Nitrogen-defective carbon nitride nanorod arrays for continuous-flow photosynthesis of H<sub>2</sub>O<sub>2</sub>

Nitrogen-defective carbon nitride nanorod arrays for continuous-flow photosynthesis of H₂O₂

PDF

Bin Lei, Zongxing Tu, Dou Chen, Qin Luo, Xiaoying Peng, Suqin Wu, Guiming Peng

Chinese Chemical Letters | 2026, 37(6) : 112004

Less

Chinese Chemical Letters | 2026, 37(6): 112004

• Communication •

Nitrogen-defective carbon nitride nanorod arrays for continuous-flow photosynthesis of H₂O₂

Full

Bin Lei, Zongxing Tu, Dou Chen, Qin Luo, Xiaoying Peng, Suqin Wu, Guiming Peng

Affiliations

College of Chemistry and Materials, Key Laboratory of Fluorine and Silicon for Energy Materials and Chemistry of Ministry of Education, Jiangxi Normal University, Nanchang 330022, China

About Author:

¹ These authors contributed equally to this work.

CRediT authorship contribution statement

Bin Lei: Methodology, Investigation, Data curation.

Zongxing Tu: Investigation, Formal analysis.

Dou Chen: Methodology, Investigation.

Qin Luo: Investigation.

Xiaoying Peng: Investigation, Formal analysis.

Suqin Wu: Writing – original draft, Supervision, Conceptualization.

Guiming Peng: Writing – review & editing, Writing – original draft, Supervision, Software, Project administration, Funding acquisition, Formal analysis, Data curation, Conceptualization.

Published: 2026-06-15 doi: 10.1016/j.cclet.2025.112004

Outline

Abstract

Less

The surface activity and continuous photosynthesis with easy product separation are two main concerns for practical H₂O₂ photogeneration which calls for integration of efficient photocatalyst with specifically designed reactor. Taking advantage of the intrinsic superior charge separation and electron transfer of the one-dimensional CN, herein, N-vacancy abundant carbon nitride nanorod arrays (CNNRs) were demonstrated as an efficient photocatalyst either in convention photocatalysis or in emerging flow photocatalytic H₂O₂ production. The N-vacancies multifunctionally enhance light harvesting, elevate conduction band, adsorb and activate O₂ molecules. As a result, the N-defective CNNRs delivered 5.2 and 2.1 times higher H₂O₂ yield in its powder form than common CN nanosheets and defect-free CNNRs, reaching 18.8 mmol g^-1 h^-1 with AQY of 13.4% at 400 nm in ethanol-containing solution. Moreover, the CNNRs on glass, as a photocatalyst panel, exhibits remarkable fixed-bed and continuous flow photocatalytic H₂O₂ production performances, with yield up to 23.6 and 14.2 mmol m^-2 h^-1, respectively. Mechanism investigation discloses the sequential two-step 1e oxygen reduction pathway that accounts for H₂O₂ photogeneration, with further thermal-dynamical confirmation by DFT calculation. This work introduces a new paradigm by combining photocatalyst design with advanced flow reactors to achieve continuous photosynthesis of H₂O₂, which may inspire many other photo-to-chemical conversions as well.

Key words

Photocatalysis / Hydrogen peroxide / Carbon nitride / Continuous-flow / Electronic modulation

Cite this Article

Bin Lei, Zongxing Tu, Dou Chen, Qin Luo, Xiaoying Peng, Suqin Wu, Guiming Peng. Nitrogen-defective carbon nitride nanorod arrays for continuous-flow photosynthesis of H₂O₂[J]. Chinese Chemical Letters, 2026 , 37 (6) : 112004 - . DOI: 10.1016/j.cclet.2025.112004

Full Text

Less

As a green oxidant, hydrogen peroxide (H₂O₂) is extensively used in medical sterilization [1], organic synthesis [2], environment remediation [3-5], and many other fields [6-8]. It is also well known as a promising energy carrier due to its high energy density and ease for transportation [9-12]. However, current H₂O₂ production mainly depends on the energy-consuming anthraquinone (AQ) process, which generates substantial harmful byproduct [13-15]. Searching for sustainable and low-cost H₂O₂ production methods has become an urgent and necessary task to scientists. Comparing to direct H₂O₂ synthesis from H₂ and O₂ in presence of noble metal catalysts which suffers from high catalyst cost and potential explosion risks [16-19], photocatalytic H₂O₂ production via two-electron oxygen reduction reaction (ORR) has been emerging as a promising alternative option [20-27]. It has significant advantages in terms of renewable energy driven characteristics, green reagents (H₂O and O₂), onsite-feasible production processes, low infrastructure requirements [28-32].

