MOF/polymer hybrids through in situ free radical polymerization in metal-organic frameworks

We use the free radical polymerization initiator 4,4′-azobis(cyanovaleric acid) coordinated to the open metal sites of metal-organic frameworks (MOFs) to give rise to highly uniform MOF/polymer hybrids. We demonstrate this strategy on two robust zirconium MOFs (NU-1000 and MOF-808), which are the most effective catalysts for degradation of chemical warfare nerve agents. The resulting hybrid materials maintain their hydrolytic catalytic activity and have substantially improved adhesion to polypropylene and activated carbon textile fibers, yielding highly robust MOF/polymer/textile hybrid systems. These composites are suitable for the green production of active protective clothing and filters capable of detoxifying organophosphorus warfare agents.


S2. Instrumentation
Powder X-ray diffraction (PXRD) data were collected on a Bruker D8 ADVANCE diffractometer equipped with a copper lamp (CuKα radiation, λ = 1.5406 Å) at 30 kV and 40 mA with a slit of 0.1°. Standard measurements were done in 2θ range of 4°-40° with a 2θ step of 0.008° and a counting time of 0.5 s. 1 H nuclear magnetic resonance (NMR) spectra in solutions were recorded using a Bruker Avance 500 spectrometer at 298 K and were calibrated on the residual solvent signal (DMSO-d6: 2.50 ppm, CDCl3: 7.26 ppm). The NMR samples of obtained MOFs before and after functionalization were prepared by digesting approximately 1 mg of dried material in D2SO4 and then diluting it with 0.6 ml of DMSO-d6. Solid-state MAS NMR 1 H-13 C CP MAS spectra were recorded at room temperature on a Bruker Avance III 11.7 T (125.75 MHz), Bruker Avance Neo 14 T (150.75 MHz), and Bruker Avance Neo 20 T (213.79 MHz) spectrometers equipped with 3.2 mm CPMAS probes and referenced to solid adamantane (38.48 ppm for the CH2 signal) 2 . 55-80 kHz 1 H decoupling was used. Further experimental details are given in Table S1 (section S6). Diffuse reflectance infrared Fourier transform (DRIFT) spectra were collected on Nicolet iS50 FT-IR Spectrometer (Thermo Scientific) with a Praying Mantis DRIFT accessory. The spectra were collected in a 4000-400 cm -1 range with number of scans set to 128. Samples were prepared under air atmosphere by grinding in a mortar with KBr and then placed under N2 purge for collection time. Variable-temperature DRIFTS were collected using high temperature reaction chamber (Harrick Scientific Products Inc), accessory with temperature control performed with EZ-ZONE software under the N2 atmosphere. Nitrogen sorption isotherms were measured at 77 K on a Micromeritics ASAP 2020. Prior to the measurements, the samples were degassed at 120 °C for 24 h (the initiator@MOF samples were activated at 30 °C for 24 h). The Brunauer-Emmett-Teller (BET) theory was used to calculate the specific surface areas of obtained materials based on N2 sorption measurements for collected samples. For all N2 isotherm analyses we ensured that the two consistency criteria described by Roquerol et al. 3 and Walton et al. 4 were satisfied. The pore sized distribution plots were derived from sorption data by DFT calculations using a carbon slit pore model with a N2 kernel. Scanning electron microscopy (SEM) images were collected on a Hitachi S-3400N-II variable-pressure scanning electron microscope. Samples were sputter-coated with 7 nm Au to facilitate viewing by SEM. Energy dispersive X-ray spectra (EDS) were obtained using an EDS Thermo Scientific Ultra Dry system. Optical emission spectrometry with excitation by argon inductively coupled plasma (ICP-OES) was performed on the Thermo Fisher Scientific iCAP 7400 DUO instrument. The analyzed samples were dissolved in 1 ml of piranha solution (H2SO4/H2O2, v/v = 3/1) and diluted with distilled water. Thermogravimetric analyses (TGA) were recorded on a Setaram SETSYS 16/18 instrument. Samples for thermogravimetric characterization were placed in alumina crucibles in synthetic air (O2:N2 = 20:80) (flow rate: 1 dm 3 h) at heating rate 5 ºCmin, samples were studied between 30 and 1000 ºC. Matrix-assisted laser desorption/ionisation mass spectrometry (MALDI-MS) spectra were acquired in linear mode on JEOL JMS-S3000 SpiralTOF™-plus Ultra-High Mass Resolution MALDI-TOFMS. The 2 mg of NU-1000/PMMA was soaked in K3PO4 solution for 24 hours (to decompose the MOF network), the obtained residue was then dispersed in THF. The collected sample was then mixed with 2,5-dihydroxybenzoic acid (DHB, dissolved in THF, used as a matrix).

