Enormous Promotion of Photocatalytic Activity through the Use of Near-Single Layer Covalent Organic Frameworks

Abstract

Open AccessCCS ChemistryRESEARCH ARTICLE14 Jul 2022Enormous Promotion of Photocatalytic Activity through the Use of Near-Single Layer Covalent Organic Frameworks Xiaomin Ren†, Chunzhi Li†, Wanchao Kang†, He Li, Na Ta, Sheng Ye, Linyan Hu, Xiuli Wang, Can Li and Qihua Yang Xiaomin Ren† State Key Laboratory of Catalysis, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian 116023 University of Chinese Academy of Sciences, Beijing 100049 †X. Ren, C. Li, and W. Kang contributed equally to this work.Google Scholar More articles by this author , Chunzhi Li† State Key Laboratory of Catalysis, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian 116023 University of Chinese Academy of Sciences, Beijing 100049 †X. Ren, C. Li, and W. Kang contributed equally to this work.Google Scholar More articles by this author , Wanchao Kang† State Key Laboratory of Catalysis, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian 116023 State Key Laboratory of Applied Organic Chemistry, College of Chemistry and Chemical Engineering, Lanzhou University, Lanzhou, Gansu 730000 †X. Ren, C. Li, and W. Kang contributed equally to this work.Google Scholar More articles by this author , He Li State Key Laboratory of Catalysis, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian 116023 Google Scholar More articles by this author , Na Ta State Key Laboratory of Catalysis, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian 116023 Google Scholar More articles by this author , Sheng Ye State Key Laboratory of Catalysis, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian 116023 Google Scholar More articles by this author , Linyan Hu State Key Laboratory of Catalysis, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian 116023 Google Scholar More articles by this author , Xiuli Wang *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] E-mail Address: [email protected] State Key Laboratory of Catalysis, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian 116023 Google Scholar More articles by this author , Can Li *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] E-mail Address: [email protected] State Key Laboratory of Catalysis, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian 116023 University of Chinese Academy of Sciences, Beijing 100049 Google Scholar More articles by this author and Qihua Yang *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] E-mail Address: [email protected] State Key Laboratory of Catalysis, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian 116023 Google Scholar More articles by this author https://doi.org/10.31635/ccschem.021.202101090 SectionsSupplemental MaterialAboutAbstractPDF ToolsAdd to favoritesDownload CitationsTrack Citations ShareFacebookTwitterLinked InEmail Enhancing the charge separation efficiency is highly effective strategy in improving the photocatalytic activity of covalent organic frameworks (COFs) which have the problems of low conductivity and difficult dissociation of excitons. In this work, we report the sevenfold increase in apparent quantum efficiency resulting from the use of a near-single layer COF (SLCOF) in photocatalytic hydrogen evolution compared with bulk COF. Detected by transient absorption spectroscopy characterization, 100% of photogenerated long-lived electrons in the near-SLCOF can be extracted and participate in the photocatalytic process. However, the electron extraction efficiency declined to only about 11% when the COFs were increased to eight layers, implying the difficulty of charge migration among COFs interlayers. The near-SLCOF was prepared by deposition of self-exfoliated COFs colloids on SiO2, driven by their strong affinity. This work not only sheds light on the significant influence of COF layer thickness on the charge separation efficiency but also provides a new route to prepare and stabilize COF layers for practical applications. Download figure Download PowerPoint Introduction Photocatalytic