My Ever-expanding Analytical Toolbox

To fully understand the point I would like to make, one needs to understand how an analytical laboratory works in industry and how the university equivalent functions – and to recognize the disparity between them. In academia, analytical research largely was – and still is – very technique oriented. Indeed, the academic analytical scientist often works on a single technique with objective of making a new discovery or some developmental breakthrough. The actual use of the technique – the application – is not very relevant; the techniques and theories developed are applied to a single sample just to fill the last (and least important) figure in the manuscript.

But in industry, how different the situation is. Except for instrument manufacturers, industry is not performing analytical research. Rather, industrial analytical chemists are trying to figure out how to use existing methods to solve problems. Emphasis in the previous sentence should be heavily applied to the word ‘existing’. Unlike academics, who are terribly disappointed if they find an article describing an idea they just had and which they thought to be original, industrial researchers are very happy to find a solution to a problem in literature.

In fact, the best industrial analytical scientists are excellent thieves; immediately able to recognize relevant methods that can be “borrowed”, slightly adapted, and “sold” as new developments to their superiors. In addition, a good industrial analytical scientist is also like a skilled plumber in that both need a huge toolbox filled with relevant tools and techniques. When confronted with a novel analytical (or plumbing) question, the real professionals can dive into the toolbox, select the tool that is most appropriate to solve the problem, and successfully apply it. However, this is just the part of the industrial analytical chemist’s job that is becoming increasingly difficult. The toolbox is ever expanding. New methods are developed, with new column chemistries, separation modes, detection devices, derivatization strategies, data processing routing... need I go on? There are simply too many tools and unsurprisingly the toolbox is getting far too heavy to be easily carried around. We do not all share the skills of the magical Mary Poppins. Certainly, one needs an excellent memory to remember which tools are in the box – and an even better memory to remember how to operate them all.

The big question is, do we really need all the tools in the toolbox? Can we not make do with just a few methods, each with wide compound or problem coverage? When thinking about the answer to these questions, please also bear in mind that for most users of analytical tools, the methods do not have to deliver the best performance. Performance should be fit-for-purpose and not necessarily a World Record attempt. An unfortunate complication here is that the parameters that industry finds extremely relevant, for example, reliability and ease of use, are rarely considered by academic technique developers. Capillary electrophoresis may well be an excellent and superior method given its basic figures of merit, but it is a technique best avoided in the real world. Classical HPLC methods might be slower, less efficient – take your pick of negative features – but at least they work.

In my view, we have too many methods. And new methods that are only marginally better than existing methods are being added as I write. In reality, we do not need more methods that are slightly faster or slightly more sensitive. What we need are fewer methods with a broader coverage of compounds, matrices and, ultimately, questions. There is no one tool that can solve all problems. But, in industry, we definitely need to slim down the toolbox. Of course, these chosen few methods may not be the best method for each question. But that’s not a problem as long as they provide clear, fit-for-purpose answers – both reliably and rapidly. Generic methods and techniques are the future.