Solar‐Driven Hydrogen Generation Catalyzed by g‐C3N4 with Poly(platinaynes) as Efficient Electron Donor at Low Platinum Content

Abstract A metal‐complex‐modified graphitic carbon nitride (g‐C3N4) bulk heterostructure is presented here as a promising alternative to high‐cost noble metals as artificial photocatalysts. Theoretical and experimental studies of the spectral and physicochemical properties of three structurally similar molecules Fo–D, Pt–D, and Pt–P confirm that the Pt(II) acetylide group effectively expands the electron delocalization and adjusts the molecular orbital levels to form a relatively narrow bandgap. Using these molecules, the donor–acceptor assemblies Fo–D@CN, Pt–D@CN, and Pt–P@CN are formed with g‐C3N4. Among these assemblies, the Pt(II) acetylide‐based composite materials Pt–D@CN and Pt–P@CN with bulk heterojunction morphologies and extremely low Pt weight ratios of 0.19% and 0.24%, respectively, exhibit the fastest charge transfer and best light‐harvesting efficiencies. Among the tested assemblies, 10 mg Pt–P@CN without any Pt metal additives exhibits a significantly improved photocatalytic H2 generation rate of 1.38 µmol h−1 under simulated sunlight irradiation (AM1.5G, filter), which is sixfold higher than that of the pristine g‐C3N4.


Photocurrent and impedance tests
A working electrode was prepared as follows: 0.05 g of sample was ground with ethyl cellulose in ethanol to make a slurry. The slurry was coated onto an indium−tin oxide glass (ITO glass) by a doctor blade method, and then dried at 120 °C for 1 h to obtain working electrodes with a similar film thickness. The photocurrent was measured on a CHI 660E electrochemical workstation with a three-electrode system, in which the prepared electrode, a Pt (CH Instruments, Inc.) and a calomel (CH Instruments, Inc.) electrode were used as the working electrode, the counter electrode and the reference electrode, respectively. 0.5 M Na 2 SO 4 solution was used as the electrolyte. A Xe lamp (CEAULIGHT, CEL-HXF300) with AM 1.5 filter (CEAULIGHT, CEL-AM 1.5) was used as the light source. The light intensity of the light source was adjusted to 50 mW·cm -2 by adjusting the distance between the light source and the sample.

Electrochemical measurement
The cyclic voltammetry was measured on a CHI 660E electrochemical workstation with a three-electrode system, in which a glassy carbon electrode, a Pt (CH Instruments, Inc.) and a Ag/AgCl (CH Instruments, Inc.) electrode were used as the working electrode, the counter electrode and the reference electrode, respectively. 0.1 M NBu 4 F 6 PO 4 solution in DCM and ferrocene were used as the electrolyte and internal standard, respectively.

Photocatalytic experiment
4 10 mg of the obtained sample was dispersed in 100 ml deionized water, which contained 10 vol% of triethanolamine (TEOA). After sonication for 10 minutes, the solution was transferred to the reaction cell, and vacuumized with stirring. Xe lamp (CEAULIGHT, CEL-HXF300) with AM1.5G cut filter (CEAULIGHT, CEL-UVIRCUTAM1.5) was used as the light source (light intensity of bottom center was 100 mW·cm -2 ). The amount of generated hydrogen was monitored by gas chromatography (TCD, CEAULIGHT, GC-7920) every 30 minutes. Ultrapure N 2 was used as the carrier gas.
Characterization 1 H, 13 C and 31 P nuclear magnetic resonance spectra (NMR) were measured on an Agilent DD2-600 using CDCl 3 as solvent. The average molar mass and distribution were determined by gel permeation chromatography (GPC) (Waters e2695 Separations Module, Waters, Singapore), and the polystyrene samples were used as the calibration standards. The morphology of the prepared BHJs was imaged by a scanning electron microscope (SEM; thermo scientific, APREO S; EDS mapping; Bruker, XFlash 6I10) and a transmission electron microscopy (TEM; JEOL & Oxford Instruments, JEM-F200&Aztec Energy TEM SP X-MaxN 80T). Diffuse reflectance spectra (DRS) were measured using a UV-Vis spectrophotometer (PekinElmer, LAMBDA) equipped with an integrating sphere unit. The fluorescence spectra were performed at room temperature using a LS55B spectrophotometer (PerkinElmer, USA). The fluorescence decay dynamics was investigated by using an optical microscope (Edinburgh, FLS980) combined with a TCSPC module. Surface chemical states were investigated by X-ray photoelectron spectroscopy (XPS) measurement with Thermo Scientific ESCALAB 250Xi system and adventitious C1s peak (284.6 eV) as the reference.
The Brunauer-Emmett-Teller (BET) of the specific surface area and pore size distributions were measured using N 2 adsorption-desorption equilibrium by Quantachrome Autosorb-1MP gas adsorption analyzer.

