Science: THE SECRET LIVES OF MOLECULES

Professor Zewail is a hot favourite to win the Nobel Prize, which will be announced on Tuesday. His achievement? To unveil what lies at the heart of chemistry - exactly what happens when reactions occur. Hugh Aldersey-Williams reports

THE VERY ESSENCE of science - what precisely takes place when chemical reactions occur - is finally being revealed by a new laboratory procedure that is capable of measuring the minutiae of atomic and sub-atomic changes. The timescale of molecular shifts requires a precision such that the second must be divided a thousand million million times.

Microelectronic devices - calculators, say, or digital watches - rely upon electronic events that happen in microseconds over distances of micrometres. Nanotechnology is a new science that promises to allow molecules to be "handled" individually during chemical construction on the nanometre scale, which is a thousand times smaller than the micrometre.

Now come closer still, and enter the world of femtosecond science. "Femto" signifies a factor of one million closer in focus than "nano". At this scale, we can capture the act of molecular bonds forming and breaking, the fundamentals of chemistry.

A chemical reaction may go rapidly, like an explosion, or slowly, like rusting. The speed at which a reaction takes place has to do with the rate at which the reacting molecules find each other. But when they do find one another, things happen very fast. As for exactly what happens, that has always been something of a mystery. It's like Blind Date. The television programme introduces us to the two individuals. In a later programme, we learn whether, not to put too fine a point on it, a "reaction" took place or not. But we are not able to see how the reactants actually behaved in real time.

It is not just for reasons of voyeurism that chemists want to know the full story. In reactions where several bonds are made or broken, knowing whether this happens in a concerted manner (simultaneous), or with one stage triggering another, or completely independently, is important in planning the "synthetic routes" used to make complex chemicals such as pharmaceuticals.

During the second half of this century, chemists have been able to look closely at chemical reactions, thanks mainly to the development of the laser. Lasers allow short pulses of light to be fired at chemical samples; these can both stimulate reaction and allow changes to be monitored by detecting alterations in the laser light. Successive improvements in laser techniques have shortened these light pulses from microseconds to nanoseconds to picoseconds (a thousand times shorter than a nanosecond), and now to femtoseconds (another factor of one thousand shorter).

The shortest pulses so far obtained are six femtoseconds long. At this tiny "exposure", molecular vibrations and rotations are essentially frozen. The main action is the motion of the atomic nuclei of reacting species towards and then away from collision. Atoms move at around one kilometre per second during reactions. A chemical bond is just over a tenth of a nanometre in length. So it takes an atom one ten thousandth of a nanosecond, or 100 femtoseconds, to run the length of a bond. Chemists at last have a kind of camera that allows "snapshots" to be taken of molecules in flagrante, as it were.

The leading champion of femtosecond chemistry is Professor Ahmed Zewail. Born in Alexandria, Egypt, in 1946, Zewail started his career at the California Institute of Technology in Pasadena as an assistant professor. He now occupies the chair there once held by Linus Pauling, the double Nobel Laureate (Chemistry and Peace), who laid the foundations of our modern understanding of chemical bonding.

Nobel Prizes have been awarded for subsequent progress in using quantum theory to explain why reactions which look possible on paper can be made to work, while others cannot. Professor Zewail's research continues in this vein, looking closer than ever before at chemical reactions as they take place. He too is a strong contender for the Nobel Prize, which will be announced on Tuesday.

Like many strong candidates, he is not unaware that he is in with a chance, and has a considerable bandwagon running for him. Some scientists are beginning to balk at the "campaigning" that increasingly surrounds the prize. Two Nobel Laureates who know and admire Professor Zewail's work refused to comment on it for this article. "There has been a hell of a lot of hype," said one. Nevertheless, Professor Zewail's chances of winning either this week or next year are extremely good.

The Caltech laboratory that Professor Zewail calls "Femtoland" employs a score of graduate and postdoctoral students, and boasts five laser and molecular-beam facilities, each of which cost hundreds of thousands of dollars. This apparatus looks like something out of Jules Verne: large stainless-steel drums fitted with various pipes and portholes. The windows admit laser light, the pipes feed in beams of atoms and molecules. They meet in the near-vacuum at the heart of the drum.

A battery of laser pulses is emitted. The first femtosecond pulse agitates the molecule, let us say one of sodium iodide. The sodium and iodine atoms begin to separate. A second laser pulse, delayed in relation to the first by being bounced back and forth between mirrors for a few femtoseconds, "snaps" the action. Subsequent pulses produce further snapshots, building up a "film" of the reaction.

As early as 1889, the Swedish chemist Svante Arrhenius imagined that there was a "hypothetical body" or transition state complex in many chemical reactions - a semi-stable agglomeration of reacting chemicals that could theoretically split back into those reactants or proceed past the point of no return to yield new products. His concept is still useful today.

The timescales at which Professor Zewail is working approach the limit that is observable, according to the Heisenberg Uncertainty Principle, which states that it is impossible to know accurately both the position and the momentum of a small particle. "If you look at it terms of quantum mechanics," says Professor Zewail, "it's impossible for there to be an observable configuration of the transitional state because the Uncertainty Principle says you must have a whole distribution of structures. People thought we would never be able to localise nuclei on this timescale."

But Professor Zewail has found a way to cheat. The Uncertainty Principle becomes less uncertain about things the more massive they are. His trick is to use laser light to assemble molecules in a "cohe-rent" state so that they undergo reaction at exactly the same time. With up to 100 million molecules behaving as one under the direction of the laser, it has become possible to observe chemical action at a new level of detail.