water splitting to produce hydrogen is a desirable approach to sustainably store intermittent solar energy. Two-dimensional (2D) covalent organic frameworks (COFs) have emerged as a novel type of photocatalyst due to their unique optoelectronic properties and π-conjugated skeleton which can be designed at the molecular level.1–10 Visible-light-responsive 2D COFs for example, diacetylene functionalized COFs,11 azine COF,12 sp2-carbon-linked triazine-cored COFs,13–15 thiazolo thiazole-linked COFs,16 hydrazone-based COFs,17 and sulfone-containing COFs18 have been synthesized and used for visible-light-driven H2 production. However, most COFs show mediocre activity in photocatalytic hydrogen evolution (PHE) compared with traditional inorganic semiconductors, which is mainly related to the difficulty in dissociation of excitons and the rapid recombination of photogenerated electrons and holes during the photocatalytic process.19 Several strategies have been developed to improve the charge separation efficiency of COFs, for example, incorporating donor–acceptor (D–A) moieties20,21 or halogen atoms in COFs,22,23 constructing novel π-conjugated building blocks,24 and generating junctions with other semiconductors.25 In addition to the above strategies, reducing the particle size of COFs to nanometer scale is a more facile method to improve the charge separation efficiency due to the possibility of the charge carrier recombination being reduced in the short diffusion distance.26 Therefore, the single-layer COFs (SLCOFs) offering the minimum diffusion distance for charge carriers should be the perfect candidate for photocatalysis. Although SLCOFs can be successfully prepared on single-crystal surfaces and solid–vapor–liquid interfaces,27–30 such materials are unsuitable for application in photocatalysis due to the difficulty in preventing the stacking of freestanding SLCOFs and the scale-up synthesis. Recently, partitioning the interlayer space of COFs and the acid-aided exfoliation method have been used for the synthesis of monolayer COFs, but the yields of SLCOFs by these approaches have not been high. The facile synthesis and stabilization of SLCOFs still remains a challenge.31,32 Herein, we report the preparation of a near-SLCOF by self-exfoliating of COF colloids in the presence of SiO2 nanospheres and other supports which have strong affinity for COF colloids. COFs with near-single to multiple layers were successfully deposited on SiO2 nanospheres under fine control, providing an ideal model to study the relationship between layer thickness and charge separation efficiency. It was found that almost all the photogenerated long-lived electrons in the near-SLCOF could be used for H2 production, and this value decreased sharply with the increase of COF , which elucidated the remarkable improvement in charge separation efficiency by decreasing the diffusion distance. Experimental Methods Synthesis of TP-TTA colloids The synthesis of 1,3,5-triformylphloroglucinol (TP)-4,4′,4″-(1,3,5-triazine-2,4,6-triyl)trianiline (TTA) COF colloids was similar to the method in the literature with a slight modification.33 In a typical process, 29.4 mg (0.14 mmol) of TP was dissolved in 0.5 mL of dimethyl sulfoxide (DMSO). The solution was added dropwise to a flask containing 58 mL of 0.05 M hexadecyl trimethyl ammonium bromide (CTAB) aqueous solution. After ultrasonication, 1.8 mL of 0.05 M sodium dodecyl sulfate (SDS) aqueous solution was added to form solution A. Separately, 49.6 mg (0.14 mmol) of TTA was dissolved in 1 mL of DMSO. The solution was added dropwise to a flask containing 58 mL of 0.05 M CTAB aqueous solution. After ultrasonication, 1.8 mL of 0.05 M SDS solution was added to form solution B. Finally, the solutions A and B were mixed, and 5.8 mL of acetic acid was added to the resultant solution. After reacting at 30 °C for 48 h, a completely transparent orange colloidal solution was formed. Synthesis of TP-TTA/SiO2-x In a typical process, the pH of the TP-TTA colloids solution was adjusted to 7 with NaOH (1 M), followed