Laser-induced breakdown spectroscopy for elemental ...

Laser-induced breakdown spectroscopy for elemental characterization of calcitic alterations on cave walls Léna Bassel, Vincent Motto-Ros, Florian Trichard, Frédéric Pelascini, Faten Ammari, Rémy Chapoulie, Catherine Ferrier, et al. Environmental Science and Pollution Research ISSN 0944-1344 Environ Sci Pollut Res DOI 10.1007/s11356-016-7468-5

1 23

Your article is protected by copyright and all rights are held exclusively by SpringerVerlag Berlin Heidelberg. This e-offprint is for personal use only and shall not be selfarchived in electronic repositories. If you wish to self-archive your article, please use the accepted manuscript version for posting on your own website. You may further deposit the accepted manuscript version in any repository, provided it is only made publicly available 12 months after official publication or later and provided acknowledgement is given to the original source of publication and a link is inserted to the published article on Springer's website. The link must be accompanied by the following text: "The final publication is available at link.springer.com”.

1 23

Author's personal copy Environ Sci Pollut Res DOI 10.1007/s11356-016-7468-5

ART AND CULTURAL HERITAGE

Laser-induced breakdown spectroscopy for elemental characterization of calcitic alterations on cave walls Léna Bassel 1 & Vincent Motto-Ros 2 & Florian Trichard 2 & Frédéric Pelascini 3 & Faten Ammari 1 & Rémy Chapoulie 1 & Catherine Ferrier 4 & Delphine Lacanette 5 & Bruno Bousquet 6

Received: 6 March 2016 / Accepted: 12 August 2016 # Springer-Verlag Berlin Heidelberg 2016

Abstract Cave walls are affected by different kinds of alterations involving preservative issues in the case of ornate caves, in particular regarding the rock art covering the walls. In this context, coralloids correspond to a facies with popcorn-like aspect belonging to the speleothem family, mostly composed of calcium carbonate. The elemental characterization indicates the presence of elements that might be linked to the diagenesis and the expansion of the alterations as demonstrated by prior analyses on stalagmites. In this study, we report the use of laserinduced breakdown spectroscopy (LIBS) to characterize the elemental composition of one coralloid sample with a portable instrument allowing punctual measurements and a laboratory mapping setup delivering elemental

images with spatial resolution at the micrometric scale, being particularly attentive to Mg, Sr, and Si identified as elements of interest. The complementarity of both instruments allows the determination of the internal structure of the coralloid. Although a validation based on a reference technique is necessary, LIBS data reveal that the external layer of the coralloid is composed of laminations correlated to variations of the LIBS signal of Si. In addition, an interstitial layer showing high LIBS signals for Fe, Al, and Si is interpreted to be a detrital clay interface between the external and the internal part of the coralloid. These preliminary results sustain a possible formation scenario of the coralloid by migration of the elements from the bedrock.

Responsible editor: Philippe Garrigues * Léna Bassel [email protected] Vincent Motto-Ros [email protected]

Bruno Bousquet [email protected] 1

IRAMAT-CRP2A, UMR CNRS 5060, Maison de l’Archéologie, Université Bordeaux Montaigne, Domaine Universitaire, Esplanade des Antilles, 33607 Pessac Cedex, France

2

Institut Lumière Matière, UMR 5306, Université de Lyon 1 – CNRS, Université de Lyon, 69622 Villeurbanne Cedex, France

Faten Ammari [email protected]

3

CRITT Matériaux Alsace, 19 rue de St Junien, C. S. 80023, 67305 Schiltigheim, France

Rémy Chapoulie [email protected]

4

PACEA, UMR CNRS 5199, Université Bordeaux, Bâtiment B8, Allée Geoffroy Saint Hilaire CS 50023, 33615 Pessac Cedex, France

5

I2M, UMR CNRS 5295, Université Bordeaux, 33607 Pessac, France

6

CELIA, UMR CNRS 5107, Université Bordeaux, 351 Cours de la Libération, F-33405 Talence Cedex, France

Florian Trichard [email protected] Frédéric Pelascini [email protected]

Catherine Ferrier [email protected] Delphine Lacanette [email protected]

Author's personal copy Environ Sci Pollut Res

Keywords LIBS . Laser-induced breakdown spectroscopy . Speleothems . Alterations . Ornated cave . Conservation

