Aryl bromides not only widely exist as pharmaceutical molecules, but also are essential building blocks for cross-coupling reactions in construction of valuable natural products and artificial molecules/materials1,2,3,4,5,6,7,8. Traditional strategies to make aryl bromides heavily relied on the use of Br2 and N-bromosuccinimide (NBS), which suffered from environmental-unfriendliness, poor functionalities-tolerance, over-bromination and low atom-efficiency2,3,5,6. Therefore, developing sustainable and effective bromination strategies is highly imperative. Recently, homogeneous organophotocatalytic bromination of (hetero)arenes using HBr or NaBr as the bromination reagent has been emerged as a promising strategy9,10. However, the low-reactivity, poor photostability and difficulty of reusing the organophotocatalysts significantly restrict their widespread applications. Therefore, on one hand, full utilization of redox centers for generation of Br+ species via Br oxidation by oxidative centers and reactive oxygen species (ROS) is crucial for improving reactivity2. On the other hand, heterogenization of advanced bromination organophotocatalysts not only helps to improve their photostability and recyclability, but also may introduce synergistic sites for cooperative catalysis.

Perylene-3,4,9,10-tetracarboxylic dianhydride (PTCDA) is a promising organic photoredox catalyst11,12 with proper oxidative potential (1.73 V vs. NHE, pH = 0) and reductive potential (−0.27 V vs. NHE, pH = 0) that may realize simultaneous oxidation of Br to Br+ and reduction of O2 to reactive oxygen species (ROS) for oxidative bromination13. Additionally, the conjugated perylene core and anhydride units of PTCDA shall enable its facile immobilization on support. As a result, in this work, PTCDA was chosen as a potential oxidative bromination organophotocatalyst. Unexpectedly, we discovered that PTCDA itself and physical mixture of PTCDA/metal oxide (i. e., Al2O3) showed little reactivity in organophotocatalytic oxidative bromination, even though superoxide radicals and Br2 were generated. However, when forming strong chemical interactions between PTCDA and Al2O3 (abbreviated as PTCDA/Al2O3) verified by in-situ diffuse reflectance infrared Fourier transform spectroscopy (DRIFTs), its reactivity for photocatalytic oxidative bromination was dramatically activated. Mechanistic studies disclosed that the anhydride units of PTCDA strongly interacted with the hydroxyl groups on the surface of Al2O3 upon high-temperature treatment. The O-vacancy in Al2O3 generated during high-temperature treatment provided Lewis-acid-type anchoring sites for O2. Upon light irradiation, the excited state catalyst was engaged in Br oxidation to Br2 and itself was converted to [cat]• −, which was subsequently quenched by donating an electron to O2 to furnish superoxide radical. Owing to the strong interaction between Al2O3 and O2, the formed superoxide radical could be fixed on the surface of Al2O3 that underwent another single electron reduction (SET) catalytic cycle to release H2O2. Finally, the in-situ generated Br2 and H2O2 could work together to produce HBrO for effective bromination of arenes.

Results and discussion

Heterogenization of PTCDA on Al2O3 support

PTCDA has sufficient oxidative potential for Br oxidation (1.73 V vs. NHE, pH = 0) and reductive potential (−0.27 V vs. NHE, pH = 0) for O2 reduction14,15,16,17,18,19. To heterogenize PTCDA for providing Lewis-acid sites to adsorb organic molecules and molecular O2 to enhance catalytic activity and stability, Al2O3, an insulator, was selected as the potential cooperative support. PTCDA and two-dimensional (2D) Al2O3 nanosheets were evenly ground and calcined at 320 °C to establish strong interaction (Fig. 1a). Taking 5%PTCDA/Al2O3 (5 wt.% of PTCDA) as an example, transmission electron microscope (TEM) characterizations confirm the nanosheet morphology of Al2O3 and 5%PTCDA/Al2O3 (Fig. 1b and Supplementary Fig. 1a). Meanwhile, many channels are evenly observed in the HRTEM images of 5%PTCDA/Al2O3, similar to the structure of Al2O3 as shown in Supplementary Fig. 1b, c. Furthermore, through observation with high-resolution spherical aberration corrected TEM, PTCDA can be identified in 5%PTCDA/Al2O3, which is well-anchored on the surface of Al2O3. Besides, the 0.228 nm lattice spacings can be assigned to Al2O3 (222) plane, matching well with the X-ray diffraction (XRD) pattern (Fig. 1c)20. Energy dispersive X-ray spectroscopy (EDS) mappings show uniform distribution of the Al, C and O elements, demonstrating homogeneous attachment of PTCDA on Al2O3 surface. XRD patterns show gradual increase of PTCDA intensity with increasing PTCDA content in PTCDA/Al2O3 while the diffraction peaks of Al2O3 remain nearly constant.

