Field Ion Microscopy of the Platinum Metals By Erwin W. Miiller Research Professor of Physics, Pennsylvania State University

Platinum, iridium, rhodium and palladium are among the few metals that may be examined by je ld ion microscopy, in which it is possible to magnify the image of a prepared surface so that individual atoms are revealed. This technique, by which such factors as plastic deformation, gas adsorption, fatigue, dislocation structure and the egects of im- purities can be examined, may well lead to a better understanding of the valuable properties shown by the platinum metals in their catalytic, electrochemical and high temperature applications.

The field ion microscope is the only device capable of directly showing us the individual atoms at a metal surface (I). Its operation is based on the simple principle of radially projecting the nearly hemispherical specimen, the cap of a needle tip, on to a fluorescent screen by using gas ions. The necessary magnification, of the order of IO~, is obtained by using a tip radius between a few hundred and a few thousand bngstroms, while the projection distance is about 10 cm. The extremely high degree of geometric perfection of the specimen surface which is necessary to make the projection principle work is achieved by field evaporation. This effect balances the binding energy of the metal ions with the force of the applied electric field. At protru- sions the field is locally enhanced, and the metal field evaporates until a smooth curva- ture is obtained over the entire tip cap.

The resolution of a projection microscope improves with increasing field strength at the emitter. Helium or neon, having the highest ionisation potentials, are, therefore, preferred for imaging. Optimum ionisation occurs at 450 or 350 MV/cm (million volts per centi- metre), respectively. Obviously, only those metals that do not field evaporate too fast at these extremely high field strengths will

give sufficiently stable images. The evapora- tion field increases approximately with the sum of the vaporisation energy of the metal and its ionisation energy. Although the theory of field evaporation is not yet too well estab- lished (in part due to the inadequacy of manageable linear surface potential models for the description of the conditions at an atomic site where the potential changes by 10 to 20 volts across one atomic diameter), one can estimate, and confirm experimentally, that about twenty pure metals, from tungsten with the highest melting point over the other refractory metals and the platinum metals down to copper and gold, are suitable for field ion microscopy.

Not all of these metals are easily observed. The field exerts a mechanical stress F2/8= which amounts to I ton 'mm2 at 480 MV, cm. For the possibility of seeing the original atomic lattice structure of the specimen it is decisive whether field evaporation or yield to the field stress occurs first. This must be found out experimentally, the balance of the two effects being uncertain because of the inadequacy of the theories for both.

Surprisingly, it turns out that besides tungsten only iridium, platinum, rhodium, palladium and gold have sufficient strength

Platinum Metals Rev., 1965, 9 , (3),84-89 84

Fig. 1 Platinum rrystal of SO0 A average radius with numerous surface vacancies caused by adsorption of nitrogen followed by jield desorption

at their respective evaporation fields to resist the field stress or its shear component. The refractory metals, rhenium, tantalum, moly- bdenum and niobium, as well as other tran- sition metals such as nickel and iron, all glide in certain crystallographic areas when exposed to the field evaporation stress. With the latter metals the discrimination between intrinsic lattice defects and artifacts due to the shear stress of the field is very difficult.

So far, the outstanding suitability for field ion microscopy of the platinum metals has not been really utilised. During field evapora- tion, platinum is exposed to a stress exceeding

its technical tensile strength by a factor of 30, and yet no inelastic deformation occurs. The surface of the specimen, and by controlled field evaporation also the bulk of the metal, can be observed to determine the vacancy concentration in quenched material by direct counting of empty atoms sites (2) or to see surface vacancies or rearrangement caused by reaction with adsorbed gases (3) (Fig. I). Cycling field stresses have been used to observe fatigue cracks (4). Radiation damage can be seen in the form of vacancies and interstitials when x particles are used ( 5 ) or

(text continuad on page 89)

Platinum Metals Rev., 1965, 9 , (3), 85

Fig. 2 Plat inum crystal hemisphere of approximately 1800 A radius. Same orientation as in Fig. 1. M a n y of the more than a thousand crystal facets (high index net planes) are resolved. Arrows point to radiation damage spikes intentionally produced by the impact of individual heavy negative ions (a luminium oxide) of 25 Ice V energy

Fig. 3 Plat inum crystal in same orientation as the one in Fig. 2. Several slip bands were produced in situ by exposing a n initially perfect crystal (similar to Fig. 2 without the radiation damage and only 1000 d radius) at 450°C to a jield of 200 MVIcm

Platinum Metals Rev., 1965, 9 , (3),

Fig. 4 Section of a plati- n u m crystal of approxi- mately 1000 A radius with several dislocations. A t the core of the upper left dislocation there i s a decor- ation by a n impuri ty atom

