Fundamentals and Analytical Applications of Multiway Calibration

N. FABER; R. BOQUÉ MARTÍ, J. FERRÉ BALDRICH; M.C. ORTIZ, L.A. SARABIA, M.S. SANCHEZ, A. HERRERO, S. SANLLORENTE, C. REGUERA; H. LONG WU, Y. LI, C. KANG, R.Q. YU

Elsevier

Lugar: Amsterdam; Año: 2015 p. 591

978-0-444-63527-3

As a popular story goes, a representative of a shoe-making factory visits a town of the distant country Barefootland, with the purpose of opening up new markets, only to discover that people in the town do not wear any shoes. He then moves to other towns, finding the same hopeless reality. Greatly disappointed, he finally sent this message to the central factory: ?don?t send anyshoes here, these people don?t need them.? After some time, another shoeseller from a different country visits Barefootland, finds the same people wearing no shoes, but sends this exciting message to the headquarters: ?please urgently send as many shoes as you can, these people need them!? This is just one version of the story, which includes sellers from many different countries.The moral is that great opportunities can be missed if one does not have the right attitude.Precisely, the purpose of editing this book is to follow the example of the second shoe-seller and try to contribute to making multiway calibration popular. We strongly believe that if multiway calibration is not widespread enough among analytical chemists, it is not because they do not need it. We also think it is the duty of chemometricians and analytical chemists using multiway calibration to disseminate the multiway philosophy and its specific techniques to the analytical community. Opening up the analytical chemistry market would make a difference both ways. Analytical chemists will fully incorporate a set of powerful calibration tools into their protocols, and at the same time chemometricians will get the bonus of myriads of interesting real-world challenges. Why is multiway calibration not as popular in analytical chemistry as webelieve it should be? The late Henry A. Bent, distinguished chemistry professor and educator at the University of Pittsburgh, once compared learning chemistry with a foreign language. Learning a new language involves incorporating new words (and sometimes new symbols) to understand everyday concepts. Learning chemistry, on the other hand, implies using new words and symbols, but to represent strange concepts. In his own words, chemistry is a foreign language twice over: strange terms for strange things. This applies to analytical chemists and multiway calibration: many unfamiliar concepts have to be grasped. However, the rewards might be countless. By conducting multiway calibration, one may accomplish the dream of any analytical chemist: to be able to quantitate an analyte in a complex interfering sample only by calibrating with the pure analyte. This property has been named the ?second-order advantage? because it was first observed in the domain of second-order (matrix) data, but it is shared by multiway data of at least second order. It is amazing how profound changes take place in the properties of experimental data when their structure goes from scalars, vectors, and matrices to threeand multiway arrays. The sea change occurring when entering the multiway world has already been enjoyed by mathematicians and pure chemometricians. It is time for us chemists, and particularly analytical chemists, to jointhe banquet. Multiway data analysis not only benefits analytical chemistry. Hidden factorswithin multiway data can be uncovered in a better way than in lowerdimensional data, thanks to the power of multiway algorithms, aiding data exploration, classification, and discrimination when it comes to the study of complex samples. Just some of the many other chemical activities whichmay equally improve from multiway data analysis are digital image processing, which is becoming increasingly popular in various fields, environmental data mining, metabolomics and related subdisciplines, where large amounts of data are being produced by modern instruments, and industrial chemical processes which depend on many different variables of economic and environmental importance. We should bear in mind that the very first paper describing the secondorder advantage was published in 1978, nearly 40 years ago! Still, multiway calibration is striving to leave the scientific ghetto in which it has lived for so many years. Interestingly enough, the historic record of publications on multiway calibration shows that after emerging in the late 1970s, it practically disappeared from the literature and finally revived during the mid-1990s,with a nearly exponential growth of publications. This temporal evolution nicely fits the so-called Gartner cycle, which in computer science characterizes an initial overenthusiasm or ?hype,? followed by subsequent phases of disappointment and ?enlightenment? that typically accompany the development of new technologies. Was it so, or was it because few analytical chemistscould foresee the power of the methods at the time they were first exposed? The advantages of multiway calibration should be enough to convince analytical chemists that it is immensely more powerful than either univariate or first-order calibration. However, crowds are not knocking at the doors of multiway calibration and will not do so until the message is adequately spread.We believe that by editing multiauthor books such as the present one, we are doing our bit. In Spanish, the mother tongue of the editors, we would say that ?we are adding our small grain of sand.?