Platinum crystallite

There is little data available to quantify these factors. The loss of catalyst surface area with high temperatures is well-known (136). One hundred hours of dry heat at 900°C are usually sufficient to reduce alumina surface area from 120 to 40 m2/g. Platinum crystallites can grow from 30 A to 600 A in diameter, and metal surface area declines from 20 m2/g to 1 m2/g. Crystal growth and microstructure changes are thermodynamically favored (137). Alumina can react with copper oxide and nickel oxide to form aluminates, with great loss of surface area and catalytic activity. The loss of metals by carbonyl formation and the loss of ruthenium by oxide formation have been mentioned before. 

The scanning transmission electron microscope (STEM) was used to directly observe nm size crystallites of supported platinum, palladium and first row transition metals. The objective of these studies was to determine the uniformity of size and mass of these crystallites and when feasible structural features. STEM analysis and temperature programmed desorption (TPD) of hydrogen Indicate that the 2 nm platinum crystallites supported on alumina are uniform In size and mass while platinum crystallites 3 to 4 nm in size vary by a factor of three-fold In mass. Analysis by STEM of platinum-palladium dn alumina established the segregation of platinum and palladium for the majority of crystallites analyzed even after exposure to elevated temperatures. Direct observation of nickel, cobalt, or iron crystallites on alumina was very difficult, however, the use of direct elemental analysis of 4-6 nm areas and real time Imaging capabilities of up to 20 Mx enabled direct analyses of these transition metals to be made. Additional analyses by TPD of hydrogen and photoacoustic spectroscopy (PAS) were made to support the STEM observations. 

Samples A and B are of particular Interest because they are composed of small, uniform platinum crystallites. The fact that these crystallites are on alumina limits the techniques available for their characterization. Sample A showed what appeared to be very thin platinum crystallites, which were barely observable by Imaging techniques or measurable by EDS. An exanq>le of a bright field Image and corresponding EDS analysis Is shown In Figure 1. In order to obtain analyses of this type, focus variation at magnifications of 1 to 4 Mx was commonly used with EDS analysis at 20 Mx to confirm that the particle was platinum. 

Sample B provided platinum crystallites that were analyzed by both EDS and MAED. MAED of several 3 nm crystallites shows a wide variation of orientations with respect to the electron beam, however, many of the patterns match (111) and (110) orientations. An example of the MAED patterns observed Is shown In Figure 2. The diffraction pattern was made with a 25 pm objective aperture at a camera length of 2 m. 

In addition to obtaining Information about the size, relative mass, and structure of the platinum crystallites, the STEM can provide a qualitative evaluation of the metal distribution from support particle to support particle. In general, the distribution of platinum was more uniform on alumina than silica, however, optimal uniformity was not achieved. This observation was based on wide variations In Pt/Sl and Pt/Al ratios measured by EDS. 

Figure 1. Example of a 2nm platinum crystallite at 4 Mx and corresponding EDS analysis at 20 Mx for 150 sec.

Figure 2. MAED pattern of a 3 nm platinum crystallite on y-alumlna (111) orientation.

Figure 3 Size and mass distribution of platinum crystallites supported on y ilnmlna ...

This Study has shown that reasonably uniform platinum crystallites can be made on y-alumlna, and that platinum and palladium can be segregated and maintained In that form for the most part even after exposure to high temperature oxidation-reduction conditions. Highly dispersed clusters of palladium, nickel, cobalt, and Iron can be observed. Cluster size determination could not be accurately made because of the lack of contrast between the cluster and the support. The marginal detectability by EDS for these clusters enabled elemental Identification to be made, however, mass uniformity determinations could not be made. 

Incorporation into a Polymer Layer In recent years a new electrode type is investigated which represents a layer of conducting polymer (such as polyaniline) into which a metal catalyst is incorporated by chemical or electrochemical deposition. In some cases the specific catalytic activity of the platinum crystallites incorporated into the polymer layer was found to be higher than that of ordinary dispersed platinum, probably because of special structural features of the platinum crystallites produced within the polymer matrix. A variant of this approach is that of incorporating the disperse catalyst directly into the surface layer of a solid polymer electrolyte. 

It was seen when studying mixed systems Pt-WOj/C and Pt-Ti02/C that with increasing percentage of oxide in the substrate mix the working surface area of the platinum crystallites increases, and the catalytic activity for methanol oxidation increases accordingly. With a support of molybdenum oxide on carbon black, the activity of supported platinum catalyst for methanol oxidation comes close to that of the mixed platinum-ruthenium catalyst. 

Bregoli LJ. 1978. The influence of platinum crystallite size on the electrochemical reduction of oxygen in phosphoric acid. Electrochim Acta 23 489-492. 

Watanabe M, Saegusa S, Stonehart P. 1988. Electro-catalytic activity on supported platinum crystallites for oxygen reduction in sulphuric acid. Chem Lett 17 1487-1490. 

