Reconstructed covalent organic frameworks – Nature

Chemicals

All reagents were obtained from Sigma-Aldrich, Fisher Chemical, Adamas, Jilin Chinese Academy of Sciences–Yanshen Technology or Shanghai Tensus Bio-Chem Technology and used as received. Carbon nitride was bought from Carbodeon. Solvents were obtained from commercial sources and used without further purification. A sulfone-decorated imine COF, FS-COF, was synthesized according to previous literature²².

Liquid NMR spectroscopy

¹H and ¹³C NMR spectra were recorded in solution at 400 MHz and 100 MHz, respectively, using a Bruker Avance 400 NMR spectrometer.

High-resolution mass spectrometry

The high-resolution mass spectrometry data were obtained using a Waters LCT Premier XE spectrometer.

Powder X-ray diffraction

PXRD patterns were recorded on a Bruker D8 Advance diffractometer with Cu Kα radiation with a voltage of 40 kV. Data were collected in the 2θ range of 2–40° with steps of 0.02°.

Fourier transform infrared spectroscopy

The FTIR spectra were recorded on neat samples in the range of 4,000–650 cm⁻¹ on a PerkinElmer FTIR spectrometer equipped with a single reflection diamond ATR module.

X-ray photoelectron spectroscopy

X-ray photoelectron spectroscopy (XPS) data were measured in powder form using an ESCALAB 250Xi instrument (Thermo Fisher Scientific) with a monochromatized Al Kα line source.

Elemental microanalyses

Elemental microanalyses were measured in the Research Center of Analysis and Test of East China University of Science and Technology using the EURO EA3000 Elemental Analyzer.

Solid-state NMR spectroscopy

The solid-state ¹³C NMR spectra were recorded on a Bruker Avance 400 NMR spectrometer with CP-MAS at a ¹³C frequency of 100 MHz under 12 kHz spinning rate under MAS condition.

Thermogravimetric analyses

Thermogravimetric analyses were performed on an EXSTAR6000 by heating samples at 20 °C min⁻¹ under a nitrogen atmosphere to 800 °C.

Gas adsorption analysis

Apparent surface areas were measured by nitrogen adsorption at 77.3 K using a Micromeritics ASAP 2020 volumetric adsorption analyser. Powder samples were degassed offline at 393 K for 12 h under a dynamic vacuum (10⁻⁵ bar). Before the adsorption test, the inert gas was removed using a high vacuum provided by the turbo molecular drag pump. The specific surface areas were evaluated using the BET model. Pore size distributions of COFs were obtained from fitting the nonlocal density functional theory to the adsorption data.

Low-pressure gas adsorption measurements of CO₂ (273, 283, 293 and 308 K) were performed on MicrotacBELsorp Max and MaxII gas adsorption analysers. Ultrahigh-purity (higher than 99.999%) CO₂ in compressed gas cylinders was used throughout all experiments. Samples were degassed at 393 K for 12 h before measurement. CO₂ adsorption isotherms of each COF were then fitted with virial model equations as follows:

in which N is the amount adsorbed (or uptake) in mmol g⁻¹; p is the pressure in kPa; T is the temperature in K; and m and n are multinomial coefficients that determine the isosteric heat.

The isosteric heat of each COF was calculated from the virial fitting adsorption isotherms by using the Clausius–Clapeyron equation, in which Q_st is the isosteric heat in J mol^–1, T is the temperature in K, P is the pressure in kPa, and R is the gas constant (8.314 J K^–1 mol^–1):

Scanning electron microscopy

COF morphologies were imaged using a field-emission scanning electron microscope (Helios G4 UC, Thermo Fisher Scientific).

Transmission electron microscopy

Transmission electron microscopy (TEM) characterizations were performed on a Themis Z microscope (Thermo Fisher Scientific) equipped with two aberration correctors under 200 kV. To minimize the electron beam damage, a cryo-transfer TEM holder (Model 2550, Fischione Instruments) was used, and the temperature was set below −175 °C during TEM imaging.

Ultraviolet-visible absorption spectroscopy

Ultraviolet (UV)-visible absorption spectra of the COFs were recorded on a PerkinElmer Lambda 950 UV-vis-NIR spectrometer by measuring the reflectance of powders in the solid state.

