Just a Dream—or Future Reality?

See allHide authors and affiliations

Science  03 Apr 2009:
Vol. 324, Issue 5923, pp. 48-49
DOI: 10.1126/science.1172083

Over the past decade, vast resources have been devoted to developing proton exchange membrane (PEM) fuel cells that use hydrogen fuel and oxygen from the air to produce electricity, for example, for automotive propulsion. Vehicle test data suggest that the 2009 U.S. Department of Energy targets for hydrogen storage—a 250-mile range and a refueling time of less than 5 min—can be met with high-pressure hydrogen storage tanks (1). Now, the main challenge is the design of cheap and stable fuel cell catalysts for the oxygen reduction reaction. On page 71 of this issue, Lefèvre et al. (2) report a breakthrough in the performance of oxygen reduction catalysts based on non-precious metals.

Oxygen reduction catalysts used in current fuel cell prototypes are platinum nanoparticles supported on carbon black (Pt/C), but cost and supply constraints for large-scale applications require a factor of >4 increase in catalytic activity per mass of precious metal (3). Drawing inspiration from nature, non-precious metal catalysts using abundant transition metals have long been explored (4, 5), but to be viable, their activity would have to approach that of traditional—but more expensive—Pt/C catalysts.

Most current work on non-precious metal oxygen reduction catalysts focuses on nitrogen-coordinated iron in a carbon matrix (referred to as Fe/N/C). The nature of the active sites remains elusive, but graphene-coordinated FeN4 or FeN2 moieties were proposed (68). However, vastly differing syntheses produced virtually identical catalyst activities (9) and PEM fuel cell performance remained 150 to 200 mV below that of Pt/C—too big a gap for practical use. The results reported by Lefèvre et al. thus seem like a dream come true, showing that non-precious metal catalysts can match the performance of state-of-the-art Pt/C catalysts.

This paradigm shift in PEM fuel cell development is even more exciting when examined in terms of turnover frequency, which quantifies the number of electrons produced per active site per second at a defined operating condition. Turnover frequencies of previous non-precious metal catalysts—assuming that each iron atom produces an active site—are 0.4 s−1 (10), compared with 25 s−1 for Pt/C (3) (see the figure). Because of the lower active-site density of non-precious metal catalysts and constraints on the maximum electrode catalyst loading, the turnover frequencies for non-precious metal catalysts must be similar to those of current Pt/C catalysts to be viable for applications (2). Lefèvre et al. now report turnover frequencies for their Fe/N/C catalyst that match those of Pt/C (25 s−1, see the figure). Thus, the dream of using non-precious metal catalysts in PEM fuel cells is becoming a reality, with potentially revolutionary impact on fuel cell technologies.

Toward higher turnover frequencies.

Turnover frequencies for the electrochemical conversion of oxygen to energy and water (e O2 + 2H+ + 2e → H2O, the cathode reaction in a PEM fuel cell) for different oxygen reduction catalysts (operating conditions: 80°C, 100 kPaabs H2 and O2, at 0.8 V versus the reversible hydrogen electrode). Previous generations of Fe/N/C catalysts had much lower turnover frequencies than the novel Fe/N/C catalyst reported by Lefèvre et al. Turnover frequencies must be much higher for Pt-based catalysts than for Fe/N/C catalysts because of cost and supply issues. Turnover frequencies were extracted from the references below and calculated as shown before (3, 10), assuming ∼70 mV tafel slopes for Pt catalysts.


In parallel with the non-precious metal approach, Pt-based oxygen reduction electro-catalysis has also seen important advances. Because of inherent cost and supply constraints, Pt-based catalysts must have high mass activities, given by the product of turnover frequency and metal dispersion, D (the fraction of surface atoms in a nanoparticle). Because D is already as high as 0.35 for state-of-the-art Pt/C, viable Pt-based catalysts must have turnover frequencies that are higher by about one order of magnitude 10 times those needed for non-precious metal catalysts.

Four main strategies aim to increase the turnover frequency of Pt-based catalysts (see the figure). The traditional approach involves PtM alloy nanoparticles (where M is another transition metal), but maximum turnover frequencies are only ∼60 s−1 (3). Another strategy, the “de-alloying” of PtM nanoparticles, leads to particles with a PtM core and a Pt shell, with turnover frequencies of ∼160 s−1 (11, 12). Similar values are obtained for ∼10 monolayer-thick Pt films on nanostructured supports (13). Both strategies offer large Pt dispersions while approaching turnover frequencies of bulk Pt surfaces, resulting in mass activities for de-alloyed PtM that are four times those of current Pt/C (12). Finally, turnover frequencies of ∼2800 s−1 were reported for Pt3Ni(111) surfaces (14), which—if grown into 30-nm-diameter octahedra (D ≈ 0.03)—should yield 10 times the mass activity of current Pt/C. The synthesis of such particles was shown recently (15).

High mass activities can also be achieved by depositing Pt monolayers on large non-precious metal cores, realizing high Pt dispersions (D ≈ 1) and bulk Pt-like turnover frequencies (16). However, the low but finite solubility of Pt is a concern in this case. On the other hand, recent studies show that large nanostructured materials such as nanostructured Pt films and large PtM nanoparticles much reduce the current problem of Pt dissolution from voltage cycling (13, 17).

Little is currently known about the stability of novel non-precious metal oxygen reduction catalysts and their ability to survive in hostile electrochemical environments. Thus, the next challenge for non-precious metal catalysts will be to overcome the twofold loss in current density at constant potential over 100 hours reported by Lefèvre et al., either related to electrode design or to the catalyst itself. Little is known about the molecular processes that might lead to Fe-N bond breaking, and our insight into the stability of the active sites is based only on trial and error. However, recent durability data on Co/N/C-based catalysts are promising (18), and the fact that nitrogen-coordinated transition metals are used successfully by nature gives hope that detailed studies will lead to more durable catalysts.

Four years ago, neither Fe/N/C-based catalysts with Pt/C-like turnover frequencies, nor de-alloyed PtM catalysts with fourfold higher mass activity than Pt/C, nor surfaces with 100-fold higher turnover frequencies than Pt/C were known. Despite a number of remaining challenges, these recent successes bring us closer to completing our quest to put PEM fuel cell technology on the road.

References and Notes

  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
View Abstract

Stay Connected to Science

Navigate This Article