News and Views

First published in Nature 447 June 2007

Published online: 20 June 2007 | doi:10.1038/447914a

Chemical engineering: Hybrid routes to biofuels

Lanny D. Schmidt¹ & Paul J. Dauenhauer¹

Traditional methods for making fuels from biomass come in two forms — biological or chemical. The latest approach combines the best of both worlds, and heralds the advent of a second generation of biofuels.

With petrol prices on the rise, biofuels are big news these days. For applications in the transportation sector, perhaps the best known liquid biofuel is biomass-derived ethanol. But ethanol has its limitations: it is highly volatile, absorbs water and has a low energy density. A team from the University of Wisconsin-Madison has developed a two-step catalytic process that can convert fructose into a potentially better liquid biofuel, 2,5-dimethylfuran (DMF). This has 40%-higher energy density and a higher boiling point than ethanol, and is not water soluble. Fructose can be made directly from biomass or from glucose and although there's some work needed before DMF production can be made commercially viable, this new catalytic process looks promising.

An ethanol plant in Colorado, surrounded by fields of corn (maize).

R. WILKING/REUTERS

Carbohydrates from biomass will almost certainly provide the source of carbon-based fuels of the future. But the fuel of choice and the method of production are still uncertain^{1, 2}. In the absence of an optimal process, there is a vigorous debate over whether the biomass conversion system should be thermochemical (using heat and metal catalysts) or biological (using enzymes and microorganisms). The fuels produced reflect the process involved: ethanol is the product of biological conversion, whereas synthetic diesel (a mixture of saturated hydrocarbons known as alkanes) is that of thermochemical methods. Reporting on page 982 of this issue, Dumesic and colleagues³ blur the lines between these two routes. They apply catalytic techniques to biologically derived sugar molecules to yield a potential fuel called 2,5-dimethylfuran (DMF).

The biggest headache for those developing biofuels is the stark contrast between what we have (biomass rich in carbohydrates) and what we want (oxygen-deficient fuels). Carbohydrates, such as cellulose and starch, comprise as much as 75% of biomass sources — typically, corn (maize), trees and grasses. These carbohydrates take the form of large polymer chains assembled from thousands of sugar units; each unit contains six carbon atoms and a similar number of oxygen atoms. But optimal fuel molecules for conventional power systems, such as automobile engines, are nothing like plant carbohydrates. They must be small (only 5–15 carbons) and contain little oxygen. The challenge of producing biofuels is finding a way of breaking down long carbohydrate chains to form small, usable molecules, while simultaneously removing the oxygen and minimizing the loss of energy value of the original biomass.

The initial thermochemical approach to this problem was to partially oxidize biomass to produce 'synthesis gas' (a mixture of carbon monoxide and hydrogen), which was then reacted on metal catalysts to form synthetic diesel (Fig. 1a). In other words, the carbohydrates were broken down to single-carbon-atom components and then built up again to larger molecules. But this process generates significant entropy, and about half of the carbohydrate's energy is lost in the process⁴.

Figure 1: Conventional and hybrid biofuel production processes.

a, The conventional thermochemical route to biofuels breaks down starch (or other biomass) into a mixture of carbon monoxide and hydrogen. This mixture is then converted catalytically into synthetic diesel. b, Conventional biological routes convert starch to glucose, which is then fermented by microorganisms to produce ethanol. c, Dumesic and colleagues3 adopt a hybrid route that enzymatically converts starch into fructose. An acid-catalysed reaction converts the fructose into 5-hydroxymethylfurfural (HMF), which undergoes another catalytic reaction with hydrogen to yield the potential fuel 2,5-dimethylfuran (DMF).

Full figure and legend (21 KB)

The alternative biological approach uses enzymes to break down starch and cellulose to glucose (Fig. 1b). This sugar can then be further processed into fuel molecules using an ever-increasing array of microorganisms and enzymes⁵. The current rapid, worldwide expansion of ethanol-plant construction depends on the hard-working yeast Saccharomyces cerevisiae. This organism ferments glucose into two equivalents of ethanol, transferring some of the original carbohydrate oxygens to carbon dioxide. However, the fermentation process is relatively slow (on the order of days), and a better transportation fuel would contain less oxygen, making it more energy dense and less likely to absorb water.

Dumesic and colleagues' hybrid process³ for converting carbohydrates to DMF provides the benefits of both the older methods while avoiding their pitfalls. It does this by combining new catalytic chemistry with techniques from conventional biological systems (Fig. 1c). The process begins by enzymatically cleaving carbohydrates into fragments, which are rearranged to form the sugar fructose. Rather than fermenting this highly oxygenated sugar, the next step is an acid-catalysed reaction that expels three oxygen atoms, eliminating them as water molecules. This generates an intermediate compound, 5-hydroxymethylfurfural (HMF), which is immediately extracted to prevent further undesired reactions. HMF will probably provide a valuable biologically derived chemical building-block for the future, but its multiple chemical groups and polar nature prevent it from consideration as a fuel⁶. The authors therefore developed a carbon-supported copper–ruthenium catalyst that allowed two more oxygen atoms to be cleaved out of HMF using hydrogen gas, so yielding the desired fuel, DMF.

Compared with existing methods, the benefits of DMF production are clear. Using fructose as the starting material avoids many of the energy-intensive procedures common to thermochemical techniques — for example, the gas-compression step that is necessary to recombine carbon monoxide into synthetic diesel. By replacing biological processes, such as fermentation, with more conventional catalytic methods, the conversion of a sugar to fuel can be hundreds to thousands of times faster than before; this permits the use of much smaller refineries, and could reduce capital investment. The introduction of catalytic techniques could also spawn a second generation of biofuels with improved physical properties. Although it is too early to determine the fate of DMF as a fuel, its high energy density and hydrophobic nature certainly recommend it as an energy carrier. If it is to be used as a transportation fuel, further examination of its combustion capabilities, as well as its impact on health and the environment, is required.

Combining techniques from diverse areas is a well-known strategy in science and engineering, but Dumesic and colleagues' process³ is a ground-breaking example of interdisciplinary engineering that may well dictate the future of biomass conversion. Their process will no doubt inspire many other combinations of chemical and biological reactions for biofuel production. This work will also focus attention on the potential of other hybrid systems, such as metal catalysts that directly process carbohydrates into fuels^{7, 8}, or microorganisms that reassemble synthesis gas to biofuels⁹. Such technologies could completely change current thinking about biofuel production, so that the competition between thermochemical and biological methods ends in a tie.

Author affiliation

1. Lanny D. Schmidt and Paul J. Dauenhauer are in the Department of Chemical Engineering and Materials Science, University of Minnesota, 421 Washington Avenue SE, Minneapolis, Minnesota 55455, USA.
   e-mail: schmi001@umn.edu