Observing Orbitals

Electrons are wily particles that, according to the laws of quantum mechanics, just can't be pinned down. So scientists describe their positions in terms of orbitals--essentially regions in which electrons are most likely to be found as they whiz around atomic nuclei. The idea of orbitals has long proved useful for describing atoms and their interactions mathematically, but not physically.

Now, that's all changed. Researchers at Arizona State University recently published in Nature the first true images of atomic orbitals in Cu₂O, a crystal called cuprite. Said lead author J.M. Zuo: "It's direct, experimental proof of the quantum model."

The pictures, taken using a novel technique by Zuo, M. Kim and John Spence in the Department of Physics and Astronomy and chemist Michael O'Keefe, confirm that orbitals are indeed shaped like spheres, dumbbells, petals and doughnuts--depending on the energy and other properties of the electrons inhabiting them and interacting with them.

Moreover, the images also resolve a controversy about the types of bonds between copper and oxygen in certain crystals--collectively called copper oxides--that conduct electricity without resistance at high temperatures. "Understanding bonding in copper oxides is the key to solving the biggest unsolved problem in solid state theory--the nature of high temperature superconductivity in copper oxides," Zuo noted. Such superconductors are thought to hold great promise for future technologies because of their unusual properties.

In fact, there are three fundamental types of bonds that hold all of matter together: metal-to-metal bonds are created when the outer electrons of individual atoms merely intermingle; covalent bonds occur when two atoms share pairs of electrons between them; and ionic bonds form when one atom gives an electron to a neighboring atom. Colin Humphries of Cambridge University suggested in the 1970s that covalent bonds might exist between copper atoms in copper oxides, but without experimental proof, the prevailing theory held that only metal-to-metal bonds were possible.

The new pictures, however, showed Humphries right. These charge density maps of non-ionic bonds in cuprite reveal a dumbbell, with a doughnut and three petals around its middle, where a copper ion resides--a configuration predicted for a s-d_z² orbital hybridization. Covalent bonding appears between copper and oxygen and copper and copper. And fainter distributions of electrons floating loosely between copper atoms are suggestive of metal-to-metal bonds. "The evidence of covalent bonding between metals is likely to make them rewrite the chemistry textbooks," Spence commented. "Chemistry has always assumed that these are only possible between copper and oxygen in this material."

The team made their images at the Center for High Resolution Electron Microscopy, bombarding cuprite crystals with both electron and X-ray beams. The electron beam bounced off mostly electronic bonds in the material, whereas the X-rays rebounded from the nuclei. As the returning beams interacted with one another, they created a diffraction pattern telling of what they had just hit, which the researchers used to generate an image much in the same way as interfering light patterns are used to create photographs and holograms.

To make this method work, Zuo and company had to measure the angles at which the beams scattered from the crystal with a higher degree of precision than had been done before. They relied on the electron beam to measure small angles, which it did more accurately by avoiding the "extinction effect" that distorts X-ray images. In contrast, the X-ray beam was better at measuring larger angles. And the combination of the two made it possible to flesh out the fine details of the crystal structure. To make a sharp picture of the covalent bonds between copper and oxygen, the group manipulated the charge density map by first moving all ions to the back of the map and then subtracting the background.

Some scientists who are analyzing the pictures have slightly different interpretations. For instance, Roald Hoffmann, the Cornell chemist awarded a Nobel prize in 1981, is skeptical about covalent bonds between copper atoms, believing them to be too far apart. But everyone is convinced of the technique the Arizona researchers developed. Putting ultimate decisions about cuprite's bonds aside, combining convergent beams of electron beam diffraction and X-ray beam diffraction should help researchers better understand a variety of complex materials in the years to come.