Among the various photocatalysts, the metal-free low-cost stable visible light responsive carbon nitride (CN), has received tremendous research interest in 2e ORR for H₂O₂ photosynthesis [33-35]. Considering the pristine CN has limited 2e ORR performance due to the limited charge transfer, serious electron recombination, and limited surface reaction activity [36-39], various strategies have been adopted to enhance CN's 2e ORR photocatalytic activity, including but not limited to molecularly D-A structure design [40-43], heterojunction engineering [44-49], cocatalyst engineering [50,51], nanostructure manipulation [52,53]. Of these efforts, CN nanostructures with different morphologies possess distinct electronic structure and charge carrier transport properties. The continuous one-dimensional nanostructure has superior charge transport over randomly packed nanoparticles, owing to the reduced grain boundaries, lower charge scattering, and confined electronic structure [54]. This principle fits with CN materials. For example, CN nanorod arrays with -C≡N functionalization and K-doping show enhanced charge separation and light absorption [55]. Carbon-doped 1D carbon nitride strips possess an extensive π-electron conjugated system, while the porous architecture could enhance charge and mass transfer, and provide abundant active sites [56]. Lin et al. reported oxygen-doped one-dimensional porous CN microtubes with remarkable photocatalytic CO₂ reduction performance due to their unique structural advantages [57]. All these studies highlight the advantages of one-dimensional CN nanostructures in promoting charge transport and separation, thereby enhancing the overall photocatalytic performance of carbon nitride materials.

In addition, the complicated catalyst recovery procedures and catalyst aggregation after use are common problems for CN suspension based photocatalysis. Fortunately, due to the immobilized state of the photocatalyst in fixed-bed photoreactor or flow photoreactor which instantly separate the product from catalyst surface by the flowing solution, they are envisioned to overcome the above problems. The requisite for this strategy is immobilizing CN catalyst onto an appropriate support material. For instance, the polypentadecalactone matrix loaded with CN achieved excellent photodegradation of methylene blue under flow conditions [58]. Shalom et al. reported CN film on glass slide serving as a continuous-flow photoreactor which demonstrated good H₂O₂ production, HMF oxidation, and water purification performances [59]. Erwin Reisner et al. developed CN_x/Ni₂P photocatalyst panel via drop-casting, and achieved continuous photocatalytic water splitting and fast catalyst recovery [60].

Herein, we integrated 1D CN nanorods' (CNNRs) advantageous charge carrier dynamics with the merits of continuous-flow photocatalysis for efficient H₂O₂ photosynthesis. Uniform CN nanorod arrays on glass slide were synthesized via chemical vapor deposition (CVD) with tailored N-vacancy density and tunable photophysical properties. N-vacancies were introduced into CNNRs by elevating the synthesis temperature, which uplifted the conduction band position and increased the 2e ORR ability. Concurrently, the thermal condensation of melamine-cyanuric acid supramolecular extends the optical absorption range while synergistically promoting charge separation and electron transfer kinetics [61,62]. The CNNRs showcases multiple photocatalytic application modes, either in the powder form or as an integrity of photocatalyst panel. H₂O₂ production experiments by N-defective CNNR₆₀₀ (synthesized at 600 ℃) powder dispersion demonstrated 5.2 times-higher H₂O₂ yield than defective-free CNNRs, reaching a high value of 18.8 mmol g^-1 h^-1 in ethanol containing solution. Besides, it presents excellent fixed-bed H₂O₂ photosynthesis performance either in static or flowing photocatalysis condition, with H₂O₂ yields of 23.6 and 14.2 mmol m^-2 h^-1, respectively. Experimental investigations and DFT calculations proved the two-step 1e ORR pathway.

The uniform CNNR arrays on glass slide were synthesized by a thermal vapor deposition process in air atmosphere by using melamine-cyanuric acid supramolecular assembly as monomer (Fig. 1a). The detailed procedures are provided in Supporting information. The CNNR synthesized at 500, 550 and 600 ℃ are denoted as CNNR₅₀₀, CNNR₅₅₀, CNNR₆₀₀, respectively. It is observed that the obtained CNNR film presents darker yellowish color at higher synthesis temperature (insets in Figs. 1b-d). Scanning electron microscopy (SEM) images show the 1D nanorod morphology of all CNNRs (Figs. 1b-g). However, it appears that the diameter decreases from ~300 nm for CNNR₅₀₀ to ~130 nm for CNNR₆₀₀. The cross-section SEM shows that all CNNRs share similar length of ~1.5 μm (Figs. 1e-g). Note that the unique nanorod morphology of CNNR is different from the amorphous irregular CN powder (PCN) (Fig. S1 in Supporting information), which was synthesized by direct thermal transformation of the melamine-cyanuric acid supramolecular, implying the inducing effect of the glass slide substrate to the formation of 1D nanoarchitecture [63]. Besides, the individual use of melamine or cyanuric acid as precursor during CVD process also yielded non-1D structures but randomly distributed CN particles (Fig. S2 in Supporting information). Transmission electron microscopy (TEM) image further confirms the nanorod structure of CNNR₆₀₀ with diameter of ~130 nm (Fig. 1h).

