High-throughput and data driven strategies for the design of deep-eutectic solvent electrolytes

Abstract

Deep eutectic solvents (DES) are an attractive class of materials with low toxicity, broad commercial availability, low costs and simple synthesis, which allows for tuning of their properties. We develop...

Review of low-cost self-driving laboratories in chemistry and materials science: the “frugal twin” concept

“…Since many of these complex characterization techniques are costly and designed for humans, the time and cost of building another robot to perform sample transfer exceed the benefits gained from automating every single task in the workflow for greater efficiency. Rodriguez et al 128 provide an excellent example of automating the most effective process steps such as synthesis (with an Opentrons OT-2 liquid handling robot), melting point determination, and electrochemical characterization for discovering new deep eutectic solvent electrolytes. Rodriguez et al 128 did not automate the processes of sample transfer or handling of existing equipment such as a dehydrator and vacuum oven because of the great opportunity cost.…”

“…Rodriguez et al 128 provide an excellent example of automating the most effective process steps such as synthesis (with an Opentrons OT-2 liquid handling robot), melting point determination, and electrochemical characterization for discovering new deep eutectic solvent electrolytes. Rodriguez et al 128 did not automate the processes of sample transfer or handling of existing equipment such as a dehydrator and vacuum oven because of the great opportunity cost.…”

Emerging Trends in Machine Learning: A Polymer Perspective

“…In these works, the goal is to find the component composition or processing conditions that optimizes some material property of interest. These include optimizing the degradation behavior of organic photovoltaic films, the gelation time and bacterial activity of living silk hydrogels, the melting point and electrical properties of deep eutectic solvents, and the physical properties and cost of surfactant solutions . The inclusion of cost as an optimization variable is notable in that it ensures that the final results balance performance against the bill of materials needed to make the sample, likely making the results more useful to industrial scientists.…”

“…Somewhat analogously to the above autonomous studies, Rodriguez et al used Principle Component Analysis (PCA) and T-distributed stochastic neighbor embedding (t-SNE) as visualization and screening tools . They leveraged these methods to visualize and aid in the process of down-selecting candidates for high-throughput analysis from a material library that was too large to analyze in full.…”

The laboratory of Babel: highlighting community needs for integrated materials data management

“…3 illustrates the application of a graph data model to an experimental procedure for one sample. 31 Projects or works that use this form of data model include the Event-Sourced Architecture for Materials Provenance (ESAMP), 32 Cript, 15 and Citrine Informatics' GEMD data model. 33 Explicitly organizing experimental data in a format that follows this structure would make tracking the provenance of any piece of data straightforward: given a sample or measurement, the chain of processes and samples can be traced backwards to determine where the sample came from, or forwards to see how a future step turned out.…”