2$ at 0 V (versus RHE) under simulated AM 1.5G illumination, attributed to the combined effect of the transparent properties of ${\mathrm{NiCoSe}}_x$ and the superior optical absorption capability of the Si nanopillar array.

B-Si based photoelectrodes have also been explored for nitrogen reduction applications. Ammonia (${\mathrm{NH}}_3$), a chemical produced on a massive scale worldwide, is essential for manufacturing fertilizers and shows promise as an eco-friendly energy carrier and energy storage intermediate. Ali et al. developed a solar-driven, plasmon-enhanced PEC cell using nanostructured b-Si to convert atmospheric ${\mathrm{N}}_2$ into ammonia, achieving a production rate of $13.3\mathrm{mg}{\mathrm{m}}^{-2}{\mathrm{h}}^{-1}$ under two suns illumination.177 As shown in Fig. 13(a), the nitrogen photo-reduction cell utilizes nitrogen gas bubbled over the electrode surface, illuminated with artificial solar light. The 24-h yield data [Fig. 13(b)] reveal that bare b-Si produces a low amount of ammonia, which increases nearly four-fold after coating with gold nanoparticles (AuNPs). Adding a chromium (Cr) layer as a hole sink and an anode further prevents silicon oxidation and enhances charge separation. This configuration, the AuNP/b-Si/Cr cell, achieved an impressive yield of $320\mathrm{mg}{\mathrm{m}}^{-2}$ over 24 h. Quantum yield analysis [Fig. 13(c)] confirmed nitrogen reduction across the visible spectrum, with efficiency dropping near silicon’s bandgap. A quantum efficiency peak around 500 nm was attributed to surface plasmon excitation in the AuNPs, providing an additional photo-excitation mechanism. In addition, Qingji Xie’s group reported a Ni-doped ${\mathrm{MoS}}_2/\mathrm{Si}$ nanowire ($\mathrm{Ni}\text{-}{\mathrm{MoS}}_2/\mathrm{Si}\mathrm{NW}$) photocathode exhibiting excellent efficiency and specificity in nitrogen reduction reactions (NRR).179 In a 0.1 mol/L ${\mathrm{Na}}_2{\mathrm{SO}}_4$ solution saturated with ${\mathrm{N}}_2$ and containing PCP, this photocathode achieved an ${\mathrm{NH}}_3$ yield of $12.0\mathrm{mmol}{\mathrm{h}}^{-1}{\mathrm{m}}^{-2}$ at a low bias of 0.25 V versus RHE, along with an apparent Faradaic efficiency (AFE) exceeding 100%. This high efficiency was attributed to photocurrent-independent photocatalysis during the PEC process. The study introduced a novel classification of PEC electrons: (i) dark-current electrons, (ii) photocurrent electrons, and (iii) photocurrent-free photocatalysis electrons. This study presents an affordable and efficient strategy for developing NRR nanocatalysts and emphasizes the advantage of using solutions with enhanced ${\mathrm{N}}_2$ solubility to enable effective PEC-driven ammonia production, with potential applications in other gas-phase renewable energy conversions.

In 2024, Peidong Yang’s group presented a silicon nanowire-based device for red light applications, enabling facilitating ${\mathrm{CO}}_2$ conversion to multicarbon compounds at the biophotocathode and glycerol oxidation at the photoanode.178 As depicted in Fig. 13(d), the system combines a ${\mathrm{CO}}_2$-utilizing microorganism, Sporomusa ovata (S. ovata), with a SiNW cathode and a SiNW anode loaded with Pt–Au. Here, S. ovata functions as the biological catalyst for ${\mathrm{CO}}_2$ reduction, whereas the co-deposited Pt–Au layer serves as the catalyst for the oxidation of glycerol. The energy diagram [Fig. 13(e)] illustrates a photoactive diode operating under illumination with red light, producing a voltage of 0.4 V at the cathode and 0.45 V at the anode. This configuration integrates the reduction of ${\mathrm{CO}}_2$ (${\mathrm{CO}}_2\mathrm{RR}$) with glycerol oxidation (GOR) without requiring an external bias. Achieving a Faradaic efficiency of $\sim 80\%$ for both the reduction and oxidation products under dim red light, this self-sustaining photochemical device [Fig. 13(f)] offers a novel photosynthetic pathway to reduce ${\mathrm{CO}}_2$ emissions while producing valuable chemicals efficiently.