X-ray diffraction (XRD) patterns showed a broad (002) diffraction peak at 27.5° for all CNNRs, while the (001) diffraction peak at 13.1° is absent by comparing to CN powder (Fig. S3 in Supporting information) [64]. Fourier-transform infrared (FTIR) spectroscopy exhibits characteristic bands of CN. The peak at 812 cm⁻¹ corresponds to the out-of-plane bending vibration of heptazine units, while the peaks in the range of 1200–1600 cm⁻¹ are associated with the stretching modes of C—N in heterocycles (Fig. 2a and Fig. S4 in Supporting information). It is observed that the -NH_x peak at 3200–3500 cm⁻¹ decreases from CNNR₅₀₀ to CNNR₆₀₀ (Fig. 2a), suggesting higher temperature synthesis led to removal of -NH_x groups [65]. In addition, X-ray photoelectron spectroscopy (XPS) survey results indicate CNNR₆₀₀ has higher surficial C/N ratio (6.61 for CNNR₆₀₀ vs. 2.93, 2.34, 0.8 for CNNR₅₅₀ and CNNR₅₀₀, respectively) (Fig. 2b). All C/N ratios for CNNRs obtained from XPS survey are higher than PCN (0.8). More accurately, elemental analysis (EA) results which reflect the body element composition further confirm higher C/N atomic ratio in CNNR₆₀₀ (0.67) than CNNR₅₀₀ (0.65). The higher carbon content in CNNRs would benefit electron transfer and charge separation. The high-resolution C 1s XPS spectra (Fig. 2c and Fig. S5 in Supporting information) presents three peaks at 288.2, 286.3, and 284.8 eV, corresponding to N=C—N, C—NH_x, and C—C/C=C, respectively. In consistence with higher C/N ratio in CNNR₆₀₀, it shows stronger C—C/C=C peak in C 1s XPS (Fig. 2c). Moreover, different from C 1s XPS of PCN (Fig. S5), the CNNRs showed a dominant C—C/C=C peak in C 1s XPS rather than N=C—N peak [65]. N 1s XPS spectra (Fig. 2d) display three resolved peaks at 398.7, 400.2 and 401 eV, assigned to C—N=C, N—C₃ and C—NH_x, respectively. It is worthy to point out that the C=N—C/N—C₃ ratio for CNNR₆₀₀ (3.03) is significantly lower than CNNR₅₀₀ (6.25), CNNR₅₅₀ (5.26), and PCN (7.69) (Fig. 2d and Fig. S6 in Supporting information), suggesting higher temperature synthesis creates more C=N—C (N_2C) vacancies in CNNR surface (Fig. 2e), in consistence with temperature effect on other CN material [37,66]. Another solid evidence for more N-vacancies in CNNR₆₀₀ is its enhanced electron paramagnetic resonance (EPR) spectroscopy intensity (Fig. S7 in Supporting information). Further attempt to introduce more N_2C vacancies in CNNR by synthesis at > 600 ℃ is limited by the deformation property of glass slide substrate.

UV–vis absorption spectra show that CNNR₆₀₀ has a red-shifted absorption edge because of the introduced N-vacancies and C-rich structure (Fig. 2f) [65]. Correspondingly, their bandgaps are 2.68, 2.64, and 2.61 eV for CNNR₅₀₀, CNNR₅₅₀, and CNNR₆₀₀, respectively. Comparing to PCN synthesized at 600 ℃, CNNR₆₀₀ has significantly stronger light absorption (Fig. S8 in Supporting information), which should be a result of incident photon reflection and scattering across the ordered nanorod arrays. Combined with Mott–Schottky measurements which reflects the conduction band (flat band) positions (Fig. S9 in Supporting information), the bandgap diagrams of CNNRs are found straddling over the redox potentials of O₂/O₂^-, O₂/H₂O₂ (Fig. 2g), suggesting their potential in 2e photocatalytic oxygen reduction way to synthesize H₂O₂. Relative to that of CNNR₅₀₀, CNNR₅₅₀ and CNNR₆₀₀ demonstrated upward-lifted conduction band position (Fig. 2g), meaning their stronger photocatalytic oxygen reduction ability. Apart from modulating the energy bandgap structure, the nanorod architecture is anticipated to accelerate the charge separation dynamics. Evidently, all CNNRs demonstrate much lower photoluminescence (PL) intensity than PCN, while CNNR₆₀₀ has the lowest because of C-doping and the presence of N-vacancies (Fig. S10 in Supporting information) [52]. In consistence, transient photocurrent measurements under on/off 1-SUN illumination verify that CNNRs generate higher photocurrents when synthesis temperature gets higher and all those for CNNRs are higher than PCN (Fig. 2h and Fig. S11 in Supporting information) [67,68]. Electrochemical impedance spectroscopies (EIS) indicate their charge transfer resistances follow the trend of PCN > CNNR₅₀₀ > CNNR₅₅₀ > CNNR₆₀₀ (Fig. 2i and Fig. S12 in Supporting information). All these evidences support CNNRs have superior charge separation and electron transfer over random CN nanosheets, and CNNR₆₀₀ outperforms other two CNNRs because of the more N-vacancies and C-rich structure [69].