Fig. 5 A rhodium crystal of 1500 d radius with numerous impuri ty atoms (oxygen interstitials) ap- pearing as bright blobs. The average eoncentration of these impuri ty atoms in this photograph i s 1.8 x 1 0-3

Platinum Metals Rev., 1965, 9 , (3), 87

Fig. 6 An ordered, almost perfect, 50 atomic per cent cobalt-platinum crystal

Fig. 7 An ordered cobalt-platinum crystal with several grain boundaries

Platinum Metals Rev., 1965, 9 , (3), 88

large damage spikes when heavier ions are used (Fig. 2). Slip bands (6) are easily pro- duced in situ by applying field stress to the heated specimen (Fig. 3). In tips with not too large radii single dislocations (7, 8) often do not move under the field stress, and their core structure and possible decoration with im- purity atoms can be seen (Fig. 4). I n platinum, iridium and rhodium a certain impurity, most probably oxygen, shows up as a bright blob with an apparent size larger than the metal atoms (Fig. 5).

There is also a promising application to the study of alloys. I n random solute alloys field evaporation cannot produce the absolutely perfect surface geometry which is obtained with the pure metals. Randomness of distri- bution of the kind of nearest neighbours locally varies the binding energy at equivalent lattice sites (9). However, dilute alloys might be suitable for the study of short range order. Of the alloy systems with long range order, 50 atomic per cent cobalt-platinum is a most promising subject (9, 10) for field ion microscopy (Fig. 6).

The reason for neglecting so long the many opportunities which the platinum metals offer for field ion microscopy lies probably in the fact that this technique is fairly new and requires a skilful operator. With the recently increasing interest in this method of inquiry, the availability of an inexpensive commercial

instrument (11) and the progress in image interpretation (12, 13) useful new informa- tion on the physical metallurgy of the platinum metals should be obtained in the not too far future.

References I E. W. Miiller, “Field Ionisation and Field Ion

Microscopy”, Advances in Electronics and Electron Physics, Vol. XIII, pp. 83-179, Academic Press, New York, 1960

z E. W. Miiller, Zeit.f. Physik, 1959, 156, 399 3 J. F. Mulson and E. W. Miiller, J . Chem.

Phys., 1963, 38, 2615 4 E. W. Miiller, W. T. Pimbley and, J. F.

Mulson, “Internal Stresses and Fatigue in ihletals”, (Eds. G. M. Rassweiler and W. L. Grube), p. 189, Elsevier, Amsterdam, 1959

5 E. W. Muller, in “Reactivity of Solids”, (Ed. J. H. de Boer) p. 682, Elsevier, Amsterdam, 1960

6 E. W. Miiller, Acta Met., 1958, 6, 620 7 E. W. Miiller, IV Intern. Kongr. Elektronen-

mikroskopie, Berlin, 1958, Vol. I, p. 620, Springer, Berlin, 1960

8 E. W. Miiller, Proc. Int. Conf. on Crystal Lattice Defects, Kyoto, 1962, J. Phys. SOL. Japan, 1963, 18, Suppl. 11, p. I

g E. W. Miiller, Bull. Am. Phys. Sac., 2962, 11, 7, 27

ro. D. G. Brandon, M. Wald, M. J. Southon and B. Ralph, Proc. Int. Conf. Crystal Lattice Defects, Kyoto, 1962, J . Phys. SOC. Japan, 1963, IS, Suppl. 11, 324

I I Central Scientific Company, Inc., Chicago 13, Illinois

12 D. G. Brandon, Brit. J. Appl. Phys., 1963, 14, 474

13 E. W. Miiller, Surface Science, 1964, 2, 484

Disposal of Radioactive Residues PHOSPHATE GLASS MELTED IN PLATINUM-LINED CRUCIBLE

During the reprocessing of spent fuel elements from nuclear reactors, highly toxic radioactive fission products are accumulated as waste, which must be carefully stored and controlled for centuries. Glass is being studied as the medium for fixation at several establishments, and A. M. Platt and C. R. Cooley of Battelle-Northwest, Richland, Washington-until recently the Hanford Laboratories operated by the General Electric Company-have now constructed a pilot plant for final demonstration of the techniques involved.

In one technique phosphoric acid is added to a concentrated waste. After further concentration by evaporation of nitric acid and water, the waste is fed into a platinum- lined melter operating at about 125ooC, which further reduces the volume and con- verts the residue to a molten glass. This overflows into a storage vessel and solidifies. By this means stable nuggets of glass have been produced that resist leaching and incorporate up to 95 per cent of the radio- activity present in the original liquid waste.

F. J. S.

Platinum Metals Rev., 1965, 9 , (3),89-89 89