Fig. 1: Synthesis, characterization and photocatalytic performance.
figure 1

a Schematic illustration showing the preparation procedure of PTCDA/Al2O3. b TEM and HRTEM images, scanning transmission electron microscope (STEM) image and the corresponding elemental mappings of 5%PTCDA/Al2O3. c XRD patterns of PTCDA/Al2O3 with different PTCDA contents. d Photocatalytic bromination reaction activities of Al2O3, PTCDA, 5%PTCDA/Al2O3(physical mixture), and 5%PTCDA/Al2O3. Reaction conditions: 10 mg of photocatalyst; temperature: 30 °C; reaction time: 5 h; acetonitrile: 1 mL; anisole: 0.2 mmol; HBr: 0.2 mL; O2: 1 atm; blue LED: 100 mW cm−2. e Evaluation of photocatalytic stability.

To understand the support effect, Al2O3, PTCDA, their physical mixture, and PTCDA/Al2O3 were studied for bromination of arenes under visible light illumination and the results are shown in Supplementary Fig. 2 and Fig. 1d. PTCDA and PTCDA/Al2O3(physical mixture) show low reactivities in oxidative bromination reaction, while 5%PTCDA/Al2O3 displays excellent oxidative bromination performance with 95% yield (the actual mass percentage of PTCDA in 5%PTCDA/Al2O3 was determined to be around 5.0 wt.% by elemental analyzer as shown in Supplementary Table 1). More impressively, this organophotocatalyst shows excellent reusability and its reactivity can keep almost unchanged after multiple runs of bromination experiments (Fig. 1e and Supplementary Fig. 2).

Interaction between PTCDA and Al2O3

To explore how simple heat treatment of PTCDA and Al2O3 mixture could dramatically improve photocatalytic bromination reaction performance, Fourier transform infrared (FTIR) spectroscopy and solid-state 13C NMR measurements were performed to investigate the structure change of PTCDA before and after calcination over Al2O3. As shown in Supplementary Fig. 3, the fingerprint peaks of PTCDA in both FTIR and CNMR spectra are well kept before and after thermal treatment but with slight shift, suggesting that PTCDA remained unchanged in structure after thermal treatment and formed strong interaction with Al2O3. Thermal gravimetry analysis (TGA) (Supplementary Fig. 4) confirms thermostability of PTCDA and PTCDA/Al2O3 below 400 °C. To gain more insights, in-situ diffuse reflectance infrared Fourier transform spectroscopy (DRIFTs) were conducted (Fig. 2). Fig. 2a shows that during calcination process, the vibration peaks of C=O, C=C, C‒O and C‒H of PTCDA apparently red shift, while the broad OH stretching vibration peak originated from Al2O3 significantly blue shifts; these vibration peaks are in great contrast to that of sole PTCDA (Supplementary Fig. 5a) or Al2O3 (Supplementary Fig. 5b), validating formation of strong interaction between PTCDA and surficial hydroxyl group on Al2O321, which is further confirmed by heating PTCDA and Al2O3 mixture at 320 °C for different durations (Fig. 2b and Supplementary Fig. 5c).

Fig. 2: Interaction between PTCDA and Al2O3.
figure 2

a In-situ DRIFTs spectra of PTCDA and Al2O3 mixture calcined at different temperatures. b In-situ DRIFTs spectra of PTCDA and Al2O3 mixture calcined for different time durations.

