Effect of particle size on laser-induced breakdown spectroscopy analysis of alumina suspension in liquids

https://doi.org/10.1016/j.sab.2012.06.021Get rights and content

Abstract

The analysis by Laser Induced Breakdown Spectroscopy (LIBS) was proposed for the detection and the quantification of different elements in water even when the analyte is composed of particles in suspension.

We have studied the effect of particle size on the LIBS signal during liquid analysis. In our study we used different particle sizes (from 2 μm to 90 μm) of Al2O3 in suspension in water. The results were compared to the signal obtained in the case of dissolved aluminum.

In the case of particles, a linear correlation between the LIBS signal versus concentration was found but a significant decrease in the slope of the calibration curve was found when the particle size increased. Several hypotheses have been tested and only a partial ablation of the particles might explain this decrease in signal intensity. This effect probably does not occur at smaller particle size. We estimated 860 nm/pulse as ablated thickness from the top of the particle. A statistical analysis over all data obtained allowed us to calculate 100 μm as ablated water column depth.

Highlights

► We have identified a decrease of calibration curve when particle size increases. ► Partial particle ablation has been identified as the origin of this effect. ► The ablation rate on Al2O3 particles in suspension in water has been estimated. ► We can determine the deepness of the interaction volume into the liquid.

Introduction

The laser-induced breakdown spectroscopy (LIBS) technique is unique in the sense that it does not require sample preparation and is increasingly implemented for the capacity of remote detection and in situ analysis in any phase (solid, liquid or gas). These capacities are undoubted advantages compared to other conventional analytical techniques. It should be noted that the majority of LIBS developments published to date are for solid samples. Less attention has been paid to LIBS analysis of liquids although this technique shows great potential for in-situ analysis of industrial processes or monitoring of water in the environment, particularly for on-line sewage and mine drainage quality monitoring.

First applications were developed for the analysis of dissolved elements in the liquid phase. Cremers et al. [1] reported the detection capability from ionized and neutral atoms and simple dissolved molecules with limits of detection (LOD) at levels above 1 ppm for Li, Na, K, Rb, Cs, Be, Mg, Ca, B, and Al. Regarding the importance of geometry interaction with the laser beam, Fang et al. [2] proved, in the case of calcium analysis, that the temporal characteristics of plasma emissions differ in the two different samples configurations. i) In the case of water bulk, the emission is characterized by two different temporal regimes: principally from the multi-photon-induced breakdown of water molecules and an atomic emission delayed by about 200 ns. ii) In the case of surface water jet, if the emission is characterized by a maximum, almost coinciding with the laser pulse then the LOD are improved. This difference is attributed to two distinct dynamic processes. Ng et al. [3] and Ho et al. [4] performed spectroscopic studies of plasma generated from a stable water jet. They were able to calculate plasma excitation temperature and electron density for a delay time of up to 1 μs by the use of different excitation wavelengths. At the same fluence the signal-to-background ratio was enhanced a thousand-fold compared to that obtained with a wavelength of 197 nm than 532 nm.

In recent years other improvements have been carried out, not only for dissolved elements, but also for suspensions and colloids by using jet liquid laminar flows. Ito et al. [5], added a coaxial flow system which made it possible to change the atmosphere around the plasma. They reported obtaining a good detection of colloidal FeO(OH) in water when helium gas flowing through the outer nozzle was used to increase the intensity ratio of Fe emission lines to the background emission (S/B ratio). These emission lines were observed at about 3.5 μs after the laser pulse and the LOD for iron was estimated at 0.6 ppm. From the same laboratory, Nakamura et al. [6] reported an improved LOD equal to 16 ppb for Fe using a similar experimental setup but with double-pulse excitation. Haisch et al. [7] decided to opt for a prior separation of the liquid and solid phases by ultra filtration. The filter was analyzed by LIBS by scanning the surface. They concluded that phase separation was preferable in terms of sensitivity for heavy metal colloid analysis. Knopp [8] applied LIBS for the detection of metal ions (Li+, Na+, Ca2 +, Ba2 +, Pb2 +, Cd2 +) in aqueous solutions with LOD of the order of ppm. However, for Er3 + the LOD result obtained was not as good as in the dissolved state but they reported that the LOD was 103 times lower when ErBa2Cu3O particles were in suspension in water.

Kovalchuk et al. [9] proved that, under their experimental conditions, the mechanism of breakdown in water is initiated by inclusion particles. Samek et al. [10] proved that LIBS is suitable for quantitative analysis of elements such as Al, Cr, Pb, Tc, Cu, and U in liquids at high and moderate concentrations from 10 to 100 ppm. On the other hand, localized concentration gradients can easily be traced using laminar water jet. They proved that for samples of mineral waters, river-water, and seawater samples LIBS is a fast real-time technique for remote analysis.

