Trojan Horse Thiocyanate: Induction and Control of High Proton Conductivity in CPO-27/MOF-74 Metal–Organic Frameworks by Metal Selection and Solvent-Free Mechanochemical Dosing

Proton-conducting metal–organic frameworks (MOFs) have been gaining attention for their role as solid-state electrolytes in various devices for energy conversion and storage. Here, we present a convenient strategy for inducing and tuning of superprotonic conductivity in MOFs with open metal sites via postsynthetic incorporation of charge carriers enabled by solvent-free mechanochemistry and anion coordination. This scalable approach is demonstrated using a series of CPO-27/MOF-74 [M2(dobdc); M = Mg2+, Zn2+, Ni2+; dobdc = 2,5-dioxido-1,4-benzenedicarboxylate] materials loaded with various stoichiometric amounts of NH4SCN. The modified materials are not achievable by conventional immersion in solutions. Periodic density functional theory (DFT) calculations, supported by infrared (IR) spectroscopy and powder X-ray diffraction, provide structures of the modified MOFs including positions of inserted ions inside the [001] channels. Despite the same type and concentration of proton carriers, the MOFs can be arranged in the increasing order of conductivity (Ni < Zn < Mg), which strongly correlates with amounts of water vapor adsorbed. We conclude that the proton conductivity of CPO-27 materials can be controlled over a few orders of magnitude by metal selection and mechanochemical dosing of ammonium thiocyanate. The dosing of a solid is shown for the first time as a useful, simple, and ecological method for the control of material conductivity.


Synthetic procedures and control experiments
All chemicals and solvents (of analytical grade) were purchased from commercial sources (Merck, ,TCI, Avantor, Polmos) and were used without further purification. Ethanol (Polmos) contained water (8% by volume).

Synthesis of CPO-27(Zn)
CPO-27(Zn) was synthesized according to a method described in literature. 1 H4DOBC (0.25 g, 1.3 mmol) and Zn(NO3)2·4H2O (1.0 g, 3.8 mmol) were dissolved in 50 mL of DMF with stirring, followed by the addition of 2.5 mL of deionized water. The mixture was heated in an oven at 100 o C for 20 hours yielding the yellow microcrystalline, porous material, CPO-27(Zn).

Synthesis of CPO-27(Mg)
CPO-27(Mg) was synthesized according to a method described in literature. 2 To a solid mixture of H4DOBDC (0.111 g, 0.559 mmol) and Mg(NO3)2·6H2O (0.475 g, 1.85 mmol) was added a 15:1:1 (v/v/v) mixture of DMF-ethanol-water (50 mL) in a 100 mL screw cap jar. The suspension was mixed and ultrasonicated until homogeneous and placed in an oven at 100°C. After 20 hours, the samples were removed from the oven and allowed to cool to RT. The mother liquor was decanted from the yellow microcrystalline material and replaced with methanol. The methanol was decanted and replenished four times over two days, yielding the dark yellow microcrystalline, porous material.

Synthesis of CPO-27(Ni)
CPO-27(Ni) was synthesized according to a method described in literature. 2 To a solid mixture of H4DOBDC (0.0956 g, 0.48 mmol) and Ni(NO3)2·6H2O (0.4756 g, 1.64 mmol) was added a 1:1:1 (v/v/v) mixture of DMF-ethanol-water (40 mL) in a 100 mL screw cap jar. The suspension was mixed and ultrasonicated until homogeneous and placed in an oven at 100°C. After 20 hours, the samples were removed from the oven and allowed to cool to RT. The mother liquor was decanted from the yellow microcrystalline material and replaced with methanol. The methanol was decanted and replenished four times over two days, yielding the yellow-brown microcrystalline, porous material.

Synthesis of CPO-27(Co)
CPO-27(Co) was synthesized according to a method described in literature. 2 To a solid mixture of H4DOBDC (0.0964 g, 0.47 mmol) and Co(NO3)2·6H2O (0.4754 g, 1.73 mmol) was added a 1:1:1 (v/v/v) mixture of DMF-ethanol-water (40 mL) in a 100 mL screw cap jar. The suspension was mixed and ultrasonicated until homogeneous and placed in an oven at 100°C. After 20 hours, the samples were removed from the oven and allowed to cool to RT. The mother liquor was decanted from the dark red-purple microcrystalline material and replaced with methanol. The methanol was decanted and replenished four times over two days, yielding the dark red-purple microcrystalline, porous material.  Figures S6 and S7 below). Quantitative formation of CPO-27(Mg, Zn Ni)-NCS was not formed.

Reversibility of CPO-27-NCS formation
In all experiments CPO-27-NCS (50 mg) were immersed in methanol (5.0 mL) for approx. 5 days at 60°C. During this time methanol was changed 4 times a day. All solid samples have been screened with IR and PXRD ( Figure S8 below). In all the trials CPO-27(Mg, Zn, Ni) were recovered. Analogous trials carried out in methanol at approx. 25 o C did not lead to removal of coordinated thiocyanate after 7 days.

