Inorganic Menagerie

In the world of nanotubes, there's no denying that carbon is king. Carbon nanotubes claim the lion's share of high-profile journal articles, reports in the popular press, and presentations at major scientific meetings. But a number of inorganic chemists have also fallen under the nanotube's spell. Delving into the periodic table, these scientists are finding that nanotubes made from inorganic materials have intriguing properties quite different from those of their all-carbon cousins and a range of potential applications.

Although the investment in financial and human resources devoted to inorganic nanotubes lags behind that of carbon nanotubes, a number of reviews suggest that inorganic nanotube research is increasing rapidly (Adv. Mater. 2004, 16, 1497; Dalton Trans. 2003, 1; Angew. Chem. Int. Ed. 2002, 41, 2446).

According to physicist Maja Remkar of the Jožef Stefan Institute in Ljubljana, Slovenia, as of last year more than 50 different varieties of inorganic nanotubes had been reported in the literature. Their compositions span the periodic table. Nanotubes made from transition-metal chalcogenides, oxides, and halides have all been synthesized, as well as mixed-phase, metal-doped, boron-based, silicon-based, and pure metal nanotubes.

"I think that carbon nanotubes are fantastic systems. The fact that their properties can be modified with small changes in diameter and chirality makes them fascinating materials," says Reshef Tenne, a materials science professor at Weizmann Institute of Science, Rehovot, Israel. He was the first to report the synthesis of inorganic nanotubes (Nature 1992, 360, 444). "But the richness of inorganic systems and their chemical versatility is also important, especially in applications where high loads, temperatures, or pressures are required."

Tenne rattles off a list of possible technologies that could capitalize on the unique properties of inorganic nanotubes: bullet-proof materials, high-performance sporting goods, specialized chemical sensors, smart windows, solar cells, catalysts, and rechargeable batteries.

LIKEWISE, Remkar envisions myriad possibilities for inorganic nanotubes. "The use of single nanotubes in special devices, such as a pointlike source of coherent electrons in the field-emission microscope, seems promising," she says. Remkar also mentions several applications for nanotubes that have few structural defects: quantum electronics with a single tube as a component, nanotubes as small electromagnets, nanopipes that could interact with a living cell without damaging the cell membrane, ultrastrong fibers, and safe containers for macromolecules.

"It is difficult to predict which kinds of inorganic nanotubes will be the most applicable in the future," Remkar notes. "After the first wave of enthusiasm following the successful synthesis of cylindrical crystals of a new compound, additional demands appear with respect to their properties," she explains. Scientists seek to control the tube dimensions and the degree of order and to achieve reproducible physical properties, which are crucial for some applications.

As carbon nanotubes move closer to commercial applications, Remkar adds, researchers studying inorganic nanotubes are feeling pressure to come up with applications quickly. Most applications of inorganic nanotubes, however, are still in the early stages of development, with commercial products a distant goal.

Some inorganic nanotubes, though, are already realizing some commercial success. Tenne's research is one good example. For more than a decade, he and his colleagues have been studying nanotubes and fullerene-type materials made of WS₂ and MoS₂ (J. Mater. Chem. 2005, 15, 1782). In 2002, the research spawned ApNano Materials, a company that is commercializing these nanoscale lubricants, known as NanoLub, for automotive and aerospace applications.

Tenne admits that he didn't set out to become an inorganic-nanotube baron. Commercial applications couldn't have been further from his mind back in the early 1990s when he was working and traveling in Japan. In the course of his travels, Tenne had the opportunity to attend a few presentations on the work that Sumio Iijima, a scientist with NEC Corp. in Tsukuba, was doing with carbon nanotubes (Nature 1991, 354, 56).

Tenne realized that the graphitic materials that Iijima was working with were very similar to certain inorganic materials. "I hypothesized that if we reduce these materials to a very small dimension, they will spontaneously form tubes and onions," he says. He approached his colleague Lev Margulis, an electron microscopist, about the possibility that these nanostructures might form.

