Improving Nature's Top Recyclers

aarondubrow sends in this snippet from an article at the Texas Advanced Computing Center: "Over billions of years, fungi and bacteria have evolved enzymes to convert abundant cellulosic plant matter into sugars to use as energy sources to sustain life. It's a great trick, but unfortunately, these enzymes don't work fast enough...yet. So computational scientists at NREL, in collaboration with a large experimental enzyme engineering group, set about trying to understand and design enhanced enzymes to ... lower the cost of biomass-derived fuel to serve the global population (abstract)."

Byllshit bingo and then some simulation

[Anonymous Coward]

I'm going to fire a full salvo here, as an industrial chemist acutely involved in this area of research.

>abundant cellulosic plant matter

Oh yeah. I can't not remember a 30-year-old book I read that the most pressing problem in converting biomass into fuel is not processing, but the lack of availability of cheap biomass. The price of oil has to go up significantly and remain at that price for years in order for biomass to become competetive. The situation has not changed.

>into sugars

Converting cellulose into sugars makes little sense by enzymatic processes given that land suitable for growing trees is rather efficiently maintained for timber and pulp production. Real "arable land" suitable for food production is used for that already. In both cases, the key word is "added value": timber, pulp and food are all higher value than energy.

>these enzymes don't work fast enough

And what about elementary reaction kinetics, transport processes and these things called physical absolutes? As we nicely see from the simulation, the enzyme does not attack the very crystallinity of cellulose directly. Cellulose has hard crystalline domains that have an intermolecular hydrogen-bonded structure not unlike that of kevlar. I can't see how this research shows practical ways to overcome this in a energetically feasible way. Nor does the bacteria do anything lignin, which would be analogous to the human attack on lignin: pump the mix full of base (lye) and nucleophiles (sulfide). Lignin forms a rather hard-to-deal with network of giant polymer, if you don't destroy it. Rather, the enzyme seems to "peel off" a cellulose strand at a time, which is necessarily slow. Furthermore, bacteria-based processes generally work at slow speed at high dilutions in water, which is not generally cost-wise acceptable in the energy or even bulk chemicals business. I've seen processes thrown out, out of hand, for this very reason.

Another underappreciated fact in this business is that even trivial-sounding operations are expensive in relation to the added value of the whole process. A dollar a kilogram here or there isn't that much of problem in fine chemicals, but cents a kilogram can make or break an energy-producing process. In this case, I am very concerned of the pretreatments suggested. It's no secret how to degrade biomass into a more processable form - we've seen rather good technologies for years and decades already - but the dealbreaker is whether it's actually profitable, not if it's simply technologically feasible.