LIBS for Liquid Samples

Although laser-induced breakdown spectroscopy (LIBS) potentially can be used for practically any kind of sample, most applications have focused on solid sample analysis. Montserrat Hidalgo, a professor in the Department of Analytical Chemistry and Food Sciences and the University Institute of Materials at the University of Alicante in Alicante, Spain, has been working with various approaches to extend the applicability of LIBS to trace-elemental analysis of liquid samples. She recently spoke to us about this research.

Your recent research has involved the use of LIBS to analyze liquid samples (1–3), as opposed to the solid samples most frequently studied by LIBS. What are the inherent difficulties associated with analyzing liquid samples with LIBS?

In fact, LIBS analysis of liquids is, in general terms, slightly more problematic than that of solids from the experimental point of view. Nevertheless, in my opinion the inherent difficulties are not associated with the LIBS technique but with the specific experimental strategy used for LIBS analysis of the liquid sample. There are different ways by which a liquid sample can be analyzed with LIBS. For instance, excluding all those strategies involving sample preparation procedures, LIBS analysis can be performed in the bulk liquid; on the surface of the liquid; or in jets, aerosols or isolated droplets generated from the liquid sample. Bulk liquid analysis is relatively simple from the experimental or operational point of view, but sensitivity is the main limitation of this modality because most of the plasma energy is lost in vaporization of the liquid and conversion into mechanical effects-such as the generation of a shock wave and a cavitation bubble-and only a small fraction is converted into radiative energy. Liquid surface analysis is comparatively more sensitive, because the laser-induced plasma expands in air. In this case, however, the strong mechanical effects arising from the laser–surface interaction cause experimental difficulties such as liquid splashing on the optic components of the system, and generation of aerosol above the liquid sample or ripples on the liquid surface, which can affect both precision and sensitivity of the method. These experimental difficulties can be minimized or even eliminated using jet, aerosol, or isolated droplets strategies for LIBS analysis, but the experimental complexity increase stemming from the need of additional devices to transform the liquid sample into the corresponding form for analysis and, in some cases, the applicability to viscous or turbid samples, can be problematic.

I think that when speaking about LIBS analysis of liquids the most critical point is not the experimental difficulty, which can be avoided by selecting the most adequate experimental strategy and experimental setup for a given application, but the analytical capability of LIBS detection for quantitative applications. Quantification limits obtained with all the aforementioned strategies are in most cases at the parts-per-million or high parts-per-billion level, which is inadequate for the quantification of most of the regulated contaminants in typical liquid samples such as environmental or food samples. Therefore, even if these strategies can be perfectly valid when trace-elemental analysis is not a requirement, other analytical methodologies should be investigated if we want to extend LIBS applicability to a higher number of applications requiring the analysis of elements at trace levels, such as the monitoring of contaminants in liquid samples of interest. To this end, the development of analytical methodologies involving sample preparation procedures before LIBS detection seems to be a promising alternative.

What are some of the sample preparation techniques you have used to enrich liquid samples for trace-element analysis with LIBS?

Our research on liquid samples analysis with LIBS has been mainly focused on the use of microextraction methodologies as sample preparation techniques for analyte enrichment, in both liquid- and solid-phase modalities. Microextraction methodologies usually lead to high enrichment factors, are more environmentally friendly than the corresponding conventional extraction modalities, and, in many cases, are also faster and more easily automatable. These two latter characteristics are very important when hyphenating sample preparation with LIBS detection if we want to use it for in situ and online applications. Among the existing liquid–liquid microextraction (LLME) modalities, we have tested single-drop microextraction (SDME) and dispersive liquid–liquid microextraction (DLLME). In the former technique, a microdrop of extraction solvent is suspended from the tip of a syringe and introduced in the sample for a certain period of time. After extraction, the analyte-enriched microdrop is retracted into the syringe and analyzed by LIBS (2). In the latter approach, the extraction solvent is dispersed in the sample in the form of fine droplets. After extraction, the analyte-enriched organic phase is separated from the aqueous phase by centrifugation, retrieved with a syringe, and analyzed by LIBS (1,3). In both cases, the analyte-enriched organic solvent is placed on a solid substrate and heated to dryness before LIBS analysis. Even if detection limits obtained with the application of LLME procedures are quite good, at the parts-per-billion level, they involve several experimental steps and therefore are not easily automatable. This is why we have also tested the use of solid-phase microextraction (SPME) procedures. SPME is a very efficient procedure for analyte separation and enrichment, with the advantage that analytes are retained on a solid sorbent, which can be considered the best matrix for LIBS analysis. One of the SPME modalities we have tested is dispersive micro solid-phase extraction (D-µ-SPE) using graphene oxide as the sorbent, in which the extraction procedure is similar to that described for DLLME. This procedure also leads to limits of detection at the parts-per-billion level, but still the analyte-enriched sorbent needs to be separated from the sample before LIBS analysis. To avoid this step, we are currently working on a promising SPME modality called thin-film microextraction (TFME). In TFME, the sorbent is coated on the surface of a sheet-like base material, which is then immersed into the liquid sample for extraction of the analytes. After extraction, the thin film with the analyte-enriched sorbent is removed from the solution and analyzed by LIBS, therefore simplifying the sample preparation procedure, which, in this case, has good possibilities of automation.