Introduction In the South West of France, the Vézère Valley represents an important region of prehistoric caves. This region is indeed rich with ornate caves, either painted or engraved, and represents therefore an area where rock art cave conservation is a serious issue. Curators are facing different kinds of calcitic coatings on the walls that make the underlying rock art disappear. In the present study, we focused on two types of alterations characterized by two types of facies. The first one is the so-called calcitic moonmilk, characterized by the presence of thin needle fiber calcite, which is frequently found in caves and has already been extensively studied; see for example (Verrecchia 1994, Hill and Forti 1997, Borsato et al. 2000). The second one, referred as coralloid, presents a popcorn-like aspect (Hill and Forti 1997). In order to investigate the mechanisms responsible for the genesis and development of such alterations, a cave without major archaeological relevance and containing both moonmilk and coralloid types of alterations has been selected since 2008 along the Vézère Valley area to become a laboratory cave (Lacanette et al. 2013). This cave is thus considered to be representative of the ornate cavities from the area from both morphological and geological aspects. In cave context, calcitic moonmilk has been much more studied than coralloids, essentially due to its ubiquitous nature (Curry et al. 2009, Baskar et al. 2011). Regarding its genesis mechanism, two scenarios have been proposed, namely either physico-chemical precipitation or biomineralization process due to fungi or bacteria (James 1972, Phillips and Self 1987, Borsato et al. 2000, Canaveras et al. 2006, Cailleau et al. 2009, Bindschedler et al. 2012, Sanchez-Moral et al. 2012). None of these two scenarios have been identified yet as the most likely. It should be noticed that the results reported about moonmilk are essentially based on morphological and mineralogical analyses such as scanning electron microscopy (SEM), environmental scanning electron microscopy (ESEM), and X-ray diffraction (XRD). Other analyzes involve microorganism culture experiments coupled with DNA and RNA determination, and isotopic analyzes have also been reported in order to understand the link between the carbonate precipitation and the local environmental conditions (Lacelle et al. 2004, Sanchez-Moral et al. 2012). In addition, moonmilk samples have been analyzed by Raman spectroscopy and X-ray fluorescence spectroscopy (XRF) (Martinez-Arkarazo et al. 2007, Cacchio et al. 2014). At the opposite, coralloids have been far less studied than moonmilk samples. A recent study based on stable isotope analysis (Caddeo et al. 2015) proposed a scenario for the formation of coralloids supporting the role of capillary water. This scenario was already proposed among

others like leaching from the groundwater through the bedrock (Hill and Forti 1997). In the present work, the LIBS technique (Miziolek et al. 2008, Noll 2012, Cremers and Radziemski 2013) is applied to perform qualitative elemental analysis of the coralloids present in the laboratory cave described earlier. Indeed, LIBS has been identified as a powerful analytical technique in various fields for both classification and quantitative measurements (Hahn and Omenetto 2012, El Haddad et al. 2014). In particular, LIBS has already been widely used in the field of geology either with a portable or with laboratory instrument (Fortes and Laserna 2010, McMillan et al. 2014, Spizzichino and Fantoni 2014) and a series of studies have also been reported on speleothems (Vadillo et al. 1998, Cuñat et al. 2008, Ma et al. 2010, Galbács et al. 2011). In the work presented in this paper, two different LIBS systems were used and evaluated in the laboratory, a portable device, and a laboratory instrument allowing to perform multi-elemental images with high spatial resolution. Both setups strongly differ from each other, and the aim of this study was thus to demonstrate at first the relevance of direct LIBS analyses operated with a handheld system and secondly to go into further detail with a laboratory LIBS instrument providing multi-elemental images.

Materials and methods Sample: collection and preparation A coralloid was sampled in the laboratory cave, situated in Marquay, Dordogne, France. Figure 1a is a picture of coralloids spread as white concretions on the cave walls. The sampling area is clearly visible in this image and a part of the extracted coralloid is shown in Fig. 1b. It should be emphasized that this present study is only based on one single sample and consequently doesn’t allow general conclusion. However, it clearly illustrates the complexity of the coralloid samples. The aim was to evaluate the portable system in the laboratory, allowing to make a diagnosis with surface measurements. Direct LIBS measurements operated by the portable LIBS were made on the unprepared sample. However, to allow measurements along a section in order to access to more information, as well as for imaging purposes, the sample was specifically prepared. For the measurements along a section, the sample was simply cut with a wire saw ensuring a neat section. The observation of the section of the coralloid reveals that it is composed of two distinct layers: the external part that seems laminated and the internal part made of large crystals. For imaging, the characterized cross-section was additionally embedded in an epoxy resin, cut, and finally polished with SiC paper under water solvent to obtain a planar section. This preparatory step was necessary prior to LIBS imaging analysis in order to get good sample flatness and remove

