Strategies for production and purification of polyhydroxyalkanoates using mixed microbial consortia: Study on fermented crude glycerol as a substrate

Polyhydroxyalkanoates (PHA) are a group of natural polyesters synthesised as storage polymers in prokaryotic microorganisms, and with applications as biodegradable and biobased plastics. Nowadays, commercial production is based on first-generation feedstocks, and is associated with high production costs and limited sustainability benefits over conventional fossil-based plastics.

One of the strategies to reduce operational demands is the use of mixed microbial consortia (MMC), which are cultivated under non-sterile conditions. Similarly, the use of second-generation feedstocks could lower the costs and life-cycle impacts related to the raw materials. The main goal of this thesis was to advance the knowledge on PHA production from second-generation feedstocks using MMC.

The study was based on crude glycerol as a feedstock, an abundant by-product from the biodiesel industry with current limited market applications. Yet, the research provided insights beyond this substrate, regarding both the production and purification of PHA. Crude glycerol can be directly converted to PHA in MMC.

Nonetheless, it presents some limitations, such as reduced PHA yields derived from the co-production of other storage polymers (glycogen). The strategy of the project was to perform a fermentation of crude glycerol prior to its use as a substrate for PHA production. Hence, the conversion of fermentation products - volatile fatty acids (VFA) and 1,3-propanediol (1,3-PDO) – to PHA was investigated.

The research showed for the first time in MMC, that 1,3-PDO could be converted to PHA. However, several limitations were identified, such as the relatively low yield of this conversion. Some of these limitations might be overcome with further research, but from the results obtained here, the use of 1,3-PDO and VFA did not show a clear advantage over a direct conversion of crude glycerol to PHA.

On the basis of these observations, an alternative approach was considered, where only the VFA fraction was converted to PHA, while 1,3-PDO was recovered as an additional high value product. This approach led to much higher PHA yields from the carbon consumed. Given that VFA represented a minor fraction of the fermentation products, the overall PHA yield from glycerol was not higher than by directly converting glycerol to PHA, but the combined process showed an overall higher carbon recovery to valuable products.

This fact, together with the higher rates of the process, could make this approach an interesting alternative to the conversion of crude glycerol to only PHA. Some of the possible limitations are discussed in this dissertation. The results above were obtained by investigating different culture enrichment strategies with distinctive selective pressures, which are the key to direct the metabolism to desired products when using MMC.

In the case of PHA, this is generally achieved with repeated cycles of availability and absence of substrate, which favour microorganisms that can store the carbon in the form of PHA and use it during fasting periods. An additional selective pressure was investigated in this study, consisting in limiting the nitrogen during substrate availability.

The application of the second strategy led to a net production of PHA from 1,3-PDO, while this substrate was mostly derived to growth under the first enrichment strategy. In a similar manner, selective conversion of VFA to PHA with 1,3-PDO recovery was attained by adaptation to the microbial community to only VFA.

Such observations might be applicable for other substrates besides 1,3-PDO. A major barrier during PHA production from second-generation substrates is their dilute carbon concentration, which leads to low values of productivity due to increased reactor volumes. In this PhD, membrane bioreactors were tested as a way to allow an exchange of bioreactor broth while keeping the cells in the bioreactor.

More specifically the study evaluated immersed pressure-driven and diffusion-based membrane bioreactors (iMBRs and dMBRs). In the dMBR configuration, the membranes tested did not provide enough VFA diffusion to meet the substrate consumption of the culture. Possible research directions to increase substrate diffusion are suggested.

On the other hand, iMBRs using hollow fibers and ceramic filters resulted in very high values of PHA productivity compared to current operations. The two iMBR filters offered similar results during fedbatch operation, but presented different limitations and advantages. In relation to the PHA purification, ammonia digestion was investigated as a method to solubilise non-PHA cell material.

Given the possibility to reuse ammonia as a nitrogen source for PHA production, this method had previously been recognised as an interesting alternative to present practices, which involve expensive solvents or generate large amounts of wastewater. However, earlier research showed high levels of PHA degradation.

The results obtained here demonstrated that the outcome of the digestion is very dependent on digestion conditions (especially in regards to the temperature), and that high levels of PHA purity, PHA recovery and thermal stability can be obtained with this method.