Together with postdoc Menachem Genut, currently the chief executive officer of ApNano Materials, the researchers examined a sample of nanoparticulate WS₂ and found that, indeed, small pieces of the compound's planar form were unstable. The reactive edges of these sheets would fold upon themselves, producing multiwalled nanotubes and onionlike cage structures, analogous to fullerenes. The discovery, Tenne says, opened up a new avenue for his research. In the course of one year, he says, they synthesized four different inorganic compounds--WS₂, MoS₂, WSe₂, and MoSe₂--in the form of nanostructures.

Novel inorganic nanotubes in hand, Tenne started thinking about applications. Tenne knew that MoS₂ had been used as a solid lubricant for a number of years. "I said to myself, 'These nanostructures would be ideal nano ball bearings,' " he remembers, "but I didn't have the faintest idea how to synthesize the material in large amounts."

Tenne consulted his colleague Gary Hodes, a professor at Weizmann Institute. The researchers developed an inorganic nanoparticle synthesis that employs the reaction of H₂S with the appropriate metal oxide at elevated temperatures and under reducing conditions. The metal oxide serves as a template, Tenne explains, and the reaction proceeds from the outside of the particle inward.

Just as Tenne predicted, the properties of the WS₂ and MoS₂ nanotubes and fullerene-like particles make them excellent solid lubricants. Unlike typical metal disulfide lubricants, the curved surfaces of Tenne's nanoparticles have no exposed, reactive edges. Furthermore, because the structures are layered, they can exfoliate without losing lubricity.

"The fact that they can support a very high load without breaking makes them better than liquid lubricants," he explains. The materials can withstand high temperatures, making them ideal lubricants for the automotive industry, Tenne says. He adds that they are inexpensive to produce--roughly $100/kg, depending upon the quality of the material needed. Practically speaking, Tenne thinks that a car would probably need no more than half a kilogram of the lubricant. Also on the plus side, ApNano Materials announced in April that NanoLub had been declared nontoxic by an independent lab in Israel.

Other properties make the nanoparticles attractive solid lubricants for other industries as well. Tenne says that aerospace agencies have expressed interest in the materials because of their stability in low-pressure environments. Also, the nanoparticles' low vapor pressure suggests that they could be useful in ultraclean environments, such as biomedical devices, semiconductor fabrication facilities, and food and pharmaceutical manufacturing equipment.

In addition to their lubricating properties, WS₂ and MoS₂ nanotubes have proven in tests to be remarkably resistant to compression. "Carbon nanotubes are only very strong if you put them under tensile stress," Tenne says. "Our nanotubes are heavier and not as strong under tensile stress, but they are very strong under compression." So, while not perfect for tensile applications like the hypothetical space elevator, these inorganic nanotubes may have good down-to-Earth applications, such as in bullet-proof vests and automotive materials that won't crumple in a collision.

WITH SO MUCH interest evident in the WS₂ and MoS₂ nanotubes and fullerene-like nanoparticles, Tenne says the biggest obstacle he faces is simply making enough of them. "If we wish to use these materials industrially, then we have to produce them by the ton." In January, ApNano Materials announced that it had secured funding to build a semi-industrial manufacturing plant that can produce about 150 kg of NanoLub on a daily basis. That, Tenne says, is a far cry from the submilligram amounts he was making back in 1992.

Tenne wasn't the only one to see similarities between graphite's layered structure and inorganic compounds. The possibility of creating nanotubes from carbon's nearest neighbors on the periodic table has attracted the attention of theorists and synthetic inorganic chemists alike. Theoretical studies indicate that boron and boron nitride nanotubes are likely to have properties comparable with those of carbon nanotubes, with one key exception: Nanotube chirality--which helps determine whether a tube is metallic or semiconducting--complicates carbon nanotube research but is not a problem with either boron-based structure. Undoped BN nanotubes, first prepared a decade ago (Science 1995, 269, 966), have proven to be uniformly insulating. Computational studies have shown that all-boron nanotubes are likely to be uniformly metallic.

There is a vast amount of literature on the synthesis and preparation of BN nanotubes. "BN and BN-based nanotubes have been a subject of interest to our group because of their uniform electronic structure that is independent of geometry," explains Yoshio Bando, a scientist at Japan's National Institute of Materials Science in Tsukuba and one of the world's most prolific inorganic-nanotube makers.