What are the technique’s advantages compared with other methods commonly used for elemental analysis of liquid samples, such as inductively coupled plasma–mass spectrometry (ICP-MS) or inductively coupled plasma–optical emission spectrometry (ICP-OES)?

It depends on the kind of measurements we are talking about. For instance, in my opinion, no advantage is gained by using LIBS for liquid analysis in routine laboratory measurements. ICP-OES and ICP-MS, among other elemental analysis techniques, are very sensitive and precise standard techniques that can be used to analyze many different kinds of liquid samples with or without the need of a previous sample preparation step, depending on the sample matrix and the concentration of the target analytes. The main advantages offered by LIBS are not for laboratory measurements, but for in situ and online measurements. LIBS analysis is very fast; the instrumentation can be of small size and the setup needs no gas supply system, vacuum condition, or bulky coolant device so it can be portable; and moreover, only optical access to the sample is needed to perform a LIBS measurement, which therefore allows stand-off analysis. All these characteristics make LIBS ideal for in situ and online applications such as oceanographic research and continuous environmental or industrial process monitoring, among others, which cannot be addressed with the conventional techniques commonly used for elemental analysis of liquid samples.

What are the limitations of LIBS for liquid sample analysis?

Compared with other techniques, the main limitation of LIBS is its low quantitative capability for trace-elemental analysis in liquids. It is possible to obtain very low absolute quantification limits with LIBS-for instance, at the picogram or even femtogram level when using experimental strategies such as isolated droplet analysis-due to its ability to sample extremely small amounts of liquid. However, as I have already discussed, relative quantification limits are usually at the parts-per-million or high parts-per-billion level, which makes the LIBS technique useless for applications requiring trace analysis. The use of sample preparation procedures can be a way to solve this limitation, but to improve the quantitative capabilities of LIBS analysis while maintaining its inherent advantages with respect to in situ and online capabilities, these sample preparation procedures should be efficient, simple, fast, and easy to automate and to hyphenate with LIBS detection. Finding a sample preparation procedure satisfying these requirements is not an easy task, and this is why it is still a challenge in LIBS research.

What are the primary samples or applications you have targeted? What makes them suitable for analysis by LIBS?

We have targeted many different applications since I started my research activity with LIBS technique, more than 20 years ago. During these years I have been studying both fundamental and applied aspects of LIBS, and the targeted samples and applications have been many and varied, so it is difficult to select the primary ones. Aerosols, solids, and liquids have been the sample types on which I have focused my research.

Studies on aerosols were mainly dedicated to sulfuric acid aerosols analysis, with the goal to detect the aerosols produced during the oxidation of atmospheric dimethyl sulfide (DMS) in laboratory chemical kinetic studies on the formation of atmospheric aerosols. In this case, the suitability of LIBS arose from the ability of the technique to analyze sulfuric acid aerosols in situ and in quasi real-time, avoiding the conventional off-line procedure consisting of collection on filters and analysis by wet chemistry and ion chromatography, which is not adequate for the continuous monitoring of the reaction products of the DMS oxidation in the course of a kinetic experiment. Another great part of my research work has been dedicated to solid samples, which have been also of variate nature. Among others, soils, archaeological samples, human hair, steel, photovoltaic cells, and halloysite nanotubes have been some of the targeted samples. Usually, these solid samples were considered suitable for analysis with LIBS because of the possibility of obtaining useful and fast analytical information by direct analysis of the sample, therefore avoiding the need for any sample preparation procedure. To conclude, in the last few years the primary samples I have targeted have been liquid samples, specifically aqueous samples. As I have already mentioned, liquids cannot be considered the most suitable samples to be analyzed with LIBS, especially if trace analysis is the pursued goal. In this case, my research activity has been precisely focused on the development of analytical methodologies able to convert these kinds of samples into appropriate forms for analysis with LIBS technique.

What are the next steps in your research?

My intention is to continue the research on hyphenation of LIBS detection with modern and efficient microextraction methodologies for liquid sample analysis. There is nowadays an increasing demand for portable analytical systems able to act, for instance, as early warning systems for environmental pollution control or industrial products quality control. I think that LIBS is a potential candidate to satisfy some of these demands, but, to this end, LIBS problems regarding the low quantitative capability in liquid sample analysis need to be solved, and therefore my intention is to focus the research on this critical point.

In view of the promising results we have obtained to date with the use of the thin-film microextraction technique combined with LIBS detection, the next step in the research will be to go deeper in the study of this sample preparation procedure. This research will include the study of new and efficient sorbent materials for TFME, different designs of films, evaluation of the developed methods for the analysis of different liquid matrices, and, a fundamental point, evaluation of possible ways for hyphenation and automation of both microextraction and detection processes in a single analytical system able to work in-field and to perform online measurements.

References

1.    M.A. Aguirre, E.J. Selva, M. Hidalgo, and A. Canals, Talanta131, 348–353 (2015).

2.    M.A. Aguirre, H. Nikolova, M. Hidalgo, and A. Canals, Anal. Methods 7, 877–883 (2015).

3.    I. Gaubeur, M.A. Aguirre, N. Kovachev, M. Hidalgo, and A. Canals, J. Anal. At. Spectrom.30, 2541 (2015).