The interaction mode of PTCDA and Al2O3 was further characterized by X-ray photoelectron spectroscopy (XPS) (Fig. 3a, b). After calcining PTCDA with Al2O3, the O 1s binding energy of C=O and C‒O in PTCDA slightly shifted from 531.5 to 531.2 eV and from 533.4 to 532.4 eV, respectively22. The C 1s binding energy of C=O in PTCDA accordingly shifted from 288.4 to 287.0 eV16, but one of the C 1 s for C‒C/C=C bond had little change. Moreover, the binding energy of Al 2p shifted from 73.9 to 74.1 eV (Fig. 3c)20. These results indicate that the O or C charge density of C‒O and C=O increases while the Al charge density decreases after forming a strong interaction between PTCDA and Al2O3 via thermal treatment, which can be attributed to the electron withdrawing effect of anhydride units in PTCDA. It also suggests that the interaction between PTCDA and Al2O3 is mainly formed via O of C‒O and C=O in PTCDA with surface OH on Al2O3. To further probe the electron state, the PTCDA adsorption configuration on Al2O3 was calculated and differential charge analysis is performed (Fig. 3d). After attached on Al2O3 surface, PTCDA molecule exhibited an increase in charge density, which is consistent with the results obtained by XPS. Meanwhile, other adsorption modes, differential charge and Bader charge as well as negative-projected crystal orbital Hamilton population (−pCOHP) for the O and H atoms with the shortest bond length were also investigated (Supplementary Fig. 6). The calculated adsorption energy is −4.808 eV for model 2 and −3.356 eV for model 3, more positive than that for model 1 (−5.744 eV), and the integral COHP values are −2.13324 for model 1, −1.85124 for model 2, and −1.44866 for model 3. These results indicate the strongest interaction between PTCDA and Al2O3 in model 1.

Fig. 3: Anchoring mode between PTCDA on Al2O3 surface.
figure 3

High-resolution XPS spectra of (a) O 1s and (b) C 1 s for 5%PTCDA/Al2O3 and PTCDA. c High-resolution Al 2p XPS spectra for 5%PTCDA/Al2O3 and Al2O3. d The adsorption mode of PTCDA on Al2O3 surface and the differential charge of PTCDA on Al2O3 surface. The isovalue is 1.25 × 10−3 eÅ−3. The yellow/cyan region represents increase/decrease in charge density.

Supplementary Fig. 7a compares the UV-vis absorption spectra for PTCDA, Al2O3, 5%PTCDA/Al2O3(physical mixture), and 5%PTCDA/Al2O3. It can be found that the absorption spectrum of 5%PTCDA/Al2O3 blue shifted in the visible range, while there was almost no change for the absorption spectrum of 5%PTCDA/Al2O3(physical mixture) as compared to that of PTCDA. Since Al2O3 only absorbs UV light (<400 nm)23, the blueshift of light absorption for 5%PTCDA/Al2O3 might result from hybridization of electronic orbitals formed between PTCDA and Al2O3. The absorption edge of PTCDA and 5%PTCDA/Al2O3(physical mixture) are at around 639 nm (bandgap: 1.94 eV) and 631 nm (bandgap: 1.96 eV), respectively17, while the absorption edge of 5%PTCDA/Al2O3 is at about 681 nm (bandgap: 1.82 eV). According to the published data17,24,25, the reduction potential of PTCDA is at −0.27 V vs. NHE (pH = 0) and the conduction band edge of Al2O3 is at −0.30 V vs. NHE (pH = 0), which are more negative than the potential of O2/O2•− [−0.13 V vs. NHE (pH = 0)] and O2/O22− [0.68 V vs. NHE (pH = 0)], implying that 5%PTCDA/Al2O3(physical mixture) and 5%PTCDA/Al2O3 are able to realize O2 activation and thus produce ROS under light illumination (Supplementary Fig. 7b). To further understand the O2 activation process, femtosecond transient absorption spectroscopy spectra were recorded under the condition of air saturation (Supplementary Fig. 7c, d)26. There shows an obvious broad positive signal in the wavelength range from 550 to 750 nm over 5%PTCDA/Al2O3 (Supplementary Fig. 7c). As time went from 0 to 1 ps, the intensity of the positive signal gradually increased to a maximum. Then, the signal intensity slightly decreased at t = 10 ps and maintained stable during 10–30 ps. Afterwards, the signal gradually reduced to 0 in the time range from 30 to 1000 ps. The increase of the positive signal results from accumulation of photogenerated electrons in the excited states. After reaching maximum, the slight signal intensity decrease (1–10 ps) and short maintenance of balanced states (10–30 ps) can be attributed to O2 activation by the excited electrons. The signal in the wavelength range from 550 to 750 nm was much weaker over 5%PTCDA/Al2O3(physical mixture) (Supplementary Fig. 7d).