In his pioneering experiment with sodium iodide in 1988, Professor Zewail's main tool was the spectroscope, which measures the distinctive signature, the spectra produced when atoms and molecules absorb and emit electromagnetic radiation. "The key concept there was that the atoms have different spectra when they are in close proximity from when they separate to infinite distances. So we can examine their passage through the transition state," he explains. He discovered that the two atoms separate and approach each other 12 times, always exactly 12, before the chemical bond between them is broken. "Transition state complexes have never been seen before in the history of chemistry."

More recent work in Femtoland has been possible with the help of new equipment. Mass spectrometers can now catch and "freeze" an intermediate species, whose normal lifetime might only be a few femtoseconds, and weigh it in order to determine its brief-lived chemical composition. This allows the Caltech scientists to explore the chemistry of molecules with up to 50 atoms at a time.

But with larger molecules, new factors become important. One of them is orientation: two molecules that are capable of reacting will only do so if they come together in the right way (as in our Blind Date analogy). Using polarised laser light, Professor Zewail can also detect a reactant molecule's orientation.

The ultimate step will be to obtain visual images. At this scale, smaller than a wavelength of light, these are not direct photographs. Instead, the patterns created by the diffraction of electrons fired in a short burst and recorded by a highly sensitive camera may give us our first pictures of a chemical reaction as it takes place. So far, the Caltech scientists have clocked chemical structures at intervals of half a picosecond, and are working towards femtosecond times.

Arrhenius's concept of the transition state has served chemists so well that some regard actually seeing chemical reactions as an irrelevance. For others, however, it is like stepping on to the moon for the first time. But it is not just nice to see verification of ideas long familiar in theory: in some reactions, there are a number of possible transition states. Knowing which one a particular reaction passes through, and how long that state lasts, will help us better to understand natural reactions and plan synthetic ones.

Synthetic chemists use well-trodden pathways (and know some good short- cuts) on their way to making complex substances such as drugs or plastics. But it can still take dozens of individual reactions performed in sequence to get to a desired compound. Techniques such as those in use in Femtoland promise "quantum control". If, for example, it is possible to observe transition state complexes with femtosecond bursts of laser light, it may also be possible to use the light energy to nudge the reaction in a desired direction.

And a transition state that presently leads to useless products could be modified to give more useful results, perhaps a product only otherwise obtainable via a long sequence of reactions. Likewise, the polarised light used to reveal the orientation of reacting molecules can be used to select correctly oriented molecules from a randomly oriented burst and increase the reaction's efficiency.

In the case of breaking bonds in larger molecules, it was formerly thought that the molecule would absorb energy (ie from light or other electromagnetic radiation) by distributing it around the molecule as vibrations of its bonds until a particular bond broke. But Professor Zewail's techniques have enabled scientists to show that this "statistical" flow of energy - like champagne pouring into all the glasses in a pyramid - does not always hold true; sometimes a more efficient flow of energy goes straight to the bond that breaks, as if the champagne were to find its way directly to a particular glass and fill it to overflowing.

Rhodopsin, a red-sensitive pigment found in the retina of the eye, is a good example of this. Drs Richard Mathies and Charles Shank at the University of California at Berkeley found that 80 per cent of the energy absorbed by the molecule is quickly channelled into twisting one bond to produce the chemical transformation that leads to a nerve signal. "They demonstrated the effect wonderfully. If the rhodopsin molecule didn't behave this way we wouldn't be able to see as efficiently," says Professor Zewail.

All this new understanding also promises greater control over chemical reactions. In an article in the American journal Science, Professor Zewail argued that examination of how molecules "waste" much of the energy loaded into them by heating during industrial chemical processes might suggest ways to channel that energy into breaking only those bonds chemists want to break. By intervening with a well-timed femtosecond-long burst of laser light before the molecule has time to spread the energy around, it may be possible to direct a reaction more economically to a desired end-point.

Femtosecond chemistry may also shed new light on the transmission of the genetic code. The structure of DNA uncovered in 1953 by James Watson and Francis Crick (who beat Linus Pauling to it) described the "specific pairing" of particular molecules on the respective coils of the double helix to form units called base pairs. Each of these embraces involves two weak, long-distance links in which hydrogen atoms tightly bonded to one molecule are also weakly attached to the molecule on the other coil. Repeated along the helix, these pairs provide the "zipper" mechanism by which DNA divides and replicates.

Watson and Crick observed that if these pairs were to exchange their hydrogens, it might disrupt the genetic code. A kind of mis-print would be recorded and reproduced every time the DNA replicates, just like a photocopied spelling mistake.

Researchers have since tried to establish whether the exchange of the two hydrogens is concerted, like an exchange of hostages, or whether one moves before the other. The former would involve an electrically neutral transition state, the latter one with two electrical charges. In 1995, Zewail and his colleagues observed in a chemical model for the DNA base pairs an intermediate state which came into existence after one hydrogen had been exchanged but before the other was. The exchange process, in other words, was not concerted. The first hydrogen switched allegiance in just 200 femtoseconds; the second took 10 picoseconds, 50 times as long, to make its move.

Earlier this year, scientists at Pennsylvania State University managed to arrest the intermediate. "This was very interesting," says Professor Zewail. "Because it moves in a non-concerted fashion, it means you have an ion pair, a positive and a negative charge, being formed which live for a tiny period of time. This may be a signal for an enzyme to correct the process. The key thing now is to show it happening in DNA." Errors in our genes are corrected by enzymes. What is not yet known is whether these enzymes are given the signal to go to work by the brief existence of this local electric charge.

But scientists' growing ability to observe biological processes in real time at atomic levels of detail should ensure that it is not too many femtoseconds before we are able to find out by sitting down to watch DNA: The Movie.