"BN has many unique properties. It is chemically inert and has a high-temperature resistance to oxidation," Bando adds. Researchers think that BN nanotubes may have promising applications in hydrogen storage.

Less is known about all-boron nanotubes. Theorists Alexander Quandt and Ihsan Boustani, at Germany's University of Greifswald and University of Wuppertal, respectively, have been using computational methods to predict the properties of boron nanotubes. They've written a review on the topic that will appear in an upcoming issue of ChemPhysChem.

In contrast to their BN cousins, all-boron nanotubes have proven to be considerably more difficult to make. "It's like structural engineering, basically," Quandt jokes. "It is clear how to make boron nanotubes on a computer, but it is not so clear how to make them in the lab."

Last year, however, Dragos Ciuparu and colleagues in Yale University's chemical engineering department synthesized the first pure, single-walled boron nanotubes (J. Phys. Chem. B 2004, 108, 3967).

In terms of applications, Quandt does not think that boron nanotubes will ever be a substitute for carbon nanotubes, despite their structural similarity. He does think that boron nanotubes could make excellent junctions between carbon nanotubes and silicon components in electronic devices, thanks to the tubes' similar size and architecture, as well as boron's natural affinity for silicon.

Not all inorganic-nanotube makers set out to mimic the layered structure of multiwalled carbon nanotubes. Some, like Peidong Yang, want to defy it.

In 2003, when Yang, an associate chemistry professor at the University of California, Berkeley, wanted to create an artificial version of the ion channels in cell membranes, nanotubes seemed like an ideal place to start. But he knew right away that the nanotubes that had been reported wouldn't be of any use to him.

"To make an artificial ion channel, you have to create nanotubes that have the ability to run liquid through themselves," Yang explains. The nanotubes, he says, need to be regular, free of pinholes, and moderately hydrophilic. To achieve a regular, defect-free structure, Yang wanted to use single-crystal nanotubes. Layered nanotubes, like WS₂ and carbon nanotubes, wouldn't work because they are too hydrophobic to transport ionic solutions.

Yang wondered if it was possible to create single-crystal nanotubes based on nonlayered materials. "When I started to look into this research program, my group had been working on nanowires," Yang says. It occurred to him that he could use the nanowires as a template around which he could cast some other material to create a nanotube.

Yang chose to make his first nanotubes out of gallium nitride because the dimensions of GaN's crystal lattice are similar to the ZnO nanowires his group had been making. Yang figured that this similarity would coax GaN into growing in a crystalline alignment. He and his coworkers used chemical vapor deposition (CVD) to coat an array of ZnO nanowires with GaN. They then removed the ZnO by heating the material in the presence of hydrogen (Nature 2003, 422, 599). They found that this procedure gave them excellent control over the nanotubes' dimensions. The size of the nanowire determined the nanotube's inner diameter, and CVD growth time determined the tube's outer diameter.

"This technique presents a way to make nanotubes without a need to use layered materials," Yang says. His research team has subsequently used the casting technique to make silica nanotubes. Others have expanded upon the technique to make a wide range of inorganic nanotubes.

Working with his collaborator Arun Majumdar, a professor in UC Berkeley's engineering school, Yang has used these nanotubes as conduits for ions and charged biological molecules in nanofluidic devices (Nano Lett. 2005, 5, 943).

When an ionic solution is introduced into one of Yang's silica nanotubes, the structure's small size and inherent negative surface charge lead to an environment within the tube that's enriched in the solution's cation, Yang explains. Using an external electrode, Yang's group can control the flow of the cations--as well as charged biological molecules, such as DNA--through the tube (Phys. Rev. Lett. 2005, 95, 086607; Nano Lett., published online June 28, dx.doi.org/10.1021/nl0509677).

"For the first time, we've clearly demonstrated that these nanotubes can be used to regulate ionic transport. This is the exact function we want to mimic in ion channels in the cell membrane," Yang says. He imagines that these ion-transport systems could eventually be used to sense, transport, and manipulate biological molecules. "I think some of the most important implications are in sorting, separation, and in situ reactions on the single-molecule level," he notes.

