Experimental demonstration of dynamic temperature-dependent behaviour of UiO-66 metal-organic-framework: Compaction of hydroxylated and deydroxylated forms of UiO-66 for high pressure hydrogen storage.

High-pressure (700 MPa or ~100 000 psi) compaction of dehydroxylated and hydroxylated UiO-66 for H2 storage applications is reported. The dehydroxylation reaction was found to occur between 150 - 300 oC. The H2 uptake capacity of powdered hydroxylated UiO-66 reaches 4.6 wt% at 77 K and 100 bar, which is 21% higher than that of dehydroxylated UiO-66 (3.8 wt%). On compaction the H2 uptake capacity of dehydroxylated UiO-66 pellets reduces by 66% from 3.8 wt% to 1.3 wt%, while for hydroxylated UiO-66 the pellets show only a 9% reduction in capacity from 4.6 wt% to 4.2 wt%. This implies the H2 uptake capacity of compacted hydroxylated UiO-66 is at least three times higher than that of dehydroxylated UiO-66, and therefore hydroxylated UiO-66 is more promising for hydrogen storage applications. The H2 uptake capacity is closely related to compaction induced changes in the porosity of UiO-66. The effect of compaction is greatest in partially dehydroxylated UiO-66 samples that are thermally treated at 200 and 290 oC. These compacted samples exhibit XRD patterns indicative of an amorphous material, low porosity (surface area reduces from between 700 and 1300 m2/g to ca. 200 m2/g and pore volume from between 0.4 and 0.6 cm3/g to 0.1 and 0.15 cm3/g) and very low hydrogen uptake (0.7 - 0.9 wt% at 77 K and 100 bar). The observed activation temperature-induced dynamic behaviour of UiO-66 is unusual for MOFs and has previously only been reported in computational studies. After compaction at 700 MPa, the structural properties and H2 uptake of hydroxylated UiO-66 remain relatively unchanged, but are extremely compromised upon compaction of dehydroxylated UiO-66. Therefore, UiO-66 responds in a dynamic manner to changes in activation temperature within the range in which it has hitherto been considered stable.

simulations of Vandichel et al 16,17 also detailed structural changes that may occur in UiO-66 during dehydroxylation processes up to 320 o C and proposed the possibility of two transition states involving bond rearrangements within the zirconium nodes. In addition, the study emphasized the role of defects on the extent of dehydroxylation, and that defects may result in significant compromising of the mechanical strength of the UiO-66 framework. The hydroxylated and dehydroxylated UiO-66 forms have further been shown to exhibit significant differences in their gas adsorption capacities for CO2, CH4, and H2 26,27 , with dehydroxylated UiO-66 generally showing less adsorption capacity compared to hydroxylated UiO-66.
To our knowledge, there are hardly any experimental studies on the highpressure compaction of dehydroxylated UiO-66, particularly as it relates to H2 storage applications, and, furthermore, post-synthesis thermal treatment conditions vastly differ in previous reports 22,23,28,29 . In a sense, the importance of the thermal treatment Aldrich, 99.8 %). All the chemicals were purchased and used without further purification.

Preparation of UiO-66
The growth of UiO-66 crystallites was specifically done via an acid-modulated solvothermal method with formic acid (FA) as the monocarboxylic modulator. The synthesis method was adopted from our previous study 12 and typically involved sonicating a mixture containing 1:1:100 ZrCl4:BDC:HCOOH in 300 mL DMF. The mixture was transferred to a round-bottom flask and maintained at 120 o C under reflux for 6 hours. After the synthesis, the white solid product was collected under centrifugation and washed in DMF for 3 hours. In order to remove DMF molecules possibly remaining within the pores, the product was further washed 3 times in acetone for 1 hour prior to recollection and drying under vacuum at room temperature for 24 hours.

Post-synthesis heat treatment and compaction
UiO-66 powder samples were heat-treated in a Micromeritics SmartVac at specified temperatures for 16 hours under vacuum (~10 -7 ). The temperatures of choice were as follows: 80,110,140,170,200,230,260,290, and 320 o C in order to obtain fully hydroxylated, partially dehydroxylated and fully dehydroxylated UiO-66 samples. For each of the heat-treated powder samples, a 400 mg portion was compacted at~700 MPa using a Specac Manual Hydraulic Press and held at that pressure for 5 minutes.

