The beauty of chemistry

Can a chemistry experiment be beautiful? Philip Ball gives his opinion and invites us to see beauty in everything

Can we call a chemistry experiment ’beautiful’? Chemists evidently think so; but what exactly do they mean by it? And which are then the most beautiful experiments of all?

One thing that does set chemists talking about beauty is molecules. The football-shaped carbon cluster C₆₀ has been called ’the most beautiful molecule’ on account of its delightful symmetry. The master of organic synthesis Robert Woodward had a clear appreciation of ’molecular beauty’, and his daughter Crystal felt that ’the delight and aesthetic pleasure’ he took from his products ’contributed to his skill’. A sense of Platonic beauty was what motivated many chemists (including Woodward) to try to make the polyhedral hydrocarbon dodecahedrane (C₂₀H₂₀), a challenge that Leo Paquette at Ohio State University, US, first conquered in 1980.

The criterion of ’beauty’ applied by chemists to the molecules they make is indeed of a Platonic nature: it typically identifies beauty in a high degree of symmetry. This notion was advanced by Plato in his book Timaeus, and it articulated his feeling that beauty in nature is connected to orderliness and regularity. It is tempting - and chemists often fall for this temptation - to conflate Platonic beauty with an artistic aesthetic. But Plato’s idea of beauty was never a part of any theory of art, and in fact he spoke disapprovingly of the ’deceptiveness’ of painting and sculpture. So it’s not particularly meaningful to regard symmetrical, ’beautiful’ molecules as objets d’art, although they surely do have about them an ’artistry’ that is connected to the older idea of art as techne, a skill in fabrication.

That a part of the beauty of a chemistry experiment can lie with the product distinguishes chemistry from some of the other sciences, for physics, biology and geology are less often concerned with producing an artefact. But it also seems fair to argue that an experiment can itself embody a kind of beauty, whether in concept, strategy or mode of realisation.

Yet the notion of ’experiment’ too has a rather unique character in chemistry. To many scientists, particularly physicists, an experiment is a procedure that is designed to ask a question - and elicit an unambiguous answer - about the natural world. This, certainly, was the template adopted by philosopher and historian of science Robert Crease in his marvellous book The prism and the pendulum: the ten most beautiful experiments in science (Random House, 2003), in which he describes 10 classic experiments that probed and quantified nature, such as Galileo’s studies of the laws of motion and Robert Millikan’s determination of the charge on the electron.

But chemistry doesn’t concern itself so much with ’experiments’ of this nature. More often, the experimental science conducted in the chemistry laboratory is all about making new compounds and molecules. Some scientists might want to draw a distinction between this kind of ’experimental science’ and the notion of an enquiry into the natural world. Is the total synthesis of a natural product, involving months or even years of painstaking preparative work, really the same kind of undertaking as, say, Ernest Rutherford’s study of alpha-particle scattering that revealed the nature of the atomic nucleus, or Isaac Newton’s decomposition and reconstitution of sunlight (two more from Crease’s list)?

Crease’s book prompted the RSC to ask me last year if I would consider writing a book on the ’10 most beautiful experiments in chemistry’ - as a kind of rejoinder, perhaps, although I don’t see it in adversarial terms. The RSC, like chemists everywhere, was dismayed to see Crease apparently claim that the ’10 most beautiful experiments in science’ were all about physics. This selection was understandable, however, for The prism and the pendulum stemmed from a readers’ poll that Crease conducted in the magazine Physics World, where he asked explicitly for nominations of beautiful physics experiments. When he adapted the results for his book, Crease justified the generalisation from ’physics’ to ’science’ partly on the grounds that his selections are all ’textbook’ examples of scientific experiments.

Fair enough; but the ’textbook’ definition of science is itself typically awry. More often than not, it supposes science to be an investigation into nature: a definition that excludes the strongly creative, artificial character of much of chemistry. At all times before the 20th century, scientific experimentation was closely linked to techne. There was, for example, a convergence between applied chemistry - the manufacture of dyes and pigments, say, and of soaps and detergents and the brewing of beverages - and ’academic’ science only during the 19th century.

We now have an image of ’science’ that is largely at odds with the way it is actually practised. Philosopher of science Joachim Schummer has estimated that there are many more scientific papers published in chemistry than in any other scientific discipline. What’s more, the overwhelming majority of those papers report the results of experiments. ’Chemistry’, says Schummer, ’has always been the laboratory science per se. Although chemistry is no longer the only experimental science, it is by far the biggest one and historically the model for all others. Thus, if we want to know what scientists mean by "experiment", chemical papers are the right point to start with.’

Schummer points out that roughly a third of all scientists worldwide are engaged not in the experimental testing of theories, but in producing and characterising new substances - in chemical synthesis. Chemistry, as the French chemist Marcelin Berthelot recognised, creates its own object: it is not necessarily an enquiry into nature, but sets synthetic goals that are shaped by the considerations of the engineer, in particular by the issues of function and design. Synthetic chemistry, as I’ve indicated, has its own aesthetic: the ’unnatural’ molecules that chemists try to make are no less a ’designed’ product than motor vehicles or buildings, and as such their structures are not inevitable.