by the addition of the desired amounts of SiO2 colloid solution, which was diluted to 15 mg/mL with water. After stirring at room temperature for 24 h, the mixture was heated at 110 °C to evaporate the solvent to afford solid products. After thoroughly washing the mixture with ethanol five times and degassing at 120 °C for 12 h under vacuum, TP-TTA/SiO2-x was obtained. TP-TTA/MOx (MOx = TiO2, WO3, Nb2O5, ZrO2) and TP-TTA/CdS were synthesized in a similar procedure to TP-TTA/SiO2-8 with the exception that CdS/MOx supports were used instead of SiO2 colloid solution (for details, see Supporting Information). PHE A flask charged with 50 mg of photocatalyst and 30 mL of 0.1 M ascorbic acid water solution was degassed by three freeze–pump–thaw cycles. An aqueous solution of H2PtCl6 (3 wt % of COF content) was injected into the flask under inert gas. The reaction mixture was illuminated with a 300 W Xenon lamp (PLS-SXE300/300 UV, Perfect Light, China) with a cutoff filter of 420 nm. The temperature of the reaction solution was maintained at 25 °C. Gas samples were taken with a gas-tight syringe (Hamilton 1700) and run on an Agilent 6890 gas chromatograph with a thermal conductivity detector (TCD) referencing against standard gas with a known concentration of hydrogen. Hydrogen dissolved in the reaction mixture was not measured, and the pressure increase generated by the evolved hydrogen was neglected in the calculations. Results and Discussion The TP-TTA colloids confined in CTAB/SDS micelle were prepared according to the method in the literature33 using TP and TTA as monomers. The size of TP-TTA colloids is ∼28 nm as measured by the dynamic light scattering (DLS) method ( Supporting Information Figure S1 and for synthesis details, see Supporting Information). TP-TTA/SiO2-x (x denotes the layer number of TP-TTA) samples with different COF layers were prepared by dispersing commercial SiO2 nanospheres (particle size ∼26 nm) in TP-TTA colloid solution (Scheme 1). The layer number of TP-TTA on SiO2 was facilely controlled by varying the mass ratio of TP-TTA colloids and SiO2 nanospheres in the initial mixture. Scheme 1 | The illustration of preparation of TP-TTA/SiO2-x (x denotes the layer number of TP-TTA) by self-exfoliating of TP-TTA colloids. Download figure Download PowerPoint The TP-TTA content of TP-TTA/SiO2-1, TP-TTA/SiO2-5, and TP-TTA/SiO2-8 was, respectively, 1.0, 4.1, and 7.1 wt % determined by 1H NMR analysis of digested TP-TTA/SiO2-x (for details, see Supporting Information and Supporting Information Figure S2). The Fourier transform infrared (FT-IR) spectra of TP-TTA/SiO2-x clearly showed the vibrations assigned to C=C, C=O and the aromatic ring respectively at 1578, 1625, and 1598 cm−1, together with the vibrations from triazine ring at 1370 and 1510 cm−1, implying the existence of TP-TTA with β-ketoenamine linkage34–36 (Figure 1a and Supporting Information Figures S3a, S3b, S3e, and S3f). No obvious absorption peaks attributed to CTAB and SDS were observed in the FT-IR spectrum of TP-TTA/SiO2-8, indicating no or a lesser amount of residue. The 13C cross-polarization total suppression of sidebands (CP-TOSS) NMR spectrum of TP-TTA/SiO2-8 exhibited characteristic chemical shifts at 183 and 106 ppm representing –C=O of the keto form and –C=C of the aryl ring. The chemical shifts at ∼170 and ∼131 ppm were assigned to the C atoms of triazine units and the C atoms directly connected to the triazine units37 (Figure 1b). The signals in the range of 14–35 ppm assigned to alkane carbons of CTAB and SDS appeared in the NMR spectrum of TP-TTA/SiO2-8. Thermogravimetric analysis (TGA) showed ∼6.9±1 wt % of organic content in TP-TTA/SiO2-8 ( Supporting Information Figure S4), in agreement with the content of TP-TTA determined by 1H NMR, showing the low amount of surfactant in the sample. The relatively stronger signals of the surfactant were mainly due to the cross-polarization experiment. For the 13C atom with a 1H atom directly connected to it, stronger coupling between 1H and 13C enhanced the intensity of the signal