Author's personal copy Environ Sci Pollut Res Fig. 1 a Picture of a cave wall from the laboratory cave covered by coralloids showing the sampling area b Zoom in on the coralloid extracted from the wall displayed in a

any potential surface contaminations resulting from cutting or handling. Portable LIBS system A commercial portable LIBS system (EasyLIBS from IVEA Solution, France) was used to perform direct analysis of the coralloid sample. This equipment was composed of a Nd:YAG laser at 1064 nm operating at 1 Hz and delivering 5 ns pulses with 43 mJ per pulse. LIBS measurements were performed in ambient conditions (air and pressure) without contact between the hand-piece and the sample thanks to a focusing lens of 8-cm focal length. Thus, a 250-μmdiameter area was analyzed after laser ablation and plasma expansion. The following emitted light was then collected through a bundle of optical fibers connected to three compact Czerny-Turner spectrometers, and finally, a LIBS spectrum— resulting in the concatenation of the three spectra—was recorded in the range 200–1000 nm. The spectrometers were not time-resolved and a gate of 7 ms was set. Ten laser shots were recorded at the same point on the sample surface and the average spectrum was calculated. LIBS imaging instrument The instrument used to provide LIBS images has already been described in details elsewhere (Motto-Ros et al. 2014, Sancey et al. 2014) and only its main components are described in the following. The instrumental setup was based on a homemade optical microscope that combined a LIBS laser injection line, a standard optical-imaging apparatus and a three-axis motorized platform used for the sample positioning. The laser source was a Nd:YAG emitting 8 ns pulses at 1064 nm with a 100 Hz repetition rate (Centurion GRM, Quantel) and a typical pulse energy of 40 mJ. A computer-controlled attenuator (ATT1064, Quantum Composers) was used to reduce the laser energy at typically 1.5 mJ. Laser pulses were focused

onto the sample by a ×15 magnification objective (LMM15X-P01, Thorlabs) resulting in a crater size close to 8 μm. During the experiments, the sample could be translated along three axes by a xyz motorized stage with a travel range of 50 mm in each direction and a maximum speed of 3 mm/s. The measurements were performed at room temperature and under ambient pressure conditions. The light emitted by the plasma plume was collected by a quartz lens and focused onto the entrance of an optical fiber bundle composed of 19 fibers with a 200-μm core diameter each, with a round to keyed linear arrangement. The output of the fiber bundle was connected to a Czerny-Turner spectrometer (Shamrock 500, Andor) equipped with a 600-l/mm grating blazed at 300 nm and an intensified camera (iStar 340T, Andor Technology). The intensified camera was synchronized with the Q-switch of the laser and the spectrum acquisition was performed with a delay of 900 ns, a gate of 2 μs, and a gain of 1350. The width of the entrance slit of the spectrometer was set to 50 μm. In this configuration, a spectral range of 80 nm was accessible with a spectral resolution of approximately 0.2 nm. The homemade software, developed in the LabVIEW environment, controlled the entire system and allowed for the performance of automated sequences to scan the region of interest of the sample with a specific lateral resolution.

Results and discussion Portable LIBS analysis The measurements performed with the portable LIBS system were done on the raw sample of coralloid, without any preparation. Considering that the coralloid sample was mostly composed of calcium carbonate, it was not surprising to observe in the LIBS spectra the saturation of several atomic lines related to calcium. These lines were removed from data analysis during a preprocessing step since they do not carry


