Dinuclear Copper Sulfate-Based Square Lattice Topology Network with High Alkyne Selectivity

Porous coordination networks (PCNs) sustained by inorganic anions that serve as linker ligands can offer high selectivity toward specific gases or vapors in gas mixtures. Such inorganic anions are best exemplified by electron-rich fluorinated anions, e.g., SiF62–, TiF62–, and NbOF52–, although sulfate anions have recently been highlighted as inexpensive and earth-friendly alternatives. Herein, we report the use of a rare copper sulfate dimer molecular building block to generate two square lattice, sql, coordination networks which can be prepared via solvent layering or slurrying, CuSO4(1,4-bib)1.5, 1, (1,4-bib = 1,4-bisimidazole benzene) and CuSO4(1,4-bin)1.5, 2, (1,4-bin = 1,4-bisimidazole naphthalene). Variable-temperature SCXRD and PXRD experiments revealed that both sql networks underwent reversible structural transformations due to linker rotations or internetwork displacements. Gas sorption studies conducted upon the narrow-pore phase of CuSO4(1,4-bin)1.5, 2np, found a high calculated 1:99 selectivity for C2H2 over C2H4 (33.01) and CO2 (15.18), as well as strong breakthrough performance. Across-the-board, C3H4 selectivity vs C3H6, CO2, and C3H8 was also observed. Sulfate-based PCNs, although still understudied, appear increasingly likely to offer utility in gas and vapor separations.

Synthesis of Cu 2 (SO 4 ) 2 (pyridine) 6 CuSO4 .5H2O (300 mg, 1.2 mmol) is stirred and heated in a capped vial to 80 °C for 24 h in 10 mL Pyridine.The solution is then allowed to cool to room temperature, then filtered under reduced pressure.Characterization data from PXRD match with a simulated PXRD of the reported material (Refcode: CEVJUL). 2 Yield: 298.6 mg, 48% yield.
Synthesis of closed phase of CuSO 4 (1,4-bib) 1.5 , 1cp The closed phase, 1cp, was formed by heating the sample to 100 °C under high vacuum or under N2.However, given the water vapor uptake at low relative humidity the sample must be maintained under a dry atmosphere to maintain the closed phase.Single crystals of 1cp could be formed from 1op by heating the crystal to 70 °C for 1h under a flow of N2.
Synthesis of 1,4-bis(1-imidazolyl)naphthalene (1,4-bin)   A mixture of 1,4-dibromonaphthalene (10.0 g, 35.0 mmol, 1 eq), imidazole (14.3 g, 209.8 mmol, 8.3 eq), K2CO3 (14.5 g, 104.9 mmol, 3 eq), and CuSO4 (111.6 mg, 0.7 mmol, 0.02 eq) were stirred at 180 °C for 24 h in a round-bottom flask (100 mL) under N2 atmosphere.After cooling down to room temperature, the crude product was washed with H2O (50 mL × 3).The solid residue was then dissolved in MeOH then filtered then the resultant filtrate was concentrated to give 1,4-di(1H-imidazol-1-yl)naphthalene as a light yellow solid (7.3 g, yield: 80.2%).Layering: In a test tube, a solution of CuSO4 .5H2O (25.0 mg, 0.1 mmol, 1 eq) is dissolved in 2 mL H2O.Above this, a buffer layer of 4 mL 1:1 (v/v) MeOH:Ethylene glycol solution is slowly added.A final layer is then slowly added of 1,4-bis(1-imidazolyl)naphthalene (26.0 mg, 0.1 mmol, 1 eq) in 2 mL MeOH and left to stand for ca. 1 month.The product is extracted as a dark blue powder that deposits as a layer towards the middle of the test tube which is extracted, filtered under reduced pressure, and washed with MeOH.Solvent exchange with MeOH is then performed by adding 1 mL MeOH before decanting and replacing with fresh MeOH three times per day over the course of 3 days.Yield: 15.7 mg activated, 43% yield. 2 samples made through this layering method were, unless otherwise stated, used for all further experiments.Slurrying: Alternatively, CuSO4 .5H2O (1.25 g, 5.0 mmol, 0.66 eq) and 1,4-bis(1-imidazolyl)naphthalene (1.955 g, 7.5 mmol, 1 eq) is added to 250 mL MeOH and stirred for 4 days.The resultant product is filtered and washed with 20 mL MeOH then left to dry in air yielding a light blue powder.Yield: 2.576 g activated, 94% yield.In a test tube, a solution of Cu2(SO4)2(pyridine)6 (20.0 mg, 0.025 mmol, 1 eq) is dissolved in 2 mL distilled water H2O.Above this, a layer of 4 mL 1:1 (v/v) MeOH:H2O solution is slowly added.This is carefully layered with a solution of 1,4-bis(1-imidazolyl)naphthalene (6.5 mg, 0.025 mmol, 1 eq) in 2 mL MeOH and left to stand for ca. 1 week.From this, large dark blue wedge-shaped crystals that were suitable for single-crystal X-ray diffraction analysis are collected.
Synthesis of narrow pore phase of CuSO 4 (1,4-bin) 1.5 , 2np The closed phase was formed by heating 2op to 60 °C under atmospheric conditions, or exposure to low humidity (<20% RH) air, high vacuum or under N2.Single crystals of 2np could be formed from 2op by placing the crystal at ca. 275 K for 1h under a flow of N2 (or in one instance at 250 K for 1h under a flow of N2), however careful control is needed to prevent the single crystal from cracking extensively during the phase change.