S-6
Samples preparation for EIS measurements Prior to a series of EIS measurements at a given relative humidity (RH), approx. 30 mg of the material was equilibrated in a humidity chamber for 1 day at a specified RH (90 or 75 or 60 or 45 or 30%) at 25 o C.

Details of physical measurements
Carbon, hydrogen and nitrogen were determined using an Elementar Vario MICRO Cube elemental analyzer. Thermogravimetric analyses (TGA) were performed on a Mettler-Toledo TGA/SDTA 851 e instrument with a heating rate of 10°C min -1 in a temperature range of 25 -600°C (approx. sample weight of 10 mg). The measurements were performed at atmospheric pressure under argon flow. FTIR spectra were recorded on a Thermo Scientific Nicolet iS10 FT-IR spectrophotometer equipped with an iD7 diamond ATR attachment. Powder X-ray diffraction (PXRD) patterns were recorded at room temperature (295 K) on a Rigaku Miniflex 600 diffractometer with Cu-Kα radiation (λ = 1.5418 Å) in a 2θ range from 3° to 45° with a 0.02° step and 3° min -1 scan speed. Additional measurement was carried out for CPO-27(Mg)-NCS on a Panalytical X-Pert Pro equipped with a sealed LFF Cu X-ray source operating at 40kV and 30 mA, a focusing mirror, a capillary sample holder and a PIXCEL detector. Electrochemical Impedance Spectroscopy (EIS) measurements were performed using Hioki IM3570 impedance analyzer on pre-conditioned samples (see Samples preparation above) pressed between two metallic electrodes (5.0 mm diameter) in a PTFE tube, and kept in KK 115 TOP+ or Memmert HCP 246 climatic chamber with ultrasonic humidifier and temperature control. The measurements were carried out in a quasi-four-probe setup in a frequency range from 4 Hz to 5MHz, alternating potential of 200 mV and temperature range from 25 to 60 o C (for 30-90% RH), and with pre-measurement sample conditioning for 1 h at each temperature. The proton conductivity (σ, S cm −1 ) of the sample was estimated by using the equation: σ = L/(RA) where L (cm) is the sample thickness and A (cm 2 ) is the cross-sectional area of the measured pellet; R (Ω) is the resistance as the real part of the measured impedance. Adsorption isotherms were measured using static volumetric Autosorb IQ apparatus (Quantachrome Instruments) at 77 K for nitrogen and 298 K for water. All samples were activated under vacuum for 8 h at 100°C (2°C/min).

Computational details
All the quantum-chemical calculations of energies, geometries, and vibrational analyses were performed by using of DMol3 3 code at the periodic DFT level of theory, the RPBE 4,5 correlation-exchange functional. This functional was chosen since it provides a better description of the hydrogen bonds than PBE or revPBE. The orbitals were expanded in the basis set of DND quality. The basis set spatial cutoff distance of 4.2 Å was set and, as a convergence accelerator, the Gaussian smearing with the window width of 5 ⋅ 10 -3 Ha was used. The SCF convergence criterion was set to 1 ⋅ 10 -6 Ha, whereas for geometry optimisationto the threshold value of 2 ⋅ 10 -5 Ha change between two successive iterations, the gradient norm lower than 3 ⋅ 10 −3 Ha∕Å, and the maximal atom displacement lower than 4 ⋅ 10 −3 Å.

Rietveld refinement procedure
The X-ray diffraction of the CPO-27(Mg)-NCS sample has been performed in a long collection time capillary experiment and a trial has been undertaken to determine the crystal structure of CPO-27(Mg)-NCS through the application of Rietveld refinement using the published structure as the starting model (CSD code: MOHGOI). Given the space group (R-3) of the initial CPO-27(Mg) structure, the insertion of NCSand NH4 + ions results in their sixfold replication. We have tried to refine this model with 50% site occupancies which led us to problematic electron density from NH4 + ions which caused them to drift towards the centre of the main [001] channel (resulting distances between nitrogen atoms were not physical). At the same time, the NCSions were clearly visible in the Fourier maps in the proper locations consistent with the theoretical calculations. We have also tried lowering the symmetry of the loaded structure to R3 to overcome the problem introduced by the six-fold replication inside the channel (the XRD patterns are undistinguishable for R-3 and R3 apart from the intensity statistics that cannot be reliably performed for powder data) but it did not lead to a significant improvement. Most likely our data (laboratory X-ray source) are of insufficient quality to unambiguously locate the positions of NH4 + and SCNions. Also, there is a significant possibility of disorder in terms of their positions since there is no difference between the Mg sites that would make some of them preferential over the others. It is noteworthy that in the original work reporting MOHGOI structure 6 the authors used synchrotron radiation which provided powder X-ray diffraction data with excellent signal-to-noise ratios and allowed crystallographic studies with high level of detail, even including the localization of the adsorbed CO2 molecules.                            The equivalent circuit proposal indicates the presence of components which may refer to bulk resistance, different orientation of crystal planes contribution, grain boundary resistance and electrode/electrolyte interface.