麻豆传媒

Baltimore and California

PRINCE Charming finds Cinderella thanks to her glass slipper, for in his kingdom only Cinderella has feet dainty enough to fit it. But what if everyone鈥檚 feet were equally dainty? To lead the Prince to his true love, the slipper would need to be moulded to the exact shape of Cinderella鈥檚 foot, down to the last delicious kink and curve. It would be a tall order, even for a fairy godmother-but not for several teams of chemists, who are building unique chemical slippers to fit the tiniest feet of all: single atoms and molecules.

The researchers are making these miniature slippers out of polymers that can 鈥渞emember鈥� the shape of a molecule. By moulding the polymers around a target molecule and then washing it away, the scientists create a cavity of a size and shape that hopefully only the target will fit. The molecule is imprinted into the memory of the polymer. Once they have memorised a molecule, these 鈥渕olecular imprinted polymers鈥� bind to it better than any others. The materials only recognise and reacquaint themselves with the original molecule; no others can squeeze themselves into the cavity. This selectivity means molecular imprinted polymers can be used to make highly specific sensors and detectors, or to pluck a single chemical out of a messy mixture.

These molecular glass slippers can be made for almost any chemical. Drugs, hormones, small molecules such as carbon monoxide, and even individual atoms can all be imprinted and detected using these materials. The latest applications, presented at the American Chemical Society meeting in San Francisco earlier this year, include sensors for toxic metals such as lead, detectors to measure the level of chemicals in the body, including nitric oxide and glucose, and even devices to sniff out traces of nerve gas.

Making an imprinted polymer is just like taking a plaster cast of a molecule (see Diagram). The first step is to bind special monomers containing units such as amino and carboxylic acid groups to the target molecule. The structure of the target molecule determines which monomers are used. The target molecule, with its binding monomer complex, is then mixed with molecules of a crosslinking agent such as divinylbenzene that will link together with the monomers to form the polymer.

Polymerisation is then started. The cross-linking agents react with each other and the binding monomers to form a stiff, three-dimensional structure in which the binding chemicals forming the complex with the target are held firmly in place. This ensures the complex keeps its shape after the target molecule is removed. If it didn鈥檛, the final product would have a blurred memory of the target and might bind to the wrong molecules.

Mind the gaps

Just as glass slippers have an ankle hole to let a foot slide in and out, chemical slippers must have gaps or pores so that a target molecule can reach the custom-made cavity. These gaps are created by molecules of solvent that get caught up in the growing polymer. The mixture of crosslinking agents, target-monomer complexes and solvent must be fine-tuned to create a rigid but accessible structure. Much of the pioneering work to create accurate imprints while keeping cavities accessible was carried out by G眉nter Wulff of the Heinrich Heine University in D眉sseldorf, Germany.

Once polymerisation is complete, the resulting chunk of imprinted polymer is ground up into a powder. This increases its surface area, and gives access to cavities that would otherwise be buried in the middle of the chunk. The powder is then washed to remove the solvent molecules from the pores, and to remove the target molecules from the imprinted cavities.

By choosing the right set of binding monomers, chemists can imprint polymers with many different molecules. And if the cavities are designed to give a measurable signal when the target molecules dock, then imprinted polymers can be used as sensors to detect very low levels of a particular chemical.

At the University of Maryland we have been using this approach to develop a sensor for the detection of the nerve gases sarin and soman. Together with student Amanda Jenkins, and O. Manuel Uy of the Johns Hopkins University Applied Physics Laboratory in Laurel, Maryland, we have made an imprinted polymer system that can detect as little as 0.5 parts per trillion of these agents-enough to track down the inevitable leaks from anyone making these chemicals.

Rather than imprint our polymer using sarin itself, we use the harmless chemical that is produced when the nerve gas reacts with water. The only chemical difference between the two compounds is lost when they bind in the cavity, and so to the imprinted polymer, they appear to be the same. The cavities we make in the imprinted polymer are complexes containing a europium ion (Eu³⁺), which act both as binding sites (see Diagram) and as the means for detecting the molecules.

