From Scientific Breakthrough to Business
When Ramon Gonzalez discovered the secret to coaxing industrial E. coli strains to anaerobically ferment glycerin, commonly found in ethanol and biodiesel plant waste streams, he suspected it would be significant to the biofuels, ethanol and biochemicals industries. GlycosBio, the company founded around that discovery, is now on the verge of commercializing its biochemical production capabilities.
Glycerin is a waste product of biodiesel production and is found in ethanol thin stillage, along with waste from other industries. It was in abundant supply at the time of his discovery, as it is now. "It was obviously very important because of the belief that it was not possible," says Gonzalez, now a William W. Akers assistant professor in the Department of Chemical and Biomolecular Engineering at Rice University in Houston. Another reason the discovery was significant is it came at a time when so many companies were looking for technologies that would enable them to make a profit from their glycerin. "You can't always marry those two things: finding something that is really exciting scientifically speaking, and that same thing that is exciting has an immediate application," he says, adding that, for those reasons, he sees it as the most significant discovery he's made public in his career to date. It would have been very good anyway having just one of the two benefits, he adds. "It's very appealing for me to think about."
Gonzalez's motivation came from the scientific challenge the project presented and his interest in finding a use for the abundance of glycerin, a problem he anticipated would become larger when more companies began making biodiesel.
The bottom line is that production costs of biofuels and biochemicals depend heavily on feedstock costs and Gonzalez's discovery enables cost-effective production through waste products. "Having a cheap, abundant feedstock is key in the production process," he says.
It's All About Environment
Gonzalez says his technique comes down to the microbe's environment, and compares it to human beings. "We behave in a given way based on what we have around us," he explains. "If it is hot, we try to cool off. If it is cold, we try to put a few layers of clothes on. Organisms behave depending on their environment." The proprietary environmental triggers Gonzalez discovered are valid for all E. coli strains he has tested, which are industrial and not strains linked to food poisoning. "The actual discovery is not really a specific strain," he says. "It's more the environment in which we put that strain to make it happy and able to ferment glycerol."
It was previously thought that E. coli would not ferment glycerin because of the actual conceptual model of glycerin fermentation. "Back then, when we started working on this, it was thought that in order for a microorganism to be able to ferment glycerol, it would need to be able to produce another product," Gonzalez explains. "That product was 1,3 propanediol. E. coli does not have the ability to produce that 1,3 propanediol." Because of that, most scientists thought E. coli could not ferment glycerin. Even under Gonzalez's environmental conditions, E. coli still will not produce 1,3 propanediol, he says, but it does produce something similar. "So it is not [1,3 propanediol] that actually enables glycerol fermentation," Gonzalez confirms. "The reason why they didn't see it was because they didn't use the right conditions."
Utilizing that Feedstock
GlycosBio co-founders saw promise and potential in Gonzalez's discovery and hope its expected impact can be realized through their company. "It was more a strong understanding of the genetics and the process conditions together that allowed [Gonzalez] to discover what other people had been unable to find," said Paul Campbell, GlycosBio chief science officer. "If you provide oxygen, it will eat it very quickly and that's not a problem, but that's not interesting. Because if you have oxygen present, all you make is more biomass, more cells. You don't make any interesting chemicals."
"That discovery was so unique that it actually had enough interest from the venture capital community to make an investment in that discovery to commercialize that microbe to leverage it into a strategy to make biofuels or biochemicals," says Richard Cilento, GlycosBio chairman.
"The uniqueness about the business is there hasn't been a great deal of investment or research into existing low-value or unique feedstock sources," Cilento says, adding that low-value feedstocks are what drove interest and excitement from GlycosBio founders in helping Gonzalez commercialize his discovery. Most public domain work focuses on sugars, a single-feedstock strategy, which Cilento says is a mistake.