Photoluminescence spectroscopy

Photoluminescence spectra were recorded on a Varian Cary Eclipse fluorescence spectrophotometer by measuring the powders in the solid state.

Electron paramagnetic resonance spectroscopy

Electron paramagnetic resonance (EPR) spectra were acquired at room temperature under ambient conditions using a Bruker EMX-8/2.7 spectrometer. COF powders were taken in an EPR tube and excited with a 300-W Xe lamp using a 420-nm filter.

Time-correlated single photon counting measurements

Time-correlated single photon counting measurements were performed on an Edinburgh Instruments LS980-D2S2-STM spectrometer equipped with picosecond-pulsed LED excitation sources and an R928 detector, with a stop count rate below 3%. An EPL-375 diode (λ = 370.5 nm, instrument response 100 ps, full width at half maximum, FWHM) with a 450-nm high-pass filter for emission detection was used. Suspensions were prepared by ultrasonicating the COF in water. The instrument response was measured with colloidal silica (LUDOX HS-40, Sigma-Aldrich) at the excitation wavelength without filter. Decay times were fitted in FAST software using suggested lifetime estimates.

Photoelectrochemical measurements

Indium tin oxide (ITO) glasses were cleaned by sonication in ethanol and acetone for 30 min respectively, then dried under nitrogen flow. Two milligrams of COF was dispersed in 0.2 ml ethanol with µ110 ten  Nafion solution (5 wt% in a mixture of lower aliphatic alcohols and water) and ultrasonicated for 20 min to give a homogenous suspension. ITO glass slides were covered with a copper mask giving an area of 0.28 cm². Ten microlitres of the suspension was drop-casted on the ITO glass and dried overnight at room temperature. Electrochemical impedance spectroscopy and photocurrent response were performed using a Bio-Logic SP-200 electrochemical system. A three-electrode set-up was used with a working electrode (COF on ITO glass), counter electrode (platinum plate) and reference electrode (Ag/AgCl), and the bias voltage was −0.35 V. A 300-W Newport Xe light source (model 6258, ozone-free) with a 420-nm filter was used to illuminate the samples. A solution of 0.5 M Na₂SO₄ (pH = 6.8) was used for measurement.

Photocatalytic hydrogen evolution experiments

A quartz flask was charged with the photocatalyst powder (2.5 mg), 0.1 mol l⁻¹ ascorbic acid water solution (25 ml) and a certain amount of platinum (Pt) as a cocatalyst, using hexachloroplatinic acid as a Pt precursor. The resulting suspension was ultrasonicated until the photocatalyst was well-dispersed before degassing by N₂ bubbling for 30 min. The reaction mixture was illuminated with a 300 W Newport Xe light source (model 6258, ozone-free) using appropriate filters for the time specified under atmospheric pressure. The Xe light source was cooled by water circulating through a metal jacket. The samples were first illuminated for 5 h to complete Pt photo-deposition; then the flask was degassed by N₂ bubbling for 30 min followed by the photocatalysis reaction. Gas samples were taken with a gas-tight syringe and run on a Bruker 450-GC gas chromatograph. Hydrogen was detected with a thermal conductivity detector 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 not considered in the calculations. The rates were determined from a linear regression fit. After 5 h of photocatalysis, no carbon monoxide associated with framework or ascorbic acid decomposition could be detected on a gas chromatography system equipped with a pulsed discharge detector.

For stability measurements, a flask was charged with 2.5 mg of COF photocatalyst, 0.1 mol l⁻¹ ascorbic acid water solution (25 ml) and a certain amount of Pt (3 wt%) as a cocatalyst, using hexachloroplatinic acid as a Pt precursor. The resulting suspension was ultrasonicated to obtain a well-dispersed suspension, then transferred into a quartz reactor connected to a closed gas system (Labsolar-6A, Beijing Perfectlight). The reaction mixture was evacuated several times to ensure complete removal of oxygen, and the pressure was set to 13.33 kPa . The reactor was irradiated in a 90° angle with a 300-W Xe light-source. The wavelength of the incident light was controlled using a 420-nm long-pass cut-off filter. The temperature of the reaction solution was maintained at 10 °C by circulation of cool water. The evolved gases were detected on an online gas chromatograph (Shimadzu GC 2014C) with a thermal conductive detector. After the photocatalysis experiment, the photocatalyst was recovered by washing with water then solvent exchange with methanol and tetrahydrofuran, respectively, before drying at 60 °C under a vacuum.