These interesting results show that LIBS may be a useful tool to analyze heavy metal elements. But what happens when the metal is in a particle state? Hahn et al. [11] demonstrated that LIBS yielded results for real-time and elemental analysis of single particles for aerosols in laboratory experiments. He obtained a sensitivity of the order of ng/m3. He provided an elemental quantitative analysis and an individual composition of sub-micrometer to micrometer sized aerosol particles during field monitoring of air quality.

Currently, some data explain how the particles in suspensions influence analytical signals. This technique can be expected to quantify the total amount of atomic species present in liquid phases whatever the physicochemical state of the analyte. Because it is important to distinguish between metallic elements bound on suspended particles and dissolved metal ions, principally in environment related fields, the aim of the present study is to investigate the effects of particle size in suspension on the quantitative analysis with LIBS and to explain possible deviations. For this study, alumina particles (Al2O3) were chosen when dissolved aluminum in water is considered as the reference standard.

Section snippets

Experimental setup

A commercial LIBS system was used but two different spectrometers were used for the measurement depending on what information was sought.

Particle size effect on calibration curve

Atomic emission lines from aluminum were observed at 394.4 nm and 396.1 nm. A standard procedure for spectra acquisition by LIBS was developed with the ESA-3000 Echelle spectrometer with the optimal settings mentioned in Table 2. One spectrum was obtained by the accumulation of 120 laser shots. It was repeated 25 times to have a precise estimation of the mean value of the Al 396.1 nm line intensity. We repeated this measurement for different particle sizes at different Al concentrations to draw

First hypothesis: sampling problem

The circulation of suspended particles within the water jet can be heterogeneous and not representative of the global sample. Because of sedimentation, this problem can be more significant for the largest particles. Although the detected signal and the percentage of “good shots” are constant along time, we decided to quantify aluminum concentration in different suspensions using ICP optical emission spectrometry (ICP-OES) for a direct sampling of the liquid jet. To minimize the number of ICP

Conclusions

In the present work, we have described the use of commercial LIBS apparatus dedicated to the analysis of particles in suspension in water. With this experimental configuration we realized spectra on Al2O3 particles with different sizes.

The calibration plots showed good linearity at their respective concentration. The slopes of the linear regression decrease drastically as the particle size increases. We demonstrated that the hypotheses of sampling procedure and self-absorption are not the

Acknowledgments

We would like to thank specially Dr Francois Piuzzi and the “Puya de Raimondi Association” (www.puyaderaimondi.net) without which this work could not have been realized. The assistance provided by Michel Tabarant in ICP measurements is acknowledged. We thank the Univ. Paris Sud and CEA for their financial support.

References (14)

There are more references available in the full text version of this article.

Cited by (7)

  • An overview of feasibilities and challenge of conductive cellulose for rechargeable lithium based battery

    2015, Renewable and Sustainable Energy Reviews
    Citation Excerpt :

    The design on metallic particle and cellulose can be enrolled on sensor technology [91], self-cleaning process [92,93] and antibacterial properties [94–96]. From the synthetic route, metallic particle was successfully prepared from two ways; the one called top-down approach which theoretically referred to breaking the solid or bulky material in order to obtain small part of metallic particle [97,98]. While the other one called bottom-up; this process was involved on the synthesis of novel material or it referred to the investigation on thermodynamic and kinetic approach.

  • Influence of particle size distribution on the analysis of pellets of plant materials by laser-induced breakdown spectroscopy

    2015, Spectrochimica Acta - Part B Atomic Spectroscopy
    Citation Excerpt :

    The particle decomposition and vaporization, as well as the analyte atomization and excitation efficiencies within laser-induced plasmas, are generally favored when smaller particles are ablated. There is an inverse relationship between particle size and sensitivity of the calibration curve, with the slope increasing as the particle size decreases [6,7]. The effect of particle size on the analysis of aerosol by LIBS has been exploited elsewhere [8].

  • Sampling considerations when analyzing micrometric-sized particles in a liquid jet using laser induced breakdown spectroscopy

    2014, Spectrochimica Acta - Part B Atomic Spectroscopy
    Citation Excerpt :

    Either such effect is the consequence of self-absorption or of non-complete vaporization of the particles, like that observed by Carranza and Hahn [26] when analyzing particles in air using LIBS. A similar result has recently been found by Diaz Rosado et al. [12] when analyzing micrometric-sized alumina particles in liquids using LIBS. A careful examination of the recorded data showed that non-complete vaporization was very likely to be the cause of signal saturation.

View all citing articles on Scopus
View full text