Synthesis of electron donors Fo-D, Pt-D and Pt-P
As illustrated in Scheme S1, Fo-D, Pt-D and Pt-P were synthesized in two steps by the Sonogashira coupling reaction. The structures were characterized by NMR, as presented in Figure S1S5. The molecules Pt-D and Pt-P show excellent solubility in common organic solvents due to the herringbone side chains, while the solubility of Fo-D is relatively low due to its rigid skeleton. Gel permeation chromatography ( Figure S6, Table S1) showed an average molecular weight of 37,776 Da and a polymer dispersity index of 1.79 for Pt-P. Scheme 1. Synthetic routes of molecules Fo-D, Pt-D and Pt-P.
The mixture was stirred at room temperature for 12 hours. The resulting mixture was dissolved in 120 mL dichloromethane and washed with water, brine and finally dried over anhydrous Mg 2 SO 4 to afford compound 3 or compound 6.

Synthesis of Pt-D.
Compound 2 (1.00 mmol, 1 equivalent) in 100 mL DCM and 100 mL triethylamine was degassed and backfilled with nitrogen for at least three times.
Bis(tributylphosphine)platinum(II) dichloride (1.00 mmol, 1 equivalent) and copper iodide (0.20 mmol, 2 mol% equivalents) were added to the flask. The mixture was bubbled with nitrogen for half an hour. Copper iodide (0.20 mmol, 2 mol% equivalents) was added to the flask. The mixture was stirred for 12 hours at 50 °C and then filtered. The filtrate was concentrated under reduced pressure, and the crude product was purified by column       Table S1. GPC data of Pt-P. a) Number average molecular weight, b) Weight average molecular weight, c) Peak molecular weight, d) Z average molecular weight, e) Z + 1 average molecular weight and f) Polydispersity index.

Synthesis of electron acceptor g-C 3 N 4
The electron acceptor g-C 3 N 4 was prepared by programmed pyrolysis of urea without further purification and was obtained as a light yellow powder that is insoluble in common organic solvents.

Preparation of bulk heterojunction photocatalysts
To prepare bulk heterojunction photocatalysts, g-C 3 N 4 (100 mg) was sonicated in 100 mL

15
The BET test has been performed. Figure S13 shows the nitrogen adsorption-desorption isotherms and the corresponding pore size distribution curves of the samples g-C 3 N 4 , Fo-D@CN, Pt-D@CN and Pt-P@CN. All the samples show isotherms of type IV according to the Brunauer-Deming-Deming-Teller (BDDT) classification and the hysteresis loops of type H3 outspread at a relative pressure range of 0.8-1, indicating the presence of mesopores (2-50 nm). Moreover, the observed hysteresis loops approach P/P0 = 1, indicating the presence of macropores (> 50 nm). The pore size distributions (inset in Figure S13) further demonstrate a wide distribution range from 2 to 150 nm for g-C 3 N 4 . As for the SEM images shown in Figure S8a, the g-C 3 N 4 overlap each other to form a multilayer structure, resulting in numerous meso-, microporous and macropores, which are beneficial for the transportation of reactants and products. Table S2 lists       Reaction condition: 50 mg samples in 100 mL distilled water with 10 vol% of TEOA; Light intensity is 12.00 mW•cm -2 ; Irradiation area is 1.20 cm 2 .
Calculation of AQE.
The number of incident photons: