快猫短视频

IMAGINE you discover that lightly tapping your fingers on the dinner table can make the plates and glasses leap several metres into the air. On an atomic scale, that鈥檚 what Yasunobu Inoue can do in his labs at Nagaoka University of Technology in Japan. By bouncing sound waves through a catalyst made from a thin film of palladium, he can make a chemical reaction run two thousand times faster than it does without the sound.

This is a stunning trick, especially when you remember how important catalysts are in all kinds of chemical processes. Fellow chemists greeted Inoue鈥檚 findings with scepticism but doubts have given way to fascination as Inoue and his colleagues in Japan, and our own group in Cambridge, have started to investigate exactly what is going on to produce this spectacular speed-up.

The sound waves that Inoue uses are high-frequency vibrations that travel through the atoms at the surface of a solid material. Known as surface acoustic waves, they are quite different from the waves that carry the sounds we hear. While ordinary sound waves propagate in three dimensions, these waves are confined to the surface, like the ripples that spread out across the surface of a pond.

Inoue achieved his amazing results with a reaction between ethanol and oxygen, which produces acetaldehyde. The reaction hardly takes place at all without the help of a catalyst such as palladium, and chemists have a pretty good idea of how these precious-metal catalysts work. We know, for example, that the action takes place on the surface of the metal, and that the molecules become reactive when they are stuck or 鈥渁dsorbed鈥� onto the catalyst. But none of this explains how acoustic waves help things along.

When we started our investigations we already knew that bouncing acoustic waves along a surface can heat it. And it鈥檚 common knowledge among chemists that raising the temperature usually makes a reaction run faster. So is it merely the heating effect on the surface of the catalyst that speeds up Inoue鈥檚 reaction?

Another possibility arises from the relatively high pressures Inoue used-around a twenty-fifth of atmospheric pressure. At this sort of pressure, the gaseous reactants behave more like a fluid and don鈥檛 mix as readily. So we also needed to explore-or eliminate-the possibility that the acoustic waves are simply helping to mix the reactants in the crucial region close to the surface of the catalyst. Or maybe it really is a surface effect, and the explanation lay somewhere else we hadn鈥檛 even thought of.

To tackle these questions we needed an experimental setup that would allow us to monitor the state of the palladium catalyst, and accurately measure its temperature during the reaction. To minimise the effect of interference from impurities, we used an ultra-clean catalyst and high-purity reactants. And to eliminate the possibility that the acoustic waves were simply mixing the gases, we arranged to run our reaction at very low pressures, where the efficiency of mixing is not an issue.

Instead of palladium, we used another precious metal, platinum, as our catalyst, which we prepared in the form of a single crystal on a piezoelectric substrate (see 鈥淢aking waves鈥�). Having a single crystal ensured that there were no imperfections on the surface of the catalyst where anomalous reactions might take place. And using pressures as low as a ten-billionth of an atmosphere-a vacuum approaching that found hundreds of kilometres out into space-ensured that the surfaces were atomically clean. Together these precautions gave us the best chance of revealing the underlying physical mechanism.

To match the ultra-clean conditions in our reaction chamber, we chose a well-known reaction to test the effect-the oxidation of carbon monoxide to carbon dioxide. The chemistry is simple. Molecules of CO and oxygen atoms react together on a platinum catalyst to form CO₂. It is also of enormous practical importance because it is one of the main reactions that occurs on the catalytic converters-platinum and rhodium catalysts -built into the exhaust systems of modern cars. So chemists have given the reaction plenty of attention and have a firm understanding of the mechanism involved.

When we ran our experiments, we found that switching on the acoustic waves boosted the production of CO₂ sixfold. Because we ran the reaction at a pressure of only fifty-thousandth of an atmosphere, we were confident that improved mixing was not the cause in our experiments. So was surface heating responsible? To find out, we had to be able to monitor the surface temperature of the catalyst. This was no easy task. The small heat capacity of our very thin platinum catalyst meant that conventional thermocouples were ruled out. Instead, we used the acoustic waves themselves to do the job.

At one particular frequency-the resonance frequency-the wavelength of the acoustic waves exactly matches the length of the piezoelectric slab, just as a particular note resonates inside an organ pipe. At resonance, the acoustic wave forms a standing wave: instead of moving across the surface of the catalyst, the waveform appears static. As the piezoelectric material warms up, the speed of the surface acoustic wave passing through it changes. This causes the resonant frequency of the wave to fall by 2 kilohertz for every 1 掳C rise in temperature. To pick out this shift and to measure the temperature of the catalyst, we designed a system to record the frequency spectrum of the acoustic waves during our experiments.

When we made our measurements, we finally proved that the enhancement is not simply caused by heating. So if it鈥檚 not temperature and it鈥檚 not improved mixing either, what is it about surface acoustic waves that make catalysts so much more effective?

In most chemical reactions, the product molecules are more stable than the molecules of the reactants-in other words, they contain less energy. It is this energy difference that drives the reaction. The problem is that as they react, the molecules have to pass through an unstable, high-energy intermediate state: you can look at it as a sort of energy barrier that the reactant molecules can鈥檛 jump unless they are given some extra energy. Catalysts work by stabilising the chemical intermediates in the reaction, thus lowering the energy barrier to a more easily jumpable height.

