" Or, as Zewail puts it, "Atoms and molecules have an enormously complex sociology, and for centuries chemists have been trying to understand why they sometimes like each other and sometimes hate each other. This love and hate is extremely importantÜit determines why substances can exist, and how they behave." And, like humans, the only way to find out how they behave is to watch them in action. So Zewail, Caltech's Linus Pauling Professor of Chemical Physics and professor of physics, makes movies of molecular births, weddings, divorces, and deaths with what the citation calls the world's fastest cameraÜone with a shutter speed measured in femtoseconds. A femtosecond is a millionth of a billionth of a secondÜ10-15 seconds, or 0.000000000000001 seconds; a femtosecond is to a second as a second is to 32 million years. The citation continues, "The contribution for which Zewail is to receive the Nobel Prize means that we have reached the end of the road: no chemical reactions take place faster than this. With femtosecond spectroscopy we can for the first time observe in ïslow motion' what happens as the reaction barrier is crossed."

This reaction barrier is generally pictured as a mountain separating two valleys. In one valley, or state of minimum energy, lie the reactants; the products lie in the other. The reactants have to have enough energy to hike up the mountain before they can ski down the other side. These landscapes are called potential-energy surfaces, but unlike the latitude and longitude coordinates one uses to navigate cross-country, the axes of a potential-energy surface are the distances between the atoms involved in the reaction. When only two atoms are involved, the potential-energy surface becomes a curved line on a piece of paper: a two-dimensional plot of energy versus bond length. When one bond breaks and a different bond forms, the surface is three-dimensional, like a relief map, and as additional atoms get involved, the surface can occupy still more dimensions. And complex reactions may have several intermediate product

Each summit (in two dimensions) or saddleback (in three or more dimensions) in the potential-energy surface is what chemists call a transition stateÜthat point when the molecule is betwixt and between, no longer reactants and not yet products, its bonds, like Richard III's physique, scarce half made up. The transition state is a razorback ridge, not a broad plateau, and mole-cules don't dally there. In fact, transition states are so fleeting that before Zewail's work they had never been observed directly, even though they had been postulated to exist since the 1930s. The best efforts to view them produced the spec-troscopic equivalent of a blurry daguerreotype of a busy street.

In order to shoot bonds in sharp focus, you need a shutter speed faster than the fastest atomic motion. Explains Zewail, "A femtosecond is shorter than the period of any nuclear vibration or rotation in the molecule, so we are able to freeze the system in time. The atoms in the molecules inside you are vibrating at about a kilometer per second, and they're so tiny that we measure their relative positions in a unit called the angstrom, which is 10-10 metersÜone ten-billionth of a meter. A typical chemical bond is a few angstroms long. If you combine these numbers, you see it takes about 100 femtoseconds for an atom to move an angstrom. So we have to be substantially faster than that to catch them in the act." And then there's the daunting task of getting all the molecules in step. When Eadweard Muybridge took his groundbreaking stop-motion photos of a galloping horse in 1887, one horse was ridden past a row of a dozen cameras. But each of Zewail's photographs contains millions of molecules, as if Muybridge had run a whole herd of horses in lock step past one camera. Even when they are undergoing the same reaction, molecules don't run in lock step, so how do you synchronize their gaits?

That's where matters stood when Zewail arrived at Caltech in 1976 as a brand-new, untenured professor. Born near Alexandria, Egypt, he had graduated from Alexandria University in 1967 with a degree in chemistry and first-class honors. He earned his PhD at the University of Pennsylvania in 1974, where he specialized in solid-state spectroscopy and nuclear magnetic resonance, and had continued in this vein during a two-year stint as a postdoc at UC Berkeley. Lasers were pretty hot stuff back then, so he "proposed for my assis-tant professorship to apply some of the concepts I had learned from my other fields into lasers, and use them to probe molecules at the fundamental level." The initial experiments, done with grad students Tom Orlowski (MS '76, PhD '79) and Dan Dawson (MS '78) and undergrad Kevin Jones (BS '77), proved that short laser pulses (and in those days, short laser pulses were measured in nanoseconds, or billionths of a secondÜan eternity in his lab today!) could excite simple or even complex molecules into so-called coherent states, and that such coherence could be detected during their spontaneous decay. In other words, even though each individual molecule in the sample was zipping around on its own course, banging into its fellows and tumbling like an X-wing fighter that's taken a hit, there were some spectroscopic phenomena in which a huge percentage of them acted together. If shot with a sufficiently brief laser pulse, they would shed the excess energy in a coordinated fashion. The chemistryfaculty was sufficiently impressed to give Zewail tenure in less than two yearsÜan ultrafast reaction in its own right.

"After doing that," Zewail recounts, "I was interested to see whether we could look, not at perturbations between molecules, which is what we had done with these experiments, but at what coherence goes on inside complex molecules. Why not try to isolate these molecules from the rest of the world by doing what's called molecular beams? So we built our first molecular beam in '79." A molecular beam is just what its name impliesÜa beam of molecules enclosed in a vacuum chamber that looks like a capped-off segment of stainless steel sewer pipe. You vaporize your sample and suck it up into a stream of an inert "carrier" gas such as argon, and squirt the gas through a pin-hole into the vacuum chamber at right angles to the laser's path. The carrier gas dilutes the sample so that, although the laser is exciting millions of molecules at a time, the molecules don't collide with one another. The gas also expands when it enters the vacuum, accelerating the molecules to supersonic speeds, and, paradoxically, cooling them. An isolated molecule has a fixed kinetic-energy budget, so the extra energy invested in speed has to come from the molecule's rotational- and vibrational-energy accounts. The energy overdraft protection plan kicks in automatically, leaving the fast-flying molecules flat broke and plunging their temperatures to within a few tens of degrees of absolute zero. This is a vital first step to establishing coherence, because the bankrupt molecules can afford to live in only the lowest-energy rotational and vibrational states, imposing a significant degree of order on them already.