Investigating automated chemical evolution of oil-in-water droplets

Parrilla Gutierrez, Juan Manuel (2016) Investigating automated chemical evolution of oil-in-water droplets. PhD thesis, University of Glasgow.

Full text available as:
[img] PDF
Download (21MB)
Printed Thesis Information:


One of the main unresolved questions in science is how non-living matter became alive in a process known as abiognesis, which aims to explain how from a primordial soup scenario containing simple molecules, by following a ``bottom up'' approach, complex biomolecules emerged forming the first living system, known as a protocell. A protocell is defined by the interplay of three sub-systems which are considered requirements for life: information molecules, metabolism, and compartmentalization. This thesis investigates the role of compartmentalization during the emergence of life, and how simple membrane aggregates could evolve into entities that were able to develop ``life-like'' behaviours, and in particular how such evolution could happen without the presence of information molecules.

Our ultimate objective is to create an autonomous evolvable system, and in order tp do so we will try to engineer life following a ``top-down'' approach, where an initial platform capable of evolving chemistry will be constructed, but the chemistry being dependent on the robotic adjunct, and how then this platform can be de-constructed in iterative operations until it is fully disconnected from the evolvable system, the system then being inherently autonomous.

The first project of this thesis describes how the initial platform was designed and built. The platform was based on the model of a standard liquid handling robot, with the main difference with respect to other similar robots being that we used a 3D-printer in order to prototype the robot and build its main equipment, like a liquid dispensing system, tool movement mechanism, and washing procedures. The robot was able to mix different components and create populations of droplets in a Petri dish filled with aqueous phase. The Petri dish was then observed by a camera, which analysed the behaviours described by the droplets and fed this information back to the robot. Using this loop, the robot was then able to implement an evolutionary algorithm, where populations of droplets were evolved towards defined life-like behaviours.

The second project of this thesis aimed to remove as many mechanical parts as possible from the robot while keeping the evolvable chemistry intact.
In order to do so, we encapsulated the functionalities of the previous liquid handling robot into a single monolithic 3D-printed device. This device was able to mix different components, generate populations of droplets in an aqueous phase, and was also equipped with a camera in order to analyse the experiments. Moreover, because the full fabrication process of the devices happened in a 3D-printer, we were also able to alter its experimental arena by adding different obstacles where to evolve the droplets, enabling us to study how environmental changes can shape evolution. By doing so, we were able to embody evolutionary characteristics into our device, removing constraints from the physical platform, and taking one step forward to a possible autonomous evolvable system.

Item Type: Thesis (PhD)
Qualification Level: Doctoral
Keywords: Chemical evolution, droplets, robots, liquid handling robots, microfluidic devices, genetic algorithm, evolution, surfactants, 3D printer.
Subjects: Q Science > QD Chemistry
T Technology > TA Engineering (General). Civil engineering (General)
T Technology > TP Chemical technology
Colleges/Schools: College of Science and Engineering > School of Chemistry
Supervisor's Name: Cronin, Prof. Leroy
Date of Award: 2016
Depositing User: MR Juan Manuel Parrilla Gutierrez
Unique ID: glathesis:2016-7744
Copyright: Copyright of this thesis is held by the author.
Date Deposited: 08 Nov 2016 13:05
Last Modified: 05 May 2022 08:35
Thesis DOI: 10.5525/gla.thesis.7744

Actions (login required)

View Item View Item


Downloads per month over past year