Analysis of speleothems by electron and X-ray microprobes

Abstract

A piece of stalagmite from a cave in Winthertore, Belgium, was analysed by the use of capillary tabletop and synchrotron versions of X-ray microprobes and then by the electron microprobe. The potential of these microprobes was tested for the first time in the study of such specific periodic objects. Indirect measurements of the morphology/density of samples were made by the use of Compton and Rayleigh scattered radiation in the X-ray tabletop microprobe and a selected channel in white scattered radiation in the X-ray synchrotron microprobe. In the electron microprobe, the linear profile extracted from the grey-scale transformation of the secondary electron image was used for the same purpose. The elemental analyses were superimposed on the density characteristics of the samples. The inclusions of iron and silicon, measurements of the magnesium/calcium/strontium/barium ratio, profiles of anionic species and the noticeable presence of iodine were among the most important findings. Some of the measured parameters probably allow the creation of thermometric scales for the potential estimation of the climatic conditions during the deposition of the calcite material.

---

Introduction

Speleothems (stalactites, stalagmites and similar materials) are among the most interesting natural periodic deposits. Essentially, they are composed of calcite. Additions of other calcium carbonate phases, namely aragonite and dolomite (the latter with a magnesium contribution) are sometimes noticeable. Calcite is deposited from slowly flowing water by the evaporation of the solvent from the surface. Together with the evaporation phenomena, there is a significant loss of carbon dioxide to the atmosphere. This implies the transformation of the acid form of calcium carbonate into a neutral form, adding a new portion of insoluble calcite. The level of the saturation/supersaturation of calcium carbonate in the water solutions is crucial for the formation of the deposits. The calcite structure is not an everlasting one since the weathering process leads in the opposite direction to that of speleothem growth. Climatic events such as glaciation are the time when disruption of the deposits occurs. The greater part of the older deposits is only preserved in debris.

Speleothems have been intensively investigated,¹ but mainly from the geological, mineralogical and petrological points of view. Sometimes, the archaeometric application of travertines is possible.² Chemical research has been of minor significance up to now, probably because only devices of the microprobe type could add some new knowledge to that already existing on stalagmites. The main direction of chemical studies was aimed at tracing isotopic ratios, as they are essential both for dating and for scaling of the temperature.^3–8 For some reason, interest in investigations of the elemental composition has been low. However, the results on this aspect presented in several papers were very promising⁹ and seemed to be important and complementary to the isotopic data.¹⁰ Especially the investigation of the contributions of magnesium^11,12 and strontium^13–16 to the composition and the interrelations between these elements and the calcium content were of significance.^17–19 It has been established that a 1 °C increase in water temperature above the initial level of 10 °C gives rise to about a 7% increase in the Mg content in the deposit, if the primary Mg/Ca ratio is ∼1 and the level of magnesium in the water is constant.²⁰ The temperature dependence of the exchange rate of Mg with calcite is much higher than for most other elements, especially Sr and Ba.¹⁷ The uptake of strontium is complicated by the fact that for lower concentrations of Sr in water (<0.03 mmol dm⁻³), this element is built into the structure of calcite, whereas at higher concentrations, the different strontianite structure is formed.

Many side factors can influence both isotopic (e.g., ¹⁸O, ¹³C) and elemental (e.g., Mg, Sr) contents of the speleothems, e.g., exchange of isotopes and elements between water and rocks, variations in the ground water temperature, variations of the surface temperature during deposition, distance between the places of the water supply and water evaporation, additional sources of water and even moisture level in the cave and the potential influence of the ocean/sea water (the last factor can probably be traced by the detection of iodine signals).

The growth rate of stalagmites is widely variable, but it can be estimated to be in the range from ∼0.1 mm per year, as is the case in the UK, up to >1 mm per year. The growth rate can be modelled.²¹ This makes stalagmites an ideal material for microprobe analysis. Moreover, this growth is frequently deposited in the laminar structures, observable even with the naked eye. In many cases, the laminar structure can be assigned to the annular periodicity, then the correlation of the chemical data and the optical image with some additional factors such as seasonal temperature, precipitation and pressure is possible.

Instrumentation

The samples were analysed with the use of several microanalytical devices. We were motivated by the success of our earlier investigations^22–24 of wood and petrified wood samples, where microprobe methods proved very useful in differentiating the ring structure of the wood through density measurements. Moreover, the analytical signals of the elements were traced in connection with the internal structure of the material. Thus, the general image of the sample could be obtained, where the characteristic signals of the elements were superimposed on the variability of the density of the material. Although the building material for speleothems (calcite) is different from that for petrified wood (silica), the periodic, laminar structure is clearly observable in both types of objects.²⁵

The first of the microprobes applied was IMIX (Integrated Microanalyzer for Imaging and X-Ray), made by Princeton Gamma-Tech, at the Department of Chemistry of Antwerp University (UIA). It was a scanning electron microscope supplied with an energy-dispersive Si(Li) detector. The working voltage for the electron gun was always set at 20 kV. The line profile analysis option was selected to make linear scans through the samples. This option was most appropriate to match a material such as stalagmite, in which the most important information is included in the radial and longitudinal profiles. The searched field of view was selected using the image obtained