The development of industrially-relevant lipids from Rhodotorula species as a feedstock for fuels and commodity products

Abstract

Lipids are becoming an increasingly important chemical feedstock for the manufacture of biofuels, bioplastics, care-products and as a food source. Many of these consumer products are derived from petroleum resources, and therefore finding suitable replacements is a key engineering challenge. While first generation lipid feedstocks have shown potential to displace some fossil fuel use, lipids produced from current sources such as oil crops, cannot realistically meet the demand for these uses sustainably. One alternative is to produce microbial oils from oleaginous yeast. These have many advantages such as high growth rates, year-round productivity and high lipid yield.

The fatty acid profile of the lipids is extremely important in determining their eventual use. Oils high in oleic acid such as rapeseed oil are the most suitable biodiesel feedstock and also offer the highest potential for further chemical upgrading to polymers, higher value chemicals or aviation fuels. Alternatively, to replace palm oils in the cosmetic or food industries, high levels of saturated lipids are necessary. Rhodotorula sp., can produce high yields of lipid and has a simple fatty acid profile, composed mainly of C16 and C18 fatty acids. Using Design of Experiments it was shown that the fatty acid profile of R. glutinis could be tailored towards a desired application. A high culture temperature and high nitrogen ratio yielded mainly unsaturated oil, whereas a low culture temperature and high glucose loadings gave a more saturated profile. On transesterification, the oil high in monounsaturated esters yielded biodiesel with fuel properties akin to rapeseed methyl ester (RME), whereas the oil high in saturates was found to be suitable as a substitute for palm oil. In contrast, the lipid profile for R. minuta showed no such fluctuation.

One of the drawbacks to the commercialisation of this technology is the high production costs involved. Low energy ultrasound is known to have a positive effect on both biomass and ethanol production in S. cerevisiae. In an attempt to reduce processing costs, intermittent ultrasound with R. glutinis was undertaken to aim to improve glucose conversion efficiencies. Sonication was found to have no positive effect on the biomass or lipid accumulation when applied in the exponential phase of R. glutinis growth. However, on applying the sonication in the stationary phase, a beneficial impact was observed with the lipid coefficient being increased by 24%.

While it is unlikely to be economic to produce lipids from refined sugars, inexpensive carbon sources such as lignocellulosic hydrolysates or waste streams offer a promising alternative. Microbial growth on these feedstocks can however be challenging, due to the large range of sugars present in the hydrolysates as well as toxic compounds formed during the sugar extraction process. The potential of two biomass hydrolysates: depolymerised Miscanthus and household food waste were investigated, alongside the effects of the model inhibitory compounds and sugar substrates on the growth of Rhodotorula sp. While the Miscanthus hydrolysate was deemed unsuitable as a feedstock for lipid production, acid-hydrolysed food waste produced a promising feedstock for Rhodotorula sp. Biomass yields of approximately 10 g/L were produced for R. minuta and R. glutinis, with the resulting lipid profile being approximately 65% oleic acid (18:1) for both species.

One obstacle for lipid production from oleaginous microbes are the energy costs associated with the extraction and subsequent conversion into biodiesel (FAME). A one-step method to produce FAME by combining lipid extraction from R. glutinis using a microwave reactor with acid-catalysed transesterification was developed. Over 99% of the lipid was extracted using 25 wt.% H2SO4 over 20 min at 120 °C. At higher catalyst loadings, similar yields were achieved at a reaction time of 30 s. Equivalent yields of FAME were achieved compared to the traditional method of Soxhlet extraction, run with the same solvent system for 4 h. Under the best conditions, the energy required by the microwave was less than 20% of the energy content of the biodiesel produced. Finally, the energetics of the conversion of household food waste to oil (SCO process) using Rhodotorula sp. was compared to that of the anaerobic digestion (AD) of food waste. Oil production alone was deemed energetically unfeasible. However, a coupled SCO and AD plant may have economic viability as a waste-to-energy route, especially for the production of bulk commodities such as jet fuel, in which the energy generated from the methane can be used to power the SCO process.

Date of Award	2 Jul 2015
Original language	English
Awarding Institution	University of Bath
Sponsors	Airbus Group Ltd
Supervisor	Chris Chuck (Supervisor), Roderick Scott (Supervisor) & Matthew Davidson (Supervisor)