Europium ions absorb light and luminesce at characteristic wavelengths that change depending on the positions of the groups bound to the ions. We have designed our europium-based cavity so that it changes shape when a nerve gas molecule attaches itself to the europium ion. The resulting change in wavelength of its luminescent light is easy to detect.

We are currently testing our device with a variety of molecules that are similar to nerve gases such as sarin. Some herbicides and pesticides have similar structures to sarin, and we don鈥檛 want them to produce a false positive that will set off our sensor. So far things are looking good: the system reacts only to sarin and the harmless product of its reaction with water.

But there is another problem. The equipment needed to detect luminescence is rather bulky for the applications we envisage-we would like to be able to make tiny sensors for sarin that can be implanted in sites where the making of nerve gas is suspected or where terrorist attack is a threat. These devices could then be monitored remotely. For this, the sensors would ideally be about the size of a fingernail. We hope to be able to shrink our system by using devices called surface acoustic wave sensors. These sensors pass acoustic waves along a thin wafer of material. Any change in the mass of the wafer causes a change in the velocity of the waves, which can be measured very accurately. Covering the wafer with an imprinted polymer means it will only bind to target molecules, so the only way the mass can change is if the target molecule is present. Using this method, it should be possible to produce a pocket-size version of the sarin sensor, and sensors for other gases and chemicals too.

Sensing chemicals in the environment is not the only role for molecular imprinted polymers. Several teams are working on devices that might detect chemicals inside the human body. Frances Arnold and colleagues at the California Institute of Technology, in Pasadena, have already proposed a polymer for sensing glucose levels (Technology, 12 April, p 22). Over the past four years, Klaus Mosbach of the University of Lund, Sweden, another pioneer in imprinted polymers, has been using the technique to sense immunosuppresant drugs such as cyclosporin, and hormones such as cortisol. In the future, imprinted polymers could be coated onto the tips of fibre-optic needles, and be used as sensors to monitor the levels of chemicals in the bloodstream in real time.

Another role for molecular imprinted polymers is separating one compound from a mixture of several others with similar properties or a similar structure. This is something chemists frequently have to do, using a technique known as chromatography. In conventional chromatography, a solution of the chemicals to be separated is passed through a stationary material to which the chemicals bind with varying degrees of strength. The binding is reversible, so the chemicals move back and forth between the solution and the material. Those that bind most strongly spend more time attached to the material and so take longest to pass through; those that are bound more weakly pass through the material relatively quickly.

Sweetness and light

Many chromatography materials cannot separate chemicals that are closely similar to each other because the chemicals bind equally strongly and so take the same length of time to flow through. This is almost always the case with the mirror-image pairs of molecules known as 鈥渆nantiomers鈥�, in which each member of the pair differs from its partner only as a left hand differs from a right hand. However, imprint a polymer with right-handed isomers, and left-handed isomers won鈥檛 fit into the cavities. So left-handed molecules pass through the imprinted polymer quickly, while right-handed molecules bind in cavities and pass through slowly.

In the past ten years, Wulff and Mosbach have produced a variety of materials for separating enantiomers. Wulff鈥檚 group has made polymers that separate sugar isomers. This could prove useful for making virtually calorie-free artificial sweeteners. Both mirror-image forms taste the same, but our bodies only metabolise one of these enantiomers. So if we swallow the mirror-image form of the normal sugar we get the sweetness without the calories. Mosbach has made polymers that separate amino acids from their enantiomers and other closely related derivatives. These could also be used to separate optical isomers of drugs, so that we take only the required form and not the useless or possibly harmful mirror-image form.

Even the smallest molecules can be separated by molecular imprinting. Andrew Borovik of the University of Kansas at Lawrence is developing polymers that selectively bind to nitric oxide (NO) and other small molecules. NO has a number of roles in the human body: as a neurotransmitter, helping to control blood pressure, helping cells to communicate with each other, and assisting the immune system. A sensor that could measure NO levels could help diagnose problems caused by a lack of the molecule. What鈥檚 more, NO is one of the pollutants produced by coal-burning power stations, turning into NO₂ and NO₃ in the atmosphere, so a sensor could help monitor emissions.

