Traditionally, electron microscopy has been a tool for biologists, materials scientists, and metallurgists. There are exceptions, but in general, chemists have tended to shy away from it. Typically, the subject isn't covered in chemistry courses and is not included in chemistry books.

But some researchers recognized long ago that TEM--not be confused with scanning electron microscopy or scanning tunneling microscopy--is a vital tool for gathering atomic-scale information from chemically relevant systems. In TEM, an intense and tightly focused electron beam with a few hundred thousand electron-volts of energy is transmitted through extremely thin specimens. Beams that have interacted with a specimen are detected and used for electron imaging, diffraction, and other types of studies.

Catalysts, catalyst supports, organic and inorganic crystals, semiconductors, nanostructures, and other scientifically and technologically important materials have been probed by electron microscopists. Many studies have relied upon innovations in instrumentation, development of new techniques, and advances in software for image analysis, structural computations, and instrument control.

For years, coming up with methods for probing solid catalysts while molecules react on their surfaces has been a top priority for researchers. Yet instrument limitations in many fields, not just microscopy, often have forced scientists to make do with the conventional approach: analyzing specimens before and after they catalyze reactions. Such studies can be informative. But before-and-after experiments leave open basic questions about mechanisms, intermediates, and dynamics that come into play while reactions are under way.

In contrast, in situ techniques would enable scientists to spy on catalysts in action. And in situ TEM, in particular, could shed light on atomic-scale events in heterogeneous chemistry. But how do you fit an 8-foot-tall, laboratory-sized electron microscope in a bench-top chemical reactor?

You don't. Pratibha L. Gai's solution was to redesign the microscope to accommodate a reactor. Gai heads the microstructural competency group at DuPont and is an adjunct professor of materials science at the University of Delaware.

"Our idea was to make an environmental cell that is an integral part of the microscope," Gai says. She explains that other microscopists had designed gas cells for TEM work, but either the microscope's resolution was inadequate for angstrom-level measurements or the cell was outside of the microscope, thus precluding in situ observations.

In the DuPont design, which was worked out by Gai and Edward D. Boyes, a senior fellow at DuPont, gas lines run through the objective lens's electromagnetic pole pieces. Delivering gaseous reactants to catalyst specimens in that way requires boring a hole through the lens--a risky operation, Gai explains.

ORDINARILY, the space in an electron microscope between the electron source and the specimen is maintained under high vacuum to enable electrons to travel unimpeded. Admitting gas to that region can degrade the instrument's resolution. But the ill effects of bumping up the pressure are minor compared with what might happen to a microscope if a hole is poked through one of its lenses.

"People thought we were crazy," Gai says. "What we were planning to do was like drilling a hole through someone's heart," because electromagnetic lenses are the most essential components of electron microscopes. "If we did it incorrectly and ruined the objective lens, we would have killed the microscope."

But the operation was a success. Gai notes that the redesigned DuPont microscope, which has been in regular use for more than three years, maintains atomic resolution (about 2 Å) while specimens are heated to 1,000 °C and exposed to gas pressures of up to a few torrs [Top. Catal., 8, 97 (1999)]. The pressures are on the order of 1 billion times higher than pressures under which conventional high-resolution TEM is carried out, and they approach industrial reaction conditions.

Gai points out that the environmental cell has been used to study reactions of alkanes, alkenes, carbon monoxide, hydrogen, and other reactive gases. She adds that the pressure can be boosted even higher--up to approximately 1 atm (760 torr)--but doing so compromises the resolution.

In one investigation, Gai and coworkers studied finely dispersed Pt particles supported on titania while the system was heated in hydrogen. Precious-metal catalysts, which are commonly activated by heating in hydrogen, are widely used in pollution control and other reactions.

ABLE TO KEEP track of a single nanometer-sized Pt particle as the temperature was raised from 300 to 450 °C, the DuPont researchers observed that at the higher temperature, a very thin titanium oxide film was beginning to encapsulate the Pt particle and small Pt clusters were beginning to grow nearby. Both processes are undesirable and can deactivate a catalyst. By revealing details of metal dispersion and sintering processes, as well as interactions between metal catalysts and their supports, the study enables scientists to improve catalyst design and preparation techniques.

