Chemistry: The Future of Cellulose

I am not a big believer in a commercial future for the biochemical conversion of cellulose into fuels. There are many big hurdles in place that are going to have to be overcome before cellulose is commercially converted to ethanol. In a nutshell, one is the logistical problem, which I have covered before. Beyond the logistical problem is the issue that biochemistry often starts to malfunction as the conditions in a reactor change, and with cellulosic ethanol that means that if you get a 4% solution of ethanol in water, you are doing well. But from an energy return point of view, a 4% solution is about like the trillions barrels of oil shale reserves we have. If it takes over a trillion barrels of energy to extract and process them, that largely defeats their usability.

Chemistry is a different matter, which is why I favor gasification processes over fermentation processes. But even beyond gasification, I have wondered about chemically processing cellulose in a refinery. I used to have a guy who e-mailed me all the time and told me he had invented a chemical process for reacting cellulose to hexane, which can then be turned into gasoline. If you look at cellulose (there is a graphic of a segment of cellulose at the previous link), you can envision that it could be done. (Whether he had actually done it is a different story).

But the chemistry pathway isn’t limited to fuels. With that preface, I want to thank a reader for bringing this story to my attention. In a recently published story in Applied Catalysis A: General (available online at Science Direct), scientists at Pacific Northwest National Laboratory have reported on a new process for converting cellulose directly into an important chemical building block (e.g., for plastics and fuel):

Single-step conversion of cellulose to 5-hydroxymethylfurfural (HMF), a versatile platform chemical

Now we all know that you can do lots of neat things in the lab that can’t really be done on a larger scale. But this particular process does not appear to be overly complicated. The abstract from the paper explains what they are doing:


The ability to use cellulosic biomass as feedstock for the large-scale production of liquid fuels and chemicals depends critically on the development of effective low temperature processes. One promising biomass-derived platform chemical is 5-hydroxymethylfurfural (HMF), which is suitable for alternative polymers or for liquid biofuels. While HMF can currently be made from fructose and glucose, the ability to synthesize HMF directly from raw natural cellulose would remove a major barrier to the development of a sustainable HMF platform. Here we report a single-step catalytic process where cellulose as the feed is rapidly depolymerized and the resulting glucose is converted to HMF under mild conditions. A pair of metal chlorides (CuCl2 and CrCl2) dissolved in 1-ethyl-3-methylimidazolium chloride ([EMIM]Cl) at temperatures of 80–120 °C collectively catalyze the single-step process of converting cellulose to HMF with an unrefined 96% purity among recoverable products (at 55.4 ± 4.0% HMF yield). After extractive separation of HMF from the solvent, the catalytic performance of recovered [EMIM]Cl and the catalysts was maintained in repeated uses. Cellulose depolymerization occurs at a rate that is about one order of magnitude faster than conventional acid-catalyzed hydrolysis. In contrast, single metal chlorides at the same total loading showed considerably less activity under similar conditions.

So they take cellulose and react it with two metal chlorides at 80–120°C for a direct conversion of cellulose into HMF – which can be easily converted to fuel or plastics. I would think then the important considerations would be 1). What happens to the lignin and hemicellulose in the biomass?; and 2). How much energy does it take? The second item is particularly important if fuel is the objective.

While it is too early to tell whether there is a fatal flaw, this one certainly bears watching. It also strengthens my conviction that in the long-run, the right way to process cellulose is chemically.

21 thoughts on “Chemistry: The Future of Cellulose”

  1. “But from an energy return point of view, a 4% solution is about like the trillions barrels of oil shale reserves we have. If it takes over a trillion barrels of energy to extract and process them, that largely defeats their usability.”

    Maybe the key to our energy future is to start asking the right questions.

    If we have the resources to supply the human race with 100 TeraWatts of power for a millenium or two using known nuclear fission processes — and that certainly seems like a viable estimate — then the real question is how do we convert that (nuclear) energy into forms which are convenient for everyday use?

    You are almost certainly right, RR, that the Energy Return on Energy Investment (less awkwardly, the energy amplification) from shale oil will be less than 1. On the other hand, if we use nuclear heat to mine liquid hydrocarbon transportation fuels from shale, the energy amplification is not germane.

    Maybe the same logic could be applied to cellulose conversion into useful fuels. If the process heat comes from some external source like nuclear fission, then the poor or negative energy amplification in the conversion process is not a matter of concern.

    So many technical opportunities, if we could ever dump politicial correctness and just get on with it!