S3.1. Synthesis of MOF materials
NU-1000: The material was synthesized according to the procedure described by Islamoglu et al. where benzoic acid and trifluoroacetic acid are used as co-modulators. 5 MOF-808: The material was synthesized following the literature procedure with some minor modifications. 6 Typically, 1.08 g of zirconium oxide dichloride octahydrate (ZrOCl2 × 8H2O, 3.35 mmol) and 0.745 g of 1,3,5-benzenetricarboxylic acid (H3BTC, 3.55 mmol) were mixed in 32 ml of DMF. Then, 66 ml of formic acid was added to the reaction mixture, and sonicated for 10 min. The sealed reaction vessel was heated at 120 °C for 24 h. After that time, the white solid was formed. The product was centrifuged and washed thoroughly with DMF. Then the DMF was exchanged to acetone and soaked in that solvent overnight. The product was washed additionally with acetone (2 x 30 ml) and dried at 80 °C overnight. The obtained Zr-MOFs were characterized by PXRD, 1 H NMR, DRIFTS, TGA and N2 sorption measurements. Figure S1. Structures of NU-1000 (left) and MOF-808 (right) materials with the 8-connected and 6connected Zr6-nodes, respectively. S-5
initiator@NU-1000: 250 mg of 4,4'-azobis(4-cyanovaleric acid) (0.891 mmol) was dissolved in 20 ml of ethanol in a falcon tube. Then, 200 mg (0.092 mmol) of NU-1000 was added to the solution. The reaction was shaken at 25 °C (600 rpm, Eppendorf ThermoMixer) for 72 h. After that time, the solids were centrifuged, soaked in a fresh portion of ethanol (40 ml) and kept for another 24 h (25 °C, 600 rpm). Then the solid was washed with ethanol (3 × 30 ml) and acetone (2 × 30 ml) with short incubation (approx. 1 h) between washings to afford complete removal of the non-coordinated ACPA. The resulting solid was dried at 25 °C for 24 h under vacuum.
initiator@MOF-808: The SALI procedure was analogous to the described above for initiator@NU-1000. Typically, 100 mg of MOF-808 (0.074 mmol), 200 mg of ACPA (0.714 mmol) and 10 ml of ethanol were used for one reaction.
The obtained initiator@MOF samples were characterized by PXRD, DRFITS, TGA and N2 sorption measurements. The number of incorporated ACPA molecules was determined by 1 H NMR analysis of the initiator@MOF samples digested in a D2SO4/DMSO-d6 mixture.

S3.3. Preparation of MOF/polymer hybrids
MOF/PMMA: Typically, 100 mg of initiator@MOF was placed in a Schlenk flask and degassed. Then, 4 ml of methyl methacrylate (MMA, 37.5 mmol) was added under N2. The reaction vessel was sealed, and the obtained suspension was stirred at room temperature for 30 minutes. Then, the reaction was stirred at 70 °C for 48 h. The obtained product was thoroughly washed with methanol (to remove the excess of monomer). After that, the MOF/PMMA material was stirred in 100 ml of acetone at room temperature for at least 1 h. The obtained suspension was centrifuged (600 rpm, 15 min) and the remaining solid was washed several times with acetone. The final product was dried at 60 °C overnight.

MOF/PDMAM:
Typically, 100 mg of initiator@MOF was placed in the Schlenk flask and degassed. Then, 5 ml of 2-dimethylaminoethyl methacrylate (DMAM, 29.6 mmol) was added under N2. The reaction vessel was sealed, and the obtained suspension was stirred at room temperature for 30 minutes to afford uniform dispersion of monomer in the pores of the MOF. Then, the reaction was stirred at 50 °C for 48 h. The obtained products then undergo the same washing procedure with methanol and acetone, as described above for MOF/PMMA.

S3.4. Fabrics coating procedure
MOF/polymer/fiber composites were prepared by a wet deposition in THF. A suspension of MOF/PMMA or MOF/PDMAM in THF was added dropwise over the surface of polypropylene (PP) or activated carbon (AC) fabrics. Once the suspension homogeneously coated the fabric, it was dried at 70 °C for 15 min. Afterwards the fabric was turned over and a subsequent deposition carried out on the other side of the fabric, thus completing one cycle. A coloration in the fabric was visible in some cases (Fig. S2). The amount of Zr %wt on the fabric was determined by ICP-OES.

