Development of biological platform for the autotrophic production of biofuels

Open Access
Author:
Khan, Nymul Ehsan
Graduate Program:
Chemical Engineering
Degree:
Doctor of Philosophy
Document Type:
Dissertation
Date of Defense:
March 02, 2015
Committee Members:
  • Wayne Roger Curtis, Dissertation Advisor
  • Manish Kumar, Committee Member
  • Esther Winter Gomez, Committee Member
  • John Michael Regan, Special Member
Keywords:
  • Chemolithoautotroph
  • CO2 fixation
  • Rhodobacter capsulatus
  • Ralstonia eutropha
  • triterpene
  • squalene
  • botryococcene
  • chemostat
Abstract:
The research described herein is aimed at developing an advanced biofuel platform that has the potential to surpass the natural rate of solar energy capture and CO2 fixation. The underlying concept is to use the electricity from a renewable source, such as wind or solar, to capture CO2 via a biological agent, such as a microbe, into liquid fuels that can be used for the transportation sector. In addition to being renewable, the higher rate of energy capture by photovoltaic cells than natural photosynthesis is expected to facilitate higher rate of liquid fuel production than traditional biofuel processes. The envisioned platform is part of ARPA-E’s (Advanced Research Projects Agency - Energy) Electrofuels initiative which aims at supplementing the country’s petroleum based fuel production with renewable liquid fuels that can integrate easily with the existing refining and distribution infrastructure (http://arpa-e.energy.gov/ProgramsProjects/Electrofuels.aspx). The Electrofuels initiative aimed to develop liquid biofuels that avoid the issues encountered in the current generation of biofuels: (1) the reliance of biomass-derived technologies on the inefficient process of photosynthesis, (2) the relatively energy- and resource-intensive nature of agronomic processes, and (3) the occupation of large areas of arable land for feedstock production. The process proceeds by the capture of solar energy into electrical energy via photovoltaic cells, using the generated electricity to split water into molecular hydrogen (H2) and oxygen (O2), and feeding these gases, along with carbon dioxide (CO2) emitted from point sources such as a biomass or coal-fired power plant, to a microbial bioprocessing platform. The proposed microbial bioprocessing platform leverages a chemolithoautotrophic microorganism (Rhodobacter capsulatus or Ralstonia eutropha) naturally able to utilize these gases as growth substrates, and genetically modified to produce a triterpene hydrocarbon fuel molecule (C30+ botryococcenes) native to the alga Botryococcus braunii. In addition to the genetic modification and bioreactor performance studies of these organisms for the production of botryococcene or squalene, the research examined the potential economic feasibility of the proposed platform through the use of bioreactor, microbial energetic models and experimentally measured growth yield and maintenance coefficients. In order to carry out an economic analysis, a process model was created in Aspen with the bioreactor at the center. This is presented in Chapter 2. The model looked at the effects of growth yield and maintenance coefficients of R. capsulatus and R. eutropha, reactor residence time, gas-liquid mass-transfer coefficients, gas composition and specific fuel productivity on the volumetric productivity and fuel yield on H2. It was found that the organism with the lowest maintenance coefficient performed better under very low growth rates evaluated in the model (based on residence time through the reactor) performed the best. The optimum parameter values were then used to determine the capital and operating costs for a 5000 bbl-fuel/day plant and the final fuel cost based on the Levelized Cost of Electricity (LCOE). It was found that under the assumptions used in this analysis and crude oil prices, the LCOE required for economic feasibility must be less than 2¢/kWh. While not feasible under current market prices and costs, this work identifies key variables impacting process cost and discusses potential alternative paths toward economic feasibility. This was the best case scenario of the two organisms evaluated, and an optimally suited organism with high growth yield and low maintenance coefficient should obviously improve the economics. This economic constraint will improve with the rise of fossil fuel prices, which should occur if the environmentally detrimental effects of their use are factored into the price, through higher taxation, for example. A review of the current status of metabolic engineering of chemolithoautotrophs is carried out in order to identify the challenges and likely routes to overcome them. This is presented in Chapter 3 of this dissertation. The initial metabolic engineering and bioreactor studies was carried out using a number of gene-constructs on R. capsulatus and R. eutropha. The gene-constructs consisted of Plac promoter followed by the triterpene synthase genes (SS or BS) and other upstream genes. In R. capsulatus, by genetically supplementing the methylerithrotol phosphate (MEP) pathway and supplementing the growth with glucose, it was found that the triterpene synthase enzymes were substrate-limited i.e. depended on the carbon-flux to them. A comparison of the production of triterpenes were done in the different growth modes that R. capsulatus was capable of growing – aerobic heterotrophic, anaerobic photoheterotrophic and aerobic chemoautotrophic. Small-scale testing (<50 ml) under typical (un-supplemented) growth conditions showed that the per-cell triterpene production levels were surprisingly similar in all the different growth modes (around 5 mg/gDW). However, the results were much improved when tested in controlled fed-batch bioreactors, capable of reaching significantly higher cell densities. In the heterotrophic case, production was found to increase up to 40 mg/L (~11 mg/gDW), unfortunately inhibited by some sort of toxic effect at OD660 around 12. Autotrophic growth on H2, O2 and CO2, on the other hand, showed no such effect and growth occurred well up to an OD660 of 17 (corresponding to about 7 gDW/L), limited only by the mass-transfer of the gases and triterpene productivity increased continuously to greater than 100 mg/L (16 mg/gDW) in the batch mode. Continuous autotrophic operation further increased the specific titer to 23 mg/gDW, reaching a steady state. The specific productivity was found to be around 0.5 mg/gDW-hr. This demonstrated that autotrophic productivity could likely be improved much further by increasing the available mass-transfer of the reactor. These efforts are presented in Chapter 4 of this dissertation.