Hemoglobin Components as Cathode Electrode Catalyst in Polymer Electrolyte Fuel Cells

We consider hemoglobin (Hb) components for possible use as cathode electrode catalyst in polymer electrolyte fuel cells (PEFCs). We choose iron-porphyrin (FeP) and (imidazole)FeP [(Im)FeP] as a representative heme complex of the Hb active site, and investigate their catalytic activities for O2 dissociation using ab initio calculations based on density functional theory (DFT). The imidazole ligand suppresses the donation/back-donation of electrons between Fe and O2, resulting in the weakening of the Fe-O2 bond and the strengthening of the O-O bond. Therefore, the catalytic activity of FeP is higher than that of (Im)FeP. [DOI: 10.1380/ejssnt.2004.226]

From an environmental viewpoint, much attention has been focused on polymer electrolyte fuel cells (PEFCs) as an alternative electronic power source of the future. However, PEFCs are too expensive for widespread use because of the large amount of platinum used for both anode and cathode electrode catalysts. Many researchers have tried to design alternative catalysts to platinum, e.g., platinum-based alloys and platinum-group metals, for many years [1]. Regardless of their great efforts, however, both cheaper and more active materials than platinum have not been discovered yet. New breakthroughs for catalyst-design may be possible using ideas other than such conventional ideas as mentioned above.
It is well known that hemoglobin (Hb) can transport oxygen and carbon dioxide in the vascular systems of animals. Heme, which has been considered as the active site of Hb, is constructed from at least one iron-porphyrin (FeP) complex. We think that the catalytic activity of FeP may be used for cathode electrode application of PE-FCs, where the oxygen reduction reaction (ORR) takes place. In this paper, we evaluate for the first time the catalytic activity of heme by determining the activation barrier for O 2 dissociation from the heme-O 2 adduct, as a first step toward better understanding of the ORR on heme. We choose FeP and (imidazole)FeP [(Im)FeP] as a representative heme complex. Imidazole is representative of a histidine amino acid which binds to heme.
We perform all calculations based on density functional theory (DFT) [2,3] with the Becke-Perdew-Wang (B3PW91) exchange-correlation functional [4,5] and the Dunning-Hay-Wadt basis sets (LANL2DZ) [6][7][8][9], as implemented in the Gaussian 03 [10]. All populations and charges are presented by Natural Bond Orbital Analysis (NBO) [11]. Table 1 gives structural parameters and vibrational frequencies for the optimized geometries of FeP, (Im)FeP and their O 2 adducts, without any symmetry constraints. B3PW91/LANL2DZ level calculations can account for the experimental heme-O 2 adduct geometries and vibrational frequencies [12][13][14][15] [16], however, suggested a different tendency from ours using Car-Parrinello molecular dynamics.) Table 2 gives Fe(3d) orbital populations and Fe charges of FeP and (Im)FeP before and after O 2 binds. When O 2 binds to Fe, the O 2 donates electrons to the unoccupied Fe(3d σ ) and the electrons of the occupied Fe(3d π ) back donates to O 2 . Fe(3d z 2 ) population (0.470) of (Im)FeP is more than that of FeP (0.111), and Fe(3d yz ) population (1.885) of (Im)FeP is less than that of FeP (1.950 Table 1). Figure 1 shows the calculated potential energy profile for O 2 dissociation from FeP-O 2 adduct. The reaction ini-  tiates from the optimized geometry of FeP-O 2 adduct (see Table 1). We searched for transition state using the Synchronous Transit-Guided Quasi-Newton (STQN) method [18,19], and confirmed that the transition state has only one imaginary frequency. The separated O atom directly attacks the center of the nearest C-C bond because the interaction with the π orbital of the C-C bond is more stable than the interaction with the σ orbital of the C atom.
The O-O bond is increased to 2.202Å, and the porphyrin is significantly bent in order to pick up the O atom at the transition state. The O-O bond cleavage requires an activation barrier of 1.974 eV. Figure 2 shows the same profile as Fig. 1 but from (Im)FeP-O 2 adduct. The reaction initiates from the optimized geometry of (Im)FeP-O 2 adduct (see Table 1). The O-O bond is increased to 2.333Å, and the O-O bond cleavage requires an activation barrier of 2.022 eV.
Next, we compare the calculated activation barriers for O 2 dissociation from FeP-O 2 and (Im)FeP-O 2 with that of well-known platinum catalyst. Eichler et al. [20], for example, showed that the activation barriers for O 2 dissociative adsorption on Pt(111) vary between ∼0.3 eV and ∼1.5 eV with the adsorption sites, using ab initio localspin-density calculations. The activation barriers we calculated here are 1.974 eV for FeP-O 2 and 2.022 eV for (Im)FeP-O 2 . Now we are designing new O-O bond breaking processes having lower activation barriers [21], based on the orientation and kinetic energy of the molecule [22][23][24][25][26][27][28].
In summary, we confirmed the O-O bond breaking abilities of FeP and (Im)FeP, which are active sites of Hb. Since imidazole ligand strengthens the O-O bond, the FeP cleaves the O-O bond more easily than (Im)FeP. Nevertheless, the activation barrier for O 2 dissociation from FeP-O 2 adduct is too high to utilize for cathode electrode application in PEFCs. This implies that just FeP cannot be of practical use of a catalyst. However, we think that this problem may be overcome by designing FeP-based materials, which are equivalent to, in terms of catalytic activity, well-known platinum catalyst [29].
This work is partly supported by the Ministry of Education, Culture, Sports, Science and Technology of Japan (MEXT) Special Coordination Funds for Promoting Science and Technology (Nanospintronics Design and Realization), the 21st Century Center of Excellence (COE) Program 'Core Research and Advance Education Center for Materials Science and Nano-Engineering' supported by the Japan Society for the Promotion of Science (JSPS) the New Energy and Industrial Technology Development Organization (NEDO) Materials and Nanotechnology Program, and the Japan Science and Technology Agency (JST), Research and Development Applying Advanced Computational Science and Technology Program. Some of the calculations presented here were carried out using the computer facilities of the Japan Atomic Energy Research Institute (JAERI) and the National Institutes of Natural Sciences (NINS).