  2. “On the other hand, if we use nuclear heat to mine liquid hydrocarbon transportation fuels from shale, the energy amplification is not germane.”

    This is true, and I almost put that caveat into the story. Energy return is not the entire story; there are also the economics. It may make perfect economic sense to turn 1 BTU of coal into 0.5 BTUs of ethanol. The energy return is terrible, but if you need liquid fuels and you have coal, this is what you would do. That’s one reason they keep talking about using nuclear to heat up the tar sands for processing.


  3. Well, this discussion of whether ethanol can be made profitably by a chemical process is way over my head, Usually, that doesn’t stop me from opining, but this time it does.
    I still wonder if the best biofuel is to grow switchgrass and burn to turn steam turbines. Use your battery-powered car.
    I sure hope this latest biofuel process, unveiled here by RR, works. However, we have the luxury of no needing it.
    I sense we are witnessing the beginning of the end of the Oil Era. And it will ned not with a bang, but a whimper.
    Between conservation and alternatives, we have a plethora of options. The mighty scythe of the price mechanism does not always work quickly, but it is inexorable. It will cut oil markets doen to size.

  4. Doesn’t this process fall into the general category of cooking plant material to extract water?

    It seems to me that all of the schemes based on biological material fall prey to the same problems. Low production density, and high cost of transporting and handling base materials. And, the environmental problems cause by over harvesting and soil mining.

    There is one place in our current material handling system where we could process bio-materials without incurring excess costs. That would be in municipal waste stream disposal, where we already spend a lot to gather material together. Using this type of process to turn garbage into fuel could be a two birds one stone type of win.

  5. Mother nature did all the materials handling and compaction/densification over millions of years with hydrocarbons.

    Now people suggest we do the same and generate all power by burning switch grass – the energy density is minimal and the logistics are overwhelming!

    Not gonna happen – unless good old Uncle Sam does it with borrowed tax dollars of course.

  6. I wouldn’t get TOO hung up on the “logistical” issues. Farmers have been cutting, baling, and transporting hay for hundreds of years. They know how to do it.

    On the other hand, being a low-value commodity, you wouldn’t want to have to transport it too far.

    My model is a couple of miles sq. of switch grass/poplar, etc. centered on the waste disposal site.

    Of course, you, also, have the corn cob model; and, you could probably build a waste wood model similar to the corn cob model. ie, gather the waste wood, and transport to a mobil facility at the time of logging.

  7. For instance, there are probably 80,000 vehicles in our county of 100,000. They probably use about 48 Million Gallons of gasoline/Yr. In a few years we could, maybe, get that down to 30 Mbbl.

    Ceres says they can average about 10 tons of switchgrass/acre. Let’s figure 100 gal of ethanol/ton. That’s probably a bit high, but the number’s easy to work with. 🙂 640,000 gal sq mi.

    4 mi. X 4 mi = 16 X 640,000 = 10,240,000 gallons.

    Add in 100,000 tons of MSW @ 80 gal ethanol/ton = 8,000,000 gallons.

    10,240,000 + 8,000,000 = 18,240,000 gal/gasoline, or about 60% of our fuel supply coming from an area of marginal land equal to approx. 1.6% of our county.

  8. Of course, by making that a 6 mile sq we could produce 640,000 X 36 = 23,040,000 which, when added to the MSW ethanol, would provide 100% of our liquid fuel needs. Of course, now we’ve used 3.6% of our total land. Whatever will we do?

    The other alternative would be to just harvest all the kudzu.

  9. In trying to understand this, I went back to an article you wrote on Coskata. Do you think the yield on this method would be 100 gallonsethanol per ton (or it’s HMF equivalent) simalar to what Coskata was claiming?

    Seperately: When one talk about a ton of corn turned to first generation ethanol are they talking about a ton of corn kernels? Not entire ears of corn, this is something that I am not clear on.

  10. “I would think then the important considerations would be 1). What happens to the lignin and hemicellulose in the biomass?; and 2). How much energy does it take?”

    This is really just speculation, as I don’t have enough facts to go on, but I would guess that the cellulose can be conveniently dissolved out of a pulp, either with those solvents or in a separate stage using the rayon process which has a cuprammonium ion complex in water as a solvent. Then the stuff that’s left can be burned (maybe indirectly after gasifying) to help power the remaining processes. Drying out the water would be cheap, but there’s the issue of reclaiming enough of the solvents to be cost effective. I’d guess that a cellulose dissolving process would be rather like that for sugar beet, making use of counterflow; that could easily be adapted to flush any of the solvent out of the final waste. Since copper is used in the main process, the only issue is reclaiming the ammonia – but (again guessing) low pressure distillation should be cost effective for that, though it might need some added alkali to free it up.