Turning to other types of catalysts, Gai and coworkers studied vanadium phosphorus oxides. These materials, in which the active phase is believed to be vanadyl pyrophosphate, (VO)₂P₂O₇, are used to convert butane to maleic anhydride, which is used to synthesize tetrahydrofurans.

Through a series of in situ TEM experiments, Gai and colleagues deduced that the catalytic chemistry proceeds by way of a novel glide-shear mechanism (a type of crystal transformation) that enables vanadyl pyrophosphate to provide oxygen to butane while preserving active surface sites and without the catalyst's crystal lattice collapsing. Oxygen is replenished from the gas phase. Gai says the work led to the development of longer lasting catalysts.

Success with the DuPont environmental cell design has led microscope makers, such as Philips in the Netherlands, to offer the feature to its customers. The reconfigured microscope is gaining popularity. For example, in Lyngby, Denmark, researchers at catalyst manufacturer Haldor Topsøe adopted Gai and Boyes's cell design in their new TEM facility and used it to uncover angstrom-level details about the location, chemical state, and function of barium used to promote ruthenium/ boron nitride catalysts for ammonia synthesis (C&EN, Nov. 19, 2001, page 13). The Topsøe group also showed that copper nanocrystals, which are used as methanol-synthesis and fuel-cell catalysts, undergo dynamic and reversible shape changes in response to variations in the gas environment [Science, 295, 2053 (2002)].

Not one to leave well enough alone, Gai notes that a shortcoming of the environmental cell design is its incompatibility with corrosive gases. She's currently working on a design modification. In the meantime, very low pressures of halogens have been admitted to the microscope without adverse effects.

First it was gases. Now it's liquids. Widening the path to real-time in situ observations with TEM, Gai and her team recently designed a new type of sample holder that is used to inject microliters of liquids into an electron microscope. The new apparatus permits high-resolution in situ studies of liquid-solid and liquid-gas-solid reactions. Applying the wet microscopy method to a commercially relevant system, the DuPont group studied hydrogenation of adiponitrile to hexamethylene diamine, an intermediate in nylon manufacturing, and polymerization of the diamine with adipic acid.

The new technique enabled the DuPont scientists to immerse titania-supported cobalt-ruthenium catalysts in the organic solutions under flowing hydrogen (for the hydrogenation reaction) and follow nanoscale events, as they occurred, on the catalyst surfaces [Microsc. Microanal., 8, 21 (2002)]. Coupled with other DuPont studies, the TEM work shows that cobalt-ruthenium may be an effective alternative to the commercial Raney nickel catalysts used currently.

Wet TEM methods also played a central role in another recent investigation. Gai and coworker Mark A. Harmer studied growth and formation of gold nanorods prepared by reducing HAuCl₄ in a surfactant solution. Based on in situ microscopy work, the team concludes that the rods grow by forming certain types of multiply twinned (imperfect) nanocrystals that serve as sites for metal diffusion and promote nanorod growth. The results demonstrate that surfactant molecules can stabilize ordinarily unstable surfaces such as Au(110). The study also reveals that the (110) surface is ridged--not atomically flat--and other details relevant to molecular electronics technology [Nano Lett., 2, 771 (2002)].

Meanwhile, on the other side of the Atlantic Ocean, Sir John Meurig Thomas is another chemist who has spent a career developing TEM methods and using them in chemistry research. Thomas, who is a professor of chemistry and a former director at the Royal Institution of Great Britain, London, and professor of materials science at the University of Cambridge, has investigated a range of materials including organic and inorganic solids and porous catalytic materials. In addition, Thomas and materials science senior lecturer Paul A. Midgley and coworkers have developed high-resolution electron tomography methods and spectroscopy techniques that capitalize on the rich variety of signals that are produced as a microscope's electron beam interacts with specimens.

Chemical & Engineering News
Copyright © 2002 American Chemical Society