Characterisation
To investigate the stability of the UiO-66 crystal structure, samples were analysed by powder X-ray diffraction (PXRD) whereby both powder and pellet samples were first ground into fine powder using an agate mortar and pestle. It is acceptable to grind the UiO-66 pellets after compaction since the uniaxial compaction of UiO-66 is an irreversible process 7,30-33 . The sample analysis was done on a Rigaku Ultima IV X- The Brunauer-Emmett-Teller (BET) surface areas and pore volumes were calculated from nitrogen (N2) adsorption data measured on a Micromeritics 3-flex sorptometer operating at 77 K. Excess hydrogen (H2, 6.0 grade purity) uptake was measured at 77 K and 298 K up to 100 bar using a Hiden Isochema XEMIS intelligent gravimetric analyser. The H2 uptake method was pre-set at specified time intervals between each measurement and the results obtained under non-equilibrium conditions. The H2 uptake data, for each sample, was corrected for buoyancy effects using measured skeletal densities obtained from helium (He) pycnometry measurements at standard temperature and pressure (STP). The absolute or total H2 uptake are calculated using Eqn. 2, taking into account H2 adsorbed within the pores of the UiO-66. Calculation of volumetric H2 capacity from gravimetric adsorption data has become somewhat controversial on whether the crystal density or packing density should be used to calculate volumetric H2 capacities in MOFs 6,34-40 . Our previous study 12 was able to show strong correlation of the packing density to the volumetric H2 capacity following the recommendations made in a publication by Balderas-Xicohténcatl et al. 35 . We, therefore, use the MOF packing density for both powder and compacted UiO-66 and calculate the volumetric H2 capacity according to Eqn.3.
The H2 densities at 77 K in the 0 -100 bar range were obtained from the National Institute of Standards and Technology (NIST) website 41 .
VT = pore volume obtained from N2 isotherm data.

Results and Discussion
The morphology of the UiO-66 crystals obtained in this study show typical octahedral shapes consistent with what is reported in literature (ESI Fig. S1). The crystal structure of UiO-66 has been well studied 13,42,43 and has a cubic unit cell (F m -3 m) with typical     entirely possesses pore channels of size in the micropore range (2 to 20 Å). The pore channels of UiO-66 are typically of size~6, 8, and 11 Å, which size relates to the free spaces in tetrahedral cages, triangular windows, and octahedral cages, respectively 14,63 . The major pore sizes shown in Fig. 4(b)  Cumulative pore volume (cm  Values in parenthesis are micropore surface area and percentage micropore surface area of the total surface area. b Values in parenthesis are micropore volume and percentage micropore of the total pore volume. c Tapped density of UiO-66 powder. d Calculated by multiplying the packing density and the surface area [35] . The H2 adsorption isotherms, in Fig. 5, further substantiate the observed effects on the porosity. A similar trend is observable for both the excess H2 uptake (Fig. S5, ESI) and the total/absolute H2 uptake as shown in Fig. 5(a).The gas sorption results were obtained using volumetric (N2) and gravimetric (H2) techniques, which provides robust evidence of the observed post-synthesis thermal treatment induced behaviour of UiO-66. The surface area, pore volume, and total H2 uptake of~1400 m 2 .g -1 , 0.6 cm 3 .g -1 , and 4.6 wt% (at 77K and 100 bar), respectively, for UiO-66 powder activated at 80 o C are similar to results obtained in our previous study 12 . In Fig. 5(b), the H2 uptake at 298 K was as expected lower compared to 77 K, and shows total uptake of 0.8 wt% for hydroxylated UiO-66 (80 o C) and~0.7 wt% for dehydroxylated UiO-66 (320 o C). It is typical for highly microporous materials to show near Type I gas adsorption isotherms at cryogenic temperatures, with fully Type I isotherms achievable at the boiling point of the adsorbed gas 64 , as shown in the N2 adsorption isotherms in   attractions of UiO-66, has major implications on the choice of post-synthesis treatment (i.e., activation temperature) of UiO-66 especially in applications where mechanical stability is essential. One such application is as hydrogen storage material where compaction or pelletization is essential.
As shown in Fig. 7 and summarised in Table 2, the hydroxylated UiO-66 pellet (activated at 80 o C) has a total H2 uptake of 4.2 wt%, which is only 9% lower than that of the hydroxylated UiO-66 powder (4.6 wt%). The levels of hydrogen uptake are consistent with the porosity of the samples ( Table 1). The relatively similar gravimetric hydrogen uptake means that the hydroxylated UiO-66 pellet has significantly higher volumetric H2 capacity compared to the powder due to a rise in packing density  (Table 2) is in line with the changes in porosity (Table 1). We also determined the hydrogen storage working capacity, which is the difference in uptake at 100 bar compared to 1 bar. While hydroxylated samples (both powder and compacted) and powder forms of partially or fully dehydroxylated samples have working capacity of between 1.7 and 3.3 wt%, the results in Table 2

Conclusion
In this study, the post-synthesis heat treatment of UiO-66 up to 320 o C was monitored by TG-MS analysis and the results show a three-step decomposition process involving the removal of H2O molecules via a two-step process between~150 and 300 o C, termed dehydroxylation, and the third step involves framework collapse at~500 o C.
Based on TGA data, evidence of the presence of transition states during dehydroxylation was observable and dynamic changes to certain peak positions in the UiO-66 PXRD pattern may signify the dynamic structural changes (decoordination and protonation) predicted by Vandichel et al. 17 . It was also evident that the porosity, and hence gas adsorption properties, of UiO-66 is extremely sensitive to the hightemperature activation. Based on the results obtained in this study, we recommend that optimum post-synthesis treatment for UiO-66 should ideally include solvent exchange of the as-synthesised UiO-66 with volatile solvents (such as acetone or ethanol) followed by low-temperature activation not exceeding 150 o C. The proposed conditions should maintain the UiO-66 structure in its hydroxylated form, which is favourable for hydrogen storage applications.

Supplementary Information
Seven additional figures and one table, showing TEM images, XRD patterns, excess hydrogen uptake, and nitrogen isotherms and corresponding textural properties.