On the Mechanism of MOF‐5 Formation under Cathodic Bias

T considerable interest in the chemistry and applications of metal−organic frameworks (MOFs) has led to significant synthetic advances in this area, yet the details of the self-assembly process that give rise to these materials are still unclear. This is curious, but not unexpected: the crystallization of MOFs is difficult to study because their nucleation typically requires unpredictable induction periods that can vary from minutes to days, depending on the medium and metal−ligand pair, and is usually followed by rapid growth. Mechanistic studies are further complicated by the fact that many polymorphic structures are accessible from a single metal−ligand pair, although remarkable selectivity for a single crystalline phase is often observed under a given narrow set of conditions. This is evident in the rich literature on optimizing the synthesis of Zn4O(BDC)3 (MOF-5; BDC = 1,4benzenedicarboxylate), one of the most iconic materials in the field, where precise control of water content, temperature, concentration, and pH are necessary to selectively form the Zn4O(O2C−)6 secondary building units. 1−6 Because electrochemistry confers highly reproducible reaction environments addressable by several variables that can be changed independently, it serves as a unique platform for fabricating integrated thin films, and for studying the electrochemical growth of MOFs. As a widely recognized model system, we were particularly interested in the details of MOF-5 formation and its relation to other phases made from Zn and BDC2− ions. For instance, we previously reported that MOF-5 can be deposited selectively at room temperature in 15 min on FTO electrodes under cathodic bias. More recently, we have also shown that modulating the applied potential during electrolysis allowed access to heterobilayer structures of two Zn-BDC phases with high selectivity, presumably by regulating the pH gradient near the electrode surface. Although electrochemical deposition allows the formation of complex heterostructures with high selectivity, finding the optimal conditions for depositing a given phase required significant empirical screening. Surmising that a deeper understanding of the mechanism of cathodic electrodeposition in the Zn-BDC system would facilitate the application of this method to other metal−ligand systems, we present here a more in-depth study aimed at elucidating the role of water and the supporting anions in the crystallization process and phase selection. Our initial hypothesis was that hydroxide generation in the original MOF-5 electrochemical growth medium, which contained nitrate, water, and 1,4-benzenedicarboxylic acid, occurred by aqueous reduction of nitrate (eq 1). However, in principle, nitrate reduction does not require water. An electrochemical half-reaction can be written with 1,4benzenedicarboxylic acid as the proton source (eq 2). In other words, electrodeposition should also occur from an initially anhydrous medium, in the absence of added water. Furthermore, in aqueous systems, water itself may function as a viable hydroxide source (eq 3), potentially obviating the need for nitrate. Thus, if eqs 2 and 3 become operative in the absence of water and nitrate, respectively, the role of these reagents in the electrochemical growth of MOF-5 could be systematically studied. Although water reduction could produce hydroxide (eq 3), when the zinc source was switched from Zn(NO3)2·4.2H2O to Zn(ClO4)2·6H2O while holding all other experimental parameters constant, no MOFs were deposited (Figure S2 in the Supporting Information). This suggested that (1) nitrate was essential for electrochemical deposition, and (2) as a sole probase, water had a nearly negligible effect toward crystallization of MOF-5.