PerspectiveChemistry

Toward Efficient Hydrogen Production at Surfaces

See allHide authors and affiliations

Science  02 Jun 2006:
Vol. 312, Issue 5778, pp. 1322-1323
DOI: 10.1126/science.1127180

Hydrogen is considered by many to be a promising energy currency, particularly for the transportation sector and for mobile devices (1). The combustion of hydrogen yields water as its only waste product, and hydrogen is a perfect fuel for fuel cells. In most hydrogen-producing technologies, a solid surface catalyzes the required chemical reactions. Higher efficiencies require the development of better catalysts. Recent studies have raised hopes that combined computational and experimental surface studies can aid the design of new catalysts.

To realize a hydrogen-based fuel economy, hydrogen must be produced in an efficient and sustainable manner. Today, most hydrogen is produced from fossil resources by steam reforming, a process in which steam reacts with hydrocarbons in the presence of a metal-based catalyst. Sustainable alternatives include biological or catalytic degradation of biomass and electrochemical or photochemical splitting of water. But irrespective of how the hydrogen is produced, the process is endothermic and requires a considerable amount of energy input.

In most steam-reforming processes, this energy is provided as heat (2), whereas water splitting is usually performed electrochemically or photochemically (3). These processes require bonds to be broken and new bonds to be made, chemical transformations that are activated and typically catalyzed by solid surfaces. Today, the development and optimization of catalytic surfaces is usually based on an intuitive trial-anderror approach. Rational catalyst development strategies would be greatly facilitated by an improved molecular-level knowledge of how chemical reactions proceed on solid surfaces. Recently, new detailed insights have been provided by density functional theory calculations.

Industrial steam reforming of natural gas—that is, CH4 + 2H2O → CO2 + 4H2—is conducted with nickel catalysts at temperatures of ∼800 °C and provides a convenient and cost-effective method for hydrogen production. Ideally, four hydrogen molecules are formed per methane molecule when the methane reacts with water. Today, compact hydrogen-production facilities with capacities of more than 250,000 m3 H2 per hour can be designed (2).

Steam reforming of renewable bioresources is also a viable route to large-scale hydrogen production. Glucose can be reformed in water at surprisingly mild conditions, producing more than six hydrogen molecules for each glucose molecule (4). Ethanol, available for example through fermentation of biomass, can also be steam-reformed in the presence of oxygen. The required heat is supplied directly by combusting some of the hydrogen produced; such an autothermal process yields five hydrogen molecules for each ethanol molecule (5).

Understanding catalyst activity.

This calculated potential-energy diagram (6) shows one way in which methanol decomposes into molecular H2 and adsorbed CO over a platinum surface. Elucidating this and other competing pathways for decomposition may help to devise new catalysts.

CREDIT: ADAPTED FROM (6)

The catalytic conversion of alkanes, alcohols, or carbohydrates with water into hydrogen and carbon dioxide are complex multistep chemical reactions. It is not possible to pinpoint a priori why one catalytic surface performs better than another. For even the simplest alcohol—methanol—the number of elementary reactions associated with its decomposition and the subsequent formation of molecular hydrogen is large (6). However, calculated potential-energy diagrams (see the first figure) can provide a direct identification of the highest activation barriers, providing information on which reaction intermediates need to be stabilized to improve the overall reaction rate. With this knowledge, one can attempt to modify the surface structure or composition in a search for more active catalysts.

Aiming for a shallow well.

This calculated free-energy diagram for electrochemical H2 formation (9) shows that some metal surfaces bind hydrogen too strongly to make hydrogen evolution likely at room temperature, whereas others bind it too weakly to allow hydrogen adsorption at all. The aim is a shallow well that allows both adsorption and evolution. (Inset) Scanning tunneling microscopy image of MoS2 nanoparticles (12), showing the edges where hydrogen can adsorb and H2 evolve.

CREDIT: DIAGRAM ADAPTED FROM (9); IMAGE PROVIDED BY FLEMMING BESENBACHER/UNIVERSITY OF AARHUS, DENMARK

There is also renewed interest in the interaction of water with surfaces and in the electrochemical splitting of water into molecular oxygen and hydrogen (7, 8). For the hydrogen evolution reaction, 2(H+ + e)→H2, a molecular picture of hydrogen evolution has been proposed to explain why platinum is an outstanding catalyst for this reaction, whereas other metals close to platinum in the periodic table (such as nickel or gold) are not (9). Free-energy diagrams calculated with density functional theory show that hydrogen adsorption on platinum surfaces is associated with the smallest free-energy change (see the second figure) (9).