In recent years, considerable attention has been given to optimizing solar thermal energy applications through b-Si. Due to its semiconductor properties, silicon is restricted to absorbing solar wavelengths below 1110 nm (corresponding to a bandgap ${\mathit{E}}_{\text{g}}=1.12\mathrm{eV}$) for applications in PV or PEC. To overcome this limitation, researchers have been exploring innovative strategies such as surface plasmonic effect (SPR) of metal nanoparticles, doping/defect engineering, and advanced surface texturing, to enhance the NIR light absorption and photothermal energy capabilities of b-Si. The applications of solar photothermal energy based on b-Si are illustrated in Fig. 14 and will be discussed in detail below.

Photothermal catalysis involves the activation or enhancement of catalytic reactions through photothermal effects. In some cases, catalysts also serve as photothermal agents, converting sunlight into heat to raise the reaction temperature. This increase in temperature can activate the reaction or enhance reaction kinetics, similar to the mechanisms observed in thermocatalysis.180^–182 In 2016, Ozin’s group introduced a combined photothermal and photochemical strategy for producing solar fuels.183Figure 15(a) illustrates the light absorption of nanocrystalline ${\mathrm{In}}_2{\mathrm{O}}_{3-x}(\mathrm{OH}{)}_y$ (blue line) alongside the utilization of photons by the ${\mathrm{In}}_2{\mathrm{O}}_{3-x}(\mathrm{OH}{)}_y@\mathrm{SiNW}$ hybrid structures. The yellow and red areas illustrate two separate light-harvesting mechanisms: photons with energies exceeding the bandgap of ${\mathrm{In}}_2{\mathrm{O}}_{3-x}(\mathrm{OH}{)}_y$ (yellow) create electron-hole pairs in the nanocrystals, driving the solar-powered reverse water-gas shift process; by contrast, photons below the bandgap (red) are taken up by the SiNW support and transformed into heat through photothermal effects. The hybrid material exhibited improved utilization of sunlight compared to ${\mathrm{In}}_2{\mathrm{O}}_{3-x}(\mathrm{OH}{)}_y$ on its own, benefiting from heat generated by the SiNWs to drive the reaction. Photons with energies above the bandgap of ${\mathrm{In}}_2{\mathrm{O}}_{3-x}(\mathrm{OH}{)}_y$ contribute to the ${\mathrm{CO}}_2$ reduction process through photocatalysis, while sub-bandgap photons captured by the SiNWs supply thermal energy to enhance the reaction efficiency. Uniformly coating ${\mathrm{In}}_2{\mathrm{O}}_{3-x}(\mathrm{OH}{)}_y$ onto SiNW arrays reduced reflective losses, improved light harvesting, and facilitated heat transfer, collectively boosting the ${\mathrm{CO}}_2$ photocatalytic reduction rate compared to non-uniform coatings or unsupported ${\mathrm{In}}_2{\mathrm{O}}_{3-x}(\mathrm{OH}{)}_y$. Applying a uniform layer of ${\mathrm{In}}_2{\mathrm{O}}_{3-x}(\mathrm{OH}{)}_y$ to SiNW arrays minimized light reflection, enhanced absorption, and improved thermal transfer, resulting in a higher ${\mathrm{CO}}_2$ reduction efficiency compared with uneven coatings or standalone ${\mathrm{In}}_2{\mathrm{O}}_{3-x}(\mathrm{OH}{)}_y$.

