Capture Fluorocarbon and Chlorofluorocarbon from Air Using DUT‐67 for Safety and Semi‐Quantitative Analysis

Abstract Fluoro‐ and chlorofluorocabons (FC/CFCs) are important refrigerants, solvents, and fluoropolymers in industry while being toxic and carrying high global warming potential. Detection and reclamation of FC/CFCs based on adsorption technology with highly selective adsorbents is important to labor safety and environmental protection. Herein, the study reports an integrated method to combine capture, separation, enrichment, and analysis of representative FC/CFCs (chlorodifluoromethane(R22) and 1,1,1,2‐tetrafluoroethane (R134a)) by using the highly stable and porous Zr‐MOF, DUT‐67. Gas adsorption and breakthrough experiments demonstrate that DUT‐67 has high R22/R134a uptake (124/116 cm3 g−1) and excellent R22/R134a/CO2 separation performance (IAST selectivities of R22/CO2 and R134a/CO2 ranging from 51.4 to 33.3, and 31.1 to 25.8), even in rather low concentration and humid conditions. A semi‐quantitative analysis protocol is set up to analyze the low concentrations of R22/R134a based on the high selective R22/R134a adsorption ability, fast adsorption kinetics, water‐resistant utility, facile regeneration, and excellent recyclability of DUT‐67. In situ single‐crystal X‐ray diffraction, theoretical calculations, and in situ diffuse reflectance infrared Fourier transform spectra have been employed to understand the adsorption mechanism. This work may provide a potential adsorbent for purge and trap technique under room temperature, thus promoting the application of MOFs for VOCs sampling and quantitative analysis.

Solid-state IR spectra were recorded using Nicolet/Nexus-670 FT-IR spectrometer in the region of 4000-400 cm −1 using KBr pellets.Single crystal X-ray diffraction data were collected on an Agilent Technologies SuperNova X-RAY diffractometer system equipped with a Cu sealed tube (λ = 1.54178) at 50 kV and 0.80 mA.Powder X-ray diffraction (PXRD) was carried out with a RigakuSmartLab diffractometer (Bragg-Brentano geometry, Cu Kα1 radiation, λ = 1.54056Å).Variable-temperature-dependent powder X-Ray diffraction date were collected on a RigakuSmartLab diffractometer (Bragg-Brentano geometry, Cu Kα1 radiation, λ = 1.54056Å) under air atmosphere.
In-situ diffuse reflectance infrared Fourier transform (DRIFT) spectra were analyzed on Thermo Fisher Nicolet IS20 spectrometer equipped with a high-sensitive Harric detector as well as in-situ diffuse reflectance cell (Harrick), the MCT detector was cooled down by liquid N2 (77 K).Then the gas cell was cooled to desired reaction temperature for collecting background spectra under N2 environment.collecting sample spectra, background spectra were subtracted, which were collected on 32 co-added scans with 4 cm -1 resolution.Thermogravimetric analyses (TGAs) were performed on a NETZSCH TG209 system in nitrogen and under 1 atm of pressure at a heating rate of 10 °C min -1 .Gas adsorption isotherms for pressures in the range of 0-1.0 bar were obtained by a volumetric method using a Quantachrome autosorb-iQ2-MP gas adsorption analyzer.Adsorption kinetic isotherms were obtained by the Vacuum Vapor/Gas Sorption Analyzer (BSD-VVS).High pressure gas adsorption isotherms for pressures in the range of 0-4.5 bar for 273K, 0-6.3 bar for 298K were obtained by BELSORP-HP high pressure gas adsorption instrument.All gas adsorption measurements were performed using ultra-high purity N2, O2, CO2, R22, and R134a gases.Breakthrough experiments were collected by two different instruments, BSD-MAB (Multi-component Adsorption Breakthrough Curve Analyzer) with mass spectrometer as detector, and a self-built instrument with gas chromatography (FL-9790 plus) as detector.

S2. MOF Synthesis
DUT-67 was synthesized with a modified literature procedure. 1he modulator has been changed from acetic acid to formic acid (FA) and the metal-ligand ratio has been optimized.ZrOCl2•8H2O (100 mg, 0.372 mmol), 2,5thiophenedicarboxylic acid (50 mg, 0.348 mmol), formic acid (5 mL) and DMAC (8 mL) were added into a screw-capped glass jar.The mixture was sonicated for 10 minutes then heated in a 120 ºC oven for 72 h.Cube-shaped colorless crystals appear on the wall of the glass jar.After cooling in air to room temperature, the resulting crystals were filtered and repeatedly washed with DMF.

S3. Single-Crystal X-Ray Crystallography
The single-crystal of DUT-67-R22 or DUT-67-R134a was picked and coated in para tone oil, attached to a glass silk which was inserted in a stainless-steel stick, then transferred to the Agilent Gemini S Ultra CCD Diffractometer with the Enhance X-ray Source of Cu radiation (λ = 1.54178Å) using the ω-ϕ scan technique.All of the structures were solved by direct methods and refined by full-matrix least squares against F 2 using the SHELXL programs. 2Hydrogen atoms were placed in geometrically calculated positions and included in the refinement process using riding model with isotropic thermal parameters: Uiso(H) = 1.2 Ueq(-CH).All the electrons of disordered solvent molecules which cannot be determined, are removed by SQUEEZE routine of PLATON program. 3Crystal data and refinement parameters are listed in Table S7.