The excellent photophysical properties of the CNNRs pave the way for their photocatalytic application. To demonstrate this, we firstly performed the photocatalytic H₂O₂ production in water in O₂ atmosphere under 2-SUN illumination using CNNR powders as photocatalysts which were collected by peeling off from the glass slide. The H₂O₂ concentration was monitored by the peroxidase colorimetric method coupled with a standard concentration-absorption plot (Fig. S13 in Supporting information) [33]. It was observed that all CNNRs delivered much higher H₂O₂ production activities in water than PCN (Fig. 3a). Specifically, CNNR₆₀₀ shows a highest H₂O₂ yield of 336 μmol g^-1 h^-1, higher than CNNR₅₅₀ (242 μmol g^-1 h^-1) and CNNR₅₀₀ (130 μmol g^-1 h^-1), and about five folds as that of PCN (69 μmol g^-1 h^-1) (Fig. 3a). While in 5% ethanol aqueous solution, the H₂O₂ production rate is further boosted to 18.8 mmol g^-1 h^-1 by CNNR₆₀₀, comparing to 3.6 mmol g^-1 h^-1 by PCN (Fig. 3b). To the best of our knowledge, both performances of CNNR₆₀₀ outperform most previously reported CN based photocatalysts under similar conditions (Fig. 3c and Table S1 in Supporting information). This high H₂O₂ production performance maintained over continuous successive cycling tests (Fig. 3d). The AQY tests showed high values of 13.4% and 4.8% at 400 and 420 nm, respectively (Fig. 3e and Table S2 in Supporting information). Furthermore, the formation rate constant (K_f) and decomposition rate constant (K_d) were analyzed following the equation [H₂O₂] = (K_f/K_d){1-exp(-K_dt)} [70]. Results showed that all CNNRs possess advantageous K_f than PCN, while PCN has > 2 times-higher K_d (Fig. 3f). CNNR₆₀₀ exhibited 3.4 times higher K_f (8.41 μmol L^-1 min^-1) than PCN, however, with reduced K_d (0.015 min^-1 for CNNR₆₀₀ vs. 0.03 min^-1 for PCN) (Fig. 3f). This indicates CNNRs not only exhibit accelerated H₂O₂ formation kinetics but also suppress the decomposition of H₂O₂, and accounts for the superior H₂O₂ production performance of CNNRs to PCN. In addition, the K_f increases from CNNR₅₀₀ to CNNR₆₀₀ while their K_d slightly decreases, probably because of more introduced N-vacancies. Control experiments in air led to a substantial decrease in H₂O₂ yield while in N₂ atmosphere showed almost no H₂O₂ generation (Fig. 3g), implying the underlying oxygen reduction reaction (ORR). Photocatalytic H₂O₂ production in ethanol solution of different pH demonstrated that acidic media significantly enhances H₂O₂ synthesis, achieving a yield of 22.9 mmol g^-1 h^-1 (vs. 18.8 mmol g^-1 h^-1 for pH 7 vs. 7.4 g^-1 h^-1 for pH 11) (Fig. 3h). This finding indicates the photosynthesis of H₂O₂ by CNNR₆₀₀ proceeds via a proton-coupled electron transfer process [71].

Rotating ring-disk electrode (RRDE) tests under O₂ atmosphere witness obvious reduction current increase starting 0.3 V vs. RHE and establishes a lower plateau in range of 0–0.3 V. Further lowering the potential leads to another more prominent saturated reduction current step in range of −0.6~0 V (Fig. 4a). Note that there is no obvious reduction current in N₂ atmosphere (Fig. 4a), as well as negligible current generated by bare glassy carbon electrode (Fig. S14 in Supporting information). This sharp comparison suggests ORR takes place on CNNR₆₀₀ electrode. The ring current which manifested as H₂O₂ oxidation current evolves and follows the same trend as the disk current (Fig. 4a), confirming the production of H₂O₂ on surface of CNNR₆₀₀ during this process. The higher disk and ring currents of CNNR₆₀₀ than CNNR₅₀₀ accords with CNNR₆₀₀'s higher H₂O₂ production performance. The electron transfer number (n) during ORR was obtained by the relationship between disk current (I_d) and ring current (I_r): n = 4 I_d/(I_d + I_r/N), where N denotes collection efficiency [71]. Results show that the electron transfer number for CNNR₆₀₀ is below 2.6 in the whole investigated potential range, while CNNR₅₀₀ has higher n but still below 3.2 (Fig. 4b), indicating their preferable 2e ORR pathway. The H₂O₂ production selectivity of CNNR₆₀₀ is > 65% while that of CNN₅₀₀ is below 50% (Fig. 4b).

Since ethanol is an efficient hole sacrificial agent, the soaring H₂O₂ yield in ethanol-containing solution highlights the central role of photogenerated electrons in the photocatalytic H₂O₂ production process. To better reveal the 2e ORR process, colorimetric analysis via H₂O₂ test strips revealed that both KBrO₃ and 1,4-benzoquinone (BQ), acting as electron and ·O₂^- scavengers individually, significantly inhibited H₂O₂ generation (Fig. 4c) [72,73]. EPR spectra were recorded using 5,5-dimethyl-1-pyrroline N-oxide (DMPO) as a spin trap under both visible-light irradiation and dark conditions. CNNR₆₀₀ exhibits much stronger DMPO-·O₂^- adduct signals than CNNR₅₀₀, confirming the ·O₂^- generation and stronger O₂ activation by CNNR₆₀₀ (Fig. 4d). In situ FTIR spectroscopy directly showed up the evolution of key intermediates signals during H₂O₂ photogeneration (Figs. 4e and f, and Fig. S15 in Supporting information). The increasing peaks at 888 and 965 cm⁻¹ indicate effective O₂ adsorption. Simultaneously, the ·O₂^- peak at 1100 cm⁻¹ increased over time [74], verifying the generation of ·O₂^- in process of H₂O₂ photogeneration. The tiny peak at 1280 cm⁻¹ is attributed to the vibration of ·OOH [46]. In addition, there appears a growing peak at 1435 cm⁻¹, which is ascribed to the accumulation of ·HOOH species on the catalyst surface as photocatalysis progressed [75]. Combined all the pre-discussed evidences with this sequential spectroscopic fingerprint—directly observing ·O₂^- formation followed by proton-coupled reduction to ·OOH and ·HOOH, they constitute decisive evidences for the two-step 1e oxygen reduction pathway: O₂ + e⁻ → ·O₂⁻; ·O₂⁻ + H⁺ + e⁻ → ·OOH; ·OOH + H⁺ + e⁻ → H₂O₂, which governs the H₂O₂ photosynthesis on surface of CNNR₆₀₀.