Researchers also have explored photothermal catalysis utilizing innovative metal particles on b-Si. O’Brien et al. have investigated silicon nanowires supported with Ru nanoparticles (Ru@SiNW) for the Sabatier methanation process, ${\mathrm{CO}}_2+4{\mathrm{H}}_2\to {\mathrm{CH}}_4+2{\mathrm{H}}_2\mathrm{O}$.184 This study revealed that Ru nanoparticles combined with SiNWs could absorb as much as 85% of solar irradiance over a wide wavelength range. The high efficiency stems from the low reflectivity of the vertically arranged nanowires and the broad absorption properties of the Ru nanoparticles, driven by electronic transitions across the visble-NIR spectrum. Figure 15(b) highlights the catalytic activity of Ru/SiNW, Ru/glass, and Ru/Si catalysts under solar-simulated illumination at 150°C over 6 h. Among these, the Ru/SiNW catalyst showed superior performance, achieving a ${\mathrm{CO}}_2$ methanation rate of $0.51\mathrm{mmol}{\mathrm{g}}^{-1}{\mathrm{h}}^{-1}$ without light, which increased by 94% to $0.99\mathrm{mmol}{\mathrm{g}}^{-1}{\mathrm{h}}^{-1}$ with illumination, significantly outperforming both Ru/glass and Ru/Si catalysts. Figure 15(c) displays the temperature-dependent ${\mathrm{CO}}_2$ methanation rates for Ru/SiNW in the dark, with an inset activation energy of $54.5\mathrm{kJ}{\mathrm{mol}}^{-1}$. Batch reactions conducted under different light intensities without external heating indicated a Sabatier reaction rate [yellow line, Fig. 15(c)] with a calculated activation energy of $53.1\mathrm{kJ}{\mathrm{mol}}^{-1}$. Although reaction rates improved under simulated solar illumination, the activation energy remained consistent regardless of whether heating was provided by light or a resistive element. These results demonstrate the efficiency and versatility of Ru/SiNW catalysts for photothermal ${\mathrm{CO}}_2$ methanation.

PV cells produce electricity by capturing high-energy photons. In thermophotovoltaic (TPV) systems, this principle is employed using thermal emitters that absorb heat and reach temperatures high enough to emit energetic photons. These photons are subsequently harvested by PV cells, effectively converting thermal radiation into electrical energy. In recent years, photo-thermoelectric conversion applications based on b-Si have attracted considerable interest for their promising potential in energy conversion technologies.187^–191 Novel b-Si combining ordered micropores and disordered nanopores was reported by Kaiying Wang’s group in 2019.25 The device, composed of b-Si, a generator, and a cooling fin, is shown in Fig. 15(d). The multiscale pores play distinct roles: larger pores extend the light path and offer space for incorporating functional materials, whereas smaller pores improve the interaction between incoming solar radiation and the substrate. The schematic in Fig. 15(e) illustrates the AR mechanism, where light waves travel across surfaces with diverse structural characteristics. This etching technique enhances the light-harvesting efficiency of traditional nanostructured b-Si, resulting in improved absorption over a wide wavelength range. Furthermore, introducing Au nanoparticles enhances light absorption in the NIR region (1100 to 1700 nm). As demonstrated in Fig. 15(f), the order-disordered silicon structure with Au nanoparticles produces the highest output voltage of 140 mV, confirming the effectiveness of multiscale-pore structures for photothermal conversion. This study uncovers a strong connection between the three-dimensional configuration and the photon-trapping characteristics of b-Si, providing valuable guidance for substrate designs that optimize light-to-energy conversion.

In addition, it is worth mentioning that highly conductive b-Si, fabricated using an RIE technique, is capable of absorbing light across a broad wavelength range, extending well into the NIR region ($\sim 2500\mathrm{nm}$).185 The improved b-Si featuring surface texturing exhibits a specular reflection reflectance below 0.1%, with a total average reflectance (combining specular and diffuse components) of $\sim 1.1\%$. Moreover, the reflective properties and underlying mechanisms of surface-textured b-Si have been further explored. Utilizing b-Si materials significantly enhances the efficiency of converting solar energy into electricity, effectively capturing wavelengths beyond 1110 nm. Pengfei Cheng et al. have conducted photothermal and photo-thermoelectric experiments to demonstrate the superior light absorption capabilities of b-Si, attributed to its rich surface micro- and nanostructures.185 A diagram of the photothermal experiment setup is presented in Fig. 15(g). In this configuration, the sample is exposed to solar light from above, whereas an IR camera positioned diagonally monitors surface temperature changes in real time. The data reveal that b-Si heats up more quickly during the initial 5 min and achieves a maximum surface temperature of 53.8°C, surpassing that of bare silicon [Fig. 15(h)]. This superior light-trapping ability highlights the potential of b-Si for various photothermal applications, including imaging, seawater desalination, and other advanced technologies.