information. Once a LIBS spectrum of coralloid was recorded, three spectral windows were selected, thanks to the atomic lines they contained. They are displayed in Fig. 2a–c after background subtraction of the spectrum. Five elements other than calcium were detected and identified: magnesium, silicon, iron, aluminum, and strontium. The related spectral lines were identified from the NIST database: 279.55 nm (Mg II), 280.27 nm (Mg II), and 285. 21 nm (Mg I) for magnesium; 288.16 nm (Si I) for silicon (Fig. 2a); 302.10 nm (Fe I) for iron; 308.22 nm (Al I) and 309.27 nm (Al I) for aluminum (Fig. 2b); and finally 460.73 nm (Sr I) for strontium and 468.45 nm (Ca I) for calcium (Fig. 2c). It should be noticed that strontium and magnesium have been considered as elements of major interest in the composition of speleothems (Fairchild et al. 2001, Borsato et al. 2007, Sinclair et al. 2012). Actually, the presence of strontium is directly linked to the growth rate of speleothems and has been consequently used as an indicator of the growth speed (Fairchild et al. 2001). In addition, strontium is also probably more favorable to the orthorhombic aragonite structure than to the calcite structure when precipitation occurs (Fairchild et al. 2001) while magnesium is considered to be a good indicator of hydrology in the caves responsible for the inhibition of calcite growth (Fairchild et al. 2001, White 2012). Finally, this direct LIBS analysis gave access to the list of elements detected on the surface of the coralloid. Furthermore, to validate quantitative elemental analysis based on LIBS data, the samples have to be analyzed by a reference technique. However, quantitative analysis is out of the scope of this work and only qualitative or semi-quantitative results are discussed. Analyzing the sample along a section was necessary to retrieve additional information about its in-depth properties and thus to enhance our understanding of genesis and growth of coralloids. In particular, in-depth profiling allowed to highlight the variation of the LIBS signals related to strontium and magnesium, which are two elements of major interest that were identified from surface experiment. Thus, 17 LIBS measurements were performed with the portable instrument along a transect axis aligned with the growth direction, as described in Fig. 3. The spacing between two LIBS analyses was fixed

to 0.5 mm and the location of the transect was arbitrarily chosen. The first LIBS measurement was performed at the sample surface, the second 0.5 mm deeper and so on, until the measurement n°17. In Fig. 3, one can clearly deduce the existence of two materials in addition to the underlying host rock, with a transition occurring at the point n°9, namely the external layer (points n°1 to n°8) and the internal layer (points n°10 to n°16). At each point, 10 LIBS measurements were made and the average spectrum was analyzed. More precisely, for each of the six elements under study, namely Ca, Si, Al, Mg, Sr, and Fe, one spectral line was selected and the signal was extracted, after background subtraction, as the intensity of the peak. Applying the same data processing to each of the 17 points of the transect, we obtained the profiles presented in Fig. 4. These profiles were normalized to unity and offsets were applied to make the reading easier. It should be mentioned that the data related to the point n°1 were not reported in Fig. 4 because they were not considered as reliable, due to the proximity to the edge of the sample. Considering that the coralloid sample is mostly composed of calcium carbonate, the variations of peak intensities in the LIBS spectra displayed in Fig. 4 can reasonably be linked to the variations of concentrations of the analytes. This assumption should be supported by a strict validation based on the use of a reference method in order to be able to provide quantitative results. Matrix effect could indeed play a significant role (Hahn and Omenetto 2012) and should be evaluated in order to draw quantitative conclusions (Motto-Ros et al. 2013, Negre et al. 2015). However, based on the work of Hill and Forti (Hill and Forti 1997) and more recently on some XRD analyses showing aragonite and calcite (Caddeo et al. 2015), we consider the assumption stating that the main composition of coralloids is calcium carbonate as correct. That’s why partial conclusion can be drawn regarding Fig. 4. Figure 4a clearly reveals the existence of an interface by the increase of the LIBS signal for Al, Fe, and Si corresponding to a detrital clay layer. This layer is schematically delimited by the two vertical dotted lines in Fig. 4a, separated by 1 mm. However, according to Fig. 3, this thickness appears to be overestimated due to

Fig. 2 Selected spectral windows of a LIBS spectrum of the coralloid sample displaying atomic line raw intensities of Mg and Si (a), Fe and Al (b), and Sr and Ca (c) after background subtraction

Author's personal copy Environ Sci Pollut Res Fig. 3 Cross-section of coralloid and location of the transect along which 17 LIBS measurements were performed (black circles). Three phases were identified as external layer (1), internal layer (2), and underlying host rock (3)

the sampling step of 0.5 mm. From Fig. 4a, the interface was found to be located at 4 mm from the external limit of the sample, which confirms what we can see on the photo displayed in Fig. 3. Consistently, the Ca signal is slightly lower at the same depth of 4 mm inside the sample (Fig. 4b). In addition, the LIBS intensities of Si and Sr appeared to be slightly higher in the external layer (upper part in Fig. 3) compared to the internal one (lower part in Fig. 3). The LIBS intensity of Mg is also higher in the external layer than in the internal one, and the difference between these intensity values observed in the two layers is even more pronounced for Mg than for Sr or Si. Furthermore, these LIBS measurements being performed on 250-μm-diameter surface areas, they are potentially sensitive to sample heterogeneity. This could explain non-constant trends observed in each layer on the profiles along the sample section, as it is the case for Si around 2mm depth for instance. On the contrary, if heterogeneities smaller than the laser spot are present, they might be lost in an average signal and thus not revealed by the LIBS analysis as discussed in the recent work of Kuhn et al. (2016) and recently evaluated in numerous articles (De Carvalho et al. 2015, Sheta et al. 2016). This LIBS analysis along the transect suggests probable variations of concentrations of the analytes along the growth axis. However, the spatial resolution of these series of LIBS measurements along the transect did not allow to push the conclusion further about the characteristics of the coralloid sample. Consequently, additional LIBS analysis with better spatial resolution in the two directions, namely parallel and perpendicular to the transect, was expected to be Fig. 4 LIBS signal evolution— raw intensities after background subtraction—along the growth axis of the coralloid. The profiles were normalized to unity and offsets were applied to facilitate the reading. Profiles of Al, Fe and Si (a) and Ca, Mg and Sr (b)

