The (Dark) Art of Spectroscopy

Xiaoyun (Shawn) Chen: Others may come up with some highly technical challenges facing spectroscopists. Instead, I’ll mention a rather mundane and yet important one: how to engage and cultivate the new generation of spectroscopists. The breadth of spectroscopy applications is growing – ranging from cancer and infectious disease diagnosis to microplastics. And artificial intelligence is playing a more prominent role in converting highly complex spectral data into actionable information. The danger? As more and more people start to treat spectroscopy as a black-box, they risk reaching wrong conclusions or confirming their own biases. A solid understanding of spectroscopy and the myriad of tricks associated with the art of spectroscopy is indispensable. I hope the next generation of spectroscopists will spend time and effort learning, mastering, and appreciating the art of spectroscopy.

John Wasylyk: From an industrial point of view, the biggest challenge spectroscopists face is hiring new graduates in the chemistry field who lack basic understanding how spectroscopy can aid in their research. This applies to chemical engineers and synthetic chemists who are not familiar enough with the various instruments and instead rely on traditional labor-intensive techniques, which often provide only snapshots of data.  Those chemists should be made aware of where spectroscopy fits in and who the subject matter experts are in their company.

C. Derrick Quarles Jr: I completely agree with John here… Graduate programs in the United States are not focusing on atomic spectroscopy, partly due to a shift in the type of funding that is being offered to the universities. There also seems to be a sense that these types of instruments are just used for routine analysis. For example, there is a lack of understanding of how an ICP-MS fundamentally works, leading to – as Shawn said – the black box concept. Alternatively, the universities in Europe are still doing a great job teaching the fundamentals of atomic spectroscopy, which is leading to more job opportunities when these students graduate.

Sian Sloan-Dennison: I think the biggest challenge facing spectroscopists, especially those using SERS measurements, is absolute quantification. To build reliable and quantitative assays, which use SERS analysis to produce the concentration of a biomarker present in a sample, we must investigate new approaches that standardize the measurements and minimize the variation between samples. This could be as simple as including internal standards in the assay, creating robust calibration models that take into account sample matrix and background interference, or using complementary analysis methods such as IC-MS to characterize the SERS active material. However, this all starts with synthesizing stable and monodispersed nanoparticles, which can be a dark art!

Alexis Weber: As a young spectroscopist, still in graduate school, I find that the largest challenge is engaging students/researchers to be self-motivated in developing their own skills and network. As students are approaching the end of their doctoral degrees, they have been within the academic community for 9–14 years – sometimes without ever having an industry job within their field. This can lead to a lack of external connections – connections that are extremely beneficial when finally exiting the academic environment. If students do not attend conferences, network, or join professional societies, they will be at a disadvantage when looking for a job. Meeting with and learning from professionals outside of your lab provides you with insights and connections that you would not otherwise have. However, in my discussions with fellow students, they do not see the importance of this until it is too late. What’s worse is that there are opportunities available for students in the form of professional organizations, but they do not see the value in it because it is not specifically research oriented. Finding a way to break this stigma is one of the larger challenges faced in 2023 because it affects the spectroscopists of the future. 

Roy Goodacre: Though spectroscopy can be thought of as a single science, it must work with multiple disciplines. As well as the biological question one is trying to address, there will be experimental design and data processing steps that are needed to turn spectral data into some sensible answer. Thus, I think the main challenge is to have multidisciplinary groups that are in it for the team and not the individual – so they are invested in the research question rather than the techniques, both analytical and computational. A common dialogue towards the central aim/objective is necessary and this starts with designing the experiment.

Today, there is a crisis in reproducibility in science – this has been highlighted by a series of opinion articles published by the Nature publishing group. A big problem is that many experiments are designed incorrectly; please note, this is not a deliberate act, but the samples chosen to be analyzed very rarely address the problem that one is wanting to answer. For example, cancer is in general a disease that tends to be found in the elderly (I know there are exceptions, but if we think of pancreatic, prostate or bowel cancer, these are usually found in more elderly populations), therefore, one needs to look at the match controls in studies to make sure there are no age differences, no gender differences, nor any differences in other demographic or clinical information. In addition, it is worth noting that people with disease are usually self-medicating before they seek help from a doctor and thus the use of over-the-counter medicines might also correlate with people who are ill. All of these confounders need to be considered during the sample selection and design phase – this is often difficult. But it is pointless designing a spectroscopic experiment that tells you the age of the person or about the drugs or supplements they are taking – rather than whether they actually have disease!

So my thoughts on the biggest challenge come down to capturing enough information about the patient and mining this to check if there are any confounding variables that one needs to be aware of and then adjust for.

SciX takes place on October 8–13 in Sparks, Nevada, USA. For more information about the program, visit: scixconference.org/scix2023program