because the magnetization transfer from 1H to 13C by simultaneously applying matching radiofrequency fields to both spins, according to the Hartmann–Hahn condition. The combination of FT-IR spectra and 13C CP-TOSS NMR characterization confirmed the existence of TP-TTA on SiO2. Figure 1 | (a) FT-IR and (b) 13C CP-TOSS NMR spectra of TP-TTA/SiO2-8, (c) UV–vis spectra of TP-TTA/SiO2 samples dispersed in water (A1 and A8 refer to the absorbance at 450 nm). Download figure Download PowerPoint The uniformly dispersed nanospheres with smooth surfaces identical to the parent SiO2 were observed in the transmission electron microscopy (TEM) and scanning electron microscopy (SEM) images of TP-TTA/SiO2-x samples, and the absence of other polymer phases implied that TP-TTA was mostly deposited on the surface of SiO2 (Figure 2a and Supporting Information Figures S5a–S5d). The scanning TEM (STEM) images and elemental mappings of TP-TTA/SiO2-x clearly showed a uniform distribution of C elements on the surface, suggesting that TP-TTA evenly wrapped on SiO2 nanoparticles (NPs) (Figures 2b–2d). The energy-dispersive system (EDS) mapping of nitrogen species was collected simultaneously. Due to the intrinsic low N content in the as-prepared sample, the signal-to-noise ratio of N elemental mapping was too low to generate an image with high quality. Although we had tried to obtain a better result by increasing both the dose rate and the integration time, the serious beam damage and the carbon contamination during lengthy collection times significantly degraded the spatial resolution of EDS mapping. The thickness of TP-TTA on the SiO2 increased from TP-TTA/SiO2-1 to TP-TTA/SiO2-8 which was evidenced by the increased intensity of the carbon signal. The TP-TTA layer thickness of TP-TTA/SiO2-8 was approximately 3.0 nm measured from the STEM elemental mapping. It is difficult to get the accurate thickness of the TP-TTA layer for TP-TTA/SiO2-1 and TP-TTA/SiO2-5 due to the weak signals of C elemental mapping. The thickness of the TP-TTA layer was calculated to be 0.46 and 1.81 nm, respectively, for TP-TTA/SiO2-1 and TP-TTA/SiO2-5 based on the density formula (for details, see Supporting Information). With the distance between adjacent layers of TP-TTA COFs approximately 0.37 nm by Vienna Ab-initio Simulation Package (VASP) (as shown in Supporting Information Table S1), the TP-TTA layer numbers for TP-TTA/SiO2-1, TP-TTA/SiO2-5, and TP-TTA/SiO2-8 were ∼1, ∼5, and ∼8, respectively. Figure 2 | (a) TEM image of TP-TTA/SiO2-8. STEM images and elemental mapping of (b) TP-TTA/SiO2-1, (c) TP-TTA/SiO2-5, (d) TP-TTA/SiO2-8, (e) TP-TTA/TiO2, and (f) TP-TTA/CdS. HRTEM images of (g) TP-TTA/TiO2 and (h) TP-TTA/CdS. Download figure Download PowerPoint The specific surface area of TP-TTA/SiO2-x was measured by Ar sorption at 87 K ( Supporting Information Table S2 and Figure S6) and calculated with the Brunauer–Emmett–Teller (BET) method. The specific surface area of TP-TTA/SiO2-1 was almost identical to the parent SiO2. The TP-TTA/SiO2-5 and TP-TTA/SiO2-8 showed higher BET surface area than SiO2. No micropore assigned to TP-TTA COF at 0.6 and 0.9 nm could be observed in the pore size distribution curve of TP-TTA/SiO2-1, implying a monolayer structure. TP-TTA/SiO2-5 and TP-TTA/SiO2-8 afforded the micropore from TP-TTA COF, further confirming the presence of multilayer TP-TTA ( Supporting Information Figure S6). The powder X-ray diffraction (PXRD) patterns of TP-TTA/SiO2-8 and TP-TTA displayed intense peaks at 5.7°, 9.2°, and 14.9°, respectively, attributed to the 100, 110, and 210 reflections, implying that the multilayer TP-TTA on SiO2 has a crystalline structure ( Supporting Information Figure S7). Two kinds of stacking arrangements (eclipsed AA and staggered AB) were applied to generate the hexagonal unit cells for TP-TTA COF. After geometrical optimization of the models, the experimental PXRD patterns of the TP-TTA COFs and TP-TTA/SiO2-8 were more likely to match the simulated diffraction patterns of the AA-stacking models, in good agreement with reports