helpful to consolidate this result obtained after a 1D scan along the transect direction. Also, it would allow to build point-by-point the elemental image of the cross-section of this coralloid sample. This is precisely the purpose of the next section.

Imaging with laboratory LIBS mapping system To investigate in more details the variation of the LIBS signals related to the elements of interest, elemental maps were acquired with the LIBS instrument described in the BLIBS imaging instrument^ section. An advanced spectrum treatment was developed to perform a fast extraction of the intensity value for each measurement site of the sample and for each element of interest. A single emission line was selected for each element, and the algorithm defined a baseline fit using a polynomial function and subtracted it from the raw emission signal. A two-dimension matrix was then provided for each element. Each cell of this matrix represented the intensity signal from one point on the sample surface for the given element and could then be displayed, as it is the case here, using gray scale to present a visual result in the form of elemental images. Generally speaking, elemental mapping is a powerful tool and its application to speleothem analysis has already been successful, either with micro-XRF (Borsato et al. 2007, Devès et al. 2012) or with micro-LIBS (Ma et al. 2010). In a work presented by Devès et al. (2012), the authors highlight the advantages of this methodology by combining micro-


XRF and autoradiography to better characterize the genesis of speleothems, focusing in particular on U, Sr, Si, and Ca. In Fig. 5, the most relevant elemental images for our study are presented, namely the images extracted from the peaks of Sr, Mg, and Si. Other elemental images (not shown) have not only been extracted for Ca (300.69 nm), Fe (302.06 nm), and Al (309.27 nm) but also for Ba (455.40 nm) and Ti (453.32 nm). For Si and Mg images, the same parameter settings were used since the lines of interest at 285.21 nm (Mg I) and at 288.16 nm (Si I) were recorded in the same spectral range covered by the spectrometer: 240–328 nm. Laser energy was equal to 1.5 mJ; the spacing between two points and crater size were close to15 and 8 μm, respectively and thus, the images depicted 1650 × 1100 = 1,815,000 individual laser shots (i.e. pixels). In order to measure the intensity of the Sr line at 407.77 nm, the sample was polished a second time and

Fig. 5 Elemental images of a coralloid sample by LIBS. a Mg image at 285.21 nm. b Si image at 288.16 nm. c Sr image at 407.77 nm. The intensity signal displayed corresponds to the raw emission signal after background subtraction. Intensity threshold calibration bars display the minimum value in black corresponding to the background level and the maximum value in white

a second LIBS experiment was conducted with the same instrumental parameters but with the spectrometer set in the spectral range 400–480 nm. Before inspecting the results, we should underline that the porosities are appearing in black in all elemental images since no signal was collected from these points. Thus, we can easily recognize an area in the external layer corresponding to an absence of matter. The first observation we can make is that the LIBS signal intensity for each element is weaker in the internal layer than in the external one, confirming the trends obtained from the transect measurements (cf. Fig. 4). Indeed, if we now consider the elemental images displayed in Fig. 5, intensity ratios close to 3 for Mg (Fig. 5a) and close to 7 for Sr (Fig. 5c) were obtained between the two layers. Here again, it should be underlined that these ratios do not take any potential matrix effect into account. It has been demonstrated that the presence of Sr and Mg in the cave waters was mainly due to bedrock dissolution (Fairchild et al. 2006). Our data are in line with this statement since in all the images presented in Fig. 5, the underlying host rock which is calcareous in the laboratory cave contains all the elements detected in the external layer. Soil leakage could also have played a role in the presence of Sr as revealed by isotopic data (Fairchild and Treble 2009, Bourdin et al. 2011). Even though, it is remarkable that the internal layer seems to be more depleted in the elements of interest. This could be the result of migration of the elements from the base of the coralloid to the external limit in contact with air. Indeed, this internal layer appears to be made of relatively pure calcite crystals. Regarding the spatial distribution of the LIBS signal of Si (Fig. 5b), it particularly reflects the presence of a series of laminations linked to higher LIBS signal of Si. This observation about Si is in good agreement with the result presented in Fig. 4a revealing a significant variation of the LIBS signal of Si at 4 mm from the external surface of the sample. The fact that we observed laminations in Fig. 5 is not surprising since it is consistent with the supposed formation process by successive precipitation events at the interface with the ambient air of the cave. The high contrast in LIBS signal of Si observed at the interface between the two layers in Fig. 5b suggests the presence of a detrital clay interface richer in Si than the calcium carbonate around it. However, this possible enrichment in Si at the interface should be supported by other measurements based on a reference technique to be definitely stated, as already discussed. Even if not shown here, a brief comment on the other LIBS images from our experiment can be done. In particular, LIBS images related to the elements Fe and Al clearly confirm the existence of an interstitial thin layer between the two materials of calcium carbonate. LIBS signal of Fe, Al, and Si was found to be all maximum at the same points of the sample. This discontinuity observed between two calcium carbonate materials is probably the result of a detrital deposit during a period of growth interruption. Furthermore, some elements such as