Measurement of external quantum efficiencies

The external quantum efficiencies (EQEs) for the photocatalytic H₂ evolution were measured using monochromatic LED lamps (λ = 420, 490, 515 and 595 nm, respectively). For the experiments, the photocatalyst (2.5 mg) with Pt loading was suspended in an aqueous solution containing ascorbic acid (0.1 mol l⁻¹). The light intensity was measured with a ThorLabs S120VC photodiode power sensor controlled by a ThorLabs PM100D Power and Energy Meter Console. The EQEs were estimated using the equation:

Computational methods

Periodic DFT calculations were performed within the plane-wave pseudopotential formalism, using the Vienna ab initio simulation package (VASP) code⁴⁹. The projector augmented-wave method was applied to describe the electron–ion interactions^50,51. A kinetic-energy cut-off of 500 eV was used to define the plane-wave basis set, and the electronic Brillouin zone was integrated using Γ-centred Monkhorst−Pack grids with the smallest allowed spacing between k-points (KSPACING) being 0.25 Å⁻¹. Geometry optimizations were performed using the Perdew−Burke−Ernzerhof exchange−correlation functional with the DFT-D3(BJ) dispersion correction^52,53,54. Tolerances of 10⁻⁶ eV and 10⁻² eV Å⁻¹ were applied during the optimization of the Kohn−Sham wavefunctions and the geometry optimizations, respectively.

For crystal structures of COFs, both lattice parameters and atomic positions are allowed to change during geometry optimization. The electronic structures of the optimized RC-COF-1 and Urea-COF-1 structures were then computed using a screened hybrid exchange−correlation functional (HSE06), giving key electronic properties, such as band gap and electrostatic potential, for comparison of the COFs. Within periodic boundary conditions, the electronic eigenvalues are given with respect to an internal reference. To achieve valence band alignment, using a common vacuum level, so that band energies can be compared for the different COF structures, we followed an approach devised for determining the vacuum level of porous structures⁵⁵.

For the binding model constructed for the hydrolysis products of the COF, the hydrolysis-released p-phenylenediamine monomer was assumed to be trapped in a three-layer COF model, with the first layer being decomposed. The periodic COF layer was parallel to the XY plane and separated from its periodic images along the Z direction by a vacuum of around 14 Å⁻¹. The lattice parameters were fixed, and the atomic positions were fully optimized during this process.

The binding energy was computed using the following formula:

in which E_system is the energy of COF with the first layer hydrolysed, E_monomer and E_framework are the energies of p-phenylenediamine and framework, respectively, and the corresponding conformers were kept the same as in that system.

To visualize the intermolecular interactions between the p-phenylenediamine monomer and the COF fragment, we used the independent gradient model (IGM)⁵⁶. The IGM method quantifies the net electron density gradient attenuation that is due to intermolecular interactions, identifying non-covalent interactions and generating data composed solely of intermolecular interactions for drawing the corresponding 3D isosurface representations. Here structures were extracted out from the periodic calculation result with no hydrogen atoms added to the fragment, because we used pro-molecular level electron density here. The Multiwfn program⁵⁷ was used for IGM analyses and the VMD program⁵⁸ was used for visualization.

The geometries of the complexes were fully optimized by means of the hybrid M06-2X functional⁵⁹. For all atoms, the def2-SVP basis set^60,61 was applied. No symmetry or geometry constraint was imposed during optimizations. The optimized geometries were verified as local minima on the potential energy surface by frequency computations at the same theoretical level. These calculations were performed with the Gaussian 16 suite of programs⁶². Water was used as the solvent in the SMD solvation model⁶³. A temperature of 433 K was used for thermochemistry analysis in all calculations.