Photocatalytic mechanism

Control experiments under different reaction atmospheres pinpoints the important role of O2 for photocatalytic bromination of arenes over PTCDA/Al2O3 (Supplementary Table 2, entry 1-2). To identify the key reactive species, scavenger studies were carefully designed and carried out, butylated hydroxytoluene (BHT), nitrotetrazolium blue chloride (NBT), t-butyl alcohol (t-BuOH) and 9,10-diphenylanthrene (DPA) are used as the sacrificial agents to probe the roles of carbon radical (R•), superoxide radical (•O2), hydroxyl radical (•OH) and singlet oxygen (1O2), respectively27, and the results are shown in Supplementary Fig. 8a. When NBT or t-BuOH was added into the reaction system, the yield of p-bromoanisole decreased from 95% to 28% or 43%, indicating that •O2 and •OH played critical roles in the photocatalytic bromination reaction. On the other hand, addition of BHT or DPA had little influence on the photocatalytic performance. To directly probe the ROS, electron paramagnetic resonance (EPR) experiments were conducted. Under light irradiation with continuous bubbling of O2 (Supplementary Fig. 8b), intense EPR signals of •O2 were observed over PTCDA, 5%PTCDA/Al2O3 and 5%PTCDA/Al2O3(physical mixture), but not over Al2O328. In consideration of the low reactivities of PTCDA and PTCDA/Al2O3(physical mixture) as shown in Fig. 1d, based on the results of scavenger studies, it is concluded that •O2 species is essential but not the only prerequisite for the photocatalytic bromination reaction. •O2 might work as an intermediate to generate other ROS, for example H2O2 can be generated via SET reduction of •O2. 0.009 mmol of H2O2 (quantified by iodometry method) was observed without adding HBr as the bromine source in the reaction system of 5%PTCDA/Al2O3 (Supplementary Table 3, entry 1), while no H2O2 could be detected in the reaction system of PTCDA and 5%PTCDA/Al2O3(physical mixture) (Supplementary Table 3, entry 2-3). To investigate the role of H2O2 in oxidative bromination reaction with HBr, 0.25 mmol of H2O2 was purposely added into the reaction system in the absence of photocatalyst and light, 96% yield of p-bromoanisole was obtained (Supplementary Table 2, entry 5), suggesting that H2O2 worked together with HBr in the oxidative bromination of arenes. H2O2 can be generated via 2e photocatalytic oxygen reduction reaction or 2e photocatalytic water oxidation reaction29,30,31,32,33,34,35. To verify the reaction pathway of H2O2, electron scavenger (H2PtCl6) was added into the reaction system under Ar atmosphere, and no H2O2 could be detected, which excludes the 2e photocatalytic water oxidation pathway to generate H2O2. Control experiments under O2 and Ar atmosphere as shown in Supplementary Table 2 (entries 1-2) and EPR studies as shown in Supplementary Fig. 9 verify that the H2O2 was produced via oxygen reduction reaction during the photocatalytic bromination reaction.

Most organophotocatalytic reactions undergo oxidative and reductive quenching cycles via single electron transfer (SET) processes. Herein we discover that the heterogeneous PTCDA/Al2O3 can enable a different dual-electron transfer process to generate H2O2 instead of the traditional SET. To shed light on this unusual phenomenon, in-situ DRIFTs was performed to probe the oxygen reduction reaction under visible light irradiation36,37. As shown in Fig. 4a, adsorption of •O2 at 1029 cm−1 can be clearly observed over 5%PTCDA/Al2O3(physical mixture). Interestingly, the -O-O- peak at 797 cm−1 is more intensive than the •O2 peak over 5%PTCDA/Al2O3. These results suggest that the O2 reduction behavior over 5%PTCDA/Al2O3 is different from that over PTCDA and 5%PTCDA/Al2O3(physical mixture). To experimentally probe O2 adsorption, O2-TPD (temperature programmed desorption) experiments were conducted. As shown in Fig. 4b, the O2 adsorption capacity (chemical adsorption) over 5%PTCDA/Al2O3 is much higher than that over PTCDA and 5%PTCDA/Al2O3(physical mixture). EPR measurements suggest a much higher concentration of oxygen vacancies in 5%PTCDA/Al2O3 (Fig. 4c), which were possibly generated during the heat-treatment of Al2O3 and PTCDA (demonstrated in Supplementary Fig. 10a, b). The rich surface oxygen vacancies on 5%PTCDA/Al2O3 shall provide abundant adsorption sites to chemically activate O2 molecules, thus promoting two-electron transfer from the firmly anchored photoexcited PTCDA to chemically bound O2 under light irradiation to produce H2O2. Theoretical calculations further support enhanced O2 adsorption on PTCDA/Al2O3(Model 1) (ΔG*O2 = −2.381 eV, Fig. 4e) as compared to that on PTCDA/Al2O3(physical mixture, Model 3) (ΔG*O2 = −1.735 eV, Supplementary Fig. 10c).