To better understand the role of N-vacancies (Nv) in CNNR₆₀₀ in H₂O₂ photogeneration, density functional theory (DFT) calculations were adopted to disclose the electronic structure and H₂O₂ production mechanism. The structure models in Fig. S16 (Supporting information) and Fig. 2e were used to represent CN structures without (CNNR₅₀₀) and with N_2C vacancies (CNNR₆₀₀). Density of states (DOS) analysis reveals that the conduction bands of both CNNR₅₀₀ and N-defective CNNR₆₀₀ are composed of N 2p orbitals, while the valence bands are contributed by both C 2p and N 2p orbitals (Fig. 5a) [52]. The introduction of N-vacancies reduces the bandgap and negatively shifts the conduction band of CNNR₆₀₀ (Fig. 5a), which is in consistence with the experimental finding (Fig. 2g). The photogenerated electrons on more negative conduction band thermodynamically favors ORR. Concurrently, the N-vacancy site shows favorable O₂ adsorption with lower adsorption energy (Fig. 5b and Fig. S17 in Supporting information). The Bader charge analysis indicates there is more electron transfer from N-defective CN to O₂ molecule than from defect-free CN (0.36e vs. 0.01e, Figs. 5c and d). As a result, the O₂ molecule adsorbed at the N_2C vacancy site displays longer O=O bond (1.308 Å), meaning effective activation. The Gibbs free energy calculation shows that, comparing to the defect-free CNNR₅₀₀ possesses positive energy barrier (0.19 eV) in ·O₂^- generation, N-vacancies abundant CNNR₆₀₀ displays thorough down-hill two-step one-electron ORR pathway in H₂O₂ formation (Fig. 5e). Though the free energy diagram of one-step two-electron ORR pathway also demonstrates a downhill route (Fig. 5f), the more exoergic ·O₂⁻ formation in two-step one-electron pathway (ΔG = −1.09 eV, Fig. 5e) than *HOOH formation (ΔG = −0.92 eV, Fig. 5f) in one-step two-electron pathway make it more preferable to follow the two-step one-electron pathway [76].

Besides functioning as powder catalyst, the CNNR₆₀₀ grown on glass slide is a good photocatalyst panel for fixed-bed and flow photocatalysis, which would simplify the product separation and catalyst recovery and thus greatly reduce the H₂O₂ synthesis cost. Fig. 6 shows the H₂O₂ photosynthesis by the CNNR₆₀₀ panel in static and flowing 5% ethanol solution. In static condition, the CNNR₆₀₀ catalyst panel delivered a H₂O₂ yield of 23.6 mmol m^-2 h^-1 with good long-term stability (Figs. 6a and b). This yield is equivalent to 20.93 mmol g^-1 h^-1, which is higher than that of CNNRs powder-based yield (18.74 mmol g^-1 h^-1), indicating the advantageous CNNR panel form and flow photocatalysis mode over the powder suspension thanks to the well aligned structure and fast product separation. The CNNR₆₀₀ panel can be easily lifted off the solution and then back into next round photocatalysis experiment. In addition, as a proof-of-concept, we assembled the CNNR₆₀₀ panel in a fixed-bed flow photocatalysis reactor (Fig. 6c and Fig. S18 in Supporting information). It is observed that this flow photocatalysis reactor efficiently generates H₂O₂ at yield of 14.2 mmol m^-2 h^-1 (Fig. 6d). The lower yield in flow photocatalysis condition than in static condition could be aroused by the lower O₂ concentration since there is no continuous O₂ bubbling to saturate the flowing solution over CNNR₆₀₀ surface but with merely naturally dissolved O₂. By refreshing the solution, the as assembled flow-reactor demonstrated relatively stable flow photocatalytic H₂O₂ production over cycling test, with H₂O₂ yield slightly decreased from 14.2 mmol m^-2 h^-1 to 11.1 mmol m^-2 h^-1 after 5 successive tests (Fig. 6d). In addition, we monitored the H₂O₂ concentration in the downstream solution of a continuous flow reaction system over six hours. It was found that the H₂O₂ concentration in the outlet solution almost remains constant at 12 ± 0.5 μmol/L and thus with linear increase in the overall produced H₂O₂ (Fig. 6e). Though the concentration of the output H₂O₂ solution is low, the higher concentrated H₂O₂ solution can be obtained by circulating the reaction solution (Fig. 6f). For this purpose, apparently, the concentration of H₂O₂ monotonically increases over reaction time and reaches 72.1 μmol/L after 6 h (Fig. 6f). The above H₂O₂ photogeneration occasions demonstration suggests the CNNR₆₀₀ photocatalyst panel has promising practical potential in fix-bed and continuous flow photoproduction of H₂O₂. After H₂O₂ photosynthesis, the morphology and chemical structure of CNNR₆₀₀ remained intact with no prominent change (Figs. S19-S22 in Supporting information).