S5.1. NMR spectra of ACPA and its stability studies
The 4,4'-azobis(4-cyanovaleric acid) (ACPA) contains two stereogenic centers and forms different stereoisomers. Commercially available sources of ACPA typically contain a mixture of rac-and meso-diastereoisomers (Fig. S7) 8 .  8 collected for the ACPA sodium salt, we assigned the singlet at 1.74 ppm to the protons of the methyl group of meso-ACPA and the singlet at 1.68 ppm to its racemic form (rac-ACPA) in the sample measured in deuterated water (Fig. S8a). It is worth noting, that the addition of deuterated sulfuric acid shifts the ACPA peaks to lower frequencies, which we attributed to the possible protonation of the nitrogen atoms of ACPA (Fig. S8c).

S-9
The 1 H NMR spectrum of ACPA in its native form was compared with the spectrum of the ACPA molecule after heating it at 70 °C for 24 hours. On the recorded spectrum (Fig. S9b), new sets of signals were observed which can be assigned to the different decomposition products of ACPA initiator 8 .      S-14 S7. DRIFT spectra S7.1. DRIFT spectra of the ACPA initiator and its thermal stability studies Figure S17. The DRIFT spectra of ACPA measured at 70 °C for 12 h (spectra recorded every 1 hour). Black curve represents the starting sample of ACPA, the red curve represents the sample after incubation at 70 °C for 12 h. The observed increase/decrease of selected bands over time is annotated with arrows. Inset shows a normalized intensity of the ν(C≡N) band vibration region. Figure S18. The VT-DRIFT analysis of ACPA, the spectra were measured in the temperature range of 30 -100 °C (spectra were collected every 10 °C, the heating rate was 5 °C/min, prior to measurement the sample was incubated at given temperature for 5 min). The observed changes of the intensities of selected bands are annotated. The inset shows a normalized intensity of the ν(C≡N) band region.

S7.2. DRIFT spectra of initiator@MOF samples
For initiator@NU-1000 the increase of the intensity of the band at 2250 cm -1 was observed during heating, and the shift to 2238 cm -1 of the respective band was observed, which can be assigned to the C≡N group vibration of the decomposition products of ACPA initiator. Figure S20. Normalized DRIFT spectra of MOF-808 and initiator@MOF-808: a) pristine MOF-808, b) initiator@MOF-808 dried at room temp., c) initiator@MOF-808 after heating at 120 °C for 24 h (initiator@MOF-808 ΔT ). The inset shows the zoomed spectra in the range of 2400 -1850 cm -1 (C≡N vibrations). For a better clarity, the spectra were normalized with respect to the band at 1900 cm -1 .
For initiator@MOF-808 the increase of the intensity of the band at 2245 cm -1 was observed during heating, and the shift to 2240 cm -1 of the respective band was observed, which can be assigned to the C≡N group vibration of the decomposition products of ACPA initiator.  S-22

S8.1. N2 sorption isotherms
The Brunauer-Emmett-Teller (BET) theory was used to calculate the specific surface areas of obtained materials. For all isotherm analyses we ensured that the consistency criteria described by Roquerol et al. 3 and Walton et al. 4 were satisfied. The obtained BET surfaces areas are collected in Table S2. The pore size distribution (PSD) plots were derived from sorption data by the DFT method using a carbon slit pore model with the N2 kernel.       S-27

S11. TGA-DTG analysis
Thermogravimetric analysis (TGA) was conducted in the range of 30-1000 °C under oxidizing conditions (O2/N2 = 20/80) with a heating rate of 5 °C/min, giving as the only final product zirconium oxide (ZrO2) 9 . The weight loss at the beginning of all of the TGA curves corresponds to the removal of solvent molecules (up to ca. 200 °C). The decomposition of the parent MOF occurs in the temperature range of 400-500 °C. The TGA curves presented on Figures S57-S59 were normalized in respect to the residual ZrO2 mass (100%) 9 .  Table S4).  S-40

S12. Composition of MOF/polymer hybrids from TGA and ICP-OES
The quantity of Zr (wt%) in the obtained Zr-based MOF materials and MOF/polymer hybrids was calculated from TGA (see section S11). The Zr content (wt%) in obtained materials was calculated using equation (1): where: m 900 ℃ -mass from TGA analysis at 900 °C (corresponds to ZrO2); m 190 ℃ -mass from TGA analysis at 190 °C (corresponds to solvent-free material); M Zr -molar mass of Zr (91.22 g/mol); M ZrO 2 -molar mass of ZrO2 (123.22 g/mol).
The calculated values are collected in Table S4 and are in good agreement with theoretical values. The Zr content measured by ICP-OES (Table S4) is typically lower, due to the presence of solvent trapped in the pores of the MOF materials. The amount of polymer (wt%) in MOF/polymer hybrids was estimated from the normalized TGA curves (see section S11). The weight difference (%) between initiator@MOF and MOF/polymer material at 190 °C (Figures S58-S59) corresponds to the weight of the polymer. The amount of polymer (wt%) was calculated using equation (2)

S13. Reproducibility tests of the FRaP-in-MOF protocol
To confirm the reproducibility of the FRaP-in-MOF procedure, four batches of NU-1000/PMMA were prepared and characterized with SEM (Fig. S60) and N2 sorption studies (Fig. S61). Additionally, the TGA (Fig. S62) and ICP-OES measurements were made for quantitative analysis, and the calculated NU-1000/PMMA composition is collected in Table S6.

