Diatom Menagerie

Scientists have long prized diatoms, photosynthetic algae that abound in marine and freshwater ecosystems, because they remove large amounts of a major greenhouse gas—carbon dioxide—from the atmosphere. But another, unusual trait has recently caught the attention of materials scientists and engineers: The cell wall of this unicellular organism is made entirely of glass. More precisely, diatom shells consist of silica, or silicon dioxide, the primary constituent of glass. Many shells are ornately patterned with features just tens of nanometers in size. What’s more, there are thousands of different species of diatoms, each with a unique shell design. Some look like miniature sieves, others resemble microscopic gears.

By seeking to understand how these organisms build intricate silica structures, researchers expect to learn valuable lessons for designing and manufacturing new kinds of nano- and micrometer-scale materials and devices. “Nature has been building things on the nanoscale for a long time,” says materials scientist Ken Sandhage of the Georgia Institute of Technology in Atlanta. “We’re just scratching the surface in terms of learning how to take advantage of these organisms to make all sorts of devices for biomedical applications, telecommunications, energy storage, and sensing.”

Joanna Aizenberg of Lucent Technologies’ Bell Laboratories in Murray Hill, N.J., says, “We can think of diatoms as living silicon chips.” Semiconductor-chip manufacturers carve micro- and nanoscale features out of blocks of electronic and optical materials—a costly and time-consuming endeavor. Diatoms build structures out of silicon much more efficiently.

Once researchers figure out how to engineer useful devices out of diatom shells, they could enlist the reproductive capabilities of diatoms to generate trillions of silica structures in a matter of weeks. Some species of diatoms can replicate up to eight times a day.

Sandhage says, “For a fairly small number of reproductions, you could get incredibly large numbers of the exact-same three-dimensional structure.”

Although diatoms are unlikely to put the semiconductor industry out of business in the near future, their capacity to create complex structures on a small scale could serve as the foundation of a powerful technology for churning out new materials.

Gene machine

Observing these glass artists under a microscope can stir the mind’s eye. “Diatoms can make just about any structure you can imagine,” says Mark Hildebrand, a biologist at the Scripps Institution of Oceanography in San Diego. He and other researchers are investigating the molecular mechanisms that underlie shell formation.

It begins when the algal cell divides, forcing it to split its shell into two halves. The new cells, each now bearing only half a shell, begin to reconstruct their missing halves by taking up silicic acid—a simple compound of silicon, oxygen, and hydrogen—from the surrounding water.

Each new organism deposits the silicic acid in a compartment called the silica-deposition vesicle. There, the chemical is converted into silica particles, each measuring about 50 nm in diameter. These then aggregate to form larger blocks of material. Researchers speculate that a set of special proteins guides the formation of the silica particles and their subsequent assembly into larger structures. Hildebrand says that other cellular proteins outside of the vesicle stretch and mold the compartment to shape the silica inside.

Once the half shell is complete, the vesicle merges into the cell’s membrane, exposing the newly created structure.

In the late 1990s, Hildebrand identified a gene for a protein that draws silicic acid from the environment into the cell. This is still the only gene reported to take part in the diatom’s silicon metabolism.

That won’t be the case much longer. An international team of biologists, including Hildebrand, is preparing to publish the first genome sequence of a diatom—specifically, of the marine species Thalassiosira pseudonana. “This is really going to change everything,” says Hildebrand. “Now, we can do large-scale surveys of all the genes to find those involved in the process.”

To find those genes among the diatom’s approximately 11,000 genes, Hildebrand and his colleagues grow the algae in the lab and then put them in a solution lacking silicon. This stops the cells from dividing and forming new silica structures. When the researchers add silicon back to the growth medium, the diatoms begin forming new shells. At that moment, the researchers analyze the organisms’ genetic material to see which genes have turned from off to on.

Resembling a small pillbox and lacking ornate features, the silica shell of T. pseudonana is “pretty dull,” says Hildebrand. However, he offers it as a model organism—the fruit fly of diatom research. Once researchers determine how shells are made in T. pseudonana, he says, they can move on to more-complex species.

Diatoms of the same species consistently form shells with exactly the same pattern, suggesting that the designs are genetically programmed. By surveying a range of diatoms, researchers may find genes that drive one species to form star-shaped shells with arrays of nanoscale pores and grooves, for example, while other genes create a solid structure jutting long spikes.

And that’s just the beginning. “Ultimately, we’d like to genetically modify these organisms,” he says. The main thrust of his team’s project, Hildebrand says, is to knock out or modify the activity of specific genes so that researchers can engineer diatom shells for a wide variety of applications requiring microscopic materials with nanoscale features.

For instance, a glassy material with well-ordered pores could serve as a photonic crystal for optical communications (SN: 10/4/03, p. 218: Hot Crystal), or a microfluidic chip with tiny channels could perform small-scale chemical reactions (SN: 9/28/02, p. 198: Available to subscribers at Liquid Logic: Tiny plumbing networks concoct and compute).

Silica replicator

Sandhage is leading a massive effort to exploit diatoms’ manufacturing prowess, although 5 years ago, he knew next to nothing about these algae. He says that a new way of thinking about materials design opened up when a biologist in Germany introduced him to these unicellular organisms. Sandhage has since teamed up with Hildebrand and researchers at Ohio State University in Columbus and the Air Force Research Laboratory in Dayton, Ohio, to turn diatoms into mass producers of new electronic and optical devices.

For industrial applications, one problem with diatoms is that “they have evolved to be pretty good at making things out of silica but not of much else,” says Sandhage. Many applications require metallic or semiconductor materials, so he is working on ways to convert diatom structures from silica to other materials.

Sandhage and his colleagues have developed a chemical process that preserves a diatom shell’s precise pattern while replacing the silicon in the shell with another element, atom by atom. In a first experiment reported 2 years ago, the researchers converted all the silica in a diatom shell into magnesium oxide.