Powder X-ray diffraction measurements
Powder X-Ray Diffraction (PXRD) under ambient conditions PXRD experiments were conducted using microcrystalline samples on a PANalytical Empyrean diffractometer (40 kV; 40 mA; Cu Kα1,2 λ = 1.5418Å) in Bragg-Brentano geometry.The sample was exposed for 99.5 s/step, with a step size of 0.0131303° in 2θ, at room temperature with a range of 2°-50°.Powder samples were evenly distributed on a zero-background holder after being lightly ground to minimize the effects of preferred orientation.Data analysis was carried out using X'Pert HighScore Plus 3 (version 2.2e).Powder patterns were calculated from SCXRD structures using Mercury (v.2023.1.0). 4 In-situ Variable Temperature Powder X-Ray Diffraction (VTPXRD) Diffractograms at different temperature were recorded using a PANalytical X'Pert Pro-MPD diffractometer equipped with a PIXcel3D detector operating in scanning line detector mode with an active length of 4 utilizing 255 channels.An Anton Paar TTK 450 stage coupled with the Anton Paar TCU 110 Temperature Control Unit was used to record the variable temperature diffractograms.The diffractometer is outfitted with an Empyrean Cu LFF (long fine-focus) HR (9430 033 7300x) tube operated at 60 kV and 60 mA and Cu Kα radiation (λα = 1.5418Å) was used for diffraction experiments.Continuous scanning mode with the goniometer in the theta-theta orientation was used to collect the data.Incident beam optics included the Fixed Divergences slit, with a 1/4° divergence slit and a Soller slit (0.04 rad).Divergent beam optics included a P7.5 S7 anti-scatter slit, a Soller slit (0.04 rad), and a Ni-β filter.Ca. 20 mg of freshly synthesized 1 or 2 was ground into a fine powder, and then loaded onto a zero-background sample holder made for an Anton Paar TTK 450 chamber.The data was collected from 4°-40° (2θ) with a step-size of 0.0167113° and a scan time of 200 seconds per step.Crude data were analyzed using the X'Pert HighScore Plus™ software V 4.1 (PANalytical, The Netherlands).PXRD was performed first in ambient air, then in dry N2 flow before the sample was heated, up to 473 K for 1 or up to 393 K for 2. The samples were equilibrated for 10 min at each new temperature step (or after N2 flow begins) before data collection.The samples were then cooled to room temperature, still under N2 flow, then measurements were performed, then the TTK 450 stage was opened to allow the sample to be exposed to ambient air for 10 min, followed by a final measurement.

