Exceptional adsorption-induced cluster and network deformation in the flexible metal – organic framework DUT-8 ( Ni ) observed by in situ X-ray diffraction and EXAFS †

The ‘‘gate opening’’ mechanism in the highly flexible MOF Ni2(2,6-ndc)2dabco (DUT-8(Ni), DUT = Dresden University of Technology) with unprecedented unit cell volume change was elucidated in detail using combined single crystal X-ray diffraction, in situ XRD and EXAFS techniques. The analysis of the crystal structures of closed pore (cp) and large pore (lp) phases reveals a drastic and unique unit cell volume expansion of up to 254%, caused by adsorption of gases, surpassing other gas-pressure switchable MOFs significantly. To a certain extent, the structural deformation is specific for the guest molecule triggering the transformation due to subtle differences in adsorption enthalpy, shape, and kinetic diameter of the guest. Combined adsorption and powder diffraction experiments using nitrogen (77 K), carbon dioxide (195 K), and n-butane (272.5 K) as a probe molecules reveal a one-step structural transformation from cp to lp. In contrast, adsorption of ethane (185 K) or ethylene (169 K) results in a two-step transformation with the formation of intermediate phases. In situ EXAFS during nitrogen adsorption was used for the first time to monitor the local coordination geometry of the metal atoms during the structural transformation in flexible MOFs revealing a unique local deformation of the nickel-based paddle-wheel node.


Experimental data for single crystal X-ray diffraction of DUT-8(Ni) cp
The microcrystal of activated DUT-8(Ni) was glued to the wall of the capillary in inert atmosphere.Afterwards the capillary was sealed with wax.The data collection was performed at Helmholtz Zentrum Berlin für Materialien and Energie (MX beamline BL14.2). 1 The synchrotron radiation with energy of 14 kEv (λ = 0.88561 Å) was used for the data collection.The data were collected at room temperature using φ-scan technique (Δφ = 1°).Image frames were integrated using the Mosflm 1.0.5 software. 2Obtained intensities were scaled with the Scala program.The crystal structure was solved in P1 space group by direct methods using SHELXTL program package. 3 Because of poor scattered crystal as well as disorder in the crystal structure, only two nickel atoms and their coordination environment could be localized.The unit cell parameters are a = 6.7372(24)Å, b = 8.0420(28) Å, c = 11.9468(43)Å, α = 90.339(15)°, β = 104.014(17)°, γ = 104.275(15)°.The atomic coordinates, detemined from the single crystal X-ray diffraction measurement are given in Table S1.The organic ligand molecules were simulated using Material Studio 5.0. 4 Further refinement was performed using X-ray powder diffraction data.

In situ adsorption and X-ray powder diffraction study
The sample of DUT-8(Ni) was prepared as described in ref. 5.In order to get an optimal grain size for X-ray diffraction, the sample was milled in the mortar and sieved using 45 μm sieves.The activated sample (20-30 mg) was filled into the sample holder for performing in situ experiments using nitrogen (77 K), n-butane (273 K), carbon dioxide (195 K), ethane (185 K), and ethene (169 K) as probe molecules.Measurements were carried out at the MAGS or KMC-2 beamlines at Helmholtz-Zentrum Berlin für Materialien und Energie (BESSY-II).All experiments were performed using recently constructed sample environment. 6The isothermal conditions were created using closed cycle He cryostat.The BELSORP-Max automated dosing system ensured the desired gas pressure as well as connection with the diffractometer.All X-ray powder diffraction patterns were collected using the monochromatic radiation with E = 8048 keV (λ = 1.5406Å) in transmission geometry using 2θ scans.VÅNTEC-2000 2D area detector from Bruker was used for the collection of diffraction images.The image data were integrated using Datasqueeze 2.2 software. 7The phase content in all powder XRD patterns was estimated by the Reference Intensity Ratio method using the Match 1.11g software.was loaded to the sample holder with thickness of 1 mm.The latter was put into the adsorption chamber and mounted to the closed cycled He-cryostat.The nitrogen adsorption isotherm was measured at 77 K using BELSORP-max automated dosing gas system, connected to the adsorption chamber by thin copper capillary.The EXAFS spectra were measured at the 5 points of interest.
Data processing, analyses and fits have been performed with the Demeter software package. 9he fits were performed in R-space in the 1.0 -3.0 Å range over k 2 -weighted FT of the (k) functions performed in the 3.0-15.0Å -1 interval.A single ΔE 0 and a single S 0 2 have been optimized for all SS (single scattering) paths used in fit.Whereas ΔR (change in half path length) and σ 2 (Debye-Waller factor) were fitted individually for each scattering pair.The series of electron scattering paths for the first shell of Ni atom were calculated from the crystal structures of both cp and lp phases using FEFF Version 6 of the Artemis software.In the case of cp phase the three Ni-O SS paths with similar Ni-O distance were averaged and fitted with degeneracy parameter of 3 in order to increase the data/parameter ratio.The SS paths with longer Ni-O distance as well as Ni-N and Ni-Ni paths are fitted separately with degeneracy parameter of 1.In the case of N 2 @DUT-8(Ni) lp phase, Ni-O and Ni-C carboxylate paths were fitted with degeneracy parameter of 4, whereas Ni-N and Ni-Ni paths were treated the same as in cp phase.The result of the fit is given in Table S2.

