The use of cultured frustule-based microcontainers for protecting and controlling the release of corrosion inhibitors within primers could offer improvements upon the existing diatomaceous earth system. Potential benefits include higher particle consistency, finer and more complex porosity (affecting inhibitor release), and the ability to alter said porosity via in-vivo modification techniques. Within the field of frustule microcontainer production, the research gap relates to the selection, extraction and in-vivo modification of diatom frustules, whereas extensive literature exists on frustule in-vitro modification. The first part of this work is dedicated to diatom selection for frustule microcontainer production. A set of criteria for diatom screening are drafted and a three step process to select appropriate species
is developed. This selection process is applied to a set of available species and used to select two diatoms for testing, Thalassiosira weissflogii and Discostella pseudostellignera, in addition to the already possessed Chaetoceros calcitrans and Cilindrotheca fusiformis. Criteria for diatom selection are later refined based on findings to focus mainly on degree of silicification and frustule porosity.
This is followed by work on diatom culturing. Cultures of all four diatom types are analyzed for
growth dynamics properties via a month long daily observation batch culturing experiment. Hemocytometer slides are imaged and automated cell count macros are used for analysis via ImageJ. The resulting data is fitted to the modified gompertz function to obtain estimates for parameters of interest. Adequate estimates are produced for Chaetoceros calcitrans and Cilindrotheca fusiformis, while Thalassiosira weissflogii and Discostella pseudostellignera are found to not be fully adapted to culturing
conditions. The third part of this work is devoted to frustule extraction method testing for the purpose of obtaining
unfractured and unseparated frustule microcontainers. Frustule extraction testing makes use of a
preparatory testing campaign attempting a range of methods and respective variations, followed by a more standardized test campaign containing three methods taken from literature and two methods developed
within this work. Results indicate the importance of silicification in determining frustule damage
tolerance to cleaning methods. Solvent based methods are found to be ideal in minimizing valve separation. A final procedure making use of a three step washing process, followed by a graded ethanol series and two hours of 400 ° incineration is found to produce adequate microcontainers of Discostella pseudostellignera. All other diatoms tested are found to be unsuitable for the intended application. Frustule extraction testing is followed by upscaling of the culturing and extraction procedures. Upscaled culture volume is estimated based on the required microcontainer mass for loading and release testing, and the extraction procedure is adjusted for increased volumes. Culture upscaling in bubble column photobioreactors is successful, yielding cell counts comparable to growth tracking cultures. Upscaling of the extraction procedures results in lower sample quality than those produced for extraction testing, confirming the need for additional development of extraction procedures. The fifth part of this work makes use of in-vivo modification of diatoms through the addition of trace metallic elements to the culture medium to promote changes in frustule morphology. All trace metals are bioaccumulated by the tested diatoms, even at extremely low dopant concentrations. Aluminum is found to consistently reduce the number of valve surface pores. Cerium doping results in a large data scatter, but may increase pore size. Zirconium appears to increase pore size, but reduce pore number. Ultimately, it is determined that more research is required to better understand the effect of such dopants on frustule morphology. The last part of this work focuses on cerium nitrate loading and release of Discostella pseudostellingera frustule microcontainers. Frustules of all modification types are bulk-loaded with cerium nitrate and analyzed using EDS to confirm loading. All modified frustules are found to have acceptable outer surface adsorption capability for cerium nitrate, although there is substantial variation between individual frustules within samples. The loading process is found to lead to partial frustule dissolution and substantial losses. Normalized release rates of loaded modified frustule microcontainers are found to be roughly equivalent to reference diatomaceous earth samples, although total release is much lower. It is then found that the loading procedure applied does not allow for cerium ingress and absorption loading due to entrapped air unable to escape from frustule nanopores, explaining the comparable release rate to the diatomaceous earth microcontainers. It is determined that likely with absorption and internal surface adsorption loading, likely the delay effect would be more significant.