    I’ve sometimes wondered if the rayon process could be a starting point for some other processing pathway, but that’s even more speculative.

  11. A bushel of corn (kernels) weighs 56 lbs. it will yield a little over 2.8 gallons of ethanol. 2,000/56 = 35.7 bu X 2.8 = 100 gal of ethanol/ton.

    Average yield is 155 bu/acre. Or, 155/35.7 = 4.34 tons/acre. You get about 17.5/lbs of high-protein cattle feed for every bushel distilled for ethanol. That gives a co-product of 2,712 lbs of DDGS/acre, or 625 lbs/ton of corn.

    You, also, get about 600 lbs of CO2/ton. About 1/3 of that is sold to soft drink bottlers, etc.

    Several refiners such as Plymouth are removing 1.2 of the approx 1.8 lbs of corn oil/bu.

  12. “Do you think the yield on this method would be 100 gallons ethanol per ton (or it’s HMF equivalent) simalar to what Coskata was claiming?”

    I haven’t checked, but I think the overall yield will probably lower for this process because you are only getting the cellulose. The big difference will be in the net energy. Coskata has a hybrid process; half gasification (which converts all of the biomass and not just the cellulose) and half biochemical. The biochemical part introduces water back into the system, which hurts the energy return.


  13. Is Rufus just Vinod Khohsla in disguise? His arguments are basically same as those put forth by the esteemed Mr. Khosla 3 years ago, i.e., that we can get thousands of farmers to grow millions of acres of switchgrass with average yields of 10 tons/ acre, and we extract 100 gal ethanol from every ton, but “don’t worry about the logistics” . . . As USDA and DoE have found out, it’s ALL about the logistics (and storing the biomass, and pretreating it, and grinding it, and dealing with all the dirt etc that’s present . . . . I don’t doubt that it can be done, but it would be nice for someone to actually show it can be done economically first at a very small scale before making grand plans for it to solve our transportation fuel supply needs.

  14. The lignin will no doubt be burnt to power the plant. There is pretty much no efficient way of getting lignin into anything we can use..the bonds are just too damn strong..

    Glad I found this blog by the way..I’ve been having trouble finding more people in this field..

    -Wood Engineer

  15. With regard to the PNNL work you cite, the solvent [EMIM]Cl is relatively expensive, so it is essential that it be efficiently recovered for reuse. Unfortunately compounds of this type decompose when heated with alkali. Because of this it is likely that the ash content of biomass will cause the solvent to degrade to an unacceptable extent. I would be interested to see how the process behaves when whole biomass is fed instead of pure cellulose.

  16. Interesting paper, but I wish they had shown their chromatograms because that is my line of work. Is the next step making DMF? Is this similar to what Virent is doing?

    A few years ago I was at a talk by a chemist working on biofuels for big Ag company. He was talking about biodiesel and alcohols. Then he started talking about how the next big thing was going to be furfurals and furans, as the building blocks for a host of industrial chemicals and fuels. You can make all sorts of things with this stuff.

    Thanks for posting this Robert

  17. Takchess,

    Way cool is a huge understatement. Lonnie Johnson’s thermoelectric generator is pure genius. If this thing doesn’t work he is just super smart for thinking it. If it does work, then Johnson joins Edison and the Wright brother as an historic inventor.

    Why didn’t I think of it first, I know about Hydrogen and fuel cells and proton exchange membranes, but I never thought of sealing the hydrogen in and using heat and pressure to drive it through the membranes.

    The thing I didn’t like about the article that you linked is that they talk about using this type of device to improve the efficiency of an ICE by using the waste heat. That would be silly. If you have this type of device, there is no need for an ICE. You would simply burn fuel, any fuel, and convert directly to electricity.

    I am telling you, this is one to watch.

    Benny, don’t get too excited, it might not work.

  18. This is only one of many approaches to cellulosic energy that can work. It is not a particularly new approach.

    The logistics problem is probably the easiest to solve, as long as you do not ask cellulose to solve all the world’s energy problems. Think local and regional solutions to local and regional needs. Think small.

    For the big picture, nuclear has always had the big guns. If you build around nuclear, you can spin off small reactors to help with oil sands and oil shales (kerogens and bitumnen).

    Solar can only get better, as can enhanced geothermal. Wind is the big loser, but it will take a lot of very high-profile people a long time to figure that out.

Comments are closed.