Automation, artificial intelligence, and autonomous machines are already having a significant impact across many industries and these technologies have huge potential for advancing chemical research. This thesis focuses on two complex physicochemical systems, oil-in-water droplets and chemical gardens, and how smart-automation can be a powerful tool for their research. Smart-automation is herein defined as ‘physically implemented platforms that undertake experiments autonomously as directed by an algorithm, rather than undertaking operations or experiments as directed by a human scientist’. Overall, this thesis illustrates how smart-automation is suitable for exploration, optimisation, discovery, and developing deeper understanding of oil-in-water droplet systems. It is proposed that the approach used in this work is also suitable for other systems studied within chemical and materials sciences, including chemical gardens. Initially, three new approaches for the analysis of oil-in-water droplet systems were developed, building on traditional analytical chemistry techniques and previous work exploring oil-in-water droplet systems using smart-automation.1 These analytical methods are orthogonal and complementary: a machine learning approach predicts bulk properties and correlates these to droplet behaviours; 1H NMR spectroscopy is used to measure the state of the system at a given time; and the use of pH indicators allows the visualization of spatiotemporal variations within the system. The machine learning approach enabled the prediction of droplet behaviour directly from the droplet formulation for the first time. This included the prediction of more instances of a rare cooperative ‘swarming’ droplet behaviour. Subsequently, a new smart-automated platform was developed that expands the exploration of oil-in-water droplet systems to include the variation of both the aqueous and oil phases. Thus, the aqueous and oil phase formulations could be optimised for a property of interest; here droplet movement is the targeted property. In doing so, both new and extreme droplet behaviours were identified. The newly described droplet behaviours are termed swarming, fusion, pulsing, and sorting. Next is reported a new platform for oil-in-water droplet experiments used to compare different formulation exploration methods. In undertaking this work, a previously unknown delicate temperature sensitivity of droplet behaviours was identified; this would not have happened without the use of smart-automation. This was studied in detail for one formulation which led to the identification of six phases of droplet motion and the preparation of a time-temperature phase diagram. Analytical methods including 1H NMR spectroscopy, DLS, and further droplet experiments enabled the proposal of mechanisms driving these phases of motion. Overall, this section of the work illustrates how smart-automation and the expert scientist can work together effectively. Following this, new avenues for reactive droplet constituents were explored and a formulation resulting in reliable droplet fusion was identified. Diels-Alder and carbonyl chemistry were then investigated as suitable reacting chemical systems to use in oil-in-water droplets. Amide and imine formation were both successfully realised in oil-inwater droplets with their macroscale effects on droplet behaviour visible to the naked eye. Finally, a wholly polyoxometalate based chemical garden system was identified and studied for the first time. This work bridges a significant difference between polyoxometalate based and classic chemical gardens as it has the same charge distribution as classic chemical gardens. Six growth regimes were identified and similarities and differences between this and previous systems are discussed. To finish, a smart-automated platform for exploring mixed chemical gardens is discussed and a prototype presented.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:754397 |
| Date | January 2018 |
| Creators | Points, Laurie James |
| Publisher | University of Glasgow |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Source | http://theses.gla.ac.uk/30792/ |
Page generated in 0.0022 seconds