Fig. 2. Platinum crystallite size distribution for 0.12 ng cm-1 ultrathin platinum film (Fig. 1). Full line, number distribution broken line, surface area distribution. 

Fig. 4. Platinum crystallite size distribution for 2.5% (w/w) platinum/silica catalyst (Fig. 3). Full line, number distribution broken line, surface area distribution. After T. A. Dorling, and R. L. Moss, J. Catal. 7, 378 (1967) and R. L. Moss, Platinum Metals Rev. 11 (4), 1 (1967).

X-Ray studies confirm that platinum crystallites exist on carbon supports at least down to a metal content of about 0.03% (2). On the other hand, it has been claimed that nickel crystallites do not exist in nickel/carbon catalysts (50). This requires verification, but it does draw attention to the fact that carbon is not inert toward many metals which can form carbides or intercalation compounds with graphite. In general, it is only with the noble group VIII metals that one can feel reasonably confident that a substantial amount of the metal will be retained on the carbon surface in its elemental form. Judging from Moss s (35) electron micrographs of a reduced 5% platinum charcoal catalyst, the platinum crystallites appear to be at least as finely dispersed on charcoal as on silica or alumina, or possibly more so, but both platinum and palladium (51) supported on carbon appear to be very sensitive to sintering. 

Metal oxide coatings Commercial lead dioxide coatings, for example, on titanium, have a higher stability compared with lead or lead alloy anodes with their in situ formed oxide layer. A secure contact between Pb02 and titanium has to be guaranteed, for example, by a platinum layer or at least by a sufficiently large number of platinum crystallites. 

FIGURE 17 Transmission electron micrograph of platinum crystallites on a y-alumina support. The black bar represents 100 A. 

Figure 13. Schematic representation of the surface coverage of platinum crystallites as they are used in hydrogenation catalysis The surface analysis instrument detects only the light parts and is affected by shadowing (grey zones arrows indicate direction of illumination).

Figure 2. Platinum crystallites on thick particles of 7-alumina using annular dark field.

Figure 3. Platinum crystallites on a carbon film using axial dark field.

For fuel-cell technology development, it has been important to understand the characteristics and operation of highly dispersed platinum and platinum alloy electrocatalysts. A series of papers on platinum crystallite size determinations in acid environments for oxygen reduction and hydrogen oxidation was published together by Bert, Stonehart, Kinoshita and co-workers.5 The conclusion from these studies was that the specific activity for oxygen reduction on the platinum surface was independent of the size of the platinum crystallite and that there were no crystallite size effects. 

The aforementioned paper by Bregoli6 called these earlier results into question, since it was implied from his paper that the mass activity appeared to be constant and the specific activity decreased as the crystallite size became smaller. Watanabe and co-workers 7 (their Figure 6) further examined this question and showed apparent constant specific activities for supported platinum on carbon in hot phosphoric add out to 210 m2g Pt. The subsequent work of Buchanan, el al.s showed a wide variety in the mass activities for dispersed platinum crystallites in the range of50-130 m2 g 1 Pt but they claimed they were not able to confirm the earlier results of Watanabe, el al.1 A later paper by Buchanan el al.9 identified mass activities for platinum in the range of 58-82 m2g Pt where they indicated a slope of the line for the mass activity versus the electrochemical area to be exactly the same as the slope of the line previously published by Watanabe, el aL7 i.e., the specific activity was constant at 0.6 A m 2 Pt surface. 

Figure 1. Compilation of platinum mass activities as a function of platinum B.E.T. surface area [ ] Watanabe et alJ [0] Buchanan et al.s [ ] Buchanan et al. and [0] Bregoli6. The solid line is 0.6A.m 2 constant specific activity platinum. The broad arrow on the abscissa denotes the maximum surface area for a platinum crystallite when all of the atoms are located at the surface (275 m2 g 1 Pt). Phosphoric acid at 190 °C and 0.9 V vs. hydrogen in the same electrolyte, (a) Data up 210 m2g" Pt. (b) Data below 100m2g Pt m Bregoli6 results on unsupported platinum black.

This study was on a smooth platinum foil using gravimetric analyses and it was found that the equilibrium concentrations of platinum in the phosphoric acid were achieved rapidly, i.e., within one hour. Since it is to be expected that the kinetics of dissolution would be modified with high surface-area platinum crystallites, this equilibrium would be achieved more rapidly. Consequently, it is clear that, at the higher potentials, dissolution of the platinum is of great concern so that operating procedures must be established to prevent exposure under hot open-circuit conditions. 

Initially, crystallite growth occurs rapidly either by ion dissolution/reprecipitation (Ostwald ripening) or by surface atom diffusion, due to the requirement for lowering the surface energy of any individual crystallite. This thermodynamic driving force will tend to eliminate the incomplete faces but with the drive to lower the surface energy, the crystallites also will strive towards sphericity. This means, that to all intents and purposes, the ratios ofthe (111) and (100) faces should be approximately the same. Bett et al.1 noted that as the platinum crystallite sizes grew, the size distribution increased. If this is so, then... 

---