Similar calculations have modeled the interaction of atomic hydrogen with the catalytically active sites of two classes of enzymes—hydrogenases (10) and nitrogenases (9)—which catalyze hydrogen evolution. The results indicate that the enzymes have hydrogen adsorption properties very similar to those of platinum (see the second figure). The catalytic sites in these enzymes contain no noble metals but rather sulfur complexes of nickel, iron, and molybdenum. Nature seems to have found an inexpensive way of imitating platinum for this purpose.

These calculations open up the possibility of designing surfaces with nanometer-scale structure that share some of the enzyme properties. One promising material involves nanometer-scale MoS2 particles (see the second figure, inset). This system yields reasonable hydrogen evolution rates (9), but is not as active as platinum. The biological examples indicate that there may be other metal sulfides that are better catalysts than MoS2, but very little research has been done, either theoretically or experimentally, in this direction.

More efficient hydrogen production methods will require more efficient catalysts. The challenge is to find inexpensive, active, and stable nanostructured materials designed for optimal performance, be it in the production of hydrogen from bioresources or via electrochemical or photochemical routes. The emerging molecularlevel picture of surface reactions may soon allow us to design such catalytic surfaces on the basis of insight (11).

HyperNotes Related Resources on the World Wide Web

Hydrogen Energy and the Hydrogen Economy

Hydrogen Economy An entry in Wikipedia, an online encyclopedia.

How the Hydrogen Economy Works From Michael Brain's How Stuff Works.

Hydrogen Fact Sheets and Hydrogen Links Provided by the National Hydrogen Association.

Hydrogen Topics An information portal provided by the Energy Efficiency and Renewable Energy Web site of the U.S. Department of Energy (DOE).

DOE's Hydrogen, Fuel Cells and Infrastructure Technologies Program Offers presentations on the hydrogen future, hydrogen production, and fuel cells, as well as Internet links, a glossary, and other resources.

International Association for Hydrogen Energy Provides a collection of Internet links related to hydrogen energy.

Hydrogen Production

How Fuel Processors Work A presentation from How Stuff Works about producing hydrogen with steam reformers.

Hydrogen Production A summary from DOE's National Renewable Energy Laboratory.

Hydrogen Production Information about U.S. government research efforts from the DOE Hydrogen Program.

Catalysis

Catalysis A tutorial from the UK Schoolscience Web site.

A Historical Approach to Catalysis Lecture notes by F. Jentoft, Fritz Haber Institute of the Max Planck Society, for a course on modern methods in heterogeneous catalysis research.

Heterogeneous Catalysts A study guide available from C. Chieh's Cyberspace Chemistry.

Lecture Notes on Catalysis From a Massachusetts Institute of Technology course on kinetics of chemical reactions.

Nanocatalysis A presentation by the Interdisciplinary Nanoscience Center of the University of Aarhus and Aalborg University, Denmark.

Surface Chemistry

Surface Chemistry and Reactions on Surfaces Entries in Wikipedia.

Introduction to Surface Chemistry A tutorial by R. Nix, School of Biological and Chemical Sciences, Queen Mary, University of London. Includes a section on the adsorption of molecules on surfaces and a collection of surface science Internet links.

Introduction to Surface Chemistry Lecture notes by U. Jonas, Max Planck Institute for Polymer Research, Mainz, Germany.

Density Functional Theory

Density Functional Theory An article in Wikipedia.

Density Functional Theory A presentation by W. G. Aulbu, Department of Physics, Ohio State University.

“A Bird's-Eye View of Density-Functional Theory” A paper by K. Capelle, Physics Institute, Universidade de São Paulo, Brazil.

Further Reading

Toward a Hydrogen Economy 13 August 2004 special issue of Science.

“Rethinking Hydrogen Cars” A Policy Forum by D. W. Keith and A. E. Farrell in the 18 July 2003 issue of Science.

“Hydrogen from Ethanol Goes Portable” A News of the Week article by A. Cho in the 13 February 2004 issue of Science.

“Hydrogen: The Next Generation” An article by J. Gorman in the 12 October 2002 issue of Science News.

The Hydrogen Economy: Opportunities, Cost, Barriers and R&D Needs A 2004 report available from the National Academies Press.

“Toward Computational Screening in Heterogeneous Catalysis: Pareto-Optimal Methanation Catalysts” (11) An article by M. P. Andersson et al. from the 25 April 2006 issue of the Journal of Catalysis, made available by T. Bligaard, a co-author.

The Authors

Jens K. Norskov is in the Department of Physics, Technical University of Denmark, Lyngby.

Claus H. Christensen is in the Department of Chemistry, Technical University of Denmark.

References

  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
View Abstract

Navigate This Article