LIBS has transitioned from a method found only in research laboratories, to a technique in wide use in commercial settings. Several leading LIBS experts share their views on how the technique has developed and where it is heading.
Laser-Induced Breakdown Spectroscopy (LIBS) has transitioned from a method found only in research laboratories, to a technique in wide use in commercial settings. Handheld LIBS systems are used to make measurements of metal scrap and other items in the field, while online LIBS systems are widely used in industrial high-speed sorting. This month's column polls several leading LIBS experts to give readers a sense of some of the history of, and most exciting upcoming research problems in, LIBS.
Over the past several years, I've had the pleasure of writing several columns on laser-induced breakdown spectroscopy (LIBS) for Spectroscopy. During that time, LIBS has matured immensely. It has transitioned from an experimental and lab-based method, to a method that is increasingly found in commercial and industrial use. Methods of quantification have improved, and the suite of lasers and spectrometers that are employed in LIBS systems are increasingly fit for purpose. With these changes in mind, it seems appropriate to cast a wide net and poll a panel of LIBS experts with a variety of backgrounds about their thoughts on recent developments in LIBS, and where this exciting field of spectroscopy is headed.
Our expert panel includes: Dr. Matthieu Baudelet, assistant professor in the Department of Chemistry in the National Center for Forensic Science at the University of Central Florida (Winter Park, FL), Dr. Amy Bauer, associate editor for Applied Spectroscopy, and principal scientist in the Chemlogix Division at TSI, Inc. (Shoreview, MN), Dr. David Hahn, professor and chairperson of the Department of Mechanical and Aerospace Engineering at the University of Florida (Gainesville, FL), and Dr. Steven Rehse, associate professor in the Department of Physics, University of Windsor (Ontario, Canada). We thank each of them for their time and thoughtfulness in responding to these questions.
LIBS is now a mature technique. Compared with other analytical methods, what do you think its relative strengths are, and how does this translate into applications?
Baudelet: LIBS is the best technique when the need includes the combination of speed and sensitivity. This makes it the most attractive elemental technique for chemical mapping at the moment. Its latest demonstrations of 10 ppm at kHz, or even at 100 Hz, rate make it very interesting for biomedical studies, where a fast multi-elemental image can help in diagnosis. LIBS can also overcome spectral interferences much more easily than mass spectrometry, which can make it the "go-to" technique for complex samples (when provided the right tools for data analysis).
Hahn: The long-standing strength of LIBS remains the lack of a need for rigorous sample preparation, such as acid digestion. LIBS allows for direct analysis of solids, liquids, and gaseous or aerosol samples, which is unmatched by other analytical schemes. Calibration does require care, as matrix effects are present with LIBS as with many other analytical schemes.
Bauer: Herbert Laitinen, in his original Analytical Chemistry editorial, "The Seven Ages of an Analytical Method," declared a method "mature" when it moved into the hands of the non-specialist. In the case of LIBS, this would seem, at first blush, to apply to the increasingly common use of handheld LIBS instruments to perform alloy identification (ID) in the metals recycling industry, and the use of handheld instruments to perform analyses on geological samples in situ. Additionally, LIBS is beginning to be applied to process control applications in various industries, but has not yet become common and is not yet seen as a "gold standard" analytical technique in any industry.
A thing that distinguishes LIBS from other analytical methods is the fact that it can perform noncontact measurements. However, this is not always as much of an advantage as is seems. LIBS is touted traditionally as requiring very little sample preparation, compared to atomic absorption (AA), or inductively coupled plasma (ICP), for example, but it is becoming more visible that care in sample preparation results in less uncertainty and greater measurement precision. There is an interesting balance that is present in LIBS analyses, one that pits easy and noncontact analysis against the need for good correlation to standard reference results and reduction of measurement-to-measurement variance.
Rehse: I believe it is the generality of the technique, and the ease with which it can be applied in so many different situations, that is its primary strength. What I mean by this is that it is not a "niche" technique; it can be applied in so very many different applications. And I mean this, not in the abstract, but in the specific. While my own specialty and area of concentration has been on bacterial analysis, over the years I have been involved in one way or another with experiments involved in such disparate applications as monitoring lead contamination in dirt from the sides of highways, analysis of ores for mineral prospecting, analysis of vegetables skins for signs of health or disease, analysis of gases from steel blast furnaces, analysis of fish bones to deduce invasiveness patterns, and several other applications. LIBS can and has made fruitful and significant contributions in many fields like this. There are such a wide range of applications, the significance is compounded by being "spread around" all over the analytical world.
Do you have a favorite "Aha" moment of your own, from your lab or data analysis that made you more excited about LIBS? Share a result that you are particularly proud of, if you don't mind.