Fig. 4: Photocatalytic mechanism.
figure 4

a In-situ DRIFTs spectra recorded over 5%PTCDA/Al2O3 and 5%PTCDA/Al2O3(physical mixture) under visible light irradiation for different time durations. b O2-TPD and c EPR spectra of Al2O3, PTCDA, 5%PTCDA/Al2O3(physical mixture), and 5%PTCDA/Al2O3. d The influence of H2O2 or Br2 on photocatalytic bromination reaction. temperature: 30 °C; reaction time: 5 h; acetonitrile: 1 mL; anisole: 0.2 mmol; Air: 1 atm; blue LED: 100 mW cm−2. In the system of Br2 + H2O, HBr was not added, but in H2O2 system, HBr was added. The amounts of H2O2 or Br2 added are 0.0042, 0.0084, 0.0253, 0.0422, 0.0633, and 0.0843 mmol, respectively. e The possible photocatalytic bromination reaction mechanism.

Next the contributions of H2O2 and Br2 in oxidative bromination reaction were studied. According to literature2,5,38, H2O2 is able to oxide HBr to generate active electrophilic HBrO. Br2 can react with water to form HBrO as well. The HBrO can subsequently react with the substrate in an electrophilic substitution reaction. First, the generation of Br2 was well demonstrated by conducting the photocatalytic reaction in the presence of electron scavenger (H2PtCl6) under Ar atmosphere, during which HBr oxidation and generation of Br2 (at 380 nm) could be clearly observed in the UV-vis absorption spectra (Supplementary Fig. 11a, b). Then, H2O2 or Br2 was directly added into the reaction system in absence of light and photocatalyst to investigate their roles in oxidative bromination. As shown in Fig. 4d, with increasing amount of added H2O2 or Br2, the yield of p-bromoanisole increases. Meanwhile, to verify the role of HBrO, HBrO was purposely added into the reaction system without photocatalyst in dark. Addition of 0.25 mmol HBrO resulted in a p-bromoanisole yield of 95%, confirming the contribution of HBrO in the bromination reaction (Supplementary Table 2, entry 6). With these results, it is concluded that the major roles of the photoredox centers in PTCDA/Al2O3 are to furnish Br2 and H2O2, which work synergistically to generate active electrophilic HBrO for sequential bromination.

Figure 4e proposes the possible photocatalytic bromination reaction mechanism over PTCDA/Al2O3 photocatalyst. Under light irradiation, the excited state PTCDA* oxidizes Br to produce Br2 and PTCDA•− via a SET process18. The resultant PTCDA•− undergoes a reductive quenching cycle with O2 to regenerate PTCDA, meanwhile O2 obtains one electron and one proton to form •OOH, which is firmly fixed on the surface of photocatalyst. Next, the •OOH takes one more electron and one more proton to make H2O2. According to the reported literatures and our experimental data2,5, the reaction between Br2 and water or H2O2 and Br can produce active electrophilic HBrO species, which will spontaneously react with arenes to produce brominated products.

Substrate scopes

The photocatalytic oxidative bromination reaction developed here shows good extensibility to a large family of arenes (Fig. 5)39,40,41,42,43,44,45,46. Electron-rich arenes bearing one, two and three substituents are found to be well tolerated with corresponding products in good to excellent yields. Bromination of heteroarenes has also been demonstrated by using 1-phenylpyrazole as the model substrate, giving the desired product in 71% yield. Arenes with sensitive functional groups including secondary amine, tertiary amine, halides and heterocycles are compatible with the photocatalytic oxidative bromination reaction system. Taking carbazoles as substrates, their brominated products, which can be used as building blocks for making organic photoelectronic materials, could be obtained in 80–92% yields. For the purpose of practical application, a gram-scale flow system4,47 using 1,2,3-trimethoxybenzene as the model substrate was designed as shown in Fig. 5, which could produce brominated product in 82% yield.

