Blowing the theory of how we smell

By: David Bradley

UPDATE: 2013-01-28 It's taken a long time since I first interviewed Luca Turin about his spectroscopic theory of how we smell at the Royal Society back in the early 1990s, but at long last he has the experimental evidence to prove he was right all along despite his detractors. Turin's team made isotopic analogues of the same molecules (same shape, different vibrational frequencies) and found that volunteers could distinguish between the odours of each. Chemistry World has more details.

An amazing theory about how we smell could put noses out of joint everywhere...read on to find out why the vibrations up your nose might be more important to the sweet smell of success than picking molecular locks. The original paper was published in Chem Senses, 1996, 21(6), 773-791.

Many molecules smell. That's a fact. What they smell of is not usually obvious from looking at their molecular structure though. For instance, some molecules that have very different shapes smell very similar, for instance hydrogen cyanide, trans-hex-2-enal and benzaldehyde all smell of bitter almonds, while others such as acetophenones, which look almost identical, can smell very different.

Some things can be predicted about a molecule's smell in general terms from its structure. Hydrocarbons found in petrol, for instance, have a nondescript odour while molecules containing amine groups have a fishy smell. Both organic and inorganic compounds containing a sulfur atom all have the sulfurous smell of rotten eggs. Bizarrely, a molecule like gardamide, which contains a CONH₂ amide group smells of grapefruit and horse fur!

The currently accepted theory of why one molecule smells like it does is based on a discovery made by British scientist John Amoore in the 1950s. He found that people's perception of smell was not always the same and some people had "blind spots" to certain smells - a condition called anosmia. Some people for instance cannot smell the male hormone androstenone others cannot sniff out musk or camphor. Amoore speculated that the receptors in the nose responsible for sending the signal from each aroma to the brain worked like a lock and key and for people with anosmia they had some locks missing.

Amoore's lock and key idea was not to be sniffed at as it was based on research in other fields that showed that some molecules could fit into proteins, such as enzymes, like a key in a lock and trigger an effect. Amoore reasoned that the combination of aroma molecules with different shapes picked the nose locks and sent a smell to the brain depending on the shape of the molecule. It all sounded very reasonable: the shape of the NH₂ group in amines fitted into the olfactory locks giving the sense of a fishy smell, the shape and size of the sulfur-hydrogen group unlocks a certain receptor. The idea has been the basically accepted theory of smell ever since.

However, another British scientist, Luca Turin a biophysicist and enthusiastic perfume expert was never satisfied with this explanation of smell. He says that the evidence just does not stack up. There are literally millions of smelly molecules many with similar shapes that smell very differently and others with wildly different shapes that smell almost the same. His favourite example, which he demonstrated at the Royal Society Exhibition in the summer of this year is that of decaborane.

Decaborane looks like the hydrocarbon camphane (the parent molecule of camphor used in cold remedies) but with all the carbon atoms swapped for boron atoms. The lock-key theory predicts that decaborane should smell like camphor. It does not. Decaborane strangely smells of sulfur.

How could that be? In the lock-key theory, the bulbous sulfur-hydrogen group (SH) is considered the only thing of the right shape and size to trigger the smell of sulfur. Decaborane has no sulfur but still has its characteristic stench! Turin has a theory to explain this and it could be set to undo the lock-key principle and revolutionise the perfumes and flavourings industry.

Turin's idea hinges not on the shape of molecules but on how they quiver and shake. Imagine a molecule made up of tiny balls connected by springs, twang the balls and the molecule vibrates with a set of frequencies which can be recorded as a vibrationalspectrum. More scientifically, he believes that the vibrational spectrum of a molecule is the real property that is detected by the nose and interpreted by the brain. The idea sounds bizarre but two of our other senses vision and sound - are based on the brain's interpretation of vibrations and spectra so why not smell? Turin reckons that the array of receptors in our olfactory bulb - the organ up our noses that detects smell - are sensitive to the vibrating springs in different molecules and can pass on a fingerprint signal to the brain.

Decaborane contains no sulfur but its vibrational spectrum is very similar to that of sulfur-containing compounds. The B-H bonds in decaborane vibrate at the same frequency as the vibrations of the S-H bond.

The ultimate test of the theory is what Turin describes as the acetophenone experiment. Acetophenone contains eight hydrogen atoms but if these are swapped for the heavier isotope of hydrogen, deuterium, the molecule is still chemically the same and is still exactly the same shape the only difference is that the eight hydrogens are heavier each by the mass of a neutron. The heavier deuterium atoms do not affect the molecules' shape but they do affect the vibrations of the molecule. This results in a marked difference in smell between normal acetophenone and the deuterated version.

The next stage is to design a molecule to produce a specific smell. This should be possible with computerised molecular modelling techniques that let the user simulate a vibrational spectrum for the molecule they design on the computer. If a compound is designed with a spectrum similar to the spectra of the components of chocolate, for instance, it might be possible to create a much more authentic chocolate smell for use in food. Emulating the spectrum of the musk or ambergris with biodegradable versions of the synthetic odour molecules could make a far more environmentally friendly perfume. Perfumes could also be designed with entirely new smells based on the vibrational spectra of different synthetic molecules.

The fact that smell, according to Turin's theory, is a spectral sense like sight and sound might also help explain a strange but very rare condition known as synaesthesia where the "sufferer's" senses are mixed up. Several musical composers claim to be able to hear in colours or smell sounds and very young babies are thought to have a mixed up sense of the world where the various inputs - sight, sound and smell - are not processed separately by the brain. The idea that these senses are based on spectra could be part of the explanation.

As if to prove there is nothing new under the sun. The idea of the nose working like a spectrometer was first proposed sixty years ago by chemist Sir Malcolm Dyson but Turin is taking the idea into the next millennium working with ICI's fragrance division Quest International to design a computer program that predicts smell based on a molecule's vibrations. The idea is surely nothing to be sniffed at.

If you want to learn how to sneeze, check out this disgusting post.

Since this article was first written, researchers at Rockefeller University have attempted unsuccessfully to reproduce Turin's predictions experimentally using deuterated molecules although Turin demonstrated and effect in fruit flies.