They accomplished the feat by removing the organic material from the diatom shells, placing the structures inside a metal tube, exposing them to magnesium gas, and heating the tube’s contents. Because magnesium is more strongly attracted to oxygen than to silicon, magnesium atoms elbow out the silicon, forming magnesium oxide. Over several hours, the metal replaces all of the silicon atoms in the shell structure.

The particular structure in that experiment was derived from a diatom called Aulacoseira. Its shell resembles a tubular capsule scored with v-shape grooves and rows of tiny pores, measuring about 200 nm each in diameter and spaced several hundred nanometers apart. After the conversion was complete, the researchers found that the shell’s features stayed within 30 nanometers of their original size and location.

In a more recent experiment, described in the April 2004 Chemical Communications, Sandhage’s team exposed diatom shells to a titanium fluoride gas. The titanium displaced the silicon, yielding a diatom structure made up entirely of titanium dioxide, a material used in some commercially produced solar cells and commonly found in paints as pigments.

The particular crystal of titanium dioxide called anatase that formed during the reaction could be used as a catalyst to split water for making hydrogen fuel, Sandhage says. It could also form the basis of a device that could detect specific gases. Carbon monoxide, for example, sticks to the surface of anatase and produces a detectable change in the material’s electrical resistance.

Gas sensors derived from diatom structures have great potential because “you want to have a very high surface area with an open structure so that you get a bigger signal,” Sandhage says. Some of the features found in diatom shells—arrays of pores or long and narrow grooves—are ideal for this kind of application, he says. What’s more, diatom shells are extremely small. “You could put lots of them in very small places,” he says.

Already, the group at Ohio State University has begun testing some of the diatom-designed titanium dioxide shells as gas sensors.

Once Hildebrand and his team have figured out how to create organisms that make specific structures, Sandhage plans to transform them into useful materials. The goal is to “genetically engineer microdevices,” he says.

Nature versus nurture

Rather than relying on diatoms to churn out inorganic structures, other groups are working to isolate specific silica-forming proteins from the diatoms and use them as templates in the assembly of the desired structures in the lab. This particular strategy is part of a worldwide effort to harness the power of biological materials to build inorganic structures for use in electronic and optical devices (SN: 7/5/03, p. 7: Microbial Materials).

The approach may prove simpler and offer greater control over the ultimate design than employing algae or other organisms to produce the materials. For example, by attaching silica-binding proteins on a polymer surface in a precise arrangement, and exposing the proteins to a solution of silicic acid, scientists at the Air Force lab in Ohio have created rows of regularly spaced silica beads. Such an arrangement could form the basis of a miniature lens.

Nils Kröger, a diatom biologist at the University of Regensberg in Germany, was the first to identify the silica-forming proteins in diatoms. The molecules of this class, which he calls silaffins, are unusual among proteins in that many of them have long side chains of organic molecules known as polyamines. The proteins are also decked out with an assortment of other molecules, including sugars and phosphates.

When Kröger and his colleagues added silaffins to a test tube containing silicic acid, tiny silica spheres formed in a matter of minutes. In contrast, a solution of silicic acid without any proteins “can take hours or even days to form hard silica,” says Kröger.

The researchers also found that combining two different silaffins from the same diatom species can yield surprising results. One of the proteins, silaffin-1, forms spheres. A second protein, silaffin-2, doesn’t by itself promote silica formation. But when the German team mixed the two silaffins in a solution of silicic acid, porous blocks of silica emerged.

“It’s something I didn’t expect to find at all, and we don’t completely understand how it works,” says Kröger.

He suspects that in the diatom, different silaffins combine to form larger molecular assemblies and that interactions between the proteins and their polyamine chains hasten the silica-formation process. Kröger has found that removing the polyamine side chains from a silaffin prevents the formation of silica. Moreover, proteins with chains of varying lengths tend to create a different array of silica structures.

Since his initial discovery of silaffins several years ago, Kröger has identified three more silica-forming proteins in T. pseudonana—the diatom whose genome was recently sequenced. Each protein does something different: One produces spheres, one makes porous shapes, and the third forms platelike structures.

Moving beyond these simple shapes will require a greater understanding of the diatom’s molecular machinery. What’s more, dozens or even hundreds of proteins may govern the shell-formation process. Mapping the myriad interactions among all the components could be a daunting task.

“It will be impossible to reproduce this process in a test tube because it’s such a complicated cellular process,” says Hildebrand.

Aizenberg adds, “The question is, ‘Will we be able to bridge the gap between what goes on in nature and what we can do in the lab?'”

She recently began investigating the silica-producing properties of Euplectella aspergillum, a deep-sea sponge that produces an intricate, cagelike glass structure (SN: 9/20/03, p. 190: Available to subscribers at Channeling light in the deep sea). Remarkably, she says, the material in this structure has optical properties that are very similar to those of telecommunication fibers.

Aizenberg looks to these organisms not only for inspiration on how to improve today’s materials and devices but also for clues as to how to make processing methods less energy-intensive and more environmentally sound. Today, commercial optical fibers are drawn inside a furnace at 2,000°C. In contrast, sponges synthesize sophisticated optical materials in a low-temperature marine environment.

Fabrication of silicon chips and other electronic devices currently requires harsh chemicals and generates much waste. “Diatoms and sponges know how to produce materials under ambient conditions without these harsh chemicals,” says Aizenberg. “And yet the end result is the same.”

It’s too early to say whether isolating silica-producing proteins to make minuscule new widgets in the lab will prove more successful than engineering microorganisms to do the job. Materials scientists are only beginning to uncover the secrets of this aquatic community of glass-sculpture artists produced over millions of years of evolution.

---