In summary, CN nanorods arrays on glass slide with tailored N-vacancies were demonstrated as good photocatalyst in conventional and emerging flow photocatalytic H₂O₂ production in this study. The one-dimensional morphology and designed N-vacancies endow with enhanced light absorption and superior charge separation and electron transfer. Besides, the introduction of N-vacancies uplifts the conduction band position and enhances the 2e ORR ability. Consequently, the N-defective CNNR₆₀₀ powder delivered a H₂O₂ yield of 336 μmol g^-1 h^-1 in water, while in ethanol solution it is up to 18.8 mmol g^-1 h^-1 with AQY of 13.4% and 4.8% at 400 and 420 nm, respectively. Moreover, the N-defective CNNRs arrays can be used as photocatalyst panel in fixed bed photocatalysis and flow photocatalysis, which effectively overcomes the conventional mass transfer limitations and achieves instant product separation and catalyst recycling. The H₂O₂ yields in static and flow photoreactor are up to 23.6 mmol m^-2 h^-1 and 14.2 mmol m^-2 h^-1, respectively. The mechanism investigations indicate the N-defective CNNRs undergo a two-step 1e ORR process, while DFT studies reveal the key role of N-vacancies in selectively O₂ adsorption and activation which subsequently proceeds with thermal-dynamically favorable step-wise proton-coupled electron addition to form H₂O₂ molecule. This work not only advances the fundamental understanding of CN nanorods in photosynthesis of H₂O₂ but also integrates material design with the emerging flow reactor for practical sustainable photosynthesis of H₂O₂.

References

Less

E. Brillas, I. Sirés, M.A. Oturan, Chem. Rev. 109 (2009) 6570–6631.

X. Sun, J. Yang, X. Zeng, et al., Angew. Chem. Int. Ed. 63 (2024) e202414417.

Z. Lu, G. Chen, S. Siahrostami, et al., Nat. Catal. 1 (2018) 156–162.

Z. Wei, M. Liu, Z. Zhang, et al., Energy Environ. Sci. 11 (2018) 2581–2589.

J. Xu, X. Zheng, Z. Feng, et al., Nat. Sustain. 4 (2021) 233–241.

S.C. Perry, D. Pangotra, L. Vieira, et al., Nat. Rev. Chem. 3 (2019) 442–458.

J.M. Campos-Martin, G. Blanco-Brieva, J.L.G. Fierro, Angew. Chem. Int. Ed. 45 (2006) 6962–6984.

K.P. Bryliakov, Chem. Rev. 117 (2017) 11406–11459.

C. Xia, Y. Xia, P. Zhu, et al., Science 366 (2019) 226–231.

S.J. Freakley, Q. He, J.H. Harrhy, et al., Science 351 (2016) 965–968.

H.B. Gray, Nat. Chem. 1 (2009) 7.

Y. Che, B. Weng, K. Li, et al., App. Catal. B: Environ. 361 (2025) 124656.

N.M. Wilson, D.W. Flaherty, J. Am. Chem. Soc. 138 (2016) 574–586.

E. Jung, H. Shin, B.H. Lee, et al., Nat. Mater. 19 (2020) 436–442.

J. Gao, H.B. Yang, X. Huang, et al., Chem 6 (2020) 658–674.

Z. Teng, W. Cai, S. Liu, et al., Appl. Catal. B: Environ. 271 (2020) 118917.

R.B. Rankin, J. Greeley, ACS Catal. 2 (2012) 2664–2672.

J.K. Edwards, B. Solsona, E.N. N, et al., Science 323 (2009) 1037–1041.

S. Chen, Z. Chen, S. Siahrostami, et al., J. Am. Chem. Soc. 140 (2018) 7851–7859.

K. Liu, Z. Wang, Y. Xie, et al., Chin. Chem. Lett. (2025), doi:10.1016/j.cclet.2025.111701.

Z. Du, Z. Zhou, N. Sun, et al., Chin. Chem. Lett. 37 (2026) 111341.

W. Wang, S. Mei, H. Jiang, et al., Chin. J. Catal. 55 (2023) 137–158.

L. Jiang, H. Du, L. Li, et al., Trans. Tianjin Univ. 29 (2023) 462–472.

M. Xu, Z. Li, R. Shen, et al., Chin. J. Catal. 70 (2025) 431–443.

X. Lin, B. Lei, Q. Luo, et al., Appl. Surf. Sci. 716 (2025) 164650.

H. Wu, Y. Wang, M. Huang, et al., Angew. Chem. Int. Ed. 64 (2025) e202420188.

X. Peng, Z. Tu, K. Wang, et al., Chem. Eng. J. 524 (2025) 169219.

Y.X. Ye, J. Pan, F. Xie, et al., Proc. Natl. Acad. Sci. U. S. A. 118 (2021) e2103964118.

