Fishing for new materials

If you want to develop high performance materials with nanometre-scale designs and environmentally friendly production processes, you should look to nature for inspiration. Michael Gross reports.

Hard materials have come a long ?ay from the stone tools of the Neolithicum to today’s high performance ceramics, diamond drills and boron nitride. Or so we would like to think. But look at the technologies we use to produce hard materials. Ceramics require exceedingly high temperatures, diamonds high pressures and strong glues use aggressive chemicals. If you had to wrap a tough coating round a sensitive little thing like a living cell today’s technology would get you nowhere. And yet, such processes happen on the megatonne scale out in the oceans. The largest group of the microscopic master-builders also produces the most aesthetically appealing microstructures: the diatom shells. 

Microscopic lace 
Diatoms are single-cell algae famous for the delicate lace-like appearance of their silica shells. The need for a hard shell that still allows nutrients to enter the cell has led to a wide range of beautifully designed sub-microscopic lacework, often highly symmetrical. Although the basic architecture of these shell structures is always that of a box with a lid, the shape of the box and the design details are highly specific for each of the many thousands of diatom species, suggesting that their blueprint must be written in their genes. This in turn suggests that gene products, most likely proteins, must be involved in constructing the shells and may even be present in the final structures. 

So all that researchers had to do to find out how the cell produces these marvellous structures was to break down the cell walls, pull out the proteins and study them. As Nils Kr?ger, Manfred Sumper, and their coworkers at the University of Regensburg in Southern Germany found out during their research into the mineralisation process of the diatom Cylindrotheca fusiformis, it wasn’t quite that easy. 

Proteins involved in biomineralisation are generally difficult to study by established methods, but the ones of the diatom cell walls proved to be particularly elusive and required some elaborate and lengthy detective work before they began to yield their secrets. 

The most widely used method of extracting protein from biological samples - boiling the cell walls in a detergent known as SDS (sodium dodecyl sulphate), which normally solubilises almost any protein - didn’t yield any result. Adding the chelating agent ethylene diamine tetra-acetic acid (EDTA) allowed the researchers to extract and characterise the first ever protein component of diatom cell walls, a group of glycoproteins (proteins with starch-like molecular branches) which they later called the frustulins (from frustulum, the Latin word for the diatom cell wall). This novel group of proteins was interesting in its own right, but they did not seem to make any difference to the chemical reactions involved in making the silica shells. 

Hoping that the most interesting proteins were still in the shell fragments, the researchers chose a more aggressive chemical that destroys even the chemical structure of silicate minerals: water-free hydrogen fluoride (HF). This treatment liberated a new group of proteins that they called silaffins, because they display a remarkable affinity towards silicon compounds. Each of the three main components found in the extracts can make a homogenous solution of silicic acid produce small solid spheres of silica, less than 1.?m in diameter, within minutes.

And yet the silica spheres resulting from the in vitro catalytic action of the purified silaffins did not resemble diatom shells at all. Parts of the jigsaw were still missing. Over several years of painstaking analysis, the researchers identified a group of polyamines involved in the process, and they figured out that the natural silaffins are indeed heavily edited and modified after their original synthesis. 

Last year, after replacing the HF extraction method with a gentler procedure involving aqueous ammonium fluoride, Kr?ger and coworkers finally presented what can count as bona fide native silaffins. Not only were the lysine residues of the peptides loaded with polyamines, including some previously unknown molecular species, but the latest research also reveals that all of the hydroxy-amino acids (seven serines and one modified hydroxylated lysine) in the peptides are phosphorylated (see below). 

Chemical structure of native silaffin 1A

The zwitterionic nature of the peptides suggested that they can self-assemble using electrostatic interactions, which was confirmed by NMR experiments. The formation of higher order structures (with the help of the phosphate groups) appears to be crucial for their function. While the previous non-native silaffins were active without being phosphorylated, they lost their activity when moved from phosphate into acetate buffer. 

With the crucial components now well characterised, researchers can now set out to try and understand just how the diatoms create those intriguing patterns. Manfred Sumper has proposed a hypothetical model of how phase separation could lead to many different yet regular patterns, but an experimental reconstruction of a diatom shell remains to be demonstrated. 

Biology to biomimetics 
Systematic checks have revealed that silaffin-type proteins are ubiquitous throughout the large family of diatoms. However, other organisms produce silica structures using different proteins that might be easier to handle and therefore more amenable to biomimetic applications. In all known cases, biomolecules are woven into or closely associated with the mineralised phase. This suggests that Nature invented composite materials several billion years before humans started manipulating materials; and also that it might be possible to replace those biomolecules with simpler molecular units working on the same principles. 

Dan Morse’s group at the University of California at Santa Barbara has discovered and studied in detail a group of proteins they called the silicateins. They are found in the middle of the glassy needles (spicules) that account for three-quarters of the dry weight of the marine sponge Tethya aurantia. The prototype of the family, silicatein alpha, turned out to be closely related to a family of proteolytic enzymes. Subtle sequence differences account for the fact that silicatein is inactive as a protease, and able to condense tetraethoxysilane molecules to polymeric silica instead. 