Ba and Ti have also been detected and localized in very small grains inside the underlying host rock but they were not detected elsewhere. Finally, considering all these results, we promote the hypothesis stating that the elements entering in the composition of the coralloid could come from the bedrock. Therefore, we think that a porosities network in the coralloid, i.e., in the two layers reported here, should be present to facilitate the transport of the elements from the bedrock to the external layer of the coralloid.

sustains the hypothesis of migration of these elements from the bedrocks to the external layer of the coralloid. Further works will thus focus on the study of porosity in order to discuss this hypothesis of migration through the material. Acknowledgments This study has been carried out with financial support from the Conseil Régional d’Aquitaine and the French State, managed by the French National Research Agency (ANR) in the frame of Bthe investments for the future^ Programme IdEx Bordeaux. The authors are also grateful to Mr. Delbos, landowner of the Leye cave, namely the laboratory cave.

Conclusion References In this work, we applied LIBS in both point and imaging modes to characterize a coralloid sample. First, from the point measurements acquired with a portable instrument, we can conclude that the two layers of the coralloid, already identified when observed with the naked eye, show two levels of LIBS signals. It might be interpreted as a juxtaposition of two chemically different materials but a validation based on a reference technique is required to support this hypothesis. In addition, the LIBS data recorded point-by-point evidenced significant heterogeneity of the materials composing the coralloid. Secondly, LIBS imaging was realized with the use of a micro-LIBS imaging setup offering spatial resolution at the micrometric scale. Although a portable LIBS system only gives access to a reduced knowledge about the samples, we underline that its use was particularly pertinent in the framework of cave wall characterization, since magnesium, silicon, and strontium were not detected on moonmilk samples and are thus strongly related to coralloid samples. This result is interesting since in the laboratory cave described here, moomilk and coralloids can be observed in close vicinity (a few meters) and appearing on the same bedrocks. Moreover, direct on-site measurements performed thanks to the use of a portable LIBS system are expected to allow collecting hundreds of data inside the caves when only very few selected samples are to be extracted and analyzed in the laboratory. It offers the possibility to perform smart sampling by selecting with more accuracy the samples that should be analyzed at the laboratory. In addition, it allows obtaining on-site important information about variation of LIBS signal and potentially about the elemental composition of the samples. However, in the present study, the observed variations of the LIBS signal could not be linked to the variation of element concentrations of the materials. Indeed, a validation based on a reference technique is required to conclude about concentration and such validation will be part of our future work. Finally, LIBS imaging applied to the cross-section of the sample of coralloid provided very interesting results and in particular revealed the existence of laminations correlated to variations of the LIBS signal of silicon. LIBS imaging also revealed that all the elements detected inside the coralloid were present in the bedrocks, which

