The standard objection to MOFs circa 2005 was blunt: beautiful crystals, dead on contact with water. Park and colleagues answered with a design argument rather than a patch. Because the metal-imidazolate-metal bond angle of about 145 degrees mirrors the Si-O-Si angle of zeolites, zinc and cobalt imidazolates crystallize into zeolite topologies, and the resulting zeolitic imidazolate frameworks inherit something of zeolite toughness.
The demonstrations were theatrical by materials-chemistry standards. ZIF-8 survived a week in boiling water, refluxing benzene and methanol, and even 8 molar aqueous NaOH at 100 degrees, while holding thermal stability to roughly 550 C. Twelve ZIFs were reported, mapping imidazolate chemistry onto sodalite and other classic nets, with ZIF-8's 3.4 angstrom windows opening into 11.6 angstrom cavities, dimensions that later made it the default material for size-selective separations and an early commercial product.
A quarter of the field's subsequent ZIF literature is footnotes to this paper, including the correction literature: later studies showed ZIF-8's apparent window is effectively larger than 3.4 angstroms because the linkers swing, the famous gate-opening effect, and that its hydrothermal armor has limits under acidic or CO2-saturated aqueous conditions. The stability claims are robust but were occasionally over-generalized to the whole ZIF family, some members of which are far frailer.
As a piece of scientific rhetoric this paper is close to perfect: identify the field's central weakness, propose a structural principle that addresses it, then torture the material publicly until skeptics concede. Every applications-minded MOF researcher should know it cold.