Using their insight into the function of silicateins, Morse and his coworkers proceeded to create a biomimetic catalyst of silica polymerisation. Their simple synthetic peptides, consisting of blocks of cysteine and lysine residues, are capable of hydrolysing the precursor molecule tetraethoxysilane and turning it into two different morphologies of silica, depending on whether or not the thiol groups of the peptide are oxidised. 

Similarly, Stephen Mann’s group at the University of Bristol also designs biomimetic catalysts for the ordered deposition of inorganic materials. Mann has shown that polyanions such as polyaspartate or polyacrylate can induce the formation of interesting nanostructures in different minerals including barium sulphate and calcium carbonate. Recently, the group has shown, in collaboration with Adriana Bigi of the University of Bologna, Italy, that the same polyanions can assemble octacalcium phosphate into hollow ?m-sized globules (spherulites) with a highly porous outer shell.

A new twist 
Clearly, a number of different composite materials can already be made by biomimetic synthesis. The next major challenge is to understand the rules to an extent that any material can be made to order and its structure controlled on a nanometre scale. One particularly challenging goal for biomimetic material synthesis is the induction of chirality. 

The conundrum is most obvious when you look at a snail’s shell, which is clearly chiral on a macroscopic level, but consists of a mineral that is achiral (typically aragonite, a form of calcium carbonate). Again, biomolecules must be directing the assembly on a nanometre level in ways that result in the macroscopic chirality. 

Two years ago, Christine Orme and coworkers at the Lawrence Livermore National Laboratory in Livermore, California, demonstrated in a simple model system how a chiral amino acid can convey chirality to growing calcite crystals, by binding to those edges that offer the best stereochemical and energetic conditions. Several laboratories have also created synthetic chiral materials using biological molecules as templates. For example, Stephen Mann’s group used a chiral lipid to create a silica structure with helical chirality. 

New materials 
To make the best possible use of the rich and varied examples of material processing in nature, researchers have to cast their nets widely and keep their eyes open for unusual phenomena. After all, there are millions of species out there producing more than 60 different kinds of minerals. When even the most common ones can still teach us lessons in material engineering, how much useful knowledge may remain untapped in the less well known biominerals? 

Apart from the spicules, Dan Morse’s lab has also looked at red abalone shells and at specific proteins conferring adhesion in the nacre of mollusc shells. His colleagues in the chemistry department at Santa Barbara have described the first example of a copper mineral conferring toughness on a mineralised tissue. 

Joanna Aizenberg and her colleagues from the Bell Labs and the Weizmann Institute recently revealed a particularly intriguing example of a clever use of biomineralisation. In the light-sensitive brittlestar Ophiocoma wendtii, the outermost microstructures of its calcite skeleton are shaped like microlenses and arranged in an array reminiscent of an insect eye. This feature is notably absent in the related, but light-indifferent species O. pumila. This finding made the cover of Nature, and it is typical of the kind of surprises that the field keeps throwing up. 

A few years earlier, another mineralised tissue was assigned an optical function. In the mid-1990s, several laboratories made observations that strongly suggest that sponge spicules serve as light guides. The silica spicules of the Antarctic sponge Rosella racovitzae end in a cross-shaped structure that may serve as a lens, collecting the light and directing it into the fibre. 

In 1996, researchers showed that these fibres indeed conduct light with satisfactory efficiency, even though they would not be able to compete with commercial glass fibres over the km distances required in optical information technology. Down in the ocean, however, where light is sparse and the fibres only a few cm long, this performance is likely to play a biological role. Some sponge species even have green algae as symbiotic guests inside their prickly bodies. 

Even though the intriguing materials discovered in nature are not necessarily ready for industrial applications, it is clear that there are many lessons we can learn from them, and that the field of biominerals still holds many surprises for us. 

Engineering the future 
While biomimetic material synthesis is still very much in an exploratory stage, one day we may be able to produce hard ceramic materials at low temperatures, without aggressive chemicals, and with exquisite control over the structures on a nm scale. 

If engineers could match nature’s performance in technical applications, they could protect soft and sensitive structures with a hard and chemically resistant shell. Proteins, for instance, can do thousands of interesting things and could become tremendously useful in future technologies if only they weren’t at risk of being eaten by bacteria or floating away from where they are needed. If one could build a diatom-like porous shell around a functional protein in a specific location, this could be harnessed for sensory, optical or electronic devices that could serve as a functional building block for nanotechnology. 

Furthermore, very small patterns such as those needed for silicon chips and similar devices, can be more easily created using molecules (soft lithography) than by conventional methods, as George Whitesides and his group at Harvard have demonstrated. If scientists were able to mimic biomineralisation, they could use the soft molecular structures readily generated by a ’rubber-stamp’ method developed at Harvard as a mould for hard ceramic structures that are impossible to shape directly using current high temperature technology. 

Molecular devices coated with biomimetic enamel might not only fit a shirt button, but they would also remain intact when you step on them or put them in the washing machine. All this we can learn from studying creatures like the humble diatoms. 

Source: Chemistry in Britain

Acknowledgements

Michael Gross is a science writer in residence at the school of crystallography, Birkbeck College, University of London. His latest book, Light and life, is available from Oxford University Press.