Fig. 5: The expanding of substrate scope for photocatalytic bromination and the implementation of fluid-phase reaction using 5%PTCDA/Al2O3 as a photocatalyst.
figure 5

The yields of brominated products with anisole, o-methylanisole and m-methylanisole as substrates are quantified by gas chromatography-mass spectrometer (GC-MS) with p-fluoronitrobenzene as the internal standard. The yields of other brominated products are quantified by isolated yield. Fluid-phase reaction conditions: room temperature; time: 4 h; acetonitrile: 90 mL; substrate: 1 g; HBr: 10 mL; O2: 1 atm; blue LED:50 mW cm−2; and flowrate: 10 mL min−1.

In summary, we have demonstrated that heterogenization of organophotocatalyst such as PTCDA can dramatically switch on the reactivity in photocatalytic oxidative bromination reaction using HBr as the atom-economic bromination reagent. This mild photocatalytic bromination strategy shows good extensibility, recyclability and scalability. Mechanism studies discover that heat treatment of PTCDA and Al2O3 can not only establish strong chemical interactions between them to facilitate electron transfer, but also introduce O vacancies on Al2O3 to adsorb molecular O2. Benefiting from these synergistic effects, an unusual dual-electron transfer reaction mode is newly presented, where O2 is reduced to H2O2 on the photoreductive centers. Meanwhile, HBr oxidation occurs on photooxidative centers to furnish Br2. Both Br2 and H2O2 contribute for the effective bromination of arenes. This work highlights that heterogenization of organophotocatalysts exhibits great potential for improving their stability and recyclability, and more importantly, it is able to trigger different reaction mode via synergistic catalysis with the well-designed supports.



Aluminum nitrate nonahydrate (Al(NO3)3•9H2O), urea, 3,4,9,10-perylenetetracarboxylic dianhydride (PTCDA, 98%), triethylamine hydrochloride, anisole, acetonitrile, and hydrogen bromide (HBr, 48% in H2O) were purchased from Energy Chemical and used without further purification.

Preparation of 2D Al2O3 nanosheets

2D Al2O3 nanosheets were prepared through a hydrothermal method21,48,49. In a typical synthesis, 4.5 g of Al(NO3)3•9H2O and 5.1 g of urea were added to a 250 mL round bottom flask, followed by adding 200 ml of ultrapure water. Then the obtained mixture was maintained at 150 °C for 24 h. After cooling, the white product was washed to neutral using ultrapure water, and then dried in an oven at 75 °C for 24 h to obtain AlO(OH). AlO(OH) was calcined under air atmosphere for 8 h at 420 °C to make Al2O3 nanosheets.

Preparation of PTCDA/Al2O3

For a typical synthesis of 5%PTCDA/Al2O3, 0.3 g of Al2O3 and 0.0158 g of PTCDA were ground evenly and then added into a quartz boat, which was calcined at 320 °C for 1 h inside a tube furnace under air atmosphere at a heating rate of 5 °C/min. After cooling down to room temperature, the obtained sample was washed twice with 0.1 M triethylamine hydrochloride solution, followed by three times with ultrapure water and once with ethanol, and then dried at 60 °C in a vacuum oven for 12 h to obtain a composite of PTCDA and Al2O3, labeled as 5%PTCDA/Al2O3. PTCDA/Al2O3 with different PTCDA mass percentages were obtained using the same method by only changing the mass of PTCDA.


Transmission electron microscopy (TEM) measurements were carried out on a JEOL F200 microscope with accelerating voltage of 200 kV. The high-resolution transmission electron microscope images were obtained on a JEOL JEM-ARF200F TEM/STEM equipped with a spherical aberration corrector. The XRD patterns were recorded on a Rigaku Ultima IV X-RAY diffractometer with Cu Kα radiation (λ = 1.5418 Å) in the 2θ range from 5 to 80° at a scanning rate of 8° min−1. X-ray photoelectron spectroscopy (XPS) and in-situ XPS under light irradiation were conducted on a Escalab 250Xi spectrometer at room temperature using an Al Kα X-ray source ( = 1486.6 eV). The C 1s peak at 284.8 eV was used as the reference for the calibration of the binding energy. UV-vis diffuse reflectance spectra were measured on an Agilent Technologies Cary Series UV-vis-NIR spectrometer. In-situ diffuse reflectance infrared Fourier transform spectroscopy (DRIFTs) measurements were conducted on a Bruker Tensor II spectrometers. The percentage of PTCDA was determined by an organic element analyzer on elementar EL III. Thermogravimetric/differential thermal (TG/DTA) analysis was carried out on a Thermo plus EV2 in air with a heating rate of 10 °C min−1. 1H and 13C NMR spectra were recorded by a Bruker AVANCE III spectrometer (frequencies of 600 and 100 MHz) and the chemical shift was referenced to TMS (tetramethylsilane).

