Enhanced surface area and reduced pore collapse of methylated, imine-linked covalent organic frameworks

Covalent Organic Frameworks (COFs) are thermally and chemically stable, nanoporous materials with high surface areas, making them interesting for a large variety of applications including energy storage, gas separation, catalysis and chemical sensing. However, pore blocking and pore collapse may limit their performance. Reducing the capillary forces by using solvents with low surface tension, like supercritical CO2, for activation, and the introduction of bulky isopropyl/methoxy groups were found to reduce pore collapse. Herein, we present an easy-to-use alternative that involves the combination of a new, methylated building block (2,4,6-trimethylbenzene-1,3,5-tricarbaldehyde, Me3TFB) with vacuum drying. Condensation of Me3TFB with 1,4-phenylenediamine (PA) or benzidine (BD) resulted in imine-linked 2D COFs (Me3TFB-PA and Me3TFB-BD) with higher degrees of crystallinity and higher BET surface areas compared to their non-methylated counterparts (TFB-PA and TFB-BD). This was rationalized by density functional theory computations. Additionally, the methylated COFs are less prone to pore collapse when subjected to vacuum drying and their BET surface area was found to remain stable for at least four weeks. Within the context of their applicability as sensors, we also studied the influence of hydrochloric acid vapour on the optical and structural properties of all COFs. Upon acid exposure their colour and absorbance spectra changed, making them indeed suitable for acid detection. Infrared spectroscopy revealed that the colour change is likely attributed to the cleavage of imine bonds, which are only partially restored after ammonia exposure. While this limits their application as reusable sensors, our work presents a facile method to increase the robustness of commonly known COFs.


General Information
Materials 1,4-Phenylenediamine (>98 % (GC)(T)) and 1,3,5-benzenetricarboxaldehyde (96%) were purchased from TCI Europe N.V., benzidine (98%) was purchased form Abcr and all chemicals were used without further purification. Mesitylene (99%, extra pure) was purchased from Fisher Scientific and 1,4-dioxane (99%) was purchased from Acros Organics B.V.B.A. All solvents and glacial acetic acid (AR) were purchased from commercial sources and used without further purifications. Instrumentation 1 H and 13 C NMR spectra were recorded on a Bruker AVANCE III NMR spectrometer at 400 MHz and 100 MHz, respectively. The spectra were referenced with respect to the deuterated solvents (CDCl 3 : 7.26 ppm, 77.16 ppm, DMSO-d 6 : 2.5 ppm, 39.52 ppm). 1 H and 13 C{ 1 H} cross-polarization magic angle spinning (CPMAS) solid-state NMR (ssNMR) spectra were recorded on a Bruker AVANCE III HD spectrometer at 700.13 MHz (16.4 T) and 176 MHz, respectively. Solid-state NMR samples were packed into 4 mm zirconia rotors and spun at MAS frequencies of 11 kHz and 14 kHz at 298 K. The 13 C CPMAS spectra were obtained with a recycle delay of 3 s and a contact time of 3 ms unless stated differently. The 13 C ssNMR spectra were referenced with respect to adamantane ( 13 C, 29.456 ppm). The spectra were analyzed using MestReNova (version 14.1.0).
Mass spectrometry data was collected using an Exactive high-resolution MS instrument (Thermo Scientific) equipped with an ESI probe or a DART probe. Pierce™ LTQ ESI Positive/Negative Ion Calibration Solution was used for calibration. Thermo XCalibur software (version) was used for instrument control, data acquisition and data processing. The theoretical mass was calculated with an online calculator (https://www.envipat.eawag.ch/). 1 FT-IR spectra were obtained on a Bruker Tensor 27 spectrometer with platinum attenuated total reflection accessory. The samples were applied as powder on top of the crystal. 64 scans were performed with a resolution of 4 cm -1 .
Scherrer analysis was carried out with Diffrac.Eva (version 5.2.0.5) from Bruker with an instrumental width of 0.050 and a Scherrer constant of 0.89. The peak with the lowest angle was used to carry out the analysis.
Nitrogen adsorption-desorption measurements were performed on a MicroActive for Tristar II Plus 2.01 at 77.350 K. Before the measurement, the samples were outgassed at 120 °C overnight. Surface areas were calculated from the adsorption data using Brunauer-Emmet-Teller (BET) methods and Rouquerol criteria. The pore-size distribution curves were obtained from the adsorption branches using the method "HS-2D-NLDFT, Carb Cyl Pores (ZTC) N2@77K". An optimum between goodness of fit and S4 smoothness of the pore size distribution was aimed for. The average of three different COF batches was used to determine the BET surface areas.
Thermogravimetric analysis was performed on a Perkin Elmer STA 6000. The sample was heated to 30 °C, the temperature was hold for one minute and afterwards the sample was heated with 10.00 °C/min to 700 °C in a nitrogen atmosphere (20 ml/min).