Baskar S, Baskar R, Routh J (2011) Biogenic evidences of moonmilk deposition in the Mawmluh Cave, Meghalaya, India. Geomicrobiol J 28(3):252–265 Bindschedler S, Millière L, Cailleau G, Job D, Verrecchia EP (2012) An ultrastructural approach to analogies between fungal structures and needle fiber calcite. Geomicrobiol J 29(4):301–313 Borsato A, Frisia S, Jones B, Van der Borg K (2000) Calcite moonmilk: crystal morphology and environment of formation in caves in the Italian Alps. J Sed Res 70(5):12 Borsato A, Frisia S, Fairchild IJ, Somogyi A, Susini J (2007) Trace element distribution in annual stalagmite laminae mapped by micrometer-resolution X-ray fluorescence: implications for incorporation of environmentally significant species. Geochim Cosmochim Ac 71(6):1494–1512 Bourdin C, Douville E, Genty D (2011) Alkaline-earth metal and rareearth element incorporation control by ionic radius and growth rate on a stalagmite from the Chauvet Cave, Southeastern France. Chem Geol 290(1-2):1–11 Cacchio P, Ferrini G, Ercole C, Del Gallo M, Lepidi A (2014) Biogenicity and characterization of moonmilk in the Grotta Nera (Majella National Park, Abruzzi, Central Italy. J Cave Karst Stud 76(2):15 Caddeo GA, Railsback LB, De Waele J, Frau F (2015) Stable isotope data as constraints on models for the origin of coralloid and massive speleothems: the interplay of substrate, water supply, degassing, and evaporation. Sediment Geol 318:130–141 Cailleau G, Verrechia EP, Braissant O, Laurent E (2009) The biogenic origin of needle fiber calcite. Sedimentology 56(6):1858–1875 Canaveras JC, Cuezva S, Sanchez-Moral S, Lario J, Laiz L, Gonzalez JM, Saiz-Jimenez C (2006) On the origin of fiber calcite crystals in moonmilk deposits. Naturwissenschaften 93(1):27–32 Cremers, D. A., L. J. Radziemski (2013). Handbook of laser-induced breakdown spectroscopy. Wiley Cuñat J, Fortes FJ, Cabalín LM, Carrasco F, Simón MD, Laserna JJ (2008) Man-portable laser-induced breakdown spectroscopy system for in situ characterization of karstic formations. Appl Spectrosc 62(11):1250–1255 Curry MD, Boston PJ, Spilde MN, Baichtal JF, Campbell AR (2009) Cottonballs, a unique subaqeous moonmilk, and abundant subaerial moonmilk in Cataract Cave, Tongass National Forest, Alaska. Int J Speleol 38(2):17 De Carvalho GGA, Santos D Jr, Da Silva Gomes M, Nunes LC, Guerra MBB, Krug FJ (2015) Influence of particle size distribution on the analysis of pellets of plant materials by laser-induced breakdown spectroscopy. Spectrochim Acta B105:130–135 Devès G, Perroux A-S, Bacquart T, Plaisir C, Rose J, Jaillet S, Ghaleb B, Ortega R, Maire R (2012) Chemical element imaging for speleothem geochemistry: application to a uranium-bearing corallite

Author's personal copy Environ Sci Pollut Res with aragonite diagenesis to opal (Eastern Siberia, Russia). Chem Geol 294-295:190–202 El Haddad J, Canioni L, Bousquet B (2014) Good practices in LIBS analysis: review and advices. Spectrochim Acta B 101:171–182 Fairchild IJ, Treble PC (2009) Trace elements in speleothems as recorders of environmental change. Quaternary Sci Rev 28(5-6):449–468 Fairchild IJ, Baker A, Borsato A, Frisia S, Hinton RW, McDermott F, Tooth AF (2001) Annual to sub-annual resolution of multiple traceelement trends in speleothems (English). J Geol Soc London 158(5): 831–841 Fairchild IJ, Smith CL, Baker A, Fuller L, Spötl C, Mattey D, McDermott F, E.I.M.F (2006) Modification and preservation of environmental signals in speleothems. Earth Sci Rev 75(1-4):105–153 Fortes F, Laserna J (2010) The development of fieldable laser-induced breakdown spectrometer: no limits on the horizon. Spectrochim Acta B 65:975–990 Galbács G, Kevei-Bárány I, Szőke E, Jedlinszki N, Gornushkin IB, Galbács MZ (2011) A study of stalagmite samples from Baradla Cave (Hungary) by laser induced plasma spectrometry with automatic signal correction. Microchem J 99(2):406–414 Hahn DW, Omenetto N (2012) Laser-induced breakdown spectroscopy (LIBS), part II: review of instrumental and methodological approaches to material analysis and applications to different fields. Appl Spectrosc 66:347–419 Hill, C. A. and P. Forti (1997). Cave minerals of the world, National speleological society James NP (1972) Holocene and Pleistocene calcareous crust (caliche) profiles: criteria for subaerial exposure. J of Sediment Res 42(4): 817–836 Kuhn K, Meima JA, Rammlmair D, Ohlendorf C (2016) Chemical mapping of mine waste drill cores with laser-induced breakdown spectroscopy (LIBS) and energy dispersive X-ray fluorescence (EDXRF) for mineral resource exploration. J Geochem Explor 161:72–84 Lacanette D, Large D, Ferrier C, Aujoulat N, Bastian F, Denis A, Jurado V, Kervazo B, Konik S, Lastennet R, Malaurent P, Saiz-Jimenez C (2013) A laboratory cave for the study of wall degradation in rock art caves: an implementation in the Vézère area. J Archaeol Sci 40(2):894–903 Lacelle D, Lauriol B, Clark ID (2004) Seasonal isotopic imprint in moonmilk from Caverne de l'Ours (Quebec, Canada): implications for climatic reconstruction. Can J Earth Sci 41(12):1411–1423 Ma QL, Motto-Ros V, Lei WQ, Boueri M, Zheng LJ, Zeng HP, BarMatthews M, Ayalon A, Panczer G, Yu J (2010) Multi-elemental mapping of a speleothem using laser-induced breakdown spectroscopy. Spectrochim Acta B 65(8):707–714 Martinez-Arkarazo I, Angulo M, Zuloaga O, Usobiaga A, Madariaga JM (2007) Spectroscopic characterisation of moonmilk deposits in