Single-Crystal X-ray Data Collection and Structure Determination
For 1, two suitably large plate crystals (dimensions: Crystal 1: 0.38 x 0.295 x 0.075 mm or Crystal 2: 0.19 x 0.168 x 0.02 mm) were selected for single-crystal analysis.Using a nitrogen-flow Oxford cryostream attachment, the first crystal was cooled to 100 K at a rate of 360 K/h from room temperature.A full data set of 1op was collected at this temperature.The crystal was then heated to room temperature (~299 K) at a rate of 360 K/h and a full data set was collected.The crystal was then heated to 333 K at a rate of 60 K/h and then held at this temperature for 1 h.A full data set was then collected at this temperature.A second crystal was heated from room temperature to 343 K at 60 K/h and held at this temperature for 1 h to form 1cp. A short data collection at this point resulted in a diffraction pattern with sharply reduced diffraction intensity and a smaller unit cell indicating the phase change.A full data collection was repeated at this temperature from which the closed-phase crystal structure was determined.
For 2, a suitably large crystal (0.3 × 0.15 × 0.05 mm) was selected for single crystal refinement.Using a nitrogen-flow Oxford cryostream attachment, the crystal was cooled to 100 K at a rate of 360 K/h from room temperature.A full data set of 2op was collected at this temperature.A second large crystal (0.42 × 0.324 × 0.194 mm) was heated from 100 K to 250 K at a rate of 360 K/h, and then steadily raised to 276 K at a rate of 30 K/h until signs of phase transition to 2np became visible (discoloration and cracking).Note, in the vast majority of cases, this phase transition resulted in crystals too disfigured for SCXRD.The crystal was held at this temperature for 30 min, then heated to 333 K at a rate of 240 K/h, then to 373 K at a rate of 240 K/h to ensure full desolvation, then cooled to 100 K at a rate of 360 K/h.Data collection was performed at this temperature from which the activated crystal structure was determined.A third large crystal of 2op (0.69 x 0.25 x 0.17 mm) was subjected to data collection at room temperature.To prevent conversion to 2np through exposure to dry air, the nitrogen flow Oxford cryostream attachment was turned off, thus inhibiting thermal control.A fourth large crystal of 2op (0.306 x 0.27 x 0.15 mm) was placed on the preheated diffractometer for 1 h at 250 K using the nitrogen flow Oxford cryostream attachment.Minor cracking and discoloration indicated conversion to 2np, which was confirmed through unit cell determination.The crystal was heated to 298 K at a rate of 60 K/h and a data set was collected.
Data collection was performed with a Mo Kα (λ = 0.71073 Å) radiation source on a Bruker D8 Quest fixed-chi diffractometer equipped with a Bruker APEX-II CCD detector.Data was indexed, integrated, and scaled in APEX4. 5Absorption corrections were performed by the multi-scan method using SADABS 6 .Space groups were determined using XPREP 7 as implemented in APEX4.The SHELX-2014 program package, implemented in OLEX2 v1.5, 8 was used for structure solution and refinement.Structures were solved using the intrinsic phasing method (SHELXT) 9 and refined with SHELXL 10 using the least-squares method.All non-hydrogen atoms were refined anisotropically.Hydrogen atoms were placed in calculated positions from the molecular geometry and assigned isotropic thermal parameters that depended on the equivalent displacement parameters of their carriers.Selected crystallographic data and refinement parameters for the crystal structures are given in Tables S3 and  S5.

Specific Refinement Details -1
As 1 is heated from 100 K to 333 K, the open phase is retained, however water molecules in the pore (O5, O6, and O7) occur with progressively lower occupancy or are no longer present entirely.
For 1op at RT, this results in a low occupancy for O6 which can only be modelled isotropically, without bonded hydrogen, and with large Uiso value.As such the position of this atom is inaccurate resulting in an unusually close O3-O6 contact.
For 1op at 333 K, when the OLEX2 implementation of SQUEEZE 11 is used to account for the remaining electron density in the pore only 4 electrons were found, which could only account for ca.20% of a single water molecule per Asymmetric Unit.A sufficient model was therefore chosen without SQUEEZE wherein the pore was modelled as empty, which is supported by Dynamic Vapor Sorption (DVS) experiments which show the weakly-bound water is removed at 298 K under dry air, and so is conceivably not present in the single crystal at 333 K under a dry N2 flow.Response: As the structure is collected at 333 K, this void space is expected to be empty as is confirmed by thermogravimetric analysis.