S14. Preparation of MOF/polymer hybrids -control experiments
For control experiments 1 ml of respective monomers (9.4 mmol of MMA, 5.9 mmol of DMAM) were stirred at 70 °C (for MMA) or 50 °C (for DMAM) for 48 hours. No visible changes of the reaction mixture were observed after that time, and the remained clear solution was analyzed with 1 H NMR analysis confirming presence of only unreacted monomers (Fig. S63, S65).
In the same reaction conditions, 2 ml of MMA (18.8 mmol) or DMAM (11.9 mmol) were stirred in the presence of 50 mg of pristine NU-1000 or MOF-808. The collected 1 H NMR spectra ( Fig. S64 and S66) of the supernatant of the reaction mixture confirm Zr-MOF inactivity in the polymerization reaction.

S-45
The reference sample of NU-1000/PMMA* before washing procedure with acetone (removal of excess of PMMA polymer chains) was analyzed with TGA and SEM analysis (Fig. S67). The determined composition of NU-1000/PMMA* sample (calculations were made following the equations (1) and (2)) was 2.9 wt% of Zr and 86 wt% of PMMA polymer. Figure S67. TGA profile (a) and SEM image (b) of NU-1000/PMMA before washing with acetone.

S15. Stability of the MOF/polymer/fiber composites
A simple water soaking test was designed to assess the stability of MOF/polymer/fiber composites. A 1x0.5 cm piece of the fabric was suspended on 2 mL of water and stirred by hand. Then, the supernatant was removed, and this process was repeated 2 more times. After that, the MOF/polymer/fiber samples were dried at 100 °C overnight. The performance of MOF/polymer/fiber composites was compared with the sample of MOF/fiber composites where instantaneous detachment of MOF was observed ( Figure S68b, S69b). The Zr content before and after water treatment was followed using ICP-OES analysis (Tables S7-S8).    Figure S69. a) Zr loss determined with ICP-OES analysis of NU-1000/polymer/fiber composites before and after water treatment (the estimated error of that analysis is 10%). b) Photographs of respective NU-1000/polymer/fiber composites immersed in 2 ml of water.

S16.1. Reaction of DIFP with MOF/polymer hybrids
The diisopropylfluorophosphate (DIFP) hydrolysis was tested in the presence of pristine Zr-MOFs or MOF/polymer hybrids. Typically, the MOF or MOF/polymer sample was placed in the vial with 0.5 ml of water and 1.25 μL of dimethylacetamide (DMA, internal standard). Then, 1.25 μL of DIFP (7.2 μmol) was added and the DIFP conversion in time was monitored using GC analysis. The amount of Zr-MOF or MOF/polymer hybrid was fixed in respect to the DIFP:MOF ratio of 2:1. The reactions were carried out at room temperature for 24 hours and the collected data are presented on Figure S70 and Table S9.

S16.2. Reaction of DIFP with MOF/polymer/fiber composites
The reactions of DIFP hydrolysis in the presence of MOF/polymer/fiber composites were carried out at 30 °C (for MOF/polymer/PP samples) or 60 °C (for MOF/polymer/AC samples) for 24 hours. For each reaction, pieces of MOF/polymer/fiber with dimensions of 6 cm × 2 cm were cut and weighted in a glass vial. Then, 2 μL of water was added, followed by 0.3 μL of DIFP (1.7 μmol) and the vial was sealed tightly. Depending on the Zr content in the MOF/polymer/fiber composites (determined by ICP-OES analysis, Tables S7-S8), the DIFP:MOF molar ratios was approximately 15:1 (for NU-1000/polymer/fiber) and 5:1 (for MOF-808/polymer/fiber). After 24 hours of incubation at 30 °C or 60 °C, the reagents were extracted with a solution of 2 ml of dichloromethane and 0.3 μL of dimethylacetamide (DMA, internal standard). This extraction was done for 2 h and at the same temperature as the catalysis was carried out. The DIFP conversion after 24 h was determined with GC analysis and the obtained results are presented in Tables S10-S11.