ring under atmospheric conditions. TEA (2.2 mL, 99+ %) was slowly added drop by drop to the solution under agitation. After 30±45 min stirring the white product was filtered off, washed with DMF, and was finally dried at 373 K for 3±5 h in an oven. After drying the yield was 71 % in weight with respect to zinc nitrate. The same synthesis was repeated adding three drops of H 2 O 2 (35 %) to the DMF solution containing dissolved BDC and zinc nitrate and then the TEA was added directly to the solution.
For the second synthesis, which did not produce pure MOF-5, 0.3 g of Zn(NO 3 ) 2 . xH 2 O and 0.227 g of BDC were dissolved in 10 mL of DMF and 2.4 mL of chlorobenzene (99.7 %, also purchased from Alfa Aesar), and three droops of H 2 O 2 were added. The vessel containing the reagent solution was put into a larger closed vessel containing 2 mL of TEA. After one week, the white crystalline product at the interface between the air and solution was filtered off and washed with DMF and then dried for 3±5 h at 373 K.
Characterization: Powder X-ray diffraction data were recorded for the as-synthesized and dried sample with a Siemens D5000 powder diffractometer using Cu Ka radiation and a secondary monochromator.
Images of MOF-5 crystals were taken with a SEM and with a fieldemission SEM.
The specific surface area (SSA) and pore size of the samples were investigated with a quantachrome Autosorb automated gas sorption apparatus using N 2 gas. To measure the SSA, the solvent incorporated in the crystalline structure during the synthesis was completely removed by heating at 473 K under a vacuum of 10 ±6 mbar.
Hydrogen-Storage Measurements: For hydrogen-storage measurements we used a volumetric setup that had been previously tested both for room-temperature measurements, using well-known metal hydrides, and at 77 K, using activated carbon.
For adsorption measurements at room temperature, the experimental set-up was immersed in a temperature-controlled water bath at 298 K. High purity (99.999 %) hydrogen gas was introduced into a reservoir of known volume and, after thermal equilibrium had been achieved, the gas was permitted to expand into the sample holder.
For measurements at 77 K, the sample holder was immersed in liquid nitrogen and the pressure drop due to cooling and to enhanced hydrogen storage was recorded. To calculate the storage capacity of the sample, the experiment was repeated accurately under the same conditions for a blind sample (sea sand) of the same volume, which does not adsorb any hydrogen. The difference in hydrogen pressure drop is attributed exclusively to hydrogen storage.
After every adsorption step, the sample was heated under vacuum and the measurement was repeated for a new hydrogen pressure. This procedure ensured that every adsorption value was measured independently from the previous one. The experiment was been performed a second time using a different sample mass. The congruency in the measured storage values for the two experiments was additional evidence of the high accuracy of our measuring system.