Experimental data and Rietveld plots for DUT-8(Ni) cp phase
The unit cell parameters for the Pawley refinement were taken from the results of single crystal X-ray diffraction measurements on a microcrystal.For the refinement of the profile parameters, the Thompson-Cox-Hasting function was used.The profile asymmetry was corrected using Berrar-Baldinozzi correction.After the refinement of the profile, the starting model was simulated in the Material Studio 5.0 visualization tool using the coordinates for the Ni atoms, obtained from the single crystal X-ray diffraction study.Two structural models in the space groups P1 and P were simulated for the Rietveld refinement.Because of ̅ 1 inappropriate ratio of observed reflection's intensity in the XRPD to the structural parameters if every atom is considered as independent, the rigid body constraints for naphtalene rings, carboxylate groups and dabco molecules were used in the refinement.The nickel atoms (as main scattering units in the structure) were refined independently.The combined Rietveld and energy refinement procedure, involving 1 % of energy contribution, was used.Although the centrosymmetric space group would be more appropriate from the symmetry point of view and data/parameters ratio, the refinement in the P does not result in the satisfactory fit.̅ 1 Therefore, the structure was refined in the P1 space group with pretty good fit for all of the reflections.It should be mentioned that non centrosymmetric space groups are often chosen for the refinement of low symmetrical contracted phases.For example, the MIL-53(Cr) lt phase was refined in the non centrosymmetric Cc space group although additional crystallographic symmetry is present.The detailed path used for the crystal structure solution and refinement is shown in Fig. S1.

Experimental data and Rietveld plot for C 2 H 6 @DUT-8(Ni) IP1 phase
The XRPD measured at p/p 0 = 0.61 was used for the pattern indexing procedure, performed using Xcell program, implemented into the Material Studio 5.0 package. 4From the variety of possible unit cells, the write one was chosen after the analysis of figure of merits, given by the program in combination with the juxtaposition of the unit cell volume with the cp phase, as made phase and correlation with the gas uptake in the isotherm.The starting model has been constructed in Material Studio visualizer using P1 unit cell setting with unit cell parameters and atom positions adopted from the cp phase structure.The structural model was optimized using geometry optimization tool of Material Studio 5.0 using Universal Force Field. 4 The obtained structural model was "filled" with 9 ethane molecules per unit cell (the amount was calculated from the adsorption isotherm) and optimized again.The created structural model was subjected to the rigid body Rietveld refinement with energy increment of 2 %.Naphtalene unit, carboxylic groupc, dabco molecules, and ethylene molecules were refined as rigid bodies.Both nickel atoms were refined independently.Because of impurity of lp phase, the 2θ ranges from 6.30° to7.00° and from 7.30° to 7.60° were excluded from the refinement.

Experimental data and Pawley plot for C 2 H 4 @DUT-8(Ni) IP2 phase
In the case of IP2 phase, the indexing of XRPD, containing predominantly single IP2 phase was performed using X-Cell utility of Material Studio software.The unit cell was chosen using two criteria: goodness of fit parameters and crystallographic information available about cp and IP1 phases.

Experimental data and Rietveld plot for C 2 H 4 @DUT-8(Ni) lp phase
Since the XRD pattern of C 2 H 4 @DUT-8(Ni) phase at p/p 0 = 0.92 does not match the theoretical patterns of "as made" DUT-8(Ni), the pattern was indexed using Xcell program in Material Studio 5.0. 4 The best match was obtained for the monoclinic unit cell with the unit cell parameters and cell volume very similar to the cell parameters of as made tetragonal phase (monoclinic angle 94°).The analysis of the systematic extinctions suggests P2 1 /m space group.The structural model for the Rietveld refinement was created in Material Studio 5.0 visualizer using "as made" structure as a starting model. 4After optimization of the geometry, rigid body Rietveld refinement was performed following the same procedure as it was used for DUT-8(Ni) cp phase with only difference that additional 9 ethylene molecules per paddle-wheel unit were placed into the pore.Crystal

Experimental data and Rietveld plot for N 2 @DUT-8(Ni) lp phase
The powder XRD pattern, measured at p/p 0 = 0.95 in the in situ adsorption experiment, does not match exactly the theoretical patterns of DUT-8(Ni) "as made" phase.Therefore, the pattern was indexed using Xcell program of Material Studio 5.0. 4 As in the previous case, the indeing results in the monoclinic cell with very similar parameters and volume.The starting model for the Rietveld refinement was created in the similar way as described for C 2 H 4 @DUT-8(Ni) lp model, but 15 N 2 molecules were added into the pore.Because of the cp impurity presented, the 2θ range from 7.3° to 7.8° was excluded from the refinement.Crystal data for N 2 @DUT-

Figure S1 .
Figure S1.The path used for the structure solution and refinement of DUT-8(Ni) cp.

Table S1 .
Atomic coordinates for DUT-8(Ni) cp determined from single crystal X-ray diffraction.

Table S2 .
Fit parameters for the EXAFS data.

Table S2 .
Results of phase analysis of powder XRDs for in situ n-butane experiment at 273K.

Table S4 .
Phase analysis of powder XRD patterns collected in situ during N 2 adsorption experiment at 77K.

Table S5 .
Phase analysis of powder XRD patterns measured during in situ CO 2 adsorption experiment at 195K.

Table S6 .
Phase analysis of powder XRD patterns measured in situ during ethane adsorption experiment at 185 K.

Table S7 .
Phase analysis of powder XRD patterns measured during in situ ethene adsorption experiment at 169K.