Shape-Shifting Perovskite Catalysts Turn Sunlight Into Fuel and Scrub Toxic Air

Photocatalysis has long promised a double dividend: using nothing but sunlight to convert carbon dioxide into useful chemical fuels while simultaneously cleansing the air of hazardous pollutants. Turning that promise into practice, however, has been hampered by the stubborn limitations of the semiconductors involved. Now, a study published in Advanced Science reports a strikingly elegant solution that hinges on one of the simplest variables in nanoscience: particle shape. By growing the same perovskite compound, cesium lead iodide (CsPbI3), either as tiny cubic nanocrystals or as ultrathin one-dimensional nanowires, and then fusing each morphology onto sheets of graphitic carbon nitride (g-C3N4), researchers created two photocatalysts that behave in fundamentally different ways under illumination. The nanowire version excels at reducing carbon dioxide to carbon monoxide, while the nanocrystal version is a champion at oxidizing nitric oxide, a common urban air pollutant, into harmless end products.

The choice of host material is no accident. Graphitic carbon nitride is a metal-free, two-dimensional polymer-like semiconductor that has become one of the most studied photocatalysts of the past decade thanks to its chemical robustness and earth-abundant composition. Yet it suffers from a well-known weakness: it absorbs only the blue portion of visible light, and ultraviolet light, which many wider-bandgap semiconductors rely on, accounts for less than five percent of the energy reaching Earth from the sun. Coupling g-C3N4 with a partner semiconductor that harvests longer wavelengths is therefore essential if solar energy is to be used efficiently. CsPbI3 is an ideal candidate because, among the all-inorganic cesium lead halide perovskites, it possesses the narrowest band gap, allowing absorption to stretch from the visible deep into the near-infrared region.

Working with CsPbI3 is not without difficulty. The compound is notoriously unstable in polar and humid environments because the cesium ion is slightly too small to fully stabilize the lead-iodide octahedral lattice, which triggers structural distortion and phase degradation. Moreover, perovskite nanostructures are typically synthesized in hydrophobic solvents that coat their surfaces with long-chain organic ligands such as oleic acid and oleylamine. These ligands are useful for controlling growth but act as insulating barriers that block intimate contact and charge transfer at a heterojunction. The research team, led by Xiao Zhang and San Ping Jiang, overcame both obstacles with a two-step fabrication strategy: a mechano-chemical pre-treatment that partially strips the ligands and promotes hydrogen-bonding between the perovskite and the nitrogen-rich nanosheet surface, followed by a solvothermal treatment at 150 degrees Celsius that locks the two components into a robust, face-to-face architecture.

The structural characterization reveals just how intimate that contact is. Transmission electron microscopy shows cubic CsPbI3 nanocrystals, averaging roughly 18 nanometers, anchored uniformly across the g-C3N4 sheets without agglomeration, with lattice fringes at 0.61 nanometers confirming the cubic (100) plane. The nanowire composites tell a different story: wires of approximately 3.8 nanometers in diameter become partially embedded within the nanosheets rather than sitting loosely on top, forming the kind of continuous, ordered interface that engineers of solar materials dream about. X-ray diffraction confirms the cubic perovskite phase in both composites alongside the characteristic (002) reflection of g-C3N4 at 27.5 degrees, while elemental mapping verifies the even distribution of cesium, lead, and iodine across the hybrid structures.

Spectroscopic measurements then demonstrate why these interfaces matter. Photoluminescence, the light a semiconductor emits when excited electrons recombine with holes, is dramatically quenched in both composites, a signature of efficient charge separation. The average fluorescence lifetime of the nanocrystal composite lengthens from about 15 to nearly 27 nanoseconds, indicating that an interfacial built-in electric field is suppressing wasteful radiative recombination. Transient photocurrent measurements reinforce the point: the nanowire composite generates the highest photocurrent density of all samples, reflecting the ability of one-dimensional structures to shuttle carriers directionally along their long axis. X-ray photoelectron spectroscopy adds further evidence, with the nitrogen, cesium, lead, and iodine core-level peaks all shifting upon coupling, a fingerprint of electronic interaction and interfacial electron redistribution between the two semiconductors.