Photocatalytic bromination reaction

Typically, 10 mg of photocatalyst, 1 mL of anisole in acetonitrile solution (0.2 M) and 0.2 mL of HBr solution were added into a 10 mL vial. The vial was then purged with O2 for 1 min to maintain the O2 pressure at 1 atm and kept at 30 °C in an oil bath for 5 h under visible light irradiation [Blue LED (center wavelength, 460 nm), 100 mW cm−2]. Afterwards, 0.01 mmol of p-fluoronitrobenzene as the internal standard and 3 mL of ultrapure water were added to the vial. The products were extracted using 1 mL of ethyl acetate. The supernatant was analyzed using Agilent Technologies 7820 gas chromatography equipped with a WondaCap 5 column. The amounts of products and reactants were calculated using an internal standard method.

H2O2 production and detection

The H2O2 concentration in solution was determined by an iodometry method as reported50. After reaction, 0.1 mL of reaction solution was centrifuged, followed by adding 1 ml of potassium titanium oxalate solution (0.1 mol L−1), 1 mL of KI solution (0.4 M) and 4 ml of deionized water. The concentration of H2O2 was determined based on the absorbance at 350 nm in the UV-vis spectra.

EPR experiments

To capture •O2, 10 mg of photocatalyst, 1 mL of methanol, and 50 μL of 5,5-dimethyl-1-pyrroline N-oxide (DMPO) solution (methanol: DMPO = 10: 1) were added into a photocatalytic reactor and then irradiated by visible light under O2 atmosphere. 20 μL of solution was taken out for EPR measurement after different durations of photocatalytic reaction. To capture •OH, 10 mg of photocatalyst, 1 mL of H2O, 0.2 mmol of H2PtCl6 and 50 μL of 5,5-dimethyl-1-pyrroline N-oxide (DMPO) solution (acetonitrile: DMPO = 10: 1) were added into a photocatalytic reactor and then irradiated by visible light under Ar atmosphere. 20 μL of solution was taken out for EPR measurement after different durations of photocatalytic reaction. Room-temperature EPR spectra were recorded on Bruker EMXPLUS10/12 EPR electron paramagnetic resonance spectrometer.

TPD experiments

Oxygen desorption was measured by temperature programmed desorption (TPD) of O2 in a micro-reactor. Typically, 100 mg of catalyst was added to a micro-reactor, pre-treated in Ar at 200 °C for 1 h and then cooled to 30 °C. The adsorption of O2 (30 mL min−1) was implemented for 10 min at 30 °C and then the catalyst was flushed with Ar (30 mL min−1) for 30 min at 30 °C to remove the physically adsorbed gas on the surface of the catalyst. Programmed desorption was performed at a heating rate of 10 °C min−1 from 50 to 800 °C probed by a TCD detector.

Computational details

The spin-polarized density functional theories (DFT) were carried out by using the Vienna Ab initio Simulation Package (VASP)51. (110) surface was chosen as the active surface to represent the as-prepared γ-Al2O3 in our calculation model because that it is estimated that (110) surface is the predominant exposed-surface according to the Gibbs–Curie–Wulff law52, which occupies 74% of the total surface area, followed by (100) surface (16%) and (111) surface (10%). The Perdew-Burke-Ernzerhof generalized-gradient approximation functional was used to describe the interaction between electrons53. The DFT-D2 method was adopted to evaluate the van der Waals (vdW) interaction54. The energy cutoff was set to 400 eV. The energy criterion was set to 10−5 eV in the iterative solution of the Kohn-Sham equation. The Brillouin zone integration was performed using a 2 × 2 × 1 k-mesh. All the structures are relaxed until the residual forces on the atoms have declined to less than 0.02 eV/Å.