Y. Zhao, P. Zhang, Z. Yang, et al., Nat. Commun. 12 (2021) 3701.

T. Liu, Z. Pan, J.J.M. Vequizo, et al., Nat. Commun. 13 (2022) 1034.

H. He, X. Wu, H. Xian, et al., Nat. Commun. 12 (2021) 6611.

M. He, X. Peng, S. Wu, et al., Inorg. Chem. Front. 12 (2025) 3237–3245.

J. Ma, X. Peng, Z. Zhou, et al., Angew. Chem. Int. Ed. 61 (2022) e202210856.

Y. Shiraishi, S. Kanazawa, Y. Sugano, et al., ACS Catal. 4 (2014) 774–780.

W. Wang, R. Liu, J. Zhang, et al., Angew. Chem. Int. Ed. 137 (2025) e202415800.

Q. Chen, C. Lu, B. Ping, et al., Appl. Catal. B 324 (2023) 122216.

F. He, Y. Lu, Y. Wu, et al., Adv. Mater. 36 (2024) 2307490.

X. Xia, Y. Jia, W. Wang, et al., J. Mater. Sci. Technol. 236 (2025) 301–309.

H. Lv, Z. Li, P. Yin, et al., Chin. Chem. Lett. 36 (2025) 110457.

C. Qin, X. Wu, L. Tang, et al., Nat. Commun. 14 (2023) 5238.

P. Sun, K. Zhong, X. Huang, et al., App. Catal. B: Environ. 366 (2025) 124998.

Q. Zhang, K. Gu, C. Dong, et al., Angew. Chem. Int. Ed. 64 (2025) e202417591.

P. Sha, L. Huang, J. Zhao, et al., ACS Catal. 13 (2023) 10474–10486.

X. Lu, Z. Chen, Z. Hu, et al., Adv. Energy Mater. 14 (2024) 2401873.

H. He, Z. Wang, J. Zhang, et al., Adv. Funct. Mater. 34 (2024) 2315426.

H. Guo, L. Zhou, K. Huang, et al., Adv. Funct. Mater. 34 (2024) 2402650.

S. Khan, M.A. Qaiser, W.A. Qureshi, et al., ACS Appl. Mater. Interfaces 17 (2025) 6249–6259.

H. Wan, Y. Xu, T. Chen, et al., Chem. Eng. J. 520 (2025) 166466.

W. Wang, Y. Wu, J. Zhang, et al., Acta Phys. Chim. Sin. 41 (2025) 100093.

K. Jinguji, M. Watanabe, R. Morita, et al., Catal. Today 426 (2024) 114400.

D. Liu, J. Shen, Y. Xie, et al., ACS Sustain. Chem. Eng. 9 (2021) 6380–6389.

Q. Li, Y. Jiao, Y. Tang, et al., J. Am. Chem. Soc. 145 (2023) 20837–20848.

H. Ou, L. Lin, Y. Zheng, et al., Adv. Mater. 29 (2017) 1700008.

C. Hu, K. Xia, C. Fu, et al., Energy Environ. Sci. 15 (2022) 1406–1422.

G. Liu, Z. Tang, X. Gu, et al., Appl. Catal. B 317 (2022) 121752.

Q. Li, L. Zhang, J. Liu, et al., Small 17 (2021) 2006622.

Q. Liu, C. Chen, K. Yuan, et al., Nano Energy 77 (2020) 105104.

N. Allasia, O. Nevskyi, M. Marelli, et al., Chem. Eng. J. 499 (2024) 155885.

V.R. Battula, G. Mark, A. Tashakory, et al., ACS Catal. 14 (2024) 11666–11674.

T. Uekert, M.A. Bajada, T. Schubert, et al., ChemSusChem 14 (2021) 4190–4197.

Y.S. Jun, E.Z. Lee, X. Wang, et al., Adv. Funct. Mater. 23 (2013) 3661–3667.

J. Barrio, M. Shalom, ChemCatChem 10 (2018) 5573–5586.

Y. Xia, P. Yang, Y. Sun, et al., Adv. Mater. 15 (2003) 353–389.

Y. Shiraishi, S. Kanazawa, Y. Kofuji, et al., Angew. Chem. Int. Ed. 53 (2014) 13454–13459.

X. Zhang, P. Ma, C. Wang, et al., Energy Environ. Sci. 15 (2022) 830–842.

G. Peng, J. Wu, M. Wang, et al., Nano Lett. 20 (2020) 2879–2885.

W. Hou, H. Guo, M. Wu, L. Wang, ACS Nano 17 (2023) 20560–20569.

Y. Che, K. Wang, C. Wang, et al., J. Mater. Sci. Technol. 243 (2026) 228–236.

Y. Cong, S. Zhang, Q. Zheng, et al., J. Colloid Interface Sci. 650 (2023) 1013–1021.

J. Cheng, W. Wang, J. Zhang, et al., Angew. Chem. Int. Ed. 63 (2024) e202406310.

Z. Li, Y. Zhou, Y. Zhou, et al., Nat. Commun. 14 (2023) 5742.

L. Yang, G. Dong, D.L. Jacobs, et al., J. Catal. 352 (2017) 274–281.

X. Lin, X. Peng, S. Wu, et al., ChemCatChem 17 (2025) e202401408.

C. Shu, X. Yang, L. Liu, et al., Angew. Chem. Int. Ed. 63 (2024) e202403926.

H. Xie, J. Fu, J. Liu, et al., Adv. Funct. Mater. 34 (2024) 2411457.

S. Zhou, W. Chen, L. Kan, et al., Angew. Chem. Int. Ed. 64 (2025) e202508436.

Appendix

Less

Year 2026 volume 37 Issue 6

PDF

Cite this Article

BibTeX

Article Info

doi: 10.1016/j.cclet.2025.112004

Receive Date：2025-08-09
Online Date：2026-10-27
Published：2026-06-15

Article Data

Affiliations

History

Received：2025-08-09
Revised：2025-10-14
Accepted：2025-10-21

Affiliations

College of Chemistry and Materials, Key Laboratory of Fluorine and Silicon for Energy Materials and Chemistry of Ministry of Education, Jiangxi Normal University, Nanchang 330022, China