Specific Refinement Details -2
For 2, both 2op and 2np displayed disorder in the naphthalene and sulfate region, with 2np also showing disorder in the imidazole region of the structure.
For 2op at 100 K, disorder in the naphthalene arises from the naphthalene of the single-ligand wall crystallizing onto a center of inversion.Additionally, solvent water molecules could not be adequately resolved and so were largely accounted for using the OLEX2 implementation of the SQUEEZE 11 routine, which found 46 electrons in the pore which approximately accounts for 4.33 water molecules.Water molecules based on O5 and O6 could not be handled with SQUEEZE due to being co-located with disordered regions of naphthalene, and so were each individually resolved with a fixed 0.33 occupancy.Disorder in the sulfate group is evident from the presence of Q-peaks with intensities of 1.6, 1.3, and 1.0 adjacent to sulfate atoms.Splitting the sulfate group into the new positions with 14.5% freely refined occupancy, as indicated by the Q-peaks, resulted in the R1 dropping from 4.21% to 3.40%.Disorder of this kind in the sulfate group is potentially supported by the existence of two known structural variants of the copper sulfate dimer MBB in the CSD wherein the Cu-O2 distance (for more information see section 2) varies between ~2.6-2.9Å and ~3.0-3.3Å.In this case, the major and minor components have Cu-O2 distances of 2.7607(17) Å and 3.170(15) Å, respectively, and so would appear to meet this trend.However, it should be noted that due to the low occupancy of the minor component, the measured distances are less reliable than would be indicated by estimated standard deviation.
For 2op at RT, similar disorder was observed in the naphthalene, sulfate, and O5/O6 water molecules of the structure.Additional pore solvent water molecules were accounted for with SQUEEZE which found 48 electrons in the pore approximately accounting for 4.33 water molecules.Disorder in the sulfate region was resolved with 14.4 % freely refined occupancy.For 2np at 100 K, cracking of the crystal upon in-situ desolvation by exposure to dry nitrogen in the cryostream results in severely reduced data quality.Attempts to form the activated phase crystals through other means, such as by gradually reducing relative humidity in a dynamic vapor sorption instrument, or exposure to reduced atmospheric pressure, or by heating, failed to yield suitable crystals.As with the as-synthesized phase, the structure presents with disorder about the naphthalene group of the single-ligand wall which crystallizes onto a center of inversion.Additionally, Q-peaks surrounding the imidazole group of the single-wall ligand and the sulfate indicate the presence of disorder in these groups, with the imidazole group rotated 88.46 (17)° between disordered forms, and the sulfate group changing from a monoatomic bridging group (μ2-η 2 ) to a diatomic bridging group (μ2-η 1 :η 1 ).Both minor components are refined with a fixed 25% occupancy.Due to poor data quality, and the extensive disorder, the structure is heavily restrained (SADI, RIGU, and ISOR) and constrained (EADP) to produce a stable refinement, and so all measurements taken from this structure should be treated with extreme caution.
For 2np at 298 K, cracking of the crystal upon in-situ desolvation reduced data quality.Similar disorder presents at 298 K as at 100 K, with the imidazole group rotated 88.8(8)° between disordered forms and sulfate group disorder between monoatomic and diatomic bridging refined with a fixed 25% occupancy.Due to poor data quality, the structure is restrained (SADI, RIGU), and so all measurements taken from this structure should be treated with significant caution.

Thermogravimetric analysis (TGA) and Differential Scanning Calorimetry (DSC)
Thermogravimetric analysis was performed under N2 flow using a TA Instruments Q50 system.A sample was loaded into an aluminum sample pan and heated at 10 °C min −1 from room temperature to 400 °C.Differential scanning calorimetry was carried out using a TA Instruments Q2000 differential scanning calorimeter.The sample and reference pans were heated at 10 °C min -1 from room temperature to 400 °C and so the heat flow, relative to the reference, was measured as a function of time and temperature under a controlled atmosphere.N2 gas flowing at a rate of 50 mL min −1 was used to purge the furnace.