Bauer: I was pretty excited when I first fiber-coupled a yttrium aluminum garnet (YAG) laser, in 1995 or so, and used the resulting plasma to decide whether there was lead in paint or not. In those early days, I was mostly involved with spark induced breakdown spectroscopy (SIBS), the electrical analog of LIBS, and my group was involved in attempts to use it to monitor bioaerosols. At some point, we started to image the plasma with a filter and an intensified charged coupled device (iCCD), and discovered that the actual plasma was much smaller than we had been predicting based on visible evidence, like a tiny lightning bolt. It was at that point that I began to understand that these plasmas were much more complicated than I previously gave them credit for. As people continue to try to understand the physics of plasma formation and evolution, each paper demonstrates this fact again and again. These physics remain my favorite part of the whole LIBS equation.
Rehse: I do. It is when I first got back "into" LIBS after a hiatus. I was lucky enough to be working at Los Alamos National Laboratory as a student in the early 1990s, when David Cremers was performing some of the seminal experiments on LIBS in a laboratory across the hall from mine. I always thought it was a fascinating technique, but my studies led me elsewhere. Fast forward to 2004, when I was performing a postdoctoral fellowship at the University of Western Ontario, and my lab was approached by a local company who wanted to talk with someone who knew LIBS to consult with them on its suitability for analyzing used engine oil for the presence of wear metals as an indicator of engine performance. Having not thought about LIBS for almost 10 years, I did a survey of the current status of the field and realized, "Hey, these people have got something here." I decided then to jump into the field on my own.
Hahn: My "aha" moment came more than 20 years ago, when I was using LIBS for emissions analysis, focusing on toxic metals. Recognizing that our targeted metals were condensed as solids, I did some simple math to model the sampling rates, and realized that only a small percentage of LIBS events actually sampled a target particle. Our signal to noise was greatly diminished by averaging many "zeros", but a careful analysis of each LIBS spectrum could greatly change the outcome. The ultimate result was an approach that is now referred to as conditional analysis.
Baudelet: Lately I've been very excited by our approach for quantifying spectral interferences in LIBS, which was recently published (1). This is an important brick in the edifice of analytical LIBS towards its recognition by the forensic practitioners, who need to be able to evaluate error rates of their measurements. Much more needs to be done to make it a staple of LIBS by including molecular emission, for example.
What do you think that the most important technical implementation or advance has been in LIBS over the past 3-5 years? How will this influence LIBS practice in the future?
Rehse: I really think the advances in high-resolution 2D mapping, as demonstrated in the papers of L. Sancey, V. Motto-Ros, and associates (2), and K. Rifai, F. Doucet, and associates (3), are incredible advances. The ability to rapidly create an elemental map on quite large two-dimensional surfaces with high resolution is a significant and import technical implementation of an idea that we all knew could be done, but no one had actually turned into a practical device. When that data is then interpreted with chemometric algorithms for autonomous classification of the surfaces, this give true utility and power to the technique.
Baudelet: Laser ablation molecular emission spectroscopy (LAMIS) and the revival of using molecular emission directly in plasma measurements started in 2010, and is still triggering enthusiasm within the LIBS community to create an optical emission technique for isotopic analysis and even halogens analysis. However, my feeling is that nanoparticle-enhanced LIBS (NELIBS) has been the most important new addition to the LIBS toolkit for better analytical performance in the last five years. It has enhanced many applications from the analysis of transparent materials to even explosive detection and protein analysis. We can see the analogy with Raman and surface-enhanced Raman spectroscopy (SERS), and should learn from that community how we can propose the best NELIBS tool for each application.
Hahn: I will offer two advances. The continuous improvement in data processing (such as chemometrics, machine learning, and artificial intelligence (AI)) schemes will continue to open up application spaces for sorting, in situ analysis and other applications. This, I believe, remains a significant advance that the LIBS community must continue to utilize. A second advancement is lower cost, high repetition rate laser systems. As noted above, sample inhomogeneity remains a challenge, notably when working in field applications. Using many thousands of laser shots can potentially help address such issues.
Bauer: The development of a whole new host of laser sources, including miniaturized millijoule and microjoule devices, have opened up an amazing new world of in situ measurements with LIBS. The most impressive technical developments, though, opened up the possibility for the development of ChemCam, the LIBS device aboard the Mars Science Laboratory rover. This device permits geological analysis and identification of hydrated minerals at a distance of 7 meters on Mars. Every time I think about the dramatic accomplishments of the ChemCam, I feel thrilled and optimistic at the same time.
What paper or papers have influenced your thinking the most in terms of your basic scientific understanding? Why have they been influential to you personally?
Bauer: Like I said earlier, I'm really very interested in understanding the basic phenomenology of sample ablation and plasma formation. My favorite papers currently are the ones that dare to try to explain the production of particles and splashing during nanosecond laser or surface encounters.
Baudelet: Many scientific contributions have influenced my basic scientific understanding of LIBS and its use for analytical chemistry. Three books in particular:
These have been in my desk for research and teaching since the beginning.
Hahn: I am a big fan of the early LIBS papers by David Cremers and Leon Radziemski. They addressed many of the fundamental issues while presenting very nice experimental data and analysis. I wore those early papers out with notations and constant reading.
