High-Resolution Electron Microscopy, Neutron Diffraction with Isotopic Substitution and X-Ray Absorption Fine Structure for the Characterisation of Active Sites in Oxide Catalysts

John Meurig Thomasjmt@ri.ac.uk    Davy Faraday Research Laboratory, The Royal Institution of Great Britain, 21 Albemarle Street, London W1S 4BS, UK
Department of Materials Science, University of Cambridge, Cambridge CB2 1QY

1 HISTORICAL INTRODUCTION

“After a long silence, for which I will not try to excuse myself, I have the pleasure of communicating to you, Sir, and through you to the Royal Society, some striking results which I have obtained in following up my experiments … … the electricity developed by the mere contact of different metals … … ”

These words were read by Sir Joseph Banks, President of the Royal Society in London in mid-April, 1800. They constitute the opening sentence of Alessandro Volta’s letter, despatched from Como on 20th March, 1800. I chose to open this lecture with these historical comments because, like others here this evening, I rejoice in being in Como, the birthplace of Plinius the Elder and Plinius the Younger, and also the place where Alessandro Volta was born and where he died. In common with millions of scientists world-wide, I have revered the memory of Volta ever since I first heard of his work. That reverence was enhanced when, nearly twenty years ago, my dear friend, the late Massimo Simonetta, with whom I had an exciting collaboration up until he passed away, took me to Pavia, where Volta was Professor. Later, more than a decade ago, my family and I came to this wonderful place, where I purchased (in the Volta Museum) a copy of Bertini’s famous painting of the occasion when Volta demonstrated his pile to the Emperor Napoleon in 1801.

My predecessor but seven as Director of The Royal Institution (RI), Michael Faraday, greatly admired Volta, who gave him a special Voltaic pole that Faraday frequently demonstrated at the RI, and which we still have (and exhibit).

In October 1842, Faraday received a letter from W.R. Grove (a Welsh lawyer-scientist working in London), describing the fuel cell, which Grove had invented. The opening of that letter is rather evocative:

“I have just completed a curious voltaic pile, which I think you would like to see ……”

When Faraday died in 1867, his successor John Tyndall (famed for the Tyndall effect, and also as the first person to describe the greenhouse effect and global warming) wrote a brilliant biography entitled “Faraday as Discoverer”[1]. Tyndall was a fine stylist and wrote with mellifluous charm. I should like to draw to your attention a passage in that book, which has particular resonance for this occasion when this Symposium honours Volta’s name:

“In one of the public areas of the town of Como stands a statue with no inscription on its pedestal, save that of a single name ‘Volta’ … … ”

Tyndall then proceeds to describe in beautiful English prose what Volta did, and the controversy that raged for some considerable time as to the origin of the electromotive force of a voltaic pile. Tyndall’s words merit repetition here, for they echo the essence of many subsequent scientific disputes.

“The objects of scientific thought being the passionless laws and phenomena of external nature, one might suppose that their investigation and discussion would be completely withdrawn from the region of the feelings, and pursued by the cold dry light of the intellect alone. This, however, is not always the case. Man carries his heart with him into all his works. You cannot separate the moral and emotional from the intellectual; and thus it is that the discussion of a point of science may rise to the heat of a battle-field”.

2 BACKGROUND

The only major ex situ technique that I propose to dwell upon is high-resolution (transmission) electron microscopy (HRTEM) which, nowadays (see later) can also be used for in situ studies. Its value as a tool in post mortem (as well as pre natal) studies of many catalysts has been highlighted on numerous previous occasions — see, for example, refs. [2–5]. I also wish to say a few words about electron crystallography[4].

So far as microporous oxide catalysts are concerned — and here I focus predominantly on zeolitic ones or their aluminophosphate analogues — the great merit of HRTEM[4–7] is that it can be used to “read off”, in real-space images, the symmetry elements of the structure, and often yield a reasonably accurate value of the number of tetrahedral sites in a pore aperture. (In general, 5-, 6-, 8-, 10- or 12-membered (T sites) can be straightforwardly discerned from the HRTEM images recorded down the appropriate zone axis). This real-space approach also proves invaluable in identifying intergrowth structures. I have given elsewhere[8] an account of how several of the ostensibly different members of the ZSM family of synthetic zeolites turned out to be no more than intergrowth variants of well defined end-members. HRTEM, twenty years ago, was shown[9] to be capable of “seeing” in real space the intergrowth variants of ZSM-5 and ZSM-11.

With CCD detectors and other instrumental advances[4,6] the age of electron crystallography is well and truly with us. Small, ultrathin specimens of zeolites (or mesoporous silicas) can be processed by a combination of electron diffraction and HRTEM imaging, so as to yield a full, hitherto unknown, structure, thanks to the pioneering work of Terasaki and his colleagues in Tohuku University.

Surface structures of complex oxides are also amenable to direct imaging and determination by HRTEM (which has the advantage over scanning tunnelling microscopy in being able, through the observation of parallel X-ray emission spectra, to yield elemental composition)[10]. There are two examples which I wish to cite.

First, La2CuO4, a material that is of great relevance in the field of warm superconductors, has been studied in detail by my former colleague Wuzong Zhou, now of the University of St. Andrews. A strictly stoichiometric powdered sample of La2CuO4 was viewed down the [110] direction by HRTEM. The outline of the [001] surface is seen in Figure 1.

Significantly, it is found that the surface composition does not correspond to La2CuO4. The images leave no doubt (and microanalysis confirms it) that the exterior surfaces (the last three layers or so) have the so-called C-La2O3 structure and composition[11]. Zhou and I have recently reported[10] that, in the warm superconductor HgBa2CuO4 + δ, the exterior and immediate sub-surface region is stoichiometrically quite different from the bulk. There are well-known reasons, discussed by Freund[12] recently, which lead us to expect polar surfaces (of oxides) to undergo reconstruction on exposure to vacuum (or an ambient gas). What is surprising here, in this powdered specimen of (nominally) La2CuO4, is that there is extensive reconstruction and reconstitution at the surface regions.

Low-energy electron diffraction (LEED) studies conducted by Freund et al[13], show that, in simple oxides belonging to the corundum family (e.g. Al2O3, Cr2O3 and Fe2O3), there is quite extensive reconstruction (in a direction perpendicular to the [0001] surface), amounting to between 50 and 60 percent change, compared with the inter-planar distances inside the bulk of the oxide. With the complex oxides that have zeolitic structure, however, the degree of reconstruction is rather minute. Thus, in zeolites L (idealised formula Na3K6[Al9Si27O72]) Terasaki et al has shown[14]...