Besides glycerin, GlycosBio's microbes can ferment feedstocks such as gums, fatty acids, crude oil extracts from animals or plants, and the system is even compatible with algae. Most of those feedstocks are waste products from certain industries such as food rendering or oleochemicals production. Once operating at commercial scale, GlycosBio will license its microbe technologies to companies such as biochemicals and biofuels producers, which will produce the desired chemicals from their waste streams for a cost savings or even a possible profit. For instance, adding fermentation equipment and microbes to a biodiesel plant can create a biochemical worth 70 cents to $1.20 per pound from crude glycerin, which is valued at 6 cents per pound, Cilento says. All biodiesel and ethanol plants are target customers, along with chemical companies that have feedstocks available to them under their umbrellas. "GlycosBio focuses on the science and technology, and the [customer] will focus on production," Cilento says. Biofuels companies most likely would sell the biochemicals they produce, while chemical companies would consume them, Cilento explains, as they are a 100 percent replacement for petrochemicals.
The palm oil industry in Malaysia also has potential to benefit from the company's technologies, Cilento says. "There's a tremendous amount of free fatty acids that come out of the refining palm oil process," he says. Malaysia also has a number of biodiesel manufacturers, opening the door for GlycosBio to ferment their glycerol waste streams, he adds. "So our market and our customer segment will definitely be more of an international flavor, and our strategy is to match where the available feedstock is that marries us to a region where they'd prefer to make green chemicals or biochemicals," he says.
GlycosBio started with native E. coli strains and, with Gonzalez's help, continues to improve the process to make more biochemicals. "Once you get the microorganism to eat glycerol, you need to tweak it," Gonzalez says. In its native form, it will produce primarily ethanol, but with metabolic engineering, it can make butanediol, hydrogen, 1,2 propanediol, and organic acids such as succinic and lactic acids. "We keep engineering to produce other products, depending on the needs of the market," Gonzalez says. "We keep adding to the product pipeline." The company uses mostly E. coli, but also dabbles in other microbes, according to Campbell.
How it's Done
"A lot of our process is fairly flexible," Campbell says. "In other words, we'll reuse a lot of our equipment even if we change the chemical we are going to make." The front end of the process is mostly the same no matter what end product the technology manufactures, according to Campbell. The difference comes in the back end when separating the chemical from the fermentation broth. "That equipment will be dependent on the chemical itself," he says. For example, if the end product is ethanol, that last step will be distillation, whereas if the end product is an organic acid, that step most likely would be ion exchange.
The glycerin requires little on-site pretreatment other than heating it up a bit to pasteurize it. The same minimal pretreatment is used for fatty acids or fatty acid-rich feedstock. After pretreatment, the glycerin is mixed with a salt solution and inoculated with the microbes. As they grow, the desired end product will be produced. Different microbes are used to produce different end products, Cilento says. "If there's a different chemical, there's a different microbe," he says. "Even for one chemical, there are several microbes."
It takes about 24 to 72 hours for the microbes to ferment the glycerin or other waste feedstock, Campbell says. After that, the fermentation "beer" is pumped into the recovery equipment. "At that point, the facility looks more similar to a traditional petrochemical plant because all you're doing is recovering your molecule from a mix of other items," he explains. Fermentation is the longest step in the process, Campbell says, and recovery takes just a few hours.
"A lot of the equipment in our lab is homemade just because of our unique techniques," Cilento says. "Our intent is to obviously make it compatible with existing fermentation equipment just because those are easier to scale at commercial levels." He adds that GlycosBio eventually will design fermentors optimized for its specific approach.
GlycosBio is funded through private investors now, but is open to federal or state funding opportunities. Cilento has looked into some of the new federal loan and grant programs, he says. "It's something that makes sense, it's just about time and energy and how likely it is to get," he adds. "Right now, from a funding perspective, we're not necessarily in need, but we can at least take a look at it."
The company has no active customers using the technology yet, but is in discussions with several companies interested in producing and using biochemicals. "We're just excited to be at that phase from a development perspective," Cilento says.
"The benefit of our strategy is there are existing feedstocks available within industries that need help," he emphasizes. "We come with a brand new idea and they don't have to change anything. That's a pretty easy story to tell." BIO
Lisa Gibson is a Biomass Magazine associate editor. Reach her at email@example.com or (701) 738-4952.