in the literature.38 The deposition of TP-TTA with varied thickness on SiO2 suggested that the exfoliation of TP-TTA colloids occurred during the deposition process. CdS, TiO2, WO3, Nb2O5, and ZrO2 were also used as supports for the deposition of TP-TTA. The high-resolution TEM (HRTEM) images showed a uniform coating of TP-TTA layers on the surface of CdS/TiO2 with a clear boundary between TP-TTA layers and TiO2/CdS, which was further confirmed by the STEM images and elemental mapping (Figures 2e–2h). The thickness of TP-TTA layers was 3.6 and 2.6 nm, respectively, for TP-TTA/TiO2 and TP-TTA/CdS, indicating the exfoliating deposition behavior of COF layers on the above solid supports. No polymer phase could be observed in the SEM image of TP-TTA/TiO2 and TP-TTA/CdS ( Supporting Information Figures S5e–S5h). The appearance of the characteristic vibrations of TP-TTA in the FT-IR spectra of TP-TTA/CdS and TP-TTA/TiO2 confirmed the formation of hybrid materials ( Supporting Information Figures S3c and S3d). Unfortunately, the SEM images and the corresponding EDS mapping results showed the coexistence of COF aggregates and metal oxide NPs when using WO3, Nb2O5, and ZrO2 as supports, implying that TP-TTA colloids can hardly be exfoliated on the surface of WO3, Nb2O5, and ZrO2 ( Supporting Information Figure S8). The interaction between adjacent layers of COFs involves the noncovalent van der Waals force.39–41 The size of TP-TTA COF colloids is ∼28 nm due to the surrounding compact surfactant layer preventing further growth and flocculation.33 Density functional theory (DFT) calculation showed that the exfoliation energy of TP-TTA COF (AA stacking) is directly related to the layer thickness ( Supporting Information Figure S9). The more layers, the more difficult it is to exfoliate. Consequently, the TP-TTA colloids with much smaller layer number than bulk TP-TTA COFs had weak interlayer strength which facilitated the stripping of TP-TTA layers. Furthermore, TP-TTA colloids synthesized at 30 °C have lower degrees of polymerization than COFs synthesized by the traditional solvothermal method, which further impaired the interaction strength among TP-TTA layers. When the interaction strength between the solid supports and TP-TTA COF colloids is stronger than the π−π staking (AA) strength between the TP-TTA layers, the exfoliation of TP-TTA COF colloids may occur. This is reasonable considering that the 2D imine-linked COF powders can be exfoliated in the presence of acid by temporarily weakening their interlayer stacking through electrostatic repulsion.42 To decrease the surface tension, the TP-TTA single layer tends to deposit on solid supports driven by the H-bonds or other interactions. In the presence of higher amounts of TP-TTA colloids in the synthesis mixture, multilayer TP-TTA was formed on the supports, possibly by the restacking of the single layer due to π–π interactions driven by the high temperature used for the solvent evaporation. Another important parameter for the exfoliating TP-TTA colloids on supports is the pH value of the colloid solution. By mixing SiO2 and TP-TTA colloids (adjusted to pH of 7), the positively charged TP-TTA colloids (zeta potential value of 60 mV, Supporting Information Figure S10) interacted with negatively charged SiO2 nanospheres (isoelectric point of 1.5–3.5) through electrostatic interactions to destroy the colloids. Thus, the TP-TTA was released from the micelles and delaminated into a single-layer, driven by the strong interaction of SiO2 and TP-TTA colloids. The control experiment was performed by mixing SiO2 and TP-TTA colloids with pH adjusted to 1. The TEM image of the resultant material showed the coexistence of SiO2 nanospheres and irregularly shaped TP-TTA ( Supporting Information Figure S5i). At pH of 1, the surface of SiO2 is positively charged, which does not favor the contact with the positively charged TP-TTA colloids. The control experiment signifies the importance of surface electrostatic interactions in successful deposition of TP-TTA layers on SiO2. In