Water evaporation, a natural process occurring everywhere from oceans to the human skin, carries significant energy. For example, sunlight-driven evaporation on the ocean’s surface absorbs vast amounts of energy globally. Recently, innovative technologies such as triboelectric nanogenerators192^,193 and hydrovoltaic generators (HGs)194^–197 have been developed to generate electricity from water or ambient moisture. Notably, HGs driven by evaporation have attracted interest due to their capacity to leverage the natural evaporation process, making them ideal for off-grid power generation. Thus, integrating HGs with solar steam generation (SSG) systems presents an exciting opportunity, which enables simultaneous electricity generation and the production of clean water by capturing water molecules from evaporation, offering a sustainable solution to water and energy challenges. Baoquan Sun’s group designed a self-floating thermal insulating layer based on b-Si to be incorporated into the interfacial SSG to enhance the conversion efficiency from solar energy to vapor and reduce thermal energy dissipation, as shown in Figs. 15(i)–15(k).186 Under standard AM 1.5G irradiation, the SiNW-enabled HG/SSG system achieves a $V_{\text{on}}$ of 0.81 V, a short-circuit current of $108.0\mu \mathrm{A}{\mathrm{cm}}^{-2}$, and a water evaporation rate of $1.31\mathrm{kg}{\mathrm{m}}^{-2}{\mathrm{h}}^{-1}$. Even with external resistances applied, the device sustains optimized power output above $80.0\mu \mathrm{W}{\mathrm{cm}}^{-2}$, establishing a new benchmark for continuous output in HGs. Furthermore, a cutting-edge 6-in. wafer-scale SiNWs device achieved a water evaporation rate of $1.06\mathrm{kg}{\mathrm{m}}^{-2}{\mathrm{h}}^{-1}$, a current of 8.65 mA, and a power output of 6 mW during outdoor simulation experiments conducted over an 8-h period. This achievement represents the highest sustained electrical output ever documented for an individual HG system, paving the way for potential large-scale applications. This design approach opens new possibilities for water-activated device applications and positions itself as a promising solution for future energy systems in off-grid buildings, capable of generating both electricity and freshwater solely from sunlight and seawater.

An ideal photodetector is designed to efficiently capture and detect photons across a broad range of wavelengths, angles, and light intensities, ensuring consistent performance under diverse optical conditions. Currently, the most advanced photodetectors are typically fabricated by integrating silicon pixel arrays (flip-chip) with “bandgap-friendly” materials such as indium arsenide (InAs) and indium antimonide (InSb). However, this process is expensive compared with using silicon alone, and other traditional photodiodes often face optical losses, with charge carriers frequently lost to recombination. As discussed above, traditional silicon does not absorb infrared (IR) energies above 1110 nm, restricting its functionality primarily to the visible region, particularly in photodetection applications. Progress in silicon-based IR photodetectors is vital for enhancing silicon-integrated optoelectronics and communication systems.36 A promising strategy to achieve this involves using b-Si, which extends the absorption edge of silicon into IR wavelengths below its bandgap.198 This method facilitates the development of IR detectors seamlessly incorporated into silicon wafers for use in optical communication and sensing applications. Compared with III–V semiconductors, b-Si offers three significant advantages: (i) silicon is significantly more cost-effective and has been extensively studied, ensuring its suitability for large-scale manufacturing; (ii) as a sensing material, b-Si aligns well with existing integrated circuit (IC) fabrication methods, facilitating its straightforward integration into upcoming fully silicon-based optoelectronic devices; and (iii) in contrast to epitaxial growth, ultrafast laser doping allows selective area processing without requiring masking. The “cold” nature of ultrashort laser pulses preserves the original physical and chemical characteristics of untreated areas, ensuring that unaffected regions in integrated systems remain intact. Given these advantages, b-Si materials created via ultrafast laser doping represent a promising pathway for addressing the challenges of developing efficient and cost-effective IR photodetectors, paving the way for enhanced Si-based optoelectronic integration.199