Together with graduate students John Krebs and Cora MacBeth, Borovik is also designing cavities for other small molecules such as carbon monoxide (CO) and oxygen (O₂) by immobilising metal complexes of manganese, iron and copper within a polymer. The team moulds the cavities around the metal complexes so that only one particular small molecule with a certain size, shape and chemistry can bind to the metal ions.

When NO binds to the metal sites in one of Borovik鈥檚 imprinted polymers, the colour changes immediately from orange to green, making it easy to detect nitric oxide. Devices based on Borovik鈥檚 polymer would be simple and cheap, and would also be reusable because the binding of NO is reversible. Polymers that selectively bind to carbon monoxide could be used to detect potentially lethal traces of CO being given off by faulty gas appliances, and oxygen sensors could monitor the health of a river, for example.

Precious metals

Imprinted polymers could also be put to work to soak up ions of toxic heavy metals such as mercury, cadmium and lead. At the Lawrence Berkeley National Laboratory (LBNL) we are investigating this technique as a way of cleaning up waste waters. The imprinting process to make metal ion-selective polymers is different from imprinting chemical molecules. The metal ion that has to be recovered is part of one of the compounds used to form the imprinted polymer. For example, to make a mercury-imprinted polymer Christine Lo, a postdoctoral fellow at LBNL, started with a sandwich-like complex with a mercury (Hg²⁺) ion at its centre (see Diagram). Fixing the complex in place in a polymer framework and then washing out the mercury ion creates a cavity that can be filled only by another mercury ion. Our mercury-imprinted polymer is very selective: it ignores cadmium, lead, silver, copper and iron ions. This selectivity occurs because mercury ions bind to the spatially regulated polymer site, and so form a stronger complex than the other ions when they enter the cavity. Polymers tailored for specific metal ions could not only remove the toxic metal contaminants from polluted groundwater, but might also recover precious metal ions for reuse.

In the Maryland group, graduate students Xiangfei Zeng and Anton Bzhelyansky are working on an imprinted polymer for lead. Our goal there is to develop systems to clean up drinking water and to treat lead poisoning. Like the LBNL researchers, we start with a compound containing the ion-lead, in this case-that we are trying to target. Our polymer is made from lead vinyl benzoate, styrene and divinylbenzene.

Although the polymer we have made so far is selective for lead, it doesn鈥檛 hold enough ions to be useful for soaking up pollution. So instead we are using it as a highly sensitive lead detection system. By combining our lead-imprinted polymer with a way of using light to detect lead ions, we have built a fibre-optic sensor that can detect lead ions at levels of just a few parts per billion. Such a sensor could be put to work measuring lead in the blood-especially of children, whose health may be badly affected by a concentration of the metal too small to be detected by other methods.

Our detector comprises a fibre-optic cable with a small blob of lead-imprinted polymer on the end. Light of a carefully chosen wavelength directed down the cable is absorbed only when lead ions are present. The lead-containing complex then phosphoresces-at a different wavelength-and this light travels back up the optical fibre to a detector. The intensity of the light gives a measure of the amount of lead bound to the polymer. As a lead detector, our system is selective in three ways. First, the polymer should only absorb lead ions. Secondly, if any other metal ion does happen to find its way into the cavity, the complex between this ion and the polymer shouldn鈥檛 absorb the light of the wavelength the lead complex absorbs. Finally, even if it does absorb light, it will not phosphoresce at the same wavelength as a lead-polymer complex.

As the memories of molecular imprinted polymers improve, it will be possible to detect, measure and separate an ever wider range of molecules and ions. Chemists playing the role of Prince Charming will find their molecular Cinderella that much more easily, thanks to these amazing chemical slippers.