Low-pressure gas sorption measurements
The sorption isotherms for N2 at 77 K, and CO2 or measured using a Micromeritics 3 Flex surface area and pore size analyzer.Bath temperatures of 77 K and 195 K were maintained using liquid nitrogen and a dry ice-acetone slurry, respectively.Before gas sorption experiments, the freshly prepared samples of 1 or 2 were placed in the quartz tube and degassed under high vacuum at 100 °C or 60 °C, respectively, for 24 h in a Micromeritics SmartVacPrep system to remove the remnant solvent molecules prior to measurements.For sorption isotherms of CO2, C2H2, C2H4, C2H6, C3H4, C3H6, C3H8 at 273 K, and 298 K, a Julabo temperature controller containing an ethylene glycol/water mixture was used to maintain a constant temperature in the bath through the duration of the experiment.
Low pressure gas adsorption of a room temperature-activated sample of CuSO 4 (1,4-bib) 1.5 , 1op A fresh, water-loaded, sample of CuSO4(1,4-bib)1.5, 1op, was evacuated directly within the micromeritics 3 Flex surface area and pore size analyser at room temperature for 1 h (to allow loosely bound water to desorb, while retaining strongly bound water and consequently the 1op phase) prior to data collection of a CO2 isotherm at 195 K. Two subsequent cycles of CO2 195 K isotherms were performed directly on the sample without reactivation, each with 1 h evacuation within the micromeritics 3 Flex surface area and pore size analyser.A final measurement was then performed after the sample was activated under high vacuum in a Micromeritics SmartVacPrep system at RT for 24 h.

High-pressure CO 2 adsorption measurements
High-pressure CO2 sorption measurements were performed using a Hiden Isochema XEMIS microbalance.A sample of 1 that was activated on a Micromeritics SmartVacPrep system at 100 °C for 24 h was transferred to the XEMIS system and then further outgassed under secondary vacuum for 3 h in situ before isotherms were collected.Excess adsorption and desorption profiles were obtained after applying buoyancy correction using the crystallographically determined skeletal density of the 1op at 100 K. Temperatures were maintained at 273 K, 298 K, 303 K, 308 K, or 313 K using a Grant LT Ecocool 150 temperature controller.

Water Vapor Sorption
Water vapor sorption was performed using an Adventure dynamic vapor sorption (DVS) instrument manufactured by Surface Measurement Systems.The instrument measures water vapor uptake gravimetrically using air as a carrier gas.A water vapor-saturated flow is created by passing dry air through a water bubbler.Specified relative humidity is generated through precise mixing of dry and saturated gas flows in desired, calibrated flow ratios.Digital mass flow controllers regulate flows of dry and saturated gases.Pure water was used to generate water vapor for these measurements and the temperature was maintained at the desired level (either 298 K, 300 K or 333 K) by enclosing the system in a temperature-controlled incubator.The mass of the sample was determined by the high-resolution microbalance Ultrabalance Low Mass with a precision of 0.01 µg.The microbalance has a symmetric configuration with both the sample pan and reference pan being exposed to the same gas and being kept at the same temperature, allowing negation of buoyancy and drag effects.The instrument is equipped with two such balances, allowing measurement of the two samples in parallel.Prior to the measurement, each sample was activated in-situ in dry air at 60-100 °C for 60 minutes using the built-in preheater and consequently cooled to sorption temperature over 90 minutes.Isotherm measurements were performed on approximately 10 mg of sample powder.For each isotherm point, dm/dt < 0.05% min -1 for a minimum of 10 minutes was used as criteria of reaching equilibrium.

Dynamic column Breakthrough (DCB)
Dynamic breakthrough experiments were performed on a custom-made rig with 2, as synthesized by slurrying.In a typical experiment, ~1.4 g of the pre-activated 2np sample was packed into quartz tubing (8 mm diameter) to form a fixed bed.The sample was activated before each experiment by purging with 20 cm 3 min -1 of He gas at 333 K for 12 h.After cooling to room temperature, the gas flow was switched to the desired composition of C2H2/C2H4 or C2H2/CO2 which was pre-equilibrated through a separate column, and regulated by mass flow controllers.The outlet composition was monitored at 5 min intervals by gas chromatography (Shimadzu, GC 2030, Flame Ionization Detector (FID) for C2H2 and C2H4 and Thermal Conductivity Detector (TCD) for CO2).