The use of ion implantation for doping in b-Si presents a versatile approach for developing IR photodetectors with tunable concentration profiles. In 2022, Tarik Bourouina’s group demonstrated that surface-doped b-Si, using phosphorus as the dopant, achieved a remarkable 98% absorptivity in the 1 to $5\mu \mathrm{m}$ range.200 They further produced a broadband absorber (spanning over $20\mu \mathrm{m}$) by creating highly doped b-Si bulk via cryogenic plasma processing. This material exhibited 99.5% absorbance up to $8\mu \mathrm{m}$ and maintained over 90% absorbance at $20\mu \mathrm{m}$ [as detailed in Fig. 2(e)].51 Their findings, supported by numerical simulations, revealed that higher doping levels effectively enhanced absorption intensity, whereas moderate doping provided improved performance across a broader wavelength range. Thus, optimizing doping levels is essential to balance performance for specific applications. The duration of the etching process, which defines the conical structure depth, was found to be a key parameter affecting absorption efficiency. In addition, Song et al. employed the SPR effects of gold nanoparticles to design an absorber that operates across the 350 to 2500 nm spectrum without relying on lithographic methods.201 By combining RIE with Au nanoparticle deposition, they achieved a high-absorption b-Si surface (96.5%). This versatile absorber offers potential in photodetectors and solar energy collection.

In 2016, Hele Savin’s group introduced an innovative photodiode design [Fig. 16(a)].202 It incorporates silicon nanostructures for enhanced light absorption, uniform ALD alumina coatings to suppress surface recombination, and an induced junction to address recombination caused by dopants and irregular junction formations. This device achieves an EQE exceeding 96% across 250 to 950 nm wavelengths and maintains performance at incident angles up to 70 deg. Building on these advancements, Zhang et al. designed a self-powered Schottky photodiode based on nanocrystal indium tin oxide and b-Si (nc-ITO/b-Si).203 As illustrated in Fig. 16(b), the device incorporates nanostalagmite structures created on high-resistivity n-Si, with ITO nanocrystals serving as the rectifying layer and aluminum functioning as the rear ohmic contact. The images in Figs. 16(c) and 16(d) reveal the nc-ITO-coated nanostalagmites and b-Si photodiodes with varying ITO contact thicknesses compared with untreated planar Si. As depicted in Fig. 16(e), the built-in electric field from the nc-ITO contact, forms a “strong-field depletion region” near the surface and a “weak-field region” extending $\sim 100\mu \mathrm{m}$ into the silicon absorber. This single-contact b-Si photodiode achieves exceptional performance, with an average broadband EQE exceeding 98% from 570 to 925 nm and over 95% from 500 to 960 nm. It operates effectively across visible and NIR wavelengths (470 to 1050 nm) and can detect nanowatt-level illumination, demonstrating its capability as a highly sensitive heart rate sensor. With near-unity EQE, a self-powered configuration and a straightforward fabrication process make this photodiode an excellent candidate for compact, lightweight optical and electronic devices. In addition, its effectiveness in monitoring heart rates under low-light conditions highlights its potential for advanced, energy-efficient applications. These advances underscore the promise of cost-effective b-Si–based optoelectronic devices for diverse and enhanced functionalities.

Beyond visible-NIR applications, b-Si has also emerged as a promising solution for UV photodetection, where conventional technologies face two critical limitations: (i) high reflectance losses and (ii) low quantum efficiency and poor UV detection sensitivity due to surface recombination. Recent innovations have addressed these challenges by employing lightly doped b-Si to suppress dark current and enhance sensitivity, coupled with conformal ${\mathrm{Al}}_2{\mathrm{O}}_3$ passivation layers to minimize recombination losses through induced junction formation. In 2020, Garin et al. developed a b-Si photodiode featuring a self-induced junction from ${\mathrm{Al}}_2{\mathrm{O}}_3$, which surpassed the theoretical one-photon–one-electron barrier to achieve a certified EQE exceeding 130% in the UV spectrum without external amplification.206 Numerical simulations attributed this exceptional performance to impact ionization-driven multiple carrier generation within the nanostructures. Furthermore, Tsang et al. extended the investigations into vacuum ultraviolet (VUV) detection using b-Si photodiodes.207 Combining conformal alumina with an induced junction, the device shows a responsivity of $0.2\mathrm{A}/\mathrm{W}$ at 175 nm under room temperature in vacuum. Notably, the findings revealed that the quantum efficiency exceeded unity and scaled linearly with photon energy above $\sim 4\mathrm{eV}$, opening new opportunities for high-efficiency UV/VUV detection. These advancements drive innovation in UV detection and pave the way for developing high quantum-efficiency b-Si photomultipliers (b-SiPMs)—critical for direct VUV photon detection in noble gas and liquid scintillation systems.

