Metal-organic framework and inorganic glass composites

Metal-organic framework (MOF) glasses have become a subject of interest as a distinct category of melt quenched glass, and have potential applications in areas such as ion transport and sensing. In this paper we show how MOF glasses can be combined with inorganic glasses in order to fabricate a new family of materials composed of both MOF and inorganic glass domains. We use an array of experimental techniques to propose the bonding between inorganic and MOF domains, and show that the composites produced are more mechanically pliant than the inorganic glass itself.

(1) Where there are N atoms in the sample, ci is the proportion of element i and fi(Q) is the X-ray atomic form factor of element i 4 . Q is the scattering vector determined by the X-ray wavelength, λ, and the scattering angle 2θ:

= 4 sin (2)
For a non-interacting mixture of M phases the total intensity is assumed to be the weighted sum of intensities of each phase (not accounting for attenuation and the presence of interfaces) 5 ; Where d dΩ is the total scattered intensity of the k th phase in the multiphase sample, which is obtained from a measurement of 1 d dΩ ⁄ from a sample of phase k on its own, i.e. where Nk is the number of atoms of phase k within the multiphase sample. However due to the difficulty of placing X-ray scattering data on an absolute scale 2 a set of additional scaling factors are introduced: Where is the approximately normalised intensity and is a constant close to unity. Equation (3) now becomes; Expt. Expt. Norm = ∑ Norm =1 (5) In order to produce the expected scattering intensity of a non-interacting mixture the experimental total scattering intensities per atom of the pure agZIF-62 and inorganic glasses are recovered from their respective S(Q) via equation (1) and added together weighted by their atomic proportions according to (5). The total scattered intensity for the composite samples, (agZIF-62)0.5(Inorganic Glass)0.5 -1 min and (agZIF-62)0.5(Inorganic Glass)0.5 -30 min, the experimental total scattering, are also recovered from their respective S(Q)s via (1).
The difference between measured intensity from a sample and the calculated intensity from a noninteracting mixture of the same chemistry therefore represents changes in the diffracted intensity resulting from interaction of the agZIF-62 and inorganic glass phases in the composite. This difference is then re-sharpened by dividing through by ∑ ( ) 2

=1
: In this paper the values of ak where determined such that the difference between the calculated mixture and the experimentally measured scattering, weighted to produce smaller differences at higher Q-values, was minimised for each different inorganic glass composition. One aExpt. scale factor was kept fixed at unity as a reference point.

S(Q)
Diff is then Fourier transformed in the usual way to produce ( ) 6 . Where the ( ) now represents a weighted histogram of the distribution of atom-atom distances in the composite sample which are due to interactions between the inorganic glass and agZIF-62, such that peaks represent new correlations formed as a result of the interaction between the two phases; either new bonds formed at the interface or changes in the structure of either phase due to the presence of the other. Finally in order to aid visualisation of new correlations across the full range of real space an alternative form of the usual D(r) function is plotted: Due to the difficulty of putting X-ray total scattering data on an absolute scale S(Q) Diff can contain residual scattering from the individual phases. The size and shape of this 'erroneous' difference is dependent on the scale factors (ak) used in equation (4). The presence of these additional features in the S(Q) Diff complicates interpretation of the G(r) Diff and D(r) Diff ; underweighting of the total scattering from a constituent phase will lead to positive peaks from that phase in the G(r) Diff and D(r) Diff as residual scattering from that phase remains in the S(Q) Diff and conversely overweighting will lead to negative peaks in the G(r) Diff and D(r) Diff as the total scattering from a phase is over-removed from the S(Q) Diff .
Supplementary Fig. 6. Thermal response of (ZIF-62)(Na-deficient)(50/50) on the first (red) and second (blue) DSC heating scans. a after 1 minute. b after 30 minutes at 450˚C. An offset has been added to improve readability.  Chemical shift (ppm) Supplementary Fig. 11. Liquid 1 H NMR spectra of (agZIF-62)(Inorganic Glass) composites treated for 1 minute. An offset has been added to improve readability.