(Invited) Development of Highly Active and Stable Pt and Pt Alloy Catalysts Evading the Draw-Back of the Nano-Sizing for PEFC Cathodes

Abstract

It is essential to increase the mass activity (MA) of the catalysts used for fuel cell (FC) reactions, particularly sluggish O2 reduction reaction (ORR), in the wide application areas such as fuel cell vehicles (FCVs) or EneFarms. Reduction of the loading amounts of Pt catalysts used could be expected by increased specific surface-area (S Pt) of nano-particles (NPs) with smaller size. However, there has been an argument to the above from viewpoints of the so-called “particle-size effect” on the loss of J k and stability from that expected at the NPs. Unlike the most of such literatures, we have proven experimentally the absence of the particle-size effect for ORR. We believe that such inconsistent results were mainly brought by non-standardized evaluations or samples not well-defined at most of the previous works. Another issue is the durability such high catalyst performance achieved. It has been accepted experimentally and theoretically that smaller NPs is less stable than larger ones, reaching a compromised value as the optimized size, e.g., 4nm. However, there is a still room to be discussed what is a major factor governing the durability. Therefore, here we will clarify the effect of the particle size on the durability at the NPs with uniform size distributions and the same inter-particle distances on the support (GCB). Precisely size-controlled Pt nanoparticles were prepared by depositing uniform Pt-skin layer(s) on the Pt-core nanoparticles with the sharp size distribution, just by bubbling 5% H2 (N2 balance) into Pt/GCB-suspending aqueous solution dissolved Pt precursor with the exact amount of Pt required for the desired number of Pt-skin layer(s). Figure 1 (A-C) shows the TEM images of as-prepared Pt/GCB and PtxAL-Pt/GCB with x=2 and 4 additional Pt-skin layers. Sizes of ca. 500 number of particles observed in several TEM images were measured to obtain the distribution histograms in Fig. 1(D). The average particle sizes (d TEM) and standard deviations of the Pt/GCB, Pt2AL-Pt/GCB, and Pt4AL-Pt/GCB were 2.4±0.2, 3.5±0.4, and 4.3±0.5 nm, respectively. The latter two are well agreed with those calculated for the particles with the skin layers of the Pt2AL and Pt4AL on the Pt(2.4nm) core particles. It must be emphasized that the mean particle-sizes increased precisely with the increase of the projected thickness of Pt-layers, i.e., ca. 0.25 nm/layer, due to the selective and uniform deposition of Pt atoms on every core Pt-particle without any additional nuclear formation of Pt on the GCB support surface. The inter-particle distances, therefore, should not be changed. This has been well supported by the experimental data shown in Table 1, i.e., the inter-particle distances remained constant within an experimental error, 10.6, 10.1 and 9.0 nm even by the change in Pt loading from 19.4, 34.7 to 58.7 wt.% or in the mean particle size from 2.4, 3.5 to 4.3 nm. Then, we evaluated strictly the effects of Pt particles size on the durability against the potential perturbations at the load-change or start/stop cycles (N) unavoidable in the practical PEFC operation by monitoring the SPt, or ECA(electrochemical active surface area). Stability tests of the ECA were performed at PtxAL-Pt/GCB electrodes coated Nafion (thickness=0.1μm) by the multi-channel flow cell, MCFC, in 0.1 M HClO4 at 65 oC, by applying the standard potential cycle protocols, which are shown in Fig. 2(A) and (B). The initial ECA values at Pt/GCB, Pt2AL-Pt/GCB, and Pt4AL-Pt/GCB were 117, 75, and 59 m2g−1, respectively, which were approximately accorded with the surface area estimated form their mean particle sizes d TEM (= 2.5, 3.5, and 4.5 nm). The ECA values slowly decreased up to ca. N=5,000. The Pt/GCB (d TEM=2.5nm) kept the highest ECA during the durability test. Loss of ECA values of Pt/GCB, Pt2AL-Pt/GCB, and Pt4AL-Pt/GCB at N=50,000 from those at N=0 were almost the same 58%, 63%, and 56%, respectively, whereas, that of c-Pt/CB decreased rapidly with increase of N and reached 82% loss at N=50,000. The similar conclusion was drawn for the start-stop durability test in Fig. 2(B). Thus, we succeed to demonstrate with Pt catalysts controlled of their particle size precisely without changing the inter-particle distance on the support and clarified that such a highly dispersed Pt catalyst with smaller particle size is advantageous not only to the catalytic activity but also the durability against the load-change or start/stop operations, which has not been a matter of common knowledge in most of the catalyst communities. We have also reported the superior ORR activity and durability of Pt2AL-PtCo*/GCB, prepared by the newly developed method, which is very much distinctive from the catalysts prepared by the other processes. Figure 1