Chemists already use high-frequency sound waves-ultrasound-to speed up the reactions of liquids on metal catalysts. What happens here is well understood: beaming powerful ultrasound with frequencies above 20 kilohertz through a liquid triggers violent events called acoustic cavitation-the formation, growth and implosive collapse of bubbles. The enormous temperatures and pressures created by this collapse provide a blast of energy to help speed the reactions on their way (see 鈥淏ubbles hotter than the Sun鈥�, 快猫短视频, 29 April 1995, p 36).

But this rather violent scenario is a far cry from what happens when acoustic waves ripple across the surface of a catalyst. Though the waves are high frequency-19.5 megahertz, corresponding to a wavelength of 0.2 millimetres-their amplitude is a tiny 10 nanometres. How can such puny waves, which cause a negligible displacement between neighbouring atoms, affect reactions occurring on top of the catalyst?

Initially, one of our favourite theories was based around a resonance mechanism. We reasoned that resonances on the surface of the catalyst might pump energy into the molecules involved in the reaction. This idea gained some support from measurements made by Andrej Boronin and his colleagues at the Boreskov Institute of Catalysis in Novosibirsk, Russia. But when we repeated the Russian experiments in our labs in Cambridge, we could find no relationship between the resonant frequency and reaction rate.

Inoue has different ideas. He believes that the reactions are speeded up by an electric field that acoustic waves generate in the piezoelectric substrate. He reasons that such fields might pass into the catalyst and influence its electronic structure in a way that increases the reactivity of the molecules on the catalyst鈥檚 surface. What we find hard to understand in this picture is how an electric field generated in the piezoelectric material could penetrate all the way through the film of catalyst to its upper surface. Being a metal, the catalyst should act as a highly effective screen, so we would expect the influence of the substrate鈥檚 electric field to fade by the time it has penetrated a few nanometres into the catalyst. Even if tiny pores perforate the catalyst down to the piezoelectric substrate, the surrounding metal would behave like a Faraday cage, stopping the electric field from reaching the catalyst鈥檚 top surface, where it could influence the reaction.

One clue we discovered is that the enhancement occurs at a specific stage of the reaction. The amount of CO₂ produced above a platinum catalyst is influenced strongly by the ratio of carbon monoxide to oxygen. When the concentration of oxygen is kept constant but the amount of carbon monoxide is slowly increased, the reaction runs through several distinct stages. At first, the surface is dominated by oxygen-each O₂ molecule splits into two oxygen atoms, which stick to the catalyst and react with adsorbed CO. Putting in more CO causes CO₂ production to rise steadily. This continues until the surface becomes covered with CO molecules, which group together into large islands and block further oxygen adsorption. At this point, the catalyst is described as 鈥淐O-poisoned鈥� and the reaction slows to a crawl. But if we switch on the surface acoustic waves at this point, the reaction leaps back to life.

Surfing islands

In our view, one likely contender follows from our calculations of the forces that the surface acoustic waves impose on the atoms on surface of the catalyst. Even with a layer of catalyst built up as a single crystal, this surface is not perfectly smooth when viewed on the atomic scale. Dotted around are tiny islands consisting of an extra layer of platinum atoms. While the top of these islands are covered with carbon monoxide molecules, the vertical edges are not, and these might provide sites on which oxygen could bind. Under the right conditions, accelerations caused by the acoustic waves could exert up to seven times the force of gravity, causing the islands to surf across the surface of the catalyst (see Diagram).

The islands鈥� violent movements could bash these oxygen atoms into regions of CO, boosting the rate of CO₂ formation. Throwing the islands around on a sea of surface acoustic waves might also cause them to break up or deform, increasing the area at their boundaries and speeding up reactions even more.

But there are other plausible explanations, too. The waves induce stresses that cause tiny cracks to appear in our single-crystal platinum catalysts. With a CO-poisoned surface, tearing the catalyst would expose fresh metal surfaces where oxygen could be adsorbed and react with CO to restart the reaction (see Diagram). However, Inoue has worked with polycrystalline catalysts that are thinner and more flexible than ours. They don鈥檛 crack, yet he still sees an enhancement.FIG-mg21345102.jpg

A million times faster

Compared to Inoue鈥檚 sensational results, our six-fold improvement may seem small. But leaving aside the fact that we are looking at a different reaction, the disparity might be due to the different pressure ranges we are working at. At the high pressures used by Inoue, the gases behave more like a liquid and under these conditions, improved mixing may play an important part.

Inoue is currently aiming for a million-fold enhancement in the reaction rate. He also believes that he will be able to use surface acoustic waves to control not only the speed of a reaction on the surface of a catalyst, but also its outcome. Where reactants can follow more than one chemical pathway, resulting in different products, it may be possible to select which path they take simply by tuning the frequency or power of the waves passing through the catalyst. In particular, Inoue hopes that adjusting the power of the acoustic wave through a palladium catalyst should change the relative proportions of ethylene and acetaldehyde produced from ethanol. He has also set his sights on an industrially important reaction-the oxidation of ethylene.

Meanwhile at Cambridge we are pursuing more basic research and aim to develop a physical model for the process. Considering that catalysts have been around in chemistry for more than a century, it鈥檚 comforting to discover that these specks of metal still have some surprises in store.