switchable solvent for lipid extraction from microalgae
Ying Du is a PhD student in the research group Sustainable Process Technology. Her supervisors are dr.ir. D.W.F. Brilman and prof.dr. S.R.A. Kersten from the faculty of Science and Technology.
Microalgae are considered one of the most promising sustainable feedstocks for lipid extraction to produce food ingredients, cosmetics, pharmaceutical products and biofuels. Reasons for the increased attention for microalgae in recent years is their rapid growth rate, and thus high productivity and high CO2 fixation rate, the absence of competition for arable land and (to lesser extent for) freshwater with terrestrial crops. Microalgae occur in a wide range of species, rich in proteins, lipids, carbohydrates, having the ability to produce high-valuable compounds, often lipid-based. For large scale production, however, developments on cost reduction are necessary. Next to the costs for cultivation, the energy intensity of drying and milling of algae prior to lipid extraction and of solvent recovery afterwards is a major obstacle. In the work described in this thesis it was aimed to find/evaluate a novel, effective, switchable solvent system for direct lipid extraction from (unbroken) microalgae in (concentrated) aqueous solutions.
In Chapter 3, several switchable solvent candidates were investigated on their lipid extraction performance and the polarity switching and phase splitting ability upon contacting with CO2. Secondary amine N-ethyl butyl amine (EBA) was selected for lipid extraction from aqueous slurries of fresh, unbroken Neochloris Oleoabundans microalgae because of its good performance on both switchability and lipid extraction performance. Moreover, this solvent allows the recovery of the lipid and of the extraction solvent via a simple phase splitting step in the presence of water, induced by contacting with pure CO2 at atmospheric pressure and ambient temperature. EBA reacts with CO2 in presence of water and forms EBA carbamate salts and EBA bicarbonate salts. The binary mixture EBA/water can form one single liquid phase, immiscible with e.g. sunflower lipid oil, even after converting only 25% of the initial amount of EBA. This latter aspect showed the possibility of lipid recovery from a semi-switched EBA solution. This semi-switching simultaneously prevents gel formation which might occur at near complete conversion of EBA into carbamate/bicarbonate salts. Solvent recovery after lipid separation is achieved by nitrogen stripping at a slightly elevated temperature (50-90°C), switching EBA back into its apolar form.
Extractions using wet and freeze dried algae as starting material were investigated for both Bligh & Dyer (B&D) extraction, commonly used as analytical technique to determine the total lipid content, and for EBA extraction. Extraction with wet algae using EBA was found to result in a higher total fatty acid (TFA) yield. Cell breaking increased the crude lipid yield to a certain extent for both methods. But extraction from non-broken algae resulted in higher TFA yield when using EBA. Therefore, for the strain Neochloris Oleoabundans, drying and cell breaking were not necessary for lipid extraction using EBA, while crude lipid yields and fatty acid yields were comparable with B&D extraction.
In Chapter 4, the lipid extraction yield from non-broken Neochloris oleoabundans slurries (~5% dry weight) using EBA was maximized for both single stage extractions, and for multistage extractions. The Neochloris oleoabundans microalgae were cultivated in fresh water (FW) and artificial seawater (ASW) under stressed and non-stressed conditions. The study showed that for wet algae a cell disruption step was not needed, although a considerable longer extraction time (12-18 h) was required to achieve the maximum yield at room temperature (22°C) and a solvent/feed ratio of 1:1 (w/w). Multiple extraction stages show that, depending on solvent to feed ratio, a second or even third extraction stage can be beneficial. For fresh water cultivated, non-stressed, non-broken Neochloris oleoabundans 13.1 wt.% of lipid extraction yield (based on dry algae mass) was obtained in a single stage extraction while 22.1 wt.% of lipid extraction yield was achieved after 4 times extraction.
For FW-stressed Neochloris oleoabundans, in one stage extraction the lipid extraction yield was 47.0 wt.% (all based on algae dry weight) of which 52.9 wt.% total fatty acids. A maximum yield of 61.3 wt.% was reached after four stages of extraction. After two extractions already 56.7 wt.% was obtained, suggesting that two stages might be optimal. The fatty acid profile of Neochloris oleoabundans was dominated by palmitic (C16:0), hexadecadienoic (C16:2), hexadecatetraenoic (C16:4), oleic (C18:1), linoleic (C18:2) and linolenic (C18:3) acids, and no significant differences in this fatty acid profile were found between the extraction methods and extraction stages tested.
The results in this work illustrate that the combination of stressing the freshwater cultivated microalgae and applying EBA as solvent in a single stage or multiple stages of extraction is very promising in the development of an energy efficient lipid extraction technology targeting non-broken, wet microalgae.