Computation Details
All DFT calculations for COF structures were performed by using the Vienna Ab initio Simulation Package (VASP, version 5.4.4). 2,3 The PBE functional based on the generalized gradient approximation was chosen to account for the exchange-correlation energy. 4 A plane-wave basis set in combination with the projected augmented wave (PAW) method was used to describe the valence electrons and the valence-core interactions, respectively. 5 The kinetic energy cut-off of the plane wave basis set was set to 500 eV. Gaussian smearing of the population of partial occupancies with a width of 0.05 eV was used during iterative diagonalization of the Kohn-Sham Hamiltonian. The threshold for energy convergence for each iteration was set to 10 -5 eV. Geometries were assumed to be converged when forces on each atom were less than 0.05 eV/Å. The Brillouin zone integration and k-point sampling were done with a Gamma centered 1*1*8 and 2*2*4 grid points for the eclipsed and staggered unit cells, respectively. The Van der Waals (vdW) interactions were included by using Grimme's DFT-D3(BJ) method as implemented in VASP. 6 Simulated XRD patterns were obtained by using VESTA (version 3.4.8). 7 Coordinates of all crystal structures are provided as separate files. Individual molecules of 1,3,5triformylbenzene, 2,4,6-trimethylbenzene-1,3,5-tricarbaldehyde, benzidine, 1,4-phenylenediamine and H 2 O were optimized in a periodic box of 20Å 20Å 20Å without thermochemical correction for the × × calculation of the formation energy.

Synthetic Procedures
Building blocks: The synthesis of 2,4,6-trimethyl-benzene-1,3,5-tricarbaldehyde is based on a procedure of Van der Made et al. 9 and Slater et al. 10  Spectroscopic data are in accordance with literature. 10 COF synthesis: The COF synthesis is based on a modified procedure of Smith et al. 11 : General procedure: The aldehyde (1 equiv.) and amine monomers (1.5 equiv.) were added to a 50 mL round bottom flask

Powder X-Ray Diffraction Analysis
All FT-IR spectra indicate that the synthesis was repeatable which is also confirmed by nitrogen sorption measurements. Therefore, only one representative PXRD and ssNMR spectrum of the triplicates will be displayed.

Nitrogen Sorption Analysis
All COFs have been synthesized three times and were then divided in two batches for different activation methods: 1. 120 °C, oven-dried, overnight 2. 120 °C, vacuum-oven-dried, overnight For all samples nitrogen sorption analysis was carried out and the BET surface area has been calculated. The range for the linear regression has been the same for all repetitions of the same COF and for both activation methods and to determine the surface area for the air stability after four weeks. The same DFT model has been used for one COF structure to determine the pore size distribution. All graphs displayed are representative for the respective COF.

Total Energy [kcal/mol]
For the rate constant of the conformation equilibrium, we took the Arrhenius equation (Equation 1) and assumed the same pre-exponential factor A, because TFB and Me 3 TFB are similar in their structure.

Vapor Experiments
COF powder is added into an empty tea bag and the tea bag is closed with tape. The tea bag is added to a flask and hydrochloric acid enriched vapor (argon bubbled through a flask filled with concentrated hydrochloric acid heated to 50 °C) is flushed through the flask for 30 min. Afterwards, the vapor stream is removed, the COF powder taken out of the tea bag and the absorption spectrum is measured.