Design of Experiments for Optimization of Polyoxometalate Syntheses

Design of experiments (DOE) is a key method for optimizing physical processes by altering multiple variables at once to assess their effect. In chemistry, DOE explores a wider parameter space than the dominant “One Factor at a Time” (OFAT) method providing greater opportunity to explore the factors that can be used to optimize yield, purity, and to explore chemical space for new compounds. One area of chemistry that suffers from low yields and poor reproducibility but is full of hard to predict and interesting materials is polyoxometalate cluster science. Herein, we developed a DOE analysis methodology to explore the parameter space of polyoxometalate cluster formation to explore the subtle input effects that are known to have an impact on the product discovery, purity, and stability under the preparation conditions. Using a Plackett–Burman screening design, we analyzed the effect of six synthetic parameters in only 12 experiments, following up with a full factorial analysis of the three most significant factors to identify the key parameters in the successful synthesis of each. Based on this, we provide a useful template that produces the input data for automated synthesis based on DOE on other synthetic procedures. In our POM test cases, redox agent stoichiometry was found in three of the four systems studied to be significant factors with pH and temperature, which also found to be commonly important. The insights derived from this analysis were applied to design optimized synthetic procedures and improve the yield of the product by on average >33% from the highest reported literature yield. Thus, the DOE methodology outlined here is shown to yield insights into reaction optimization rapidly with facile experimental design and analysis even for complex multivariate synthetic procedures.