Cambridge structural database (CSD) analysis of CuSO 4 MBB dimer
A CSD survey was conducted using ConQuest (2023.2.0) and the results were processed with Mercury (v.2023.2.0) to find all instances of the CuSO4 MBB dimer using the following search query: Two oxygen atoms were limited to being one-connected (T1) to reduce the number of unrelated hits.No further restrictions were added.This results in 48 total hits from which 24 hits contain the CuSO4 MBB dimer which are listed in Table S1.

Differential scanning calorimetry (DSC) for 1op
Figure S6: DSC trace of 1op.Two peaks observed below 150 °C corresponding to water loss events, followed by thermal decomposition events beginning at 258 °C and 370 °C, matching mass loss observed in TGA.3.12 PXRD after thermal decomposition of 1

IAST Selectivity Calculations
The selectivities for C2H2/CO2 and C2H2/C2H4 gas mixtures of 2np were calculated from single component adsorption isotherms using the Ideal Adsorbed Solution Theory (IAST) 30,31 using the software IAST++. 32Single-component adsorption isotherms for each gas at 298 K were fitted to the dual-site Langmuir equation: Where P is the total pressure (mbar) of the bulk gas at equilibrium with the adsorbed phase, q1 and q2 are the saturation uptakes (in mmol//g) for sites 1 and 2, respectively, k1 and k2 are the affinity coefficients (in mbar −1 ) for sites 1 and 2, respectively, and n(P) is the uptake (mmol/g) as a function of pressure.After the isotherms have been parameterized, mixed-gas fractional uptakes are calculated and the IAST selectivity (Si/j) is obtained through the equation: Where xiand xj are the mole fractions of components i and j, respectively, in the adsorbed phase, and yi and yj are the mole fractions of components i and j, respectively, in the gas phase.

Figure S1 :
Figure S1: Q-peaks and crystallographic disorder in 2op at 100 K before and after disorder is resolved, resulting in a reduction in R1.

Figure S2 :
Figure S2: Q-peaks and crystallographic disorder in 2np at 100 K before and after disorder is resolved, resulting in a reduction in R1.

Figure S3 :
Figure S3: Comparison of known structural variations of the CuSO4 MBB dimer in (a) CSD refcode: PIGYEK and (b) CSD refcode: CEVJUL showing that the interaction between SO4 and the Cu open metal site (OMS) is in some cases within or beyond the standard Mercury Cu-O bond distance cut-off of 2.80 Å.

Figure S4 :
Figure S4: Histogram showing long sulfate-Cu bond length in CuSO4 MBB dimer structures.A bimodal distribution is seen below 3.3 Å indicating the bimodal structural variation in MBB.Structures with long sulfate-Cu bond length above 3.3 Å are characterized by other functional groups blocking the copper binding site.

3. 5 Figure S8 :
Figure S8: Structural overlay of 1op at 100 K and 1cp at 343 K formed by overlaying the Cu2O2 of each structure: a) sql net, b) CuSO4 MBB dimer and the single-wall 1,4-bib ligand, c) sql net with 1op at 100 K shown in blue and 1cp at 343 K in red, d) CuSO4 MBB dimer and the single-wall 1,4-bib ligand with 1op at 100 K shown in blue and 1cp at 343 K in red.Hydrogen atoms and solvated water molecules have been omitted for clarity.

Figure S9 :
Figure S9:An overlay of the calculated PXRD of 1 collected at 100 K with the experimental PXRD of the as synthesized 1op material at 298 K, followed by the material under dry N2 at 298 K, 313 K, 333 K, 353 K, 373 K, 393 K, 413 K, 433 K, 453 K, 473 K.The sample was then allowed to cool to 298 K while remaining under dry N2 then exposed to air for 5 min.From this it can be seen that the open phase slowly changes as the temperature rises, then at 373 K the 1cp begins to appear and is the only phase present at 413 K up to 473 K.When the sample is cooled to 298 K the closed phase persists, and when exposed to air the sample reverts to 1op, as would be indicated by the water sorption of the sample.Each PXRD is colored to match the phase with the 1op calculated PXRD in Navy blue, and experimental PXRD patterns in blue, 1cp calculated PXRD in orange, and experimental PXRD patterns in red.PXRD patterns where a mixture of 1op and 1cp is present are shown in purple.

