The Importance of Finding Nothing Out

‘Melts in the mouth, not in the hand’: so said a chocolate advertisement from my youth for the predecessor of Minstrels (a discontinued brand called Treets). Melting temperature is of course an important consideration when it comes to the elegance of stuffing your mouth full of chocolate, and chocolate-making is a science as well as an artisanal art (judging by the incredibly expensive chocolate shops that are now springing up on our high streets). And it is the science of chocolate melting that Brian Cox is exploring in this neat video produced by the Royal Society to inspire children with an interest in science – not to mention to drill home the meaning of the ‘fair test’ beloved of school curricula. There is no doubt that it is an experiment likely to maintain children’s interest (do they get to eat the chocolate at the end of the experiment?), and it should teach them some useful concepts – even if not about the chocolate itself. The video says nothing about the difference in ingredients between dark, milk and white chocolate or why this might impact on the melting behaviour, or indeed its overall mouthfeel.

Chocolate may seem an odd topic for research but – as the advertising slogan above makes clear – getting the formulation right for physical properties as well as taste really does matter. It was a material I once studied using all the power of an X-ray synchrotron source to explore the internal packing of the triglycerides and how this changed during melting. If the triglycerides form the wrong crystal structure (a different so-called polymorph) then it is perceived as possessing the unattractive ‘bloom’ that dulls the surface of chocolate, usually occurring as a result of the chocolate being stored at too high a temperature. Different sources of the cocoa butter that is the key ingredient in chocolate have different proportions of the various triglycerides, each of which have their own melting temperature. And cheap chocolate has other ingredients added to mimic the same ‘mouthfeel’ response with inexpensive substitutes, thereby ending up with a chocolate that usually doesn’t taste anything like as nice.

However the aim of this post is not to give you a lesson in triglyceride physics and chemistry (there’s a little more of that in this piece I wrote for the Guardian), but to highlight an often overlooked aspect of research: null results. The particular problem we were looking at was a comparison of chocolate produced via the normal processing route, which involved temperatures somewhat above room temperature, with chocolate produced by ‘cold extrusion’, a process invented by Malcolm Mackley in the University of Cambridge’s Chemical Engineering Department. My student dutifully took many physical measurements to compare the end results of these two processes: melting, crystal structure, appearance in the electron microscope….I forget all the things she tried. Not a single difference did our physical measurements show up. We knew they were different: cold extruded chocolate could be shaped, wires looped for instance, whereas normal chocolate could not be so manipulated. Yet we found nothing, zilch, zero to report. All tests showed the two samples seemed exactly the same.

I won’t say this was good for my student’s morale, but there were plenty of (null) data to turn into a thesis. We did some studies on model systems for good measure. Actually those were to try out some possible ideas which might have fed into the full messy world of chocolate, but they provided solid, novel if unhelpful data. At the end of the day the student sailed through her viva. She had done original work and written it up thoroughly and rigorously, even if she hadn’t been able to solve the problem posed. But she had been able to rule out quite a large number of potential differences. It is helpful to remember that null results are important too.

However, there is bound to be a however, had she wanted to pursue an academic career she wouldn’t have got very far with not a single publishable paper out of her thesis. Luckily it was always clear she did not want to go that route and took her skills into management consulting where she clearly thrived. Had she been a determined wannabe academic I probably would have diverted her thesis work to something else early on to be sure she could pursue her dreams. At least with hindsight I hope that is the case, although the sponsor might have been unenthusiastic. (It wouldn’t have been the first time I modified a sponsor’s project on good scientific grounds).

There is another point I’d like to make about chocolate physics: that physics can indeed be done on such a material (even if not always very productively). That science is pervasive amongst the most common objects is something people may overlook in the belief that physics (perhaps more than other subject) is pigeonholed only to be concerned with the exotic and exciting not the everyday. Examples could include the Higgs Boson or the mysteries of black holes, but chocolate  – or starch, or shampoo or paint or any of the other things we are daily surrounded by – is perceived by many physicists and lay-people as ‘not suitable for physicists’. I talked a bit about this challenge in my recent Guardian podcast with Hannah Devlin . In the past I have spent an entire dinner failing to convince Andrew Cohen, head of the BBC Science Unit, that the public might be interested in things which they didn’t find mysterious. He told me the viewer wanted to be mystified; I disagree.

Food is not only an interesting topic of research, but also big business. To quote from a recent email I received from the Institute of Physics about the launch of a new Food Manufacturing Group

Food Manufacturing is the largest manufacturing sector in the UK employing more than 400000 people and supporting nearly 3 million further jobs across the whole supply chain and directly contributing £28 billion in GVA to the UK economy.

Physics is critically important to supporting the Food Manufacturing Sector in addressing the challenges it faces which include health and nutrition concerns, minimising waste and environmental impact, improving food security, responding to population growth and globalisation and improving productivity and developing and manufacturing successful products. A multidisciplinary approach is vital to solving these problems and physics has a crucial role to play.

It isn’t an area that hits the headlines – like drug discovery or driverless cars – but getting the underlying science right is likely to have significant consequences for the manufacturer and the consumer. As the country struggles with its economy, and the underlying low levels of productivity below our fellow G7 nations, as laid out in this week’s report from the Industrial Strategy Commission, we should not simply be looking at  hi-tech sorts of industry. We should worry about investing in improving productivity across the manufacturing landscape – chocolate included. After all, one of the advantages of cold extrusion of chocolate is the reduction in energy costs in its manufacture


This entry was posted in Research, Science Culture and tagged , , , . Bookmark the permalink.

2 Responses to The Importance of Finding Nothing Out

  1. bill harvey says:

    Cursed arrogant BBC people. No doubt an arts grad?!

  2. I agree entirely. Debunking mistaken results is important, but nobody ever listens.

    Last week I was at a workshop where a slightly famous result in physics got debunked in front of the 30 or so people who’re interested enough in it to have gone. Twelve years ago, Yves Couder published a paper reporting that droplets bouncing on a vibrating oil bath exhibit many of the characteristics of quantum mechanical particles, including showing interference fringes in a double slit experiment. This attracted wide interest from people who think about quantum foundations as it appeared to be an example of an analogue of quantum mechanics arising in a completely classical system. It seemed on the face of it plausible as it’s well-known that a nonlinear Schroedinger equation governs the modulation of surface waves in a fluid; what if you select the parameters so that the coefficient of the nonlinear term can be disregarded?

    Careful work by John Bush’s team at MIT and Andreas Andersen’s at Copenhagen has cast serious doubt on this. They’ve failed to replicate the experiment despite thorough exploration of the parameter ranges, showed that the relevant wave field is only present in one slit, and re-analysed the data from the original experiment to show that it’s as well-fitted by a Gaussian as by a Fraunhofer pattern. The only positive result is that the droplets’ behaviour does appear to be governed by a 2-d analogue of Maxwell’s equations (with c being the speed of surface waves); the droplets are actually an analogue of special relativity, not (as far as we can see) of quantum mechanics.

    Where does this leave the two dozen postdocs and research students who’ve worked on bouncing droplets over the past decade? I expect that the community will carry on while the smart ambitious young people will leave. I can think of several subfields of computer science (my day job) which have lost industrial relevance yet which stumble along undead for a decade or more. It’s all rather unfair on the students who get drawn in.

    What will happen to all the students rushing to work in the current blockchain bubble? What happened to the 10,000 who did PhDs in string theory? What will happen to all the quantum computing hopefuls when people give up on that?

Comments are closed.