Tuning Two-Dimensional Phthalocyanine Dual Site Metal-Organic Framework Catalysts for the Oxygen Reduction Reaction.

Metal-organic frameworks (MOFs) offer an interesting opportunity for catalysis, particularly for metal-nitrogen-carbon (M-N-C) motifs by providing an organized porous structural pattern and well-defined active sites for the oxygen reduction reaction (ORR), a key need for hydrogen fuel cells and related sustainable energy technologies. In this work, we leverage electrochemical testing with computational models to study the electronic and structural properties in the MOF systems and their relationship to ORR activity and stability based on dual transitional metal centers. The MOFs consist of two M1 metals with amine nodes coordinated to a single M2 metal with a phthalocyanine linker, where M1/M2 = Co, Ni, or Cu. Co-based metal centers, in particular Ni-Co, demonstrate the highest overall activity of all nine tested MOFs. Computationally, we identify the dominance of Co sites, relative higher importance of the M2 site, and the role of layer M1 interactions on the ORR activity. Selectivity measurements indicate that M1 sites of MOFs, particularly Co, exhibit the lowest (<4%), and Ni demonstrates the highest (>46%) two-electron selectivity, in good agreement with computational studies. Direct in situ stability characterization, measuring dissolved metal ions, and calculations, using an alkaline stability metric, confirm that Co is the most stable metal in the MOF, while Cu exhibits notable instability at the M1. Overall, this study reveals how atomistic coupling of electronic and structural properties affects the ORR performance of dual site MOF catalysts and opens new avenues for the tunable design and future development of these systems for practical electrochemical applications.

Interpretable Machine Learning Models for Practical Antimonate Electrocatalyst Performance.

Computationally predicting the performance of catalysts under reaction conditions is a challenging task due to the complexity of catalytic surfaces and their evolution in situ, different reaction paths, and the presence of solid-liquid interfaces in the case of electrochemistry. We demonstrate here how relatively simple machine learning models can be found that enable prediction of experimentally observed onset potentials. Inputs to our model are comprised of data from the oxygen reduction reaction on non-precious transition-metal antimony oxide nanoparticulate catalysts with a combination of experimental conditions and computationally affordable bulk atomic and electronic structural descriptors from density functional theory simulations. From human-interpretable genetic programming models, we identify key experimental descriptors and key supplemental bulk electronic and atomic structural descriptors that govern trends in onset potentials for these oxides and deduce how these descriptors should be tuned to increase onset potentials. We finally validate these machine learning predictions by experimentally confirming that scandium as a dopant in nickel antimony oxide leads to a desired onset potential increase. Macroscopic experimental factors are found to be crucially important descriptors to be considered for models of catalytic performance, highlighting the important role machine learning can play here even in the presence of small datasets.

Understanding the Stability of Manganese Chromium Antimonate Electrocatalysts through Multimodal In Situ and Operando Measurements.

Improving electrocatalyst stability is critical for the development of electrocatalytic devices. Herein, we utilize an on-line electrochemical flow cell coupled with an inductively coupled plasma-mass spectrometer (ICP-MS) to characterize the impact of composition and reactant gas on the multielement dissolution of Mn(-Cr)-Sb-O electrocatalysts. Compared to Mn2O3 and Cr2O3 oxides, the antimonate framework stabilizes Mn at OER potentials and Cr at both ORR and OER potentials. Furthermore, dissolution of Mn and Cr from Mn(-Cr) -Sb-O is driven by the ORR reaction rate, with minimal dissolution under N2. We observe preferential dissolution of Cr totaling 13% over 10 min at 0.3, 0.6, and 0.9 V vs RHE, with only 1.5% loss of Mn, indicating an enrichment of Mn at the surface of the particles. Despite this asymmetric dissolution, operando X-ray absorption spectroscopy (XAS) showed no measurable changes in the Mn K-edge at comparable potentials. This could suggest that modification to the Mn oxidation state and/or phase in the surface layer is too small or that the layer is too thin to be measured with the bulk XAS measurement. Lastly, on-line ICP-MS was used to assess the effects of applied potential, scan rate, and current on Mn-Cr-Sb-O during cyclic voltammetry and accelerated stress tests. With this deeper understanding of the interplay between oxygen reduction and dissolution, testing procedures were identified to maximize both activity and stability. This work highlights the use of multimodal in situ characterization techniques in tandem to build a more complete model of stability and develop protocols for optimizing catalyst performance.

Electro-assisted regeneration of pH-sensitive ion exchangers for sustainable phosphate removal and recovery.

Removal and recovery of phosphate from wastewater can minimize deleterious environmental impacts and supplement fertilizer supply. Hybrid anion exchangers (HAIX, with doped ferric oxide nanoparticles (FeOnp)) can remove phosphate from complex wastewaters and recover concentrated phosphate solutions. In this study, we integrate HAIX with a weak acid cation exchanger (WAC) to enrich phosphate and calcium in mild regenerants and precipitate both elements for recovery. We demonstrated an electro-assisted regeneration approach to avoid strong acid and base input. Based on demonstrated pH sensitivities of both materials, electrochemically produced mild electrolytes (pH 3 and pH 11), which are 100-1000 times less concentrated than typical regenerants, preserved 80% WAC and 50% HAIX capacities over five batch adsorption-regeneration cycles. FeOnp in HAIX facilitated regeneration due to pH sensitivity and their likely distribution on the resin particle surface, which reduced intraparticle diffusion path length. In column tests, repeatable phosphate removal (> 95%) from synthetic wastewater (3 mg P/L) was achieved with 20 kWh/kg P specific energy consumption. After removal, a similar 50% HAIX regeneration efficiency as batch experiments was achieved. In spent regenerant, more than 95% phosphorus was recovered as hydroxyapatite. This novel approach enhances ion exchange by minimizing chemical inputs.