Pozalagua tourist cave (Karrantza, Basque Country, North of Spain). Spectrochim Acta A Mol 68(4):1058–1064 McMillan NJ, Rees S, Kochelek K, McManus C (2014) Geological applications of laser-induced breakdown spectroscopy. Geostand and Geoanal Res 38(3):329–343 Miziolek, A.W., V. Palleschi and I. Schechter (2008). Laser induced breakdown spectroscopy. Cambridge University Press Motto-Ros V, Sancey L, Wang XC, Ma QL, Lux F, Bai XS, Panczer G, Tillement O, Yu J (2013) Mapping nanoparticles injected into a biological tissue using laser-induced breakdown spectroscopy. Spectrochim Acta B 87:168–174 Motto-Ros V, Negre E, Pelascini F, Panczer G, Yu J (2014) Precise alignment of the collection fiber assisted by real-time plasma imaging in laser-induced breakdown spectroscopy. Spectrochim Acta B 92:60– 69 Negre E, Motto-Ros V, Pelascini F, Lauper S, Denis D, Yu J (2015) On the performance of laser-induced breakdown spectroscopy for quantitative analysis of minor and trace elements in glass. J Anal At Spectrom 30:417–425 Noll R (2012) Laser-induced breakdown spectroscopy: fundamentals and applications. Springer, Berlin Heidelberg Phillips SE, Self PG (1987) Morphology, crystallography and origin of needle fiber calcite in quaternary calcretes of South Australia. Aust J Soil Res 25(4):429–444 Sancey L, Motto-Ros V, Busser B, Kotb S, Benoit JM, Piednoir A, Lux F, Tillement O, Panczer G, Yu J (2014) Laser spectrometry for multielemental imaging of biological tissues. Sci Rep 4:6065 Sanchez-Moral S, Portillo MC, Janices I, Cuezva S, Fernández-Cortés A, Cañaveras JC, Gonzalez JM (2012) The role of microorganisms in the formation of calcitic moonmilk deposits and speleothems in Altamira Cave. Geomorphology 139-140:285–292 Sheta SA, Di Carlo G, Ingo GM, Harith MA (2016) Surface morphology study of some Cu–Ni reference alloys using laser induced breakdown spectroscopy. Mater Chem Phys 173:516–523 Sinclair DJ, Banner JL, Taylor FW, Partin J, Jenson J, Mylroie J, Goddard E, Quinn T, Jocson J, Miklavič B (2012) Magnesium and strontium systematics in tropical speleothems from the Western Pacific. Chem Geol 294-295:1–17 Spizzichino V, Fantoni R (2014) Laser induced breakdown spectroscopy in archeometry: a review of its application and future perspectives. Spectrochim Acta B 99:201–209 Vadillo JM, Vadillo I, Carrasco F, Laserna JJ (1998) Spatial distribution profiles of magnesium and strontium in speleothems using laserinduced breakdown spectrometry. J Anal Chem 361:119–123 Verrecchia EPVKE (1994) Needle-fiber calcite: a critical review and a proposed classification. J of Sediment Res 64(3) White W (2012) Speleothem microstructure/speleothem ontogeny: a review of Western contributions. Int J of Speleol 41(2):329–358