SERS is an advanced approach for the detection and analysis of materials, offering exceptional sensitivity, robustness against photobleaching, narrow spectral bandwidth, and the ability to detect molecular features.21 These advantages make SERS widely applicable in fields such as food safety, biomedical research, environmental monitoring, and beyond. Traditionally, SERS substrates are constructed from rough surfaces incorporating noble metals. In recent years, b-Si has attracted notable interest as a potential SERS substrate, thanks to its extensive surface area and aspect ratio, providing an excellent platform for enhanced Raman scattering.208 Viehrig et al. introduced a b-Si–based electrochemical SERS platform that adjusts the analyte–surface interaction while maintaining the integrity of the SERS substrate, allowing the sensor to be reused effectively.204 The system incorporates uniformly Au-capped Si nanopillars with electrochemical techniques to facilitate efficient real-time reuse of the SERS substrate during assays. The mechanism of this electrochemical SERS detection system, shown in Fig. 16(f), highlights the variation in surface charges on the Au-capped nanopillars and their engagement with melamine. Figure 16(g) presents an image of the fully assembled detection chamber along with an SEM image of the Au-capped nanopillar structures utilized for SERS detection. In addition, Fig. 16(h) illustrates the tailored electrochemical-SERS system and its interface arrangement. This study demonstrates an integrated electrochemical-SERS platform with enhanced detection performance, enabling real-time calibration, sensing, and substrate reusability.

Moreover, some researchers demonstrated a realistic 3D modeling methodology to simulate b-Si–based multi-stacked nanostructures with randomly distributed gold nanoparticles on the highly roughened nonflat surface.40 This method effectively links theoretical analyses to experimental optical responses, with far-field calculations accurately reproducing reflectance spectra to reveal the dependence of light trapping on the thickness of the conformal reflector and nanoparticle size, whereas near-field analysis identifies three stochastic “hotspots.” Both factors contribute significantly to overall field enhancement, which is highly sensitive to nanoscale surface morphology. This framework enables efficient control of stack configurations, amplifying localized fields in nanoparticle gaps and nanocavities, thereby maximizing electromagnetic field enhancement and the analytical enhancement factor (AEF) for SERS detection.

In brief, SERS devices based on b-Si overcome the sensitivity limitations of conventional Raman spectroscopy, offering a powerful tool for obtaining detailed structural and interface information. It is extensively applied in surface science, adsorption studies, and the analysis of molecular orientation, conformation, and structural properties. The integration of b-Si as a substrate has further elevated SERS capabilities, significantly enhancing signal strength and detection sensitivity, and paving the way for broader and more advanced applications in various fields.

Mass spectrometry imaging (MSI) is gaining prominence in the field of omics, particularly in proteomics and metabolomics. Stefania et al. explored the fabrication and application of b-Si and gold-coated black silicon (AuBSi) substrates in surface-assisted laser desorption/ionization mass spectrometry imaging (SALDI-MSI) for analyzing animal tissues and human fingerprints.43 Cross-sectional SEM images of b-Si and AuBSi are displayed in Fig. 16(i). During the experiment, the SALDI-MS performance was improved by coating b-Si surfaces with gold nanoparticles through sputtering. As demonstrated in Fig. 16(j), the sample based on AuBSi substrates effectively detected fingerprint metabolites and analytes originating from animal tissues under both positive and negative ionization conditions, enabling high-quality imaging of mouse brain and liver tissues. Utilizing biocompatible materials such as silicon and gold allows for matrix-independent analysis of metabolites directly on biological samples, paving the way for safer and more versatile imaging techniques.

In addition, Jianmin Wu’s group introduced a method for tissue imprinting based on SiNWs capped with 2D nanoflakes ($\mathrm{NGQD}@{\mathrm{MoS}}_2/\mathrm{SiNWs}$) for MSI and cancer diagnosis [Fig. 16(k)].205 This innovative approach demonstrated superior lipid extraction, enhanced laser desorption/ionization (LDI) efficiency, and improved molecular imaging capabilities. Compared with conventional SiNW substrates, $\mathrm{NGQD}@{\mathrm{MoS}}_2/\mathrm{SiNWs}$ achieved a four- to five-fold enhancement in the detection of lipid signals and total ion intensity from animal lung tissues, underscoring its potential in advanced MSI applications and medical diagnostics.

