Semiconductive microporous hydrogen-bonded organophosphonic acid frameworks

Herein, we report a semiconductive, proton-conductive, microporous hydrogen-bonded organic framework (HOF) derived from phenylphosphonic acid and 5,10,15,20‐tetrakis[p‐phenylphosphonic acid] porphyrin (GTUB5). The structure of GTUB5 was characterized using single crystal X-ray diffraction. A narrow band gap of 1.56 eV was extracted from a UV-Vis spectrum of pure GTUB5 crystals, in excellent agreement with the 1.65 eV band gap obtained from DFT calculations. The same band gap was also measured for GTUB5 in DMSO. The proton conductivity of GTUB5 was measured to be 3.00 × 10−6 S cm−1 at 75 °C and 75% relative humidity. The surface area was estimated to be 422 m2 g−1 from grand canonical Monte Carlo simulations. XRD showed that GTUB5 is thermally stable under relative humidities of up to 90% at 90 °C. These findings pave the way for a new family of organic, microporous, and semiconducting materials with high surface areas and high thermal stabilities.


Synthesis of GTUB5
All the reagents and solvents employed were commercially available and used as received without further purification. As can be seen in Supplementary Figure 1 (1.36:0.24, v/v) were added to a 5-mL glass vial. The reaction mixture was ultrasonically dissolved and then heated to 80 °C in an oven for 48 h. After cooling down to room temperature, dark purple block crystals of GTUB5 formed, which were then isolated by filtration, washed with DMF and acetone, and finally air-dried. The yield of GTUB5 was ~5 mg.

Molecular simulations
The accessible pore volume, pore size distribution, and surface area of GTUB-5 were calculated by forcefield based atomistic simulations, which were performed with the RASPA molecular simulation package 2 . For these simulations, the GTUB-5 unit cell was replicated by 1 x 2 x 4 times in the x, y, and z directions, respectively, and the replicated framework atoms were fixed in their crystallographically determined positions. Lennard-Jones (LJ) and Coulomb potentials were employed to determine the nonbonded interaction energies between atoms: where rij is the distance between atoms i and j, εij and σij are the LJ well depth and diameter, respectively, qi is the partial charge of atom i, and ε0 is the dielectric constant. In all simulations, the LJ parameters between different types of sites were calculated using the Lorentz-Berthelot mixing rules, and the Ewald summation method was employed to compute the electrostatic interactions. The LJ interactions were shifted to be 0 at a cutoff distance of 12.0 Å. For the real part of the Ewald summation, the cutoff was also set to 12.0 Å.
LJ parameters for the GTUB-5 atoms (See Supplementary Table 1) were taken from the DREIDING 3 force field. Partial atomic charges for the framework atoms were obtained with the REPEAT method 4 , which fits point charges against the electrostatic potential. The electrostatic potential of GTUB-5 was obtained from a single-point energy calculation using periodic plane-wave DFT with the CASTEP 17.21 software 5 and by employing the PBE 6 functional and ultrasoft pseudopotentials 7 with a 550 eV cutoff. Accessible pore volume. The accessible pore volume of GTUB-5 was computed with the Widom insertion method using a helium probe 8 , and estimated to be 0.176 cm 3 g -1 . This calculation involved averaging over 100,000 random insertions of a single helium atom into the framework. Then, the specific pore volume, i.e., pore volume available per unit mass, was determined by

Supplementary
where ϕ is the helium-solid interaction potential for a single helium atom, dr is a differential volume element, and ms is the mass of the solid adsorbent in the simulation box. The LJ parameters for helium were taken from Hirschfelder et al. 9 , and are σHe= 2.640 Å and ε He /kB= 10.9 K. In the GC ensemble, the chemical potential, volume, and temperature of the system are fixed; however, the number of molecules fluctuates. For all GCMC simulations, a 100,000 cycle initialization and a 100,000 cycle production run were performed. Each cycle is N steps, where N is equal to the number of molecules in the system. Random insertions, deletions, translations, rotations, and reinsertions of the N2 molecules were sampled with equal probability. The TraPPE force field was used to model the N2 molecules 11 , which was originally fit to reproduce the vapor-liquid coexistence curve of N2. In this force field, the N2 molecule is rigid with the N-N bond length fixed at its experimental value of 1.10 Å. This model reproduces the experimental gas-phase quadrupole moment of the N2 molecule by placing partial 8 charges on nitrogen atoms and on a point located at the center of mass (COM) of the molecule.
Supplementary Table 2 shows the LJ parameters and partial charges for the N2 molecule.
Supplementary Table 2. LJ parameters and partial charges for the sites in the N2 molecule Using GCMC simulations, one can compute the absolute adsorption (Ntotal); whereas, in adsorption experiments, the excess adsorption (Nexcess) is measured. Therefore, the simulated excess adsorption of N2 was calculated using the following expression where ρgas is the bulk density of the gas at simulation conditions which were calculated using the Peng- Electronic structure. The geometry optimization of GTUB5 was performed using density functional theory (DFT) and the conjugate gradient method 13 within the Quickstep-CP2K program 14,15 , starting from the experimental crystal structure and with the lattice vectors set to their experimental values.
Since GTUB5 is a bulk material, periodic boundary conditions were applied to a reoriented 1x1x1 cell

X-ray data collection and structure refinement
Data for GTUB-5 was obtained with a Bruker APEX II QUAZAR three-circle diffractometer. Indexing was performed using APEX2 25 . Data integration and reduction were carried out with SAINT 26 .    Table 5. Hydrogen bond parameters (in Å and °) for GTUB-5.

FT-IR spectroscopy
IR spectra of H8TPPA and GTUB-5 were recorded between 4000 and 550 cm -1 using a Perkin Elmer Spectrum 100 FT-IR spectrometer with an attenuated total reflection (ATR) accessory featuring a zinc selenide (ZnSe) crystal.

UV-Vis spectroscopy
The solid-state diffuse reflectance ultraviolet-visible (UV-Vis) spectrum of GTUB-5 crystals was collected on a Varian Cary 300 UV-Vis Spectrophotometer and the corresponding solution spectrum was collected using a Varian Eclipse spectrofluorometer with 1-cm path length cuvettes at room temperature in DMSO.
The HOMO-LUMO gap of GTUB-5 was extracted using cyclic voltammetry (See Supplementary Figure   14) 32 . From the measurement, the first oxidation and reduction potentials of GTUB-5 in DMSO were determined to be 0.42 V and -1.23 V, which give rise to a HOMO-LUMO gap of 1.65 eV.

Proton conductivity measurement
The proton conductivity of GTUB-5 was determined by electrochemical impedance spectroscopy. A Zahner Zennium electrochemical workstation was used with an oscillation voltage of 10 mV over a frequency from 1 to 10 6 Hz. The needles were compressed between two glassy carbon electrodes by a torque of 30 cNm to obtain pellets of 82 mm in diameter and ca. 0.114 mm thickness. The stack was placed in a PTFE sample holder. The sample holder was placed in a stainless-steel chamber with an attached water reservoir. The relative humidity (%rh) was determined by the Clausius-Clapeyron relation and controlled by heating the cell and water reservoir. The sample is held overnight at the desired %rh and temperature before measuring each data point. To ensure reproducibility, each data point was measured three times.