comparison with previously reported methods,43 the self-exfoliation of COF colloids could precisely control the layer thickness, and the SiO2 support could prevent the stacking of COF layers during the practical application process. More importantly, this method is easy for the scale-up synthesis. The color changed gradually from light yellow to yellow brown when TP-TTA content increased ( Supporting Information Figure S11). The UV–vis spectra of TP-TTA/SiO2-x water suspension gradually showed red shifts of the absorption edge from 427 to 440 nm with increased layers (Figure 1c). This can be attributed to an increased conjugation length with layer thickness and/or the J-type aggregation with the chromophores between adjacent layers.44,45 The UV–vis reflectance of solid TP-TTA/SiO2-x showed a similar tendency ( Supporting Information Figure S11). Calculated by Tauc plots,46,47 the optical band gaps of TP-TTA/SiO2-x varied from 2.28 to 2.41 eV, showing a slight increase of band gap with the decrease of the TP-TTA layer ( Supporting Information Figure S11). Mott–Schottky tests were performed to determine the conduction band minimum (CBM) of the materials ( Supporting Information Figure S12). The CBM for TP-TTA COF and TP-TTA/SiO2-x was −0.62 eV versus normal hydrogen electrode (NHE), which is negative than the thermodynamic proton reduction potential ( Supporting Information Figure S13). The above results indicate the theoretical feasibility for PHE by TP-TTA/SiO2-x and TP-TTA COFs under visible light. The PHE activity of TP-TTA/SiO2-x and corresponding bulk TP-TTA COFs ( Supporting Information Figures S4j and S4k) was tested in a H2 evolution reaction under visible light irradiation (λ > 420 nm) with ascorbic acid as a hole scavenger and photodeposited Pt NPs from H2PtCl6 as a cocatalyst (for details, see Supporting Information). Control experiments showed no H2 evolution occurred without visible light and with a bare photocatalyst, revealing that the PHE reaction occurs only under light irradiation and in the presence of a cocatalysts. As shown in Figure 3a, the yield of H2 increased linearly with the irradiation time over TP-TTA COF and TP-TTA/SiO2 samples. The H2 evolution rate was relatively slow during the first 0.5 h, possibly due to the activation of Pt cocatalyst, which is a common phenomenon for photodeposited cocatalyst.48,49 TP-TTA/SiO2-1 produced 270 μmol H2 in 4 h, much higher than the H content in the photocatalysts, showing that the evoluted H2 is not from the decomposition of the photocatalyst. The isotope labeling experiment confirmed that the H2 indeed originated from water ( Supporting Information Figure S14). The H2 evolution rate was calculated after 1 h reaction to exclude the influence of the induction period. The H2 evolution rate decreased from 76.6 to 19.8 μmol/h when increasing the layer numbers from 1 to 8 (Figure 3b). TP-TTA/SiO2 samples were more active than TP-TTA COF with the H2 evolution rate of 6.6 μmol/h even though the former samples had much lower content of photoactive TP-TTA. Therefore, it is reasonable to conclude that the size, more precisely the thickness, of TP-TTA significantly affects its photocatalytic activity by altering the diffusion distance of charge carriers. The H2 evolution rate of TP-TTA/SiO2-1 was as high as 153.2 mmolgCOF−1h−1. And this remarkable catalytic performance achieved the highest rank for COF-based photocatalysts ever reported ( Supporting Information Table S3). Figure 3 | (a) PHE as a function of reaction time with 50 mg TP-TTA/SiO2-x and TP-TTA COFs under visible-light irradiation (≥420 nm, in the presence of 30 mL 0.1 M ascorbic acid and 3 wt % H2PtCl6). (b) The relation of layer numbers of TP-TTA with PHE rate and AQE. (c) Wavelength-dependent AQE of H2 production over TP-TTA/SiO2-1. (d) Long-term course of H2 evolution over TP-TTA/SiO2-1. Download figure Download PowerPoint The apparent quantum efficiency (AQE) of TP-TTA/SiO2-1 was measured using a 300 W Xenon lamp with a 440, 480, 500, and 520 nm band-pass filter (Figure 3c, for details, see