Note for DUT-67-R22 Refinement
The R22 molecules are disordered in pores.Thus, the occupancy of F1, F2, Cl1 and H1 were refined as 33.33%,F3 and F4 were refined as 8.33% and 4.17%, Cl2 and H5 were refined as 25% and 4.17%.C4 and C5 were refined as 33.33% and 4.17%.while the occupancy of other set of atoms was 100%.AFIX and DFIX were used to restrain the atoms.ISOR, DELU and SIMU were used to restrain the ADP refinement.
The 97 restraints caused 149 refine parameters.The large pores contain highly disordered gas molecules which cannot be determined.SQUEEZE treatment was applied and the squeezed void volume is 29964 Å 3 , equivalent to 51.7% of the unit cell.
The R1 value is 0.113 without SQUEEZE treatment and 0.0732 with SQUEEZE treatment.

Note for DUT-67-R134a Refinement
The R134a molecules are disordered in pores.Thus, the occupancy of F1, F2, F3, F4, C4, C5, H5A and H5B were refined as 16.67%.while the occupancy of other set of atoms was 100%.AFIX, DFIX and SADI were used to restrain the atoms.ISOR, DELU and SIMU were used to restrain the ADP refinement.The 118 restraints caused 140 refine parameters.The large pores contain highly disordered gas molecules which cannot be determined.SQUEEZE treatment was applied and the squeezed void volume is 31360 Å 3 , equivalent to 53.5% of the unit cell.The R1 value is 0.1087 without SQUEEZE treatment and 0.0665 with SQUEEZE treatment.Fitting error: 0.377 %.

S5. Adsorption Kinetics
The adsorption kinetics was analyzed by simplified kinetic models such as the pseudo first order and pseudo second order, through the following two equations.The calculation and fitting results show that the linear correlation coefficient (R 2 ) of the pseudo first order is very low, so we choose the pseudo-second-order mode to calculate the sorption rate constant of R22 and R134a.Where qe (mg g -1 ) and qt (mg g -1 ) are the quantity of the R22 and R134a adsorbed at equilibrium and at t time, respectively, and the k2[g (mg min) -1 ] is the pseudo second order sorption rate constant that is deduced from the slope of the plot of t/qt versus t. qe, exp qe, cal k2 R 2 (cm 3 g -1 ) (cm 3 g -1 ) (g•mg -1 min -1 ) (cm 3 g -1 ) (cm 3 g -1 g - 1 ) (g•mg

Limit of detection calculation (LOD):
The LOD was inferred by eq 1. the LOD is defined as 3 times the standard deviation (3σ) of the zero determinations, in which the values of the sensitivity are 56869 and 56752 for R22 and R134a (see Figure S17 a and Figure S17a) [15] .
The LOD of this method for R22 and R134a were estimated to be 1.533 × 10 -5 mL and 4.772 × 10 -5 mL according to Figure S16 a and Figure S17a.

In-Situ Diffuse Reflectance Infrared Fourier Transform (DRIFT) Spectra
In-situ DRIFT spectra were collected on Thermo Fisher Nicolet 6700 spectrometer with a high-sensitive MCT detector, which was cooled down by liquid N2 (77 K).
Activated DUT-67 was placed into an IR gas cell (PIKE), heated at 100 o C in a heated rate of 2 o C min -1 while purge with high purity N2 gas (50 mL min -1 ) 40 min to remove adsorbed water on the surface of samples.Then the gas cell was cooled down to the required temperature for collecting background in high purity N2 atmosphere.
Before collecting sample spectra, background spectra were subtracted, which were collected on 32 co-added scans with 4 cm -1 resolution.Then pure R22 and R134a were injected into the sample cell for tests, respectively.

Figure S2 .
Figure S2.Thermogravimetric analyses (TGAs) of the fresh DUT-67 in nitrogen and under 1 bar of pressure at a heating rate of 10 °C min -1 .

Figure S3 .Figure S4 .
Figure S3.The variable-temperature PXRD patterns of DUT-67 in air and under 1 bar of pressure.

Figure S14 .Figure S15 .
Figure S14.R134a virial fitting (lines) of the adsorption isotherms (points) of DUT-67 measured at 273 and 298 K.S7.Breakthrough ExperimentsTransient breakthrough experiments for the separation of R22/R134a/air and R22/R134a/air (1:1:98, v/v/v), R22/R134a/air (10:10:80, v/v/v) were carried out in a fixed bed.The flow rates of gases were regulated by mass flow controllers.The column (6 mm inner diameter × 150 mm) contained 1.0 g of pre-activated sample for the experiment using a binary /ternary component.Before filled in the column, the samples were activated at 333 K for 12 hours under vacuum conditions.After filling the column, the column was purged with a He flow (30 mL min -1 at 298 K and 1 bar) for 2 h.Then the gas mixture was introduced to the column.The outlet composition was continuously monitored by a mass spectrometry until a complete breakthrough was achieved.The sample was regenerated with a He flow (30 mL min -1 ) at 373 K until all gases signal disappeared before each cyclic experiment.For the separation of R22/R134a/air (0.1:0.1:99.8,v/v/v) and (0.002:0.002:99.996,v/v/v) the outlet composition was continuously monitored by a gas chromatograph (FULI GC9790 Plus) until a complete breakthrough was achieved.

Figure S18 .
Figure S18.The working curve of R134a.a) R134 injection volume vs corresponding peak area in GC, b) R134 peak area in GC vs R134 injection volume (in order to directly calculate the volume of R134a in an unknown sample).

Table S1
Summary of reported MOFs and molecule sieves with BET surface areas, R22/R134a gas uptake capacities, and adsorption enthalpy of R22 and R134a.

Table S2 .
Kinetic parameters of quasi-second order dynamics of R22 and R134a adsorption by DUT-67.

Table S4 .
Relative peak area records and calculations of R22 concentrations in the enriched

Table S6 .
Summary of the semi-quantitative analysis results for R22 and R134a.

Table S8 .
Summary of DFT calculation results.