Notably, extensive studies have explored the potential applications of b-Si in sensing209^–211 and antibacterial materials212^–215 in recent years. However, as these topics fall outside the scope of photonics, they are not discussed in this article.

Based on the discussion above, b-Si exhibits significant potential for solar energy conversion and photonics applications; however, its implementation encounters several challenges. Although its nanostructured surface significantly reduces reflectance and enhances light absorption—particularly in the ultraviolet and infrared ranges—the inherent surface roughness and associated defects can increase carrier recombination, shortening the lifetime of non-equilibrium charge carriers. In addition, the rough surface morphology may introduce high and sometimes non-uniform doping concentrations, posing significant challenges for forming a reliable silicon/metal contact.20 These factors collectively diminish solar energy conversion efficiency, reduce the signal-to-noise ratio in photodetectors, introduce unwanted background noise in SERS, and degrade image resolution in optical sensing applications. For SERS application, further functionalization with plasmonic nanoparticles (e.g., Au or Ag) is necessary to achieve high enhancement factors, but this introduces fabrication complexity and may compromise stability under laser irradiation. Thus, optimizing the trade-off between optical absorption and surface recombination remains critical for photonic devices based on b-Si.

Second, b-Si structures suffer from the bottleneck of hasty surface photo-oxidation, photo-corrosion, and indigent acid stability, where the high surface area accelerates surface corrosion in the PEC system. This instability limits quantum efficiency and long-term performance.29^,216 Passivation techniques, such as protective coatings, chemical treatments, or catalyst integration, are often needed to mitigate these issues and improve surface charge transfer but also increase manufacturing complexity and cost.

Third, the irregular morphology of b-Si can enhance plasmonic effects, but achieving uniform and reproducible signal amplification remains difficult due to inconsistent nanostructuring for SERS application. Across these applications, precise control over nanostructure geometry and seamless integration with existing optoelectronic systems complicate fabrication and scalability. Balancing optical performance with device reliability and manufacturability is an ongoing challenge in advancing b-Si–based technologies.

Finally, although some aforementioned fundamental limitations related to b-Si applications have been addressed in laboratory settings, industrial viability in scalable fabrication presents additional challenges. More comprehensive cost and lifecycle analysis would be required for b-Si to transition into the mainstream technology for solar energy conversion. Techniques such as MACE and RIE have obtained high expectations in industrial production. MACE, in particular, is favored for commercial solar cells due to its compatibility with the existing manufacturing lines based on wet chemical etching, and there are pioneers in the field already mass-producing b-Si solar cells.217^,218 However, challenges such as metal contamination and waste etching solution treatment remain critical concerns, thus reducing the metal consumption during silicon surface texturing should be recommended as a reliable route. RIE, on the other hand, offers advantages for industrial applications, including large-scale preparation, rapid processing, and the absence of noble metal contamination.132 However, RIE faces its own challenges, such as high equipment costs, the need for precise plasma control to avoid excessive ion-induced surface damage, and potential limitations in achieving ultra-deep nanostructures with uniform aspect ratios.

B-Si nanostructures have demonstrated immense potential for a variety of solar energy conversion and photonic applications. Various techniques are available for creating b-Si with diverse structures, including nanoscale spikes, cones, vertical cylinders, and porous silicon. These methods are typically more cost-effective than many advanced fabrication approaches, making b-Si a promising candidate for commercialization in various applications. The remarkable unique optical and electronic properties imparted by the high-aspect-ratio nanostructures, including broadband light absorption, enhanced carrier collection, and tailorable optical responses, position b-Si as a highly appealing material for pushing the boundaries of solar energy conversion and various photonic devices.

In the realm of solar energy conversion to electrical, chemical, and thermal energy, b-Si has demonstrated superior capabilities in improving light absorption and carrier collection compared to planar silicon, leading to record-breaking solar cell efficiencies. Further optimizations in nanostructure design, passivation schemes, and device architectures hold promise for continued advancements in b-Si solar cell performance. Integrating b-Si into tandem and multi-junction solar cell configurations could also unlock new pathways for boosting overall conversion efficiencies. In PEC systems and thermal energy conversion, the development of hybrid systems that combine the strengths of b-Si with other nanostructured materials may pave the way for innovative solutions in renewable energy technologies.