Supporting Information). The trend of AQE at different wavelengths was in good agreement with its optical absorption spectrum. The AQE at 440 nm for TP-TTA/SiO2-1 reached 7.0%, which is seven times higher than TP-TTA COF with AQE of 1.0%. With increased layer thickness, the AQE decreased intensively (Figure 3b). In a long-term PHE test, no appreciable activity decay was observed over 20 h using TP-TTA/SiO2-1 as the model photocatalyst (Figure 3d). The reused TP-TTA/SiO2-1 had similar morphology and optical absorption properties as the fresh one, revealing the high stability of TP-TTA/SiO2-1 during the photocatalysis process ( Supporting Information Figure S15). Figure 4 | Transient absorption decays of (a) TP-TTA/SiO2-1 and (b) TP-TTA/SiO2-8 in deionized water (blue line), in 0.1 M ascorbic acid (red line) and in 0.1 M ascorbic acid with photodeposited Pt cocatalyst (orange line) (pumped at 450 nm, probed at 508 nm). (c) Normalized decay profiles for TP-TTA/SiO2-1 and TP-TTA/SiO2-8 in the presence of ascorbic acid. (d) The electron extraction efficiency of TP-TTA/SiO2-1 and TP-TTA/SiO2-8. Download figure Download PowerPoint The CBM of TP-TTA/SiO2 samples was almost the same, suggesting that the high PHE activity of TP-TTA/SiO2-1 was not due to the different driving force for H+ reduction ( Supporting Information Figure S13). According to previous reports,50,51 the surface hydrophilicity/hydrophobicity of COFs has a big influence on the PHE activity. The water contact angle of TP-TTA/SiO2-x was measured, and the results showed that TP-TTA/SiO2 samples and TP-TTA COFs had a similar water contact angle of ∼20° ( Supporting Information Figure S16), implying that the surface hydrophilicity/hydrophobicity is not responsible for the difference in PHE activity. Electrochemical impedance spectroscopy (EIS) and photocurrent measurement of TP-TTA/SiO2-x samples were performed and compared with TP-TTA COFs. As shown in Supporting Information Figure S17a, the EIS arc radius of TP-TTA/SiO2-x increased with the increase of the layer number of TP-TTA, suggesting that thinner TP-TTA could improve charge mobility. The photocurrent response of TP-TTA/SiO2 samples also increased with the decrease of the number of TP-TTA layers, indicating that the shortened transmission distance was beneficial for the separation of photogenerated charge carriers ( Supporting Information Figure S17b). As a result, the PHE rate of TP-TTA/SiO2-1 sample was greatly enhanced relative to pure TP-TTA COF. To further understand the charge separation efficiency of TP-TTA/SiO2 samples, transient absorption spectra (TAS) of the water suspensions of TP-TTA/SiO2-1 and TP-TTA/SiO2-8 were studied comparatively. In the presence of the hole scavenger (ascorbic acid), the bleach signal at 508 nm was observed for both samples, and the bleach amplitude was increased considerably at the μs-s time scale in the presence of the hole scavenger (Figures 4a–4b and Supporting Information Figure S18), confirming that in such intense, long-lived bleach signal in μs-s is from free electrons and the hole scavenger promotes exciton dissociation. The initial bleach amplitude could be used to estimate the number of free electrons due to the positive relation between the electron numbers and the intensity of the bleach signal. The initial bleach amplitude increased from 0.27 mΔOD for TP-TTA/SiO2-1 to 0.36 mΔOD for TP-TTA/SiO2-8 in ascorbic acid. The relative ratio (0.36/0.27 = 1.33) of the initial bleach amplitude of TP-TTA/SiO2-8 and TP-TTA/SiO2-1 was in good accordance with the relative ratio (0.856/0.489 = 1.73) of their absorbance at 450 nm excitation under the same excitation intensity (Figure 1c). This good agreement demonstrates that all the absorbed light by TP-TTA/SiO2-x will generate long-lived electrons in the presence of ascorbic acid, demonstrating their similar exciton dissociation efficiency.52 Moreover, TP-TTA/SiO2-1 and TP-TTA/SiO2-8 have long lifetimes of free electrons (τ50%, the time when the bleach amplitude decays to half of the initial amplitude) of 719 and 887