04 August 2015 On sugar in coffee.

It's obvious that adding sugar to coffee makes it sweeter. What is not so obvious is that it also masks the bitter taste of caffeine. The explanation for the masking is simple: caffeine as a single molecule is quite bitter, but it tends to self-associate to form a dimer and the dimer is less bitter.

So the interesting question is why the sugar causes the caffeine to dimerise. Although the question sounds of no great importance, in fact it is coupled to a whole bunch of questions that are of major importance. It is often observed that some types of additives (e.g. sugars, urea, guanadine, "hydrotropes") have a significant impact on issues such as protein folding, starch solubility or the solubility of drug molecules. Protein folding is important because enzymes can be rendered inactive if they fold badly. Starch solubility is important in the making of cakes and cookies. Many pharmaceuticals are poorly soluble in water so adding a harmless solubilizer can make all the difference between bad and good drug delivery.

For decades the explanations for these effects have been confused and confusing. But in the past few years my friend and colleague Dr Seishi Shimizu of U. York has developed (with the help of associates in Japan and the UK) a fresh way to look at all these phenomena, removing all the historical confusion. His latest paper shows that Sugar supresses as well as masks caffeine’s bitter flavour. A common feature in all these phenomena is that water is such an odd solvent and many of the explanations have claimed that it is the odd properties of water that are affected by the additives such as the sugars. Another feature often observed is some sort of self-association of the additive (not to be confused with the self-association of the caffeine which is the solute, not the additive). It is often claimed that this self-association is the key to the observed effect.

Seishi has shown that in general the water effects are negligible - and in particular there is strong evidence that "water structure" is of no significance in these effects. Similarly he has shown that when self-association of the additive occurs this actually reduces the size of the effect!

The method Seishi uses is rather basic, pure, assumption-free thermodynamics: Kirkwood-Buff theory. By crunching some basic solubility, density and vapour/osmotic pressure numbers it is possible to show unambiguously what is interacting with what as the concentration of the additive increases. Thus it can be shown that water-water, water-solute and water-additive changes are not significant. It can be shown that when additive-additive interactions are strong then (with the exception of true surfactants where the interactions take place at very low concentrations) the effect on the solute is decreased. Finally it shows that solute-additive and, even, solute-additive+additive interactions dominate the effects.

In the case of sugar and caffeine the dimerisation is encouraged by the negative sugar-caffeine interaction; basically the caffeine is excluded from the space where the sugar is, so has no choice but to self-associate.

Naturally I think that Seishi's approach is very powerful. But I have to admit I'm biased because I have the honour of being the co-author on three of his papers on the mechanisms behind hydrotropy (solubilisation of drug molecules). Being a co-author was a special privilege because it meant that I was involved in science in action. The first paper was correct, but difficult to understand. The second paper was correct and easier to understand. The third paper looks at more data and also manages to greatly simplify the whole story. What changed? Thermodynamics is powerful but dry. The first explanation looked at the effect from one point of view which was entirely reasonable but happened to lead to a complicated story. The second explanation adopted basically the same approach (with different data) but looked from another point of view which automatically simplified things. The third explanation Hydrotrope accumulation around the drug: the driving force for solubilization and minimum hydrotrope concentration for nicotinamide and urea took advantage of some deep fundamental work that Seishi carried out in intense sessions in Japan which could then be used to make the apparently complex story much simpler. That's how science works - it only looks simple in hindsight.

Seishi loves to mix the hard core science necessary to achieve simplicity with the simplicity needed to make the science accessible. So there are more food science papers in the pipeline (as Seishi is Japanese, the science of tofu features as one example). My ambition is to appify much of Seishi's work so that his approach can be adopted more generally. Those who know me are well aware that I am hopeless at thermodynamics. It is that (sad) weakness that I am trying to turn into a strength by helping to translate the deep insights of high-powered thermodynamics into tools so simple that even I can use them.