Shades of Green
Tremendous attention is now being paid to the principle of "sustainable development": the idea that we should build our civilization in a way that is in harmony with the workings of Nature rather than plunders her bounty. A set of tools that we might use to do this has been unlocked through recent advances in biology and biotechnology, which might one day let us produce the fuels and chemicals we need in a sustainable way. In its own small way my PhD research project is dedicated to this, and I'm inspired by this goal. I also believe that in the final analysis we will indeed switch to this way of building things, but the reasons will have little to do with sustainable development.
May 11, 2014
A huge array of "stuff" we use every day is made from oil. Even a partial list might surprise you: any kind of fuel, almost all plastics, nylon and polyester fabrics, certain dyes or colors, paints, cooking oil, Vaseline (literally, "petroleum jelly"). Most of these things are made from oil that we drill for in the ground ("crude oil"), but some, like the food oils or personal products, are made with oils we harvest from plants.
What makes oil so special? The word is a bit of a catch-all: an "oil" is essentially any mixture of molecules made of atoms of carbon, oxygen, and hydrogen ("hydrocarbons"). To simplify a lot of chemistry, three things that matter about these molecules are chain length (how many carbons there are), branching (whether the carbons are in a line or in a more complex arrangement), and unsaturation (whether all the carbons are connected by single bonds, or some are connected by double bonds). Varying these properties, a virtually infinite array of molecules can be produced, and it's this variety that makes oil so versatile.
This is the starting material: a soup†. Depending on what's being made, various things can be done to this: for fuels, the crude oil is separated into different "fractions" containing different types of chemicals‡; for plastics, a particular molecule or molecules are isolated from the soup and processed to harden them, etc. We are interested in the hydrocarbons (and even then, not necessarily all of them), but there are other chemicals in there, mostly unreacted detritus and side products accumulated over the millions of years it takes oil to form. Nevertheless, this is the best starting material we have so far for making the things we use every day.
It turns out that oil is useful to biology too, and a series of enzymes have evolved to produce the same hydrocarbon molecules, specifying those same three properties fairly exactly. And so palm trees produce "palm kernel oil", coconuts produce "coconut oil", etc. The differences in the oil properties come from which enzymes are present in the genomes and which ones are induced to be expressed by the particular environmental conditions the plant finds itself in.
There is nothing magical about the composition of natural oils. They were produced by random decomposition of biomass, or by plants and animals for their own ends, not ours. If we're able to engineer microbes to do this, we can specify through genetic manipulation which enzymes are present, and through fermentation control which ones are expressed and how much. So we should be able to set up a system to produce whichever molecule(s) we want, including a set that matches one of the oils we find in nature. And with better engineering there is in theory the option of custom-synthesizing a molecule or oil that's optimized for what we need it to do and better than what we could get from nature. A few examples of how we're starting to see that now:
Biologically-produced fuels are often cleaner-burning than petroleum fuels because they have fewer impurities. In recent field testing, Hitachi found that diesel produced by Solazyme — an algae biochemicals company in South San Francisco — was a higher-quality fuel, partially because it doesn't contain sulfur or aromatic compounds that lower the quality of petrodiesel. Fewer impurities also reduce emissions of carbon monoxide, sulfur oxides, and other pollutants often overlooked because of the focus on CO2: butanol-blended fuel for boats produced by Gevo (in Colorado, using yeast) showed lower carbon monoxide emissions than with the usual fuel. Because they burn cleaner, bioproduced fuels have lowered the operating temperature of jet engines by up to 135 °F, reducing wear and extending their useful life.
BioAmber (in Montreal, which produces chemicals in engineered yeast) has said foam made with its bio-produced succinic acid was stronger, less brittle, and more stable over time than foam made from petroleum.
Amyris, a yeast biochemicals company in Emeryville, produces a chemical called farnesene that replaces synthetic rubber (produced from petroleum) in tires. Farnesene is less viscous than synthetic rubber, making it easier to work with in production and reducing manufacturing costs. But the same lower viscosity makes the tires perform better, too, because there is less resistance to the small deformations caused by the forces on a tire in motion; this lower "rolling resistance" actually improves fuel economy.
This theme of "what's good for one is good for many" comes up often when thinking about these biological technologies. This might be part of the reason that DuPont's management have decided to convert its entire chemicals operation from a petroleum-based one to biological production.
There are many other reasons to be excited about using biology to make chemicals: the impacts on our environment are milder, in some cases it can be cheaper, and it frees us from dependence on a resource that isn't available everywhere in the world. But in a way none of this matters as much as how good the thing itself is. With the ability to engineer precisely the chemicals we want, we now have a shot at not only replicating the "stuff" we make from oil, but actually improving it. This is even more interesting because scientists only understood the molecular basis of how this works in the 1950s and 1960s, and started to engineer simple organisms in the 70s and 80s.
Sang Yup Lee, a professor at KAIST who pioneered the biobased chemicals field, came to Stanford to give a talk on the subject a few years ago. Afterwards, I asked him if he could see anything that fundamentally limits what natural molecules we can make using biology in this way. He gave me the shortest answer an academic has ever given to any question: "no".