I was relieved to find that I was not the first to wrestle with these issues about beauty and experiment in chemistry. In late 2002 the American Chemical Society (ACS) canvassed readers of Chemical and Engineering News to submit proposals for precisely the same kind of list that the RSC asked me to compile. A shortlist of 25 experiments was then ranked by a panel of chemists and science historians. I discovered this only after I’d drawn up my own ’top 10’, and so I was keen to see how they compared.

While I was encouraged to find some coincidences between the two lists, I was also stimulated to defend the differences. What struck me most about the ACS list was first how it seemed to conflate ’experiment’ with ’discovery’ - the now pervasive paradigm for historical and philosophical discussions of scientific work. ’Beautiful’, meanwhile, was often equated by the panellists with what one of them called ’conceptual simplicity’, coupled to the lingering notion that a ’beautiful’ experiment ought also to be an important one. (In practice, that latter consideration tends to take care of itself, since inevitably the experiments we tend to record and recall and analyse in sufficient detail to know what really happened are those that made an impact.)

Elegance and simplicity are surely among the key attributes that entitle an experiment to be labelled beautiful, and some of my selections have been made for that reason. But it is not at all clear that these should be the only, or even the principal, criteria for every selection. And in fact, if there was ever any intention of that being so for the ACS list, the rules were flouted more than once. For example, William Perkin’s synthesis of aniline mauve, the first aniline dye, in 1856, which fetched in at number five in the ACS ranking, was as messy and inelegant an experiment as one could imagine: the dye was the initially unpromising residue produced by a wholly misconceived attempt at the chemical synthesis of quinine.

Once we allow that ’experiment’ can imply ’experimental science’, which could involve a series of investigations, perhaps even spanning several years, compiling a list inevitably means comparing apples and oranges. How do you weigh a single, neat test of some hypothesis against a conclusion derived from the dedicated accumulation of data over a long period? The former can have the beauty of a dramatic revelation; the beauty of the latter can derive from the construction of a coherent chain of logical argument and deduction.

For example, experiment number three on the ACS list, the determination by the German chemist Emil Fischer in the early 1890s of the precise three-dimensional structure of the glucose molecule, was, in the words of science historian Peter Ramberg, ’part of a large research project involving several smaller projects on the classification of the natural monosaccharides [sugars] that gradually came together in 1891. There was therefore no one specific experiment that "determined" the configuration of glucose’.

So I have tried to take a very loose view of how one should interpret both ’beautiful’ and ’experiment’. One of the key themes in all of the cases I have chosen is that they are shaped by human attributes: invention, elegance, perseverance, imagination, ingenuity. This has tended to work against the inclusion of experiments (like Perkin’s) whose success depended on serendipity: chance discoveries are appealing and entertaining, but I find it hard to see beauty in sheer good fortune. (Admittedly, however, most serendipity is more than that.)

I have included two examples of ’classical’ organic syntheses, and for different reasons. The synthesis of vitamin B₁₂ by teams led by Robert Woodward and Albert Eschenmoser was a phenomenal project that took over 10 years to complete. It has the beauty of a chess game, a masterpiece of strategy in which each move forms an integral part of the whole, and where the protagonists could never afford for a moment to lose sight of the broader picture. That was Woodward’s genius: his syntheses were always elegant, and he could see precisely where he was going even when to other chemists he might seem to be taking a curious diversion.

But Woodward’s formidable and thoroughly deserved reputation has tended to overshadow the contribution of Eschenmoser, who in fact pioneered techniques for making the ’corrin’ macrocyclic ring that forms the central framework of vitamin B₁₂. Woodward would not have succeeded alone. And when Woodward’s pathway led him, with the help of theorist Roald Hoffmann, to uncover the stereochemical principles of ’pericyclic’ bond formation now known as the Woodward-Hoffmann rules - one of the most profound examples of how synthesis can lead to fundamental chemical insights - Eschenmoser saw at once how to exploit this new understanding to create a new pathway to the corrin ring.

Neil Bartlett’s synthesis of the first xenon compound, XePtF₆, in 1962 is another example of the chemist’s compulsion to create new substances. The beauty here, however (aside from the sheer delight of the colour changes involved, as red PtF₆ vapour turns into the bright yellow xenon compound), stems from the simplicity, even the simple-mindedness, that enabled Bartlett to leap ahead of conventional wisdom.

Even though everyone ’knew’ that the inert gases were just that, Bartlett reasoned that if PtF₆ could oxidise oxygen molecules, as he decided it had in the O₂PtF₆ compound he’d made in 1956, then it should do the same to xenon, since O₂ and Xe have almost identical first ionisation potentials. And he was right, to the delight of chemists everywhere. Primo Levi celebrates this activation of ’the Alien’ (xenon) on the very first page of his famous book The Periodic Table, where he wrongly (but justifiably) imagines that the feat won Bartlett a Nobel prize.

Here, then, is my list (see box). I hope you might agree with at least some of it, but I will be content too if you are irritated or even outraged by what it includes and omits. Happily, there is room in the world for many different ideas about what is beautiful.

Philip Ball’s book, ’Elegant solutions: ten beautiful experiments in chemistry’ will be published by the RSC in September 2005.

Philip Ball is consultant editor for Nature, 4-6 Crinan Street, London N1 9XW