AS a child in wartime Prague, Josef Michl made models of animals and toy figures out of cocktail sticks and chestnuts picked up in the local parks. As long as the chestnuts were fresh and soft, he says, sticking the points in was easy. Now, fifty years on and settled at the University of Colorado in Boulder, he鈥檚 playing the same game 鈥� this time using rods and connectors the size of molecules.
His 鈥渕olecular construction kit鈥� is part of a new trend in chemistry 鈥� while nano technologists are struggling to craft ever smaller gadgets out of bulk materials, chemists are beginning to find ways to build such gadgets from scratch starting from basic molecular components (鈥淢olecules that build themselves鈥�, New 杏吧原创, 19 February 1994). And Michl has some ambitious ideas about what his kit could be used for. At the end of April at a conference in Paris organised by Nature, he presented one such vision 鈥� a microscopic wind farm, made up of molecular turbines driven by a tiny stream of gas.
This audacious scheme, though backed up by detailed computer simulations, is still some way away. At the moment, Michl is busy building up his sets of molecular rods and connectors, which he hopes will one day provide an instant, off-the-shelf scaffolding system for chemists. He likes to compare his approach to an American toy construction kit known by its trade name Tinkertoy 鈥� comprising narrow wooden rods of different lengths, and round, spool-like connectors into which rods can be inserted. The geometry is fixed and the constructions rigid and stable.
Advertisement
Michl stumbled across the rods for his molecular version of the Tinkertoy game almost by accident. Until the mid-1980s he had been mainly interested in studying the spectroscopy and structure of organic compounds. The Tinkertoy inspiration came from his studies on the spectroscopy of highly strained carbon compounds. These are molecules in which the four bonds of the carbon atoms are forced out of their normal tetrahedral arrangement 鈥� the shape of a triangular pyramid 鈥� into uncomfortable geometries. An early example was the hydrocarbon cubane, C8H8, first made by chemist Philip Eaton at the University of Chicago back in 1964. Eaton built a cube using eight CH groups, each sitting at one of the corners, by forcing the bonds between them into 90掳 angles 鈥� far from their normal orientation of 109.5掳.
In 1986, Michl鈥檚 student, Piotr Kaszynski, was trying to prepare a product of another strained molecule, C5H6 known as propellane because of its resemblance to a three-bladed propeller. But Kaszynski鈥檚 analysis showed that, instead of making the compound he was expecting, he had accidentally joined the propellane units together to make rods. Michl quickly realised that the new molecule would be straight and inflexible, and it dawned on him that with rods like these, he had the basis of a molecular Tinkertoy set. These rod molecules came to be called staffanes 鈥渂ecause their shape resembles the staffs of medieval dignitaries鈥�, says Michl.
Further rods have followed. Eaton has made some out of cubane molecules, bicyclo-octanes have been strung together by Howard Zimmerman at the University of Wisconsin, and Michl himself has devised rods using carbon and boron. Michl has even assembled rods from a mixture of units; because the units have slightly different lengths, mixing them in a single rod gives Michl fine control over the total rod length. So far, he has made rods whose lengths vary from 5 to 25 angstroms (10鈭�10 metres). Such is their diversity, that he can select one to within an angstrom 鈥� half the length of a typical bond 鈥� of any desired size.
The rods are not rigid in the sense that toy rods are but stiff and flexible, 鈥渋f anything, more like the rubber sticks that policeman carry in some countries鈥�, explains Michl. In each case, the rod strength comes from the geometry of its subunits. These are all tight, cage-like groups, with several bonds running through them (see Diagram). These bonds reinforce each other like struts in a girder. With their geometry fixed, the bonds between the units have to lie along the rod鈥檚 axis.
There are other, easier ways of making reasonably rigid rod-like molecules. Conjugated polymers, for example, have a certain amount of rigidity because they have a backbone of overlapping double or triple bonds. Where Michl鈥檚 intricate compounds score over these more conventional materials is that his molecules are very inert. They do not absorb visible or ultraviolet light so they do not decompose when subjected to light, they are stable at temperatures of at least 200掳C and often much higher, and they do not react with the oxygen in air even at high temperatures. This is why he envisages that his construction kit could be used to make an inert scaffolding on which could be hung more reactive molecules with useful electronic or mechanical properties.
Making connections
Michl is also working on connectors to join the rods together. Metal atoms would be the simplest solution. Chemists have extensive experience of binding organic molecules to metal atoms, in fact, the rather peculiar bonds involved are essential to the functioning of molecules such as haemoglobin. One quality makes these bonds particularly useful to Michl: although they are chemically strong, they can be fairly easily undone. This could help molecular engineers with a problem larger-scale engineers have long been familiar with 鈥� the need to correct errors in constructions.
When large numbers of molecules are assembled chemically without hands-on control, some will inevitably end up in the wrong place or imperfectly fixed. Michl and his team want 鈥渢her modynamic control鈥�, the ability to break and remake the bonds in their growing structures, until everything is in the right order.
Different metals give different binding geometries 鈥� square, octahedral and so on 鈥� so the trick is to choose the right ones. Michl also needs to build the right chemical groups onto the ends of his rods to make them stick to the connectors. For example, he has managed to build tiny crosses by attaching carboxylate groups to the staffanes and binding them to a connector made of two rhodium atoms. He prevented the whole solution of rods and connectors from turning into an enormous molecular tangle by 鈥減rotecting鈥� the carboxylate group at one end of the rods, converting it temporarily into an ester group, which does not bind to rhodium.