For photonic applications, the broadband and tunable optical absorption of b-Si enables high-sensitivity, spectrally versatile devices. Recent developments in b-Si photodetector have demonstrated impressive responsivities and bandwidths, opening new opportunities for their adoption in diverse sensing and imaging technologies. Exploring novel device structures and incorporating them with complementary metal-oxide-semiconductor (CMOS) electronics could further enhance the functionality and practicality of b-Si photodetectors. Moreover, the distinctive optical characteristics of b-Si nanostructures show great promise in AR coatings, optical filters, metamaterials, and so on. Harnessing the ability to precisely control the optical responses of b-Si through nanostructure engineering could lead to the development of innovative photonic technologies and components.

As the research on b-Si nanostructures continues to evolve, several key areas of investigation warrant further exploration. These include developing scalable and cost-effective fabrication techniques, optimizing nanostructure morphologies for specific applications, enhancing the thermal and mechanical stability of b-Si, and seamlessly integrating it with existing semiconductor technologies. Of these, control over the uniformity and reproducibility of nanostructures is critical for scaling up production and ensuring consistent performance in commercial applications. Furthermore, although b-Si demonstrates excellent absorption in the visible spectrum, enhancing its performance in the NIR region remains an area for further research. Addressing these challenges will be crucial for translating the promising laboratory-scale demonstrations of b-Si into practical, large-scale, and commercially viable products. Notably, recent advancements in artificial intelligence (AI) present exciting opportunities for accelerating progress, and AI-assisted approaches on b-Si have the potential to accelerate advancements by optimizing fabrication processes, predicting material properties, and identifying novel applications across energy and photonic technologies.149 This integration of AI into b-Si research holds significant promise for driving innovation in efficient and scalable solutions in the future.

Overall, the remarkable optical and electronic features of b-Si nanostructures, coupled with the ongoing refinement of fabrication processes, highlight its potential to drive innovation in solar energy conversion and a wide range of photonic devices.

Huaping Jia is currently a postdoctoral fellow in the Department of Applied Physics at the Hong Kong Polytechnic University. She received her BEng degree in communication engineering in 2013 and her PhD in electronics science and technology in 2022, both from Taiyuan University of Technology. Her current research interests cover mainly surface plasmonic effects in solar energy conversion and photonics.

Fengjia Xie obtained her PhD from the Department of Applied Physics at the Hong Kong Polytechnic University in 2024. She is currently a postdoctoral fellow focusing on photocatalysis and microfluidics.

Elyes Nefzaoui earned his PhD from the University of Poitiers, France, in 2013. Following a postdoctoral appointment at the CNRS Centre for Energy and Thermal Sciences of Lyon (CETHIL), he joined the University Gustave Eiffel where he managed the Energy Graduate Program at the ESIEE Paris School of Engineering. Currently, he is an associate professor at the University of Gustave Eiffel, CNRS, ESYCOM Laboratory, leading the energy harvesting group. His research focuses on infrared meta-materials for energy applications, heat transfer in electronic devices, thermal energy harvesting, and the development of sensors, sensor networks, and data analytics for environmental and energy monitoring.

Tarik Bourouina received his PhD in 1991 and his habilitation (HDR) in 2000 from Université Paris-Saclay. He has been a professor in physics at ESIEE Paris, Université Gustave Eiffel since 2002. He is also affiliated with the French National Center for Scientific Research (CNRS), within the CINTRA laboratory IRL 3288 in Singapore jointly with Nanyang Technological University (NTU) and THALES, and within the ESYCOM laboratory UMR9007 in France. His interests include micro-scale physics and silicon-based metasurfaces.

Heng Jiang obtained his BE and ME degrees from the School of Power and Energy, Northwestern Polytechnical University, China, in 2018 and 2021, respectively. He earned his PhD from the Department of Applied Physics at the Hong Kong Polytechnic University in 2025, where he is currently a postdoctoral fellow. His research interests include artificial compound eyes and microfluidics.

Xuming Zhang is currently a full professor at the Photonics Research Institute and Department of Applied Physics, The Hong Kong Polytechnic University. He received his BEng degree from the University of Science & Technology of China (USTC) in 1994 and his PhD from Nanyang Technological University (NTU) in 2006. His research has produced more than 160 journal papers. Recently, he set up two start-up companies. His current research interests cover mainly nanophotonics, plasmonics, microfluidics, artificial photosynthesis, biomimetics, and green energy.