Pt-skeleton surfaces

Figure 8.12 Relationships between the catalytic properties and electronic structure of Pt3M alloys correlation between the specific activity for the oxygen reduction reaction measured experimentally by a rotating disk electrode on Pt3M surfaces in 0.1 M HCIO4 at 333 K and 1600 lev/min versus the li-band center position for (a) Pt-skin and (b) Pt-skeleton surfaces. (Reprinted with permission from Stamenkovic et al. [2007b]. Copyright 2007. Nature Pubhshing Group.)...

Stamenkovic V, Mun BS, Blizanac BB, Mayrhofer KJJ, Ross PN Jr, Markovic NM. 2006a. The effect of surface composition on electronic structure, stability and electrocehmical properties of Pt-transition metal alloys Pt-skin vs. Pt-skeleton surfaces. J Am Chem Soc 137 1. 

Fig. 3.1 UHV characterization of Pt Fe alloy (a) LEIS spectra during UHV treatment annealed and sputtered surfaces, (b) LEIS spectra red) after electrochemical treatment reveals dissolution of Fe atoms from the surface and formation of Pt-skeleton surface (c) UPS spectra (d) The d-band center position relative to the Fermi level from the valence band spectra measurements on PtjM alloy surfaces (red) and calculated values of d-band center position (blue), Schematic ball models for the sputtered surface and Pt-skin are included to help visuahze the structure and composition of Pt and M atoms in the near-surface regions. Reprinted with permission from [29], copyright 2007 Nature Pubhshing Group...

The ex situ method, pioneered by Hubbard [1] and Ross [56], entails electrode preparation in UHV, clean transfer of the specimen from UHV into an electrochemical cell and transfer back to UHV for postelectrochemical reaction analysis. The methodology was developed to study the stability of Pt-skin and 3d elements in sputtered surfaces. For this purpose, we restrict our analysis to the polycrystalline PtjFe surface, which is discussed in Sect. 2.1. As summarized in Fig. 3.1b, LEIS spectra indicate that after exposing this alloy to an electrochemical environment Fe surface atoms dissolve from the near-surface layers leading to a first atomic layer composed entirely of Pt [21]. For the remaining surface, which consists of only Pt atoms after dissolution of the alloying component, we employ the term Pt-skeleton structure. The conclusions drawn from these experiments are equally valid for all sputtered PtjM alloys i.e., whenever 3d transition metal atoms are exposed to acidic environments, nonplatinum atoms dissolve leading to the formation of Pt-skeleton surfaces. 

In contrast to sputtered surfaces, UHV-annealed PtjM surfaces are rather stable in electrochemical experiments, implying that Pt-skin surfaces are structurally and compositionally identical in both UHV and electrochemical environments. Although the composition of the top most layer is the same (pure Pt) for both structures [21], there are three key differences between the Pt-skin and Pt-skeleton surfaces (1) electronic properties are different for the two surfaces i.e., for each PtjM alloy the corresponding Pt-skin surface has a larger d-band center shift from the Fermi level than the sputtered surface (2) Pt-skeleton is morphologically different than the Pt-skin i.e., the Pt surface atoms have a lower average coordination number and (3) the composition of the 3d element in the second layer of the skeleton surfaces... 

Figure 3.6 demonstrates on both the Pt-skin as well as Pt-skeleton surfaces the relationship between the specific activity and the d-band center position exhibits a volcano shape, with the maximum catalytic activity obtained for PtjCo. This behavior is apparently a consequence of the Sabatier principle discussed earlier, and published in many recent studies [77, 78]. For metal surfaces that bind oxygen too strongly, as in the case of Pt, the d-band center is too close to the Fermi level and the rate of the ORR is limited by the availability of spectator-free Pt sites. 

Fig. 3.6 Relationships between the catalytic properties and electronic structure of PtjM alloys [29, 31] (a-b) Specific activity (at 0.9 V) for the ORR on Pt M surfaces in 0.1 M HCIO at 333 K vs. experimentally measured d-band center positions for the Pt-skin and Pt-skeleton surfaces (c) The activity calculated from DFT vs. calculated d-band center for the Pt-skin slabs, (a-b) Reprinted from [29], Fig. 4...

If the d-band center is too far from the Fermi level, as in the case of PtjV and PtjTi, the intermediates and bind too weakly to the surface. For the PtjM systems, the experimental results (summarized in Fig. 3.6a) and componential screening of the same binary alloys (Fig. 3.6c) converge to the same optimal Pt to 3d ratio (e.g., PtjM) and the identity of the 3d element (e.g., Co or Ni). While it is tempting to conclude that the rationale for the variation in activity depends exclusively on the position of the metal d-states relative to the Fermi level. Fig. 3.6 clearly shows that for a similar position of the d-band center, a different activity is obtained on Pt-skin and Pt-skeleton surfaces. The relationships between the elec-trocatalytic activity of the ORR and the d-band center position were also recently demonstrated by Adzic and Mavrikakis [77, 78]. 

Figure 4.4 Trends in electrocatalysis relationships between experimentally measured specific activities for the ORR on PtsM surfaces in 0.1 M HCIO4 at 0.9 V and 333 K vs the d-band center position for (A) the Pt-skin and Pt-skeleton surfaces. Results for each alloy were collected from five independent RDE measurements and error range is expressed by the size of circles. (B) Activities and values for d-band center position predicted from (Density functional theory) calculations for (111) oriented skin surfaces. Reprinted from Science, 315,493 (2007), Vojislav R. Stamenkovic., Improved oxygen reduction activity on Pt3Ni(lll) via increased surface site availability, 493—497, 2007, with permission from Sciencemag.

K versus the d-band center position for (a) the Pt-skin and (b) the Pt-skeleton surfaces. Reprinted by permission from Macmillan Publishers Ltd Nature Mat. [21] copyright 2007. 

---