Rehse: In my particular area of investigation, I am always drawn back to the work of Dr. Jennifer Gottfried that she described in her papers (4,5). I think they were important for a reason. Firstly, they showed just how far one can push the LIBS technique to discriminate between highly similar targets. In many of these cases (such as explosives), the targets are elementally identical, only their stoichiometry is different! These papers showed that, with careful construction of a classification library and suitable algorithms, the targets can be reliably discriminated. Secondly, no one before or since has so exhaustively studied the effect that the mounting surface (material upon which a trace target is scattered) and similar interferent materials can have on degrading classification accuracy. They tested dozens of combinations of similar materials, scattered on a wide variety of surfaces, and clearly and accurately reported the results in a way that allowed some sensible conclusions to be drawn. I think they showed that LIBS has an incredible ability to differentiate similar targets, but that a lot of work has to first go into understanding the ways in which a spectrum can become "contaminated," or ways in with can algorithm may be "fooled." I find myself looking at these papers a lot.
What areas of LIBS research would you advise new students or new practitioners to work on? What is your favorite "frontier" problem in LIBS?
Hahn: As an engineer that has enjoyed fruitful collaborations with analytical chemists and physicists for nearly two decades, I encourage students to relate their underlying fundamental training and knowledge to the LIBS arena. So as a mechanical engineer, that was heat and mass transport, allowing me to explore these underlying processes within the laser-induced plasmas, and ultimately to help translate new understanding to improved analytical approaches. My favorite frontier problems remains the local effects of analyte mass (such as small particulates) within a larger LIBS plasma as related to quantitative analyte response.
Rehse: I am still a believer that LIBS-based in vivo surgery is a possibility. I would like to see surgeons or dentists ablating away tumors or caries tissue, for example, while the ablation events are analyzed and categorized spectroscopically shot-by-shot to convey real-time information about the nature of the tissue being removed. This information could be relayed to the surgeon by means of a wearable head's up display. Rather than relying solely on the surgeon's experience and intuition, I would like to see true diagnostic information being made available during the procedure to guide the surgeon's actions and decisions. I think this type of information could ultimately improve patient outcomes, while reducing pain and suffering by minimizing the removal of healthy tissue from surgical margins and ensuring the removal of all the tissue that needs to removed.
Baudelet: LIBS, as a laser-ablation based technique, still relies on matrix-matched standards, especially for biological materials. Compacted powders are often not the answer for good quantitative practice. Finding better calibration standards and methods is a field of research that my group is focusing on so LIBS can be an even more robust analytical technique. This requires the full cooperation of material scientists, analytical chemists, specialists of laser-matter interaction and the support of certification institutions.
Bauer: I would love to be guiding a student at this exciting time. I would have them focus on ablation processes and plasma evolution processes involving new laser sources. From a more practical standpoint, the community is realizing that sample processing before analysis is very helpful to yield better results. LIBS folks are starting to remember traditional analytical tools, like internal standards and standard additions, and pairing them with very sophisticated data analysis tools, like PCA and support vector machines (SVM). The intentional paring of careful sample handling and crafty data analysis will result in more robust and accurate LIBS analysis.
This is an exciting point in the scientific understanding and practical application of LIBS. We hope that this combination of scientific, retrospective, and forward-looking thoughts from our panel of experts increases your interest, and your excitement, about the possibilities for LIBS technology. Thanks to all of our panelists for their energy expended and wisdom imparted in contributing to this column! Please send us your questions and requests for future column topics at the address below.
(1) J. Chappell, M. Martinez, and M. Baudelet, Spectrochimica Acta B 149, 167–175 (2018).
(2) L. Sancey, V. Motto-Ros, B. Busser, S. Kotb, J.M. Benoit, A. Piednoir, F. Lux, O. Tillement, G. Panczer, and J. Yu, Scientific Reports 4, 1–7 (2014).
(3) K. Rifai, F. Doucet, L. Özcan, and F. Vidal, Spectrochimica Acta B 150, 43–48 (2018).
(4) J.L. Gottfried, Anal. Bioanal. Chem 400(10), 3289–3301 (2011).
(5) J.L. Gottfried, F.C. De Lucia, Jr., C.A. Munson, and A.W. Miziolek, Applied Spectroscopy 62(4), 353–363 (2008).
Steven G. Buckley, PhD, is the Vice President of Product Development and Engineering at Ocean Optics, an affiliate associate professor at the University of Washington, and has started and advised numerous companies in spectroscopy and in applications of machine learning. He has approximately 40 peer-reviewed publications and 6 patents. His work in practical optical spectroscopy, such as LIBS, Raman, and TDL spectroscopy, dovetails with the coverage in this column, which reviews methods (new and old) in laser-based spectroscopy and optical sensing. Direct correspondence to: SpectroscopyEdit@mmhgroup.com.
Related Content:Column: Lasers and Optics Interface