3. 7 1 Figure S10 :
Figure S10: Dynamic vapor sorption experiment showing water sorption and desorption of a 100 °C activated sample of 1 at 298 K (red).A second cycle (blue) was then performed without reactivation where the change in mass was normalized using the mass at the beginning of cycle 1.From this it can be seen that the strongly bound water molecule does not fully desorb in the first cycle at RT, and so cycle 2 begins with strongly bound water already present in the material.

Figure S11 :
Figure S11: H2O dynamic vapor sorption of 1 at 333 K of a 100 °C activated sample.At ca. 5% RH a sharp rise in uptake is observed, indicating a switch from 1cp to 1op structures.

3. 10
Figure S14: 298 K CO2 sorption of 1 activated at RT directly within the gas sorption instrument.Successive cycles show a decline in uptake as a result of slow conversion of 1op to 1cp.

3. 11 Slurry-based synthesis of 1 Figure S15 :
Figure S15: An overlay of the experimental PXRD of 1op synthesized from a H2O:MeOH slurry of 1,4bib and CuSO4 .5H2O, with the experimental PXRD of 1op formed from layering and the calculated PXRD of a single crystal of 1op collected at 100 K.

Figure S16 : 2 4. 1 4 . 1 Figure S17 :
Figure S16: An overlay of the experimental PXRD of 1 collected after heating a sample to 528 K.As seen in the TGA and DSC plots, the material thermally decomposes at this temperature and so the PXRD of the material changes.Experimental PXRD patterns of 1op and 1cp collected at 298 K are shown for comparison.

Figure S18 :
Figure S18: Thermogravimetric analysis of 2op from layering and slurrying synthesis.Water loss occurs < 100 °C, followed by thermal degradation events beginning at 227 °C and > 340 °C.

4. 3
Figure S19: DSC trace of 2op from layering and slurrying synthesis.Thermal event corresponding to water loss occurs < 100 °C, followed by thermal degradation events observed at 220 °C for layering and 227 °C for slurrying, then further degradation at 334 °C for layering and 358 °C for slurrying.

4. 4 2 Figure S20 :
Figure S20:An overlay of the calculated PXRD of 2op collected at 100 K and RT with the experimental PXRD of 2op in air at 298 K, followed by the material under dry N2 at 298 K, 313 K, 333 K, 353 K, 373 K, 393 K.The sample was then allowed to cool to 298 K while remaining under dry N2 then exposed to 3% RH air for 10 min, then to 53% RH air for 10 min then to 53% RH air after a total time of 30 min.From this it can be seen that the 2op phase immediately changes to the 2np phase upon exposure to dry N2 at 298 K and this phase is maintained as the temperature rises to 393 K, and as the sample is cooled to 298 K and exposed to 36% RH air for 10 min.When the sample is exposed to 53% RH air for 10 min, a mixture of 2op and 2np are observed, and after 30 min only 2op is present.Each PXRD is colored to match the phase with 2op (100 K and RT) calculated PXRD in Navy blue, and experimental PXRD patterns in blue, 2np calculated PXRD in orange (100 K and 298 K), and experimental PXRD patterns in red.The PXRD pattern where a mixture of 2op and 2np is present is shown in purple.

Table S1 : List of known structures in the CSD featuring the CuSO4 MBB dimer.
*HULTIT features a 1D polymer formed from the Copper sulfate dimer, and a separate 2D net formed from bridging sulfate groups.**Coordinates for HULTIT01 are found in refcode HULTIT.

Table S7 . Isotherm fitting parameters and R 2 values for IAST calculations.
Note, data truncated at the emergence of flexibility for C2H2 at 0.328 mbar for layering a or 0.427 mbar for slurrying b .