The MixAlco process uses an anaerobic, mixed-culture fermentation to convert lignocellulosic biomass to carboxylate salts. The fermentation broth must be clarified so that only carboxylate salts, water, and minimal impurities remain. Carboxylate salts are concentrated by evaporation and thermally decomposed into ketones. The ketones can then be chemically converted to a wide variety of chemicals and fuels. The presence of excess lime in the thermal decomposition step reduced product yield. Mixtures of calcium carboxylate salts were thermally decomposed at 450 degrees C. Low lime-to-salt ratios (g Ca(OH)2/g salt) of 0.00134 and less had a negligible effect on ketone yield. In contrast, salts with higher lime-to-salt ratios of 0.00461, 0.0190, and 0.272 showed 3.5, 4.6, and 9.4% loss in ketone yield, respectively. These losses were caused primarily by increases in tars and heavy oils; however, a three-fold increase in hydrocarbon production occurred as well. To predict ketone product distribution, a random-pairing and a Gibbs free energy minimization model were applied to thermal decompositions of mixed calcium and sodium carboxylate salts. Random pairing appears to better predict ketone product composition. For sodium and calcium acetate, two types of mixed sodium carboxylate salts, and two types of mixed calcium carboxylate salts, activation energy (EA) was determined using three isoconversional methods. For each salt type, EA varied significantly with conversion. The average EA for sodium and calcium acetate was 226.65 and 556.75 kJ/mol, respectively. The average EA for the two mixed sodium carboxylate salts were 195.61, and 218.18 kJ/mol. The average EA for the two mixed calcium carboxylate salts were 232.78, and 176.55 kJ/mol. In addition, three functions of conversion were employed to see which one best modeled the experimental data. The Sestak-Berggren model was the best overall. Possible reactor designs and configurations that address the challenges associated with the continuous thermal decomposition of carboxylate salts are also presented and discussed. Methods of fermentation broth clarification were tested. Flocculation showed little improvement in broth purity. Coagulation yielded broth of 93.23% purity. Filtration using pore sizes from 1 micrometer to 240 Daltons increased broth purity (90.79 to 98.33%) with decreasing pore size.

With increasing energy demand, decreasing oil supply, and continuously accumulating waste in landfills, the interest in converting lignocellulosic biomass to liquid fuels has grown. The MixAlco™ process requires no exogenous enzymes, no sterility, can be adapted to any biodegradable feedstock, and converts lignocellulosic biomass into valuable chemicals and transportation fuels. This work focuses on the effects different feedstocks and inocula have on mixed-acid/hydrogen fermentations. When volatile solids (VS) are digested, mixed-acid fermentations produce hydrogen gas as a secondary byproduct. Hydrogen is only produced when there is an excess of NADH within the cell and when the energy selectivity (gamma) of the system has not been met. Continuous fermentations of paper produced 16.7 g carboxylic acid/L and 15.7 mL H2/g VS digested. Continuous fermentations of pretreated bagasse produced 17.1 g carboxylic acid/L and 41.1 mL H2/g VS digested. Both fermentations produced a fraction of the theoretical amount of hydrogen. The paper fermentation had a hydrogen percent yield of 6.9 percent, whereas the bagasse fermentation had a hydrogen percent yield of 22.6 percent. Hydrogen production was capped at this level because gamma had been met for these systems. The Bioscreening Project, a joint project between three departments, sought to improve the MixAlco™ process by finding natural cultures containing high biomass converters and high acid producers. A total of 505 inoculum samples were collected from 19 sites and screened using paper and yeast extract fermentations. The best converters were analyzed with Continuum Particle Distribution Modeling (CPDM). Nine inocula were run in paper and yeast extract countercurrent fermentations in which the overall performance varied less than 13 percent. Comparisons between six countercurrent train cultures showed an average culture similarity of 0.40 (Yue-Clayton similarity). With the dissimilar microbial cultures and the very similar fermentation performance, the performance of the MixAlco™ process depends on fermentation conditions, not on the microorganisms. Batch fermentations of office paper wastes, pineapple residue, Aloe vera rinds, wood molasses, sugar molasses, extracted algae, non-extracted algae, crude glycerol, obtained from the biodiesel process, and pretreated water hyacinths produced sufficient carboxylic acids and had sufficiently high conversions to be viable substrates for the MixAlco™ process.