The band alignment explains the underlying physics. Tauc analysis of diffuse reflectance data yields band gaps of 2.78 electronvolts for g-C3N4 and 1.95 electronvolts for the CsPbI3 nanowires, while Mott-Schottky measurements place the flat-band potential of the nanowires at a more negative value than that of the nanosheets. When the two semiconductors touch, electrons spontaneously flow from the higher-Fermi-level perovskite into the carbon nitride until equilibrium is reached, creating a depletion region and an internal electric field across the junction. Electrochemical impedance spectroscopy shows that both composites have smaller semicircle radii than their parent materials, confirming lower charge-transfer resistance. Together, these measurements establish an S-scheme charge transport mechanism, in which excited electrons in the conduction band of g-C3N4 recombine with holes in the valence band of CsPbI3, preserving the most strongly reducing electrons and the most strongly oxidizing holes for chemistry.

That mechanism pays off spectacularly in carbon dioxide reduction. Under visible light, the nanowire composite produces carbon monoxide at a rate of 4.6 micromoles per gram per hour, outperforming the nanocrystal version at 3.6 and far exceeding bare g-C3N4. Remarkably, carbon monoxide is the only detectable product, corresponding to 100 percent selectivity, a feat attributed to the well-matched redox potentials of the preserved carriers. Because the two-electron pathway from carbon dioxide to carbon monoxide is both thermodynamically and kinetically favorable, and because separating mixed gas products is a costly downstream burden, this level of selectivity is highly valuable for both mechanistic studies and practical deployment. Radical scavenging experiments with ascorbic acid and isopropanol show that superoxide radicals play the decisive role, indicating that photogenerated electrons activate dissolved oxygen to form reactive intermediates that assist the reduction chemistry.

Stability and safety, the perennial Achilles heels of lead-halide perovskites, also fare well. After ten consecutive carbon dioxide reduction cycles, the nanowire and nanocrystal composites retain 91 and 94 percent of their initial activity respectively, and the X-ray diffraction pattern of the cycled nanowire sample remains unchanged. Lead ion leakage into solution after testing measures a mere 0.31 nanomolar, kept low by the small perovskite loading and by the residual hydrophobic ligands that suppress dissolution. For a class of materials often dismissed as too fragile and too toxic for real-world catalysis, these numbers represent a meaningful step toward credibility.

The air-purification results are equally compelling, and here the morphology story flips. The nanocrystal composite removes 69.5 percent of nitric oxide at a realistic concentration of 600 parts per billion, compared with 59.2 percent for the nanowire version and just 23.9 percent for pristine g-C3N4. More importantly, the fraction of nitric oxide converted into nitrogen dioxide, a toxic byproduct that contributes to acid rain, drops from 14.2 percent for bare carbon nitride to only 5.0 percent for the nanocrystal system. The S-scheme pathway generates an abundance of superoxide radicals, which oxidize nitric oxide first to nitrogen dioxide and then further to nitrate, effectively preventing the dangerous intermediate from accumulating. Hydroxyl radicals, generated through a chain involving hydrogen peroxide formed on the water-adsorbing carbon nitride surface, drive the final conversion to stable nitrous and nitric acids. The nanocrystal catalyst also retained its full activity over five successive reaction cycles.

The deeper lesson of this work lies in quantum confinement and defect engineering. The 3.8-nanometer-diameter nanowires possess a wider band gap of about 1.95 electronvolts, while the larger 18-nanometer nanocrystals sit at roughly 1.68 electronvolts, absorbing more effectively in the red. The nanocrystals also exhibit well-defined facets and far fewer surface defects than the defect-rich nanowires, whose vacancies and dangling bonds can trap carriers and diminish oxidative power. By simply choosing which morphology to grow, the researchers can dial the same chemical system toward reduction chemistry or oxidation chemistry, a strategy they describe as morphology-governed redox tuning. As the world searches for technologies that simultaneously address the carbon problem and the air-quality crisis, this study suggests that some of the most powerful answers may come not from new compounds, but from reshaping the ones we already have.

Subject of Research: Morphology-dependent CsPbI3/g-C3N4 S-scheme heterostructures for photocatalytic CO2 reduction and NO removal

Article Title: Perovskite CsPbI3 Nanocrystal and Nanowire Incorporated G‐C3N4 Nanosheets for CO2 Reduction and NO Removal

Article References: Perovskite CsPbI3 Nanocrystal and Nanowire Incorporated G‐C3N4 Nanosheets for CO2 Reduction and NO Removal. (n.d.). https://doi.org/10.1002/advs.78119

Image Credits: AI Generated

DOI: 10.1002/advs.78119

Keywords: photocatalysis, CsPbI3 perovskite, graphitic carbon nitride, CO2 reduction, NO removal, S-scheme heterojunction, nanowires, nanocrystals, quantum confinement, charge separation, solar fuels, air purification