Corresponding:

E-mail addresses: wusu.qin@163.com (S. Wu)

guiming@jxnu.edu.cn (G. Peng)

References

Share

https://castjournals.cast.org.cn/joweb/ccl/EN/10.1016/j.cclet.2025.112004

Share to

Scan QR to access full text

Cite this article

BibTeX

Citations

表12种不同金属材料的力学参数

科 Family	属数 Number of genus	种数 Number of species	占总种数比例 Percentage of total species (%)	属 Genus	种数 Number of species	占总种数比例 Percentage of total species (%)
鹅膏菌科Amanitaceae	2	11	5.26	鹅膏菌属 Amanita	10	4.78
小菇科 Mycenaceae	2	12	5.74	丝盖伞属 Inocybe	5	2.39
多孔菌科 Polyporaceae	8	14	6.70	蜡蘑属 Laccaria	5	2.39
红菇科 Russulaceae	3	23	11.00	小皮伞属 Marasmius	6	2.87
				小菇属 Mycena	11	5.26
				光柄菇属 Pluteus	5	2.39
				红菇属 Russula	17	8.13
				栓菌属 Trametes	5	2.39

关闭全屏

BibTeX
EndNote
RefWorks
TxT

Fig.1 Synthesis and morphologies of CNNRs. (a) Scheme of the synthesis of CNNR. Top-view SEM images of CNNR₅₀₀ (b), CNNR₅₅₀ (c), and CNNR₆₀₀ (d). (Insets are corresponding catalyst photos). Cross-section SEM images of CNNR₅₀₀ (e), CNNR₅₅₀ (f), and CNNR₆₀₀ (g). (h) TEM image of CNNR₆₀₀.

Fig.2 Chemical structure and photophysical properties characterizations of CNNRs. (a) FTIR spectra. (b) The C/N atomic ratios obtained from XPS survey. (c) XPS C 1s spectrum. (d) XPS N 1s spectrum. (e) The proposed N-defective chemical structure of CNNR₆₀₀. (f) UV–vis spectra. Insets are the Tauc plots. (g) Bandgap diagrams. (h) Photocurrent response. (i) Nyquist plots.

Fig.3 Evaluation of H₂O₂ photogeneration performances. (a) H₂O₂ production in water. (b) H₂O₂ production in 5% ethanol aqueous solution. (c) Statistic of H₂O₂ yields by carbon nitride based photocatalysts. (d) Cycling tests of H₂O₂ production by CNNR₆₀₀ in ethanol solution. (e) AQY in ethanol solution. (f) H₂O₂ formation rate constants and H₂O₂ decomposition rate constants. (g) H₂O₂ production in different atmosphere. (h) H₂O₂ production in ethanol solution of different pH.

Fig.4 H₂O₂ photogeneration process study. (a) RRDE voltammograms in O₂ saturated solution. (b) Electron transfer number and H₂O₂ selectivity. (c) Colorimetric strip test with addition of different scavengers during photocatalytic H₂O₂ production. (d) EPR signals of DMPO-^•O₂^-. (e) 3D color contour of in-situ FTIR. (f) In-situ FTIR spectra of CNNR₆₀₀ during H₂O₂ photogeneration.

Fig.5 DFT studies of CNNR₅₀₀ and CNNR₆₀₀ for H₂O₂ photogeneration. (a) Calculated density of states. (b) O₂ adsorption energy. (c, d) Top-view and side-view of charge density differences after O₂ absorption (isosurface value is 0.001 e/Å³) on (c) defect-free CNNR₅₀₀ and (d) N-defective CNNR₆₀₀. The yellow and cyan regions represent electron gaining and losing, respectively. Grey balls represent N atoms while brown balls represent C atoms. (e) Gibbs free energy diagrams for two-step 1e⁻ ORR pathway and (f) one-step 2e⁻ ORR pathway.

Fig.6 Application of CNNR₆₀₀ film in fixed bed and continuous-flow photocatalytic H₂O₂ production. (a) Scheme and (b) H₂O₂ production yield for fixed bed photocatalytic. (c) Scheme for continuous-flow photocatalytic H₂O₂ production. (d) H₂O₂ production test in continuous-flow photocatalysis. (e) Monitoring of the H₂O₂ concentration and total generated H₂O₂ in the outlet solution. (f) H₂O₂ concentration evolution when circulating the reaction solution in flow photocatalysis.

Articles: Latest Articles; Most Read; Collections

Updates: Events; News; Multimedia

About: About Us

Contact

No. 86 Xueyuan South Road, Haidian District, Beijing

100081

010-62662699(Operation and Technical Support)

qkjq@cast.org.cn

Copyright © 2025 China Association for Science and Technology. All rights reserved. For all open access content, the relevant licensing terms apply.
Sponsored by the Office of the Leading Group for Cybersecurity and Informatization of CAST, and supported by Science and Technology Review Publishing House