With the inevitable depletion of the petroleum supply and increasing energy demands in the world, interest has been growing in bioconversion of lignocellulosic biomass (e.g., sugarcane bagasse). Lignocellulosic biomass is an abundant, inexpensive, and renewable resource. Most of current conversion technologies require expensive enzymes and sterility. In contrast, the patented MixAlco process requires no enzymes or sterility, making it attractive to convert lignocellulosic biomass to transportation fuels and valuable chemicals. This study focuses on pretreatment and thermophilic fermentation in the MixAlco process. Ammonium bicarbonate (NH4HCO3) was discovered to be a better pH buffer than previously widely used calcium carbonate (CaCO3) in anaerobic fermentations under thermophilic conditions (55°C). The desired pH should be controlled within 6.5 to 7.5. Over 85% acetate content in the product was found in paper fermentations and bagasse fermentations. Hot-lime-water-treated bagasse countercurrent fermentations buffered by ammonium bicarbonate achieved 50–60% higher total product concentrations than those using calcium carbonate. It was nearly double in paper batch fermentations if the pH was controlled around 7.0. Ammonium bicarbonate is a “weak” methane inhibitor, so a strong methane inhibitor (e.g., iodoform) is still required in ammonium bicarbonate buffered fermentations. Residual calcium salts did not show significant effects on ammonium bicarbonate buffered fermentations. Lake inocula from the Great Salt Lake, Utah, proved to be feasible in ammonium bicarbonate buffered fermentations. Under mesophilic conditions (40°C), the inoculum from the Great Salt Lake increased the total product concentration about 30%, compared to the marine inoculum. No significant fermentation performance difference, however, was found under thermophilic conditions. The Continuum Particle Distribution Model (CPDM) is a powerful tool to predict product concentrations and conversions for long-term countercurrent fermentations, based on batch fermentation data. The experimental acid concentrations and conversions agree well with the CPDM predictions (average absolute error < 15%). Aqueous ammonia treatment proved feasible for bagasse. Air-lime-treated bagasse had the highest acid concentration among the three treated bagasse. Air-lime treatment coupled with ammonium bicarbonate buffered fermentations is preferred for a “crop-tofuel” process. Aqueous ammonia treatment combined with ammonium bicarbonate buffered fermentations is a viable modification of the MixAlco process, if “ammonia recycle” is deployed.

Using mixed-culture fermentation, the carboxylate platform produces carboxylic acids, which are chemically converted into chemicals and fuels. To optimize the mixed-acid fermentation, different bioreactor configurations and operating modes were investigated. Intermittent air exposure did not affect fermentation performance and bacterial profiles, but reduced the high-molecular-weight carboxylic acids. The microbial flora contained strict and facultative microbes, suggesting the presence of a facultative anaerobic community existing in a biofilm. Compared to countercurrent trains, propagated fixed-bed fermentations have similar selectivity and acid distribution, but lower yield, conversion, productivity, and acid concentration. One- to six-stage countercurrent fermentations were operated with similar conditions. Fewer stages increased conversion, whereas more stages increased acid concentration and selectivity. One to four stages achieved similar yield, and four to six stages achieved similar maximum acid concentration. Maximum conversion was achieved with a single stage. Recycling residual biomass retained microorganisms and nutrients and increased yield and productivity. Relative to lower biomass reflux, higher reflux increased conversion, decreased selectivity, and did not affect yield. The recommended carbon-nitrogen ratio is ~24 g carbon/g nitrogen. In four-stage fermentations, recycle to the second fermentor and in parallel to the first three fermentors was optimal. Fermentations with excess or insufficient nitrogen had higher selectivity, but decreased yield and conversion. The glucose-utilization assay is a rapid and repeatable method for determining the amount of microbial activity in a sample. This method determined ~25% efficiency of a new cell separation method. In continuous fermentation, compared to no cell recycle, recycling cellular biomass increased selectivity and yield, but decreased conversion. Compared to lower cell reflux, higher reflux increased productivity, yield, and conversion, but decreased selectivity. Compared to residual biomass recycle, cell recycle had increased selectivity and yield, but decreased conversion. A new method to screen and rank inoculum sources from natural environments was successfully developed and tested.

