Hybrid glasses from strong and fragile metal-organic framework liquids

Hybrid glasses connect the emerging field of metal-organic frameworks (MOFs) with the glass formation, amorphization and melting processes of these chemically versatile systems. Though inorganic zeolites collapse around the glass transition and melt at higher temperatures, the relationship between amorphization and melting has so far not been investigated. Here we show how heating MOFs of zeolitic topology first results in a low density ‘perfect' glass, similar to those formed in ice, silicon and disaccharides. This order–order transition leads to a super-strong liquid of low fragility that dynamically controls collapse, before a subsequent order–disorder transition, which creates a more fragile high-density liquid. After crystallization to a dense phase, which can be remelted, subsequent quenching results in a bulk glass, virtually identical to the high-density phase. We provide evidence that the wide-ranging melting temperatures of zeolitic MOFs are related to their network topologies and opens up the possibility of ‘melt-casting' MOF glasses.

, together with the limit of (equ. 1) to m15, fragility of the LDL phase was estimated to be m=182. Following earlier procedures used for obtaining T g and m from the collapse of inorganic zeolites (5,14), more accurate values for LDA were obtained from the dependences of the reciprocal heating rate (1/q h ) on the T g -scaled peak temperature T peak (Fig. 3c). These yielded m=14 and T g = 589 K, the LDA fragility being ~1/3 that of HDA. Angell plots contrast the respective fragile and ultra-strong viscosity relations of HDA and LDA in The inset shows a closer view on the various impurity peaks of low intensity (also marked in light red in the main section). The integrated intensity of the impurity peaks suggests that less than 10% of the imidazolates (partially) decompose during melting.

Supplementary Methods
To determine the heat capacity (C p ) of the samples, both the baseline (blank) and the reference sample (sapphire) were measured. In order to confirm reproducibility, the measurements for some samples were repeated to check for drift in the baseline.
The fictive temperatures (T f ) of samples were determined from the second up-scan curve of heat capacity (C p ) obtained at a heating rate equal to the prior down-scan rate using DSC. T f is assigned as the temperature at the cross point between the extrapolated linear fits to the C p of the glass and to the rapidly rising C p . The standard glass transition temperature (T g ) was determined from the second up-scan curve obtained at the rate of 10 K min -1 (Fig. 3b). The ZIF-zni crystallization exotherm and melting endotherm on fusion were obtained as shown in Supplementary Figure 3 and described in the text (Fig. 1b).
Melt fragility m for HDL was determined from Tg values using the MYEGA equation : (1)  , the viscosity at 1/T=0, was set at 10 -5 Pas, with the viscosity at the glass transition (Tg) set to 10 12 Pas (Fig. 4a), yielding m=412 and T g = 565 K. The melt-quenched glass had virtually identical m and T g values.
The 1 H NMR spectrum of digested ZIF-4 only shows signals for imidazolium (at 8.89 and 7.48 ppm) and the two signals for the utilized solvent mixture (at 7.21 ppm for HCl/water and at 2.50 ppm for DMSO). The 1 H NMR spectrum of the digested ZIF-4 glass basically shows the same signals, however, the peaks of imidazolium and HCl/water are slightly shifted, which is ascribed to their pH sensitivity. Additionally, a number of small signals in the aromatic region at 9.4-9.0 ppm and 8.6-7.6 ppm are present (see inset of Fig. 10b). We ascribe these signals to a partial decomposition of the imidazolate linker upon amorphisation/melting. However, the integrated intensity of the impurity peaks suggests that only about 10% of the ligands undergo thermal decomposition.
Elemental analysis was performed at the Department of Chemistry, University of Cambridge. Elemental Analysis on the MQG.

ZIF-4 Evacuated:
Calculated The following empirical relationship was established from the collapse dynamics of of zeolite A and zeolite Y 3 between the thermally induced amorphization temperature T A that occurs at the LDA T g and the pressure induced amorphization pressure P A viz., where ΔV A is the collapse volume 3 .
By comparing ZIF-4 and ZIF-8, Fragility from 2/3's Law Following 4,5 the melt fragility m for molecular liquids is empirically related to Cp(Tg) (Fig. 1e, Fig. 3a, Supplementary Figure 7), by where Hm is the heat of fusion (49 Jg -1 Supplementary Figure 3). Fragility values for LDL and HDL calculated from Table S1 are also listed and over-estimate those determined experimentally (Fig. 4a). Among molecular liquids, Wang et al identified similar outliers, such as for Se and PPP, which they attributed to changes in topology/connectivity 4 . Since increases in connectivity across glassy materials is inversely related to Poisson's ratio Agreement between fragility determined from eqn. 4 and measured values is much improved.
DLPOLY classic 7 was used to perform the MD simulations (Fig. 1a). The potential model was constructed from the reference 8 including bonded and non-bonded terms. The bonded parts contain stretching, bending, proper and improper potentials, while the nonbonded parts have Lennard-Jones and Coulombic potentials.
In order to achieve pressure induced amorphization of ZIF-8, a cubic super cell with 7452 atoms was studied. Simulations were performed in the constant temperature/pressure (NPT) ensemble throughout. The temperature (275K) was controlled using berendsen thermostats with a relaxation constant of 1ps. Constant pressures varied from 1atm to 1.2GPa were maintained by applying isotropic barostats with the same relaxation constant as thermostats. The final structure of amorphous ZIF8 was obtained after the compressing of 1.2GPa for 10ps.