The study in Chapter 5 advanced the understanding on the equilibrium extraction of lipid from FW-stressed Neochloris oleoabundans. With the hypothesis that after extraction, the algae cells were completely filled with the organic solvent phase, having the same composition as the organic phase outside the cells, the model was successfully fitted to the experimental crude lipid yields of the four stage extractions at various solvent to feed ratios. By modelling it was found that nearly all fatty acids were released from the cell material, but not all is recovered due to the organic phase remaining inside the cell. This mechanism also explains the incomplete lipid recovery in a single extraction stage. The extraction yields estimated by the model showed good agreement with those obtained in experiments. The developed model can predict the amount of crude lipid being recovered from any stage of a multistage extraction process. The parameters of this model may need to be adjusted before the model can be applied to other algae, or other solvent. For common applied solvent to feed ratios, two extraction stages are sufficient for recovering most of the lipid containing fatty acid.
In Chapter 6, an alternative solvent switching method (in comparison to CO2 switching) for lipid separation and solvent recovery was investigated. The lower critical solution temperature (LCST) behavior of EBA-water mixtures and temperature-dependent phase behavior in the presence of lipid oil were studied. It was proven that the temperature responsive partitioning of EBA over aqueous and apolar phases can be used for oil recovery and also for lipid recovery from microalgae, when cycling the mixture temperature (indicatively) between 4°C and 20°C. The temperature at which LCST-like behavior is observed increases slightly with the oil content. The maximum EBA concentration which can be used for oil separation is 50 wt.%. The oil recovery efficiency can reach more than 90% when EBA fraction in the aqueous phase is below 0.3 wt./wt. In order to get an efficient oil recovery, the lipid concentration with respect to the organic phase should be more than 5 wt.%. A process for wet lipid extraction from microalgae with EBA, and then using the temperature responsive partitioning behavior of EBA for lipid recovery and subsequently for solvent regeneration was proposed.
Research on lipid extraction in a microalgae biorefinery context has been mainly targeting oil content enhancement, cell disruption method investigation, extraction efficiency improvement, etc. There has not been so much attention on the treatment of biorefinery residuals, such as raffinate streams and microalgae cells after lipid extraction. The treatment of raffinate and microalgae cells after wet lipid extraction is important for the viability of this microalgae biorefinery process because solvent loss will cause pollution to the environment and increase the process cost. In Chapter 7, several solvents/solvent combinations were evaluated for EBA recovery from both the raffinate and from the residual microalgae cells after extraction.
Dodecane was selected to recover EBA from the raffinate after lipid extraction because of its lower energy usage. Besides the EBA loss in the raffinate, there was also EBA loss in the algal cells residue, both inside and outside the cells. More than 85% of this EBA loss can be recovered by using simply water in one step extraction with a solvent/feed ratio 20:1. EBA was most likely completely removed from algal cells residue. The EBA that was not recovered was because of evaporative losses during the manual operation procedures in the lab, which one should be able to prevent in a large scale process. The EBA from the wash water and the raffinate streams were combined and a liquid-liquid extraction process using dodecane as extractant was applied to extract and recover the EBA from this aqueous stream. After extracting the EBA, the EBA was recovered from the dodecane solvent by distillation. These extraction and distillation processes were simulated and optimized using Aspen Plus V8.8. When pure dodecane was used as extractant, an extraction column should have 24 stages with solvent to feed ratio 1:1.8 (w/w) in order to meet the requirement for waste water at the point of discharge (740 mg/L). In this study, the solvents will be reused after distillation, for which the distillation targets are set at a distillate that contains 99.9 wt.% EBA and a bottoms product that contains 99.9 wt.% dodecane. When this regenerated 99.9 wt.% dodecane is used as extractant, the extraction column has to be at least 39 stages in order to have the raffinate that can meet legal requirements at the point of discharge. The RadFrac block in Aspen Plus was selected for the distillation calculation and optimization, resulting in a heat duty of 2.50 MJ/kg dry algae in the reboiler (Qreb) and -0.55 MJ/kg dry algae cooling duty for condenser (Qcond).
In Chapter 8, the two process options studied in this thesis for wet lipid extraction from microalgae with EBA were evaluated, with focus on their energy use. The difference between these two processes is that for switching for lipid recovery and solvent regeneration, one method uses CO2 to transform the solvent to the hydrophilic state and the other method uses temperature as the stimulus. Process designs have been made for both options to calculate and compare the energy flows. Results showed that the energy usage due to “EBA carbamate loss” accounts for more than 80% (49.7 MJ/kg lipid) of the overall energy requirement (61.7 MJ/kg lipid), making the use of CO2 switching for energy production via lipid extraction unviable, unless this carbamate reconversion is solved. More work on the realization of complete carbamate reconversion is therefore needed for this option.
A much lower energy usage (12.4 MJ/kg lipid) and a clear positive energy balance (22.4 MJ/kg lipid) for lipid extraction was achieved with the temperature switching process, making this a very promising method for extracting lipid from algae for use in energy applications.
Finally, the advantages, weaknesses and challenges of using EBA as switchable solvent for lipid extraction from wet microalgae slurries, as microalgae biorefinery extraction technology, are discussed.