About 8 tons of dry undigested solid waste is generated by the MixAlco process for every 40 tons of food residue waste fed into the process. This MixAlco process produces liquid fuels and the sludge generated can be further converted into synthesis gas using the process of pyrolysis. The hydrogen component of the product synthesis gas may be separated by pressure swing adsorption and used in the hydrogenation of ketones into fuels and chemicals. The synthesis gas may also be catalytically converted into liquid fuels via the Fischer-Tropsch synthesis process. The auger-type pyrolyzer was operated at a temperature between 630-770 degrees C and at feed rates in the range of 280-374 g/minute. The response surface statistical method was used to obtain the highest syngas composition of 43.9 +/- 3.36 v % H2/33.3 +/- 3.29 v % CO at 740 degrees C. The CH4 concentration was 20.3 +/- 2.99 v %. For every ton of sludge pyrolyzed, 5,990 g H2 (719.3 MJ), 65,000 g CO (660 MJ) and 21,170 g CH4 (1055.4 MJ) were projected to be produced at optimum condition. At all temperatures, the sum of the energies of the products was greater than the electrical energy needed to sustain the process, making it energy neutral. To generate internal H2 for the MixAlco process, a method was developed to efficiently separate H2 using pressure swing adsorption (PSA) from the synthesis gas, with activated carbon and molecular sieve 5A as adsorbents. The H2 can be used to hydrogenate ketones generated from the MixAlco process to more liquid fuels. Breakthrough curves, cycle mass balances and cycle bed productivities (CBP) were used to determine the maximum hydrogen CBP using different adsorbent amounts at a synthesis gas feed rate of 10 standard lpm and pressure of 118 atm. A 99.9 % H2 purity was obtained. After a maximum CBP of 66 % was obtained further increases in % recovery led to a decrease in CBP. The synthesis gas can also be catalytically converted into liquid fuels by the Fischer-Tropsch synthesis (FTS) process. A Co-SiO2/Mo-Pd-Pt-ZSM-5 catalyst with a metal-metal-acid functionality was synthesized with the aim of increasing the selectivity of JP-8 (C10-C17) fuel range. The specific surface areas of the two catalysts were characterized using the BET technique. The electron probe microanalyzer (with WDS and EDS capabilities) was then used to confirm the presence of the applied metals Co, Mo, Pd and Pt on the respective supports. In addition to the gasoline (C4-C12) also produced, the synthesis gas H2:CO ratio was also adjusted to 1.90 for optimum cobalt performance in an enhanced FTS process. At 10 atm (150 psig) and 250 degrees C, the conventional FTS catalyst Co-SiO2 produced fuels rich in hydrocarbons within the gasoline carbon number range. At the same conditions the Co-SiO2-Mo-Pd-Pt/HZSM-5 catalyst increased the selectivity of JP-8. When Co-SiO2/Mo-Pd-Pt-HZSM-5 was used at 13.6 atm (200 psig) and 250 degrees C, a further increase in the selectivity of JP-8 and to some extent diesel was observed. The relative amounts of olefins and n-paraffins decreased with the products distribution shifting more towards the production of isomers.