This need to maintain careful control over the assembly process is an important feature of the Tinkertoy approach. Under the right conditions, leaving everything to react together can produce a structured crystal. But some devices, tunable lasers for example, need material made up of different components with fixed positions, which requires a more subtle approach. Michl鈥檚 main focus at present is in building large, three-dimensional structures, layer by layer, starting with a foundation layer attached to a surface. This way, he hopes to make 鈥渄esigner鈥� solids, which could have different components in each layer.
To do this, Michl is using large, prefabricated 鈥渟tar connectors鈥� in which the rods are built into the molecules as covalently bonded arms of a star. One of his successes so far has been a flat, three-arm star connector by coupling three large carboranes to a central benzene molecule. The advantage here is that a 鈥減edestal鈥� can be attached vertically to the benzene ring via a ruthenium 鈥渟andwich鈥� bond. The pedestal has sticky tentacles that should anchor it strongly to a metal surface, and the arms of the connectors can be coupled together, using mercury atoms, to make the foundation layer of a three-dimensional scaffolding.
Michl鈥檚 scaffold structures will be riddled with holes so they will lack the strength of covalently bonded carbon compounds such as diamond. 鈥淚f you step on them,鈥� he concedes, 鈥渢hey will crush.鈥� An inert filling, squeezed into the gaps and then polymerised, would reinforce the construction. But more interesting would be to fill the empty space with more active materials. For instance the scaffolding could be impregnated with metal atoms, or semiconducting material which would give the material useful electrical properties.
Optically active molecules 鈥� absorbers or emitters of light 鈥� and electronically active groups could be hung on the scaffolding, to act as components of a larger device; a nonlinear optical device, for example, to convert light from one frequency to another.
Unfilled, the voids could be just as useful, especially in more conventional chemical applications. Michl suggests, for example, that the structures would make excellent large-pore molecular sieves 鈥� the pore size could be carefully designed to let through molecules of any chosen size in a way that current molecular sieves cannot. And it may be possible to design catalysts using the construction kit, defining the shape of the pores and the activity of the chemical groups hung onto the framework.
Of course, the voids are also the space in which the molecular windmill blades would turn. Although Michl had not spoken in public about his turbines before the Nature meeting, it was not some off-the-cuff quip. The idea is mentioned in the patent Michl holds on the construction kit, and, working with a graduate student, Jaroslav Vacek, he has already done some of the groundwork by running detailed computer simulations of a single rotor.
The plan is to suspend the rigid axle of a two-bladed rotor from the vertex of the molecular scaffolding through a single bond, which would allow nearly free rotation (see Diagram). The 鈥渂lades鈥� would be made from a number of fused aromatic rings, which are impermeable to gas atoms. Even the bonding has been fixed so that the blades are slightly inclined, just like real propeller blades. In the computer simulations, the rotor spins in a stream of helium gas, as well as bending from side to side and flexing the scaffolding around it as it is buffeted by gas atoms.
When Michl reported these simulations in Paris, he took his audience by surprise. According to Fraser Stoddart, a chemist at the University of Birmingham, 鈥渋t was clear that Josef was trying to get a response鈥�. Michl doesn鈥檛 deny it. The questions came thick and fast. In particular, many participants wanted to know, how would you be able to tell that the rotors were actually spinning.
Well, says Michl, for starters, charged groups could be added to the rotor, positive on one side, negative on the other. Because moving charges generate electromagnetic radiation, the spinning molecule would radiate microwaves. The signal from a single rotor would be far too weak to be seen, but there is no reason why you couldn鈥檛 build a huge array of turbines 鈥� a microscopic wind farm 鈥� to boost the signal. But still there were arguments. Unless the rotors spin in step, you would see nothing. Michl鈥檚 reply: a second uncharged set of rotors could lie between the charged ones, coupling the whole array together like mechanical cogs.
Backward running
More realistically, if the rotor idea comes off at all, Michl first expects to run it backwards, using microwaves to spin the rotors, which would then act like tiny turbopumps, driving the helium atoms before them.
The scheme wasn鈥檛 taken too seriously, but then Michl is hardly staking his reputation on it. His real agenda is to get chemists thinking about the possibility of mechanically conceived molecular structures. There is clearly some resistance. Stoddart, for one, despite having himself made a molecular shuttle with moving parts, feels that this approach could be misguided. 鈥淚t conjures up the idea of some excitement that we are not going to deliver on,鈥� he says. 鈥淚 don鈥檛 think chemists鈥� contribution [to nanotechnology] will be making mini-mini-mini computers or mini-mini-mini cars.鈥�
But Michl holds to his belief in molecular machines 鈥� if not the turbines he is working on now then perhaps molecular waterwheels or something completely different. He says that many ingenious molecular devices, Stoddart鈥檚 shuttle for instance, have been invented, but as yet they simply float freely in solution; Michl鈥檚 construction kit could be the 鈥渂ricks and mortar鈥�, coupling such devices together to make microscopic machines that are now just pipe dreams. And if he has set his sights high, he has an answer: 鈥淚 have taught my children that a hiker who is lost in the woods and comes to a fork in the trail should always take the branch that goes more steeply uphill. I should live up to my own admonitions, right?鈥