IT WAS SIMPLE curiosity that drew Craig A. Grimes, an associate professor of electrical engineering at Pennsylvania State University, into inorganic nanotube research. Grimes and coworkers wanted to see what happened when they anodized titanium--that is, oxidized it by electrolysis.

"What we saw was this very interesting structure," he recalls. The anodization process had transformed the titanium sheet into a highly organized array of amorphous TiO2 nanotubes. Grimes's group then annealed the material to make the tubes crystalline. Crystalline tubes, Grimes explains, are important for applications.

Titanium dioxide nanotubes had been prepared by other researchers previously. But Grimes's group found that creating the nanostructures in organized arrays dramatically altered the material's properties and, consequently, its potential applications.

"The architecture that we work with is very different from a random glob of nanotubes," he says. "These things are precisely ordered." Contrary to much of the hype about nanotechnology, Grimes explains that simply going to the nanoscale doesn't necessarily endow a material with new properties that can immediately translate into novel applications. "It's not like a miracle happens," he says. "You have to have the right architecture to get somewhere.

"It's been known for decades that titania responds to different gases, but in this architecture the sensitivities are absolutely mind-boggling," he continues. In fact, Grimes reports that the sensitivity of his TiO₂ arrays to hydrogen is greater than that of any other material to any gas at any temperature (J. Nanosci. Nanotech. 2004, 4, 733).

Grimes thinks that dissociated hydrogen is adsorbed directly onto the nanotubes' walls and into the regions between the tubes. This adsorbed hydrogen, he explains, creates a layer of charge on these regions, substantially reducing the tubes' resistance.

The architecture of the array is the key to this phenomenon, Grimes says. He likens the hydrogen's movement to walking a straight path through the trees of a neatly planted Christmas tree farm. In bulk material or in a disorganized mass of nanotubes, there's no easy path for hydrogen, he adds.

Grimes's team is capitalizing on the array's unique sensitivity by using it in a hydrogen sensor. He says the system has a number of advantages. For example, the arrays are inexpensive to make, and the anodization technique is scalable. "You can take a sheet as big as you want, put it in an electrolyte bath, hook it up to a voltage, and voil?," he says. "You take it out, you rinse it, and you're ready to go."

The TiO₂ endows the sensors with the self-cleaning properties initiated by ultraviolet light, Grimes notes. "We showed that you can contaminate this hydrogen sensor in a variety of ways--dousing it in motor oil, blowing smoke on it--and under UV light it cleans itself and regains its sensitivity." Salty fluids were the only things the group found that incapacitated the sensor.

Although there are several possible applications for the hydrogen sensors, Grimes is particularly interested in using them to detect a lethal bacterial infection that afflicts premature infants. One sure sign of the infection is the release of small amounts of hydrogen from the bacteria. Being able to detect that emission early would give doctors a chance to treat the infection before it becomes dangerous.

Grimes says the technology is currently being tested on lactose-intolerant adult volunteers because their stomachs give off hydrogen after they consume lactose. Once those tests are complete, Grimes and his collaborators plan to test the sensors on premature babies--a test he hopes will eventually become standard care.

While the hydrogen-sensing applications are exciting, Grimes also has used the arrays for water photolysis (J. Nanosci. Nanotech. 2005, 5, 1158). He imagines that nanotube arrays of other metal oxides could have remarkable properties as well. "We're not only applying these titania nanotube arrays to hydrogen sensors and photolysis, we're trying to create this same architecture in other metal oxides."

Grimes tells C&EN that the world's limited supply of titanium--most of which, he says, went into making Soviet nuclear submarines and which is currently in demand for high-end racing bicycles--is another motivating factor behind his efforts to make other metal oxide nanotube arrays. "We're trying to achieve this same architecture in iron oxide," he says. "If we could do that, it would be a really great thing."

All the inorganic nanotube researchers who spoke with C&EN agree that inorganic nanotubes, long overshadowed by their carbon cousins, are finally being recognized for their potential in nanotechnology. While they point out that there is still plenty of work to be done, all expect that nanotube researchers will continue to mine treasures from the expanse of the periodic table.