The physics of boron nitride nanotubes

© 2010 American Institute of Physics, S-0031-9228-1011-020-5 Tailored materials have long been central to condensedmatter physics. They often lead to new insights about the underlying physics governing the properties of materials— natural and synthetic—and sometimes provide new opportunities for applications. Superconductors are just one example. In the quest for higher transition temperatures, researchers have learned a great deal about superconductivity and the nature of electron–electron interactions. The goal usually pursued is “the higher, the better.” In the case of nanoscience, however, the goal might be rephrased as “the smaller, the better.” Interesting physics invariably emerges as material dimensions approach the atomic scale and quantum size effects influence how electrons interact with each other (see the article by Michael Tringides, Mieczyslaw Jałochowski, and Ernst Bauer in PHYSICS TODAY, April 2007, page 50). Dimensionality and symmetry restrictions also affect electronic behavior. When electron wavefunctions are confined to one or two dimensions, for example, the spectrum of collective excitations can change (see the article by Marc Kastner in PHYSICS TODAY, January 1993, page 24), and properties that are familiar in bulk crystals can become unfamiliar in nanostructures. Synthesizing intrinsically metastable nanostructures and characterizing them structurally, optically, thermally, electronically, or mechanically is challenging, but impressive advances in the field have been made. In addition to the usual atomic-scale microscopy techniques at their disposal, researchers have recently developed approaches for attaching individual wires or probes to isolated nanotubes and molecules for transport and mechanical measurements. Such approaches also probe molecular structures using methods— such as Raman spectroscopy, optical conductivity, and techniques that elicit a magnetic response—formerly amenable only to bulk materials. The case of boron nitride nanotubes (BNNTs) is a good example of how theoretical and experimental research led to the discovery of a new material that has opened a path to exotic material properties, intriguing new phenomena, and unique applications. The 1994 prediction by one of our groups (Cohen’s) that BNNTs may exist1 and a successful synthesis the following year2 by the other (Zettl’s) gave credibility to the idea that calculation techniques developed for bulk solids should be applicable to nanoscale objects. Hence the standard theoretical models and tools work. Ever since, BNNT research has been active and is sure to grow considerably in the near future. Carbon counterparts Currently, the most widely studied nanostructures are based on carbon—carbon-60 buckyballs3 and carbon nanotubes (CNTs)4 are prime examples (see articles by Phaedon Avouris in PHYSICS TODAY, January 2009, page 34, and by Cees Dekker in PHYSICS TODAY, May 1999, page 22)—partly because its synthesis technologies are easily adaptable. Fortunately, interesting polyhedral and cylindrical structures are not limited to carbon, and the work of Reshef Tenne and his colleagues on transition-metal compounds has produced a variety of novel materials that are part of that class.5 Indeed, numerous types of nanotubes and fullerene-like nanoparticles from inorganic layered (and even nonlayered) materials have emerged during the past two decades. The addition of BNNTs to the list—either in a pure form or in one that mixes C atoms into the composition, if desired—has given researchers yet another system to mine. Structurally, BNNTs resemble CNTs: As sketched in figures 1a and 1b, the material is an (effectively) onedimensional allotrope of a two-dimensional sheet of BN arranged in a hexagonal lattice (h-BN). Not surprisingly, tightly sp2-bonded BN sheets separated by weaker van der Waals bonds resemble the layered structure of graphite. But the well-known offset stacking of graphite layers—each C above or below the center of an adjacent layer’s hexagonal ring—is absent in h-BN, whose rings stack directly atop each other. The different stacking reflects a fundamental difference in bond charge configuration between C and BN, even for a single sheet. The charge distribution is symmetric in C– C bonds and asymmetric in B–N bonds. Because some of the electron charge on B is transferred to N, the bonds are not purely covalent (as in graphite) but possess some ionic character. The charge transfer gives rise to a gap between the valence and conduction bands; BN thus behaves like a widegap semiconductor, while graphite is metallic. On the other hand, h-BN is a purely synthetic material so far as anyone is aware, having been first synthesized in 1842. But its structural similarities to graphite have earned it the nickname “white graphite.” Tons of h-BN, widely used industrially as a high-temperature lubricant, are produced each year. Beyond its electronic bandgap and striking white color, many of the material’s properties are also found in BNNTs, including thermal stability at extreme temperatures in oxidizing environments and an ability to absorb neutron radiation. Despite the relative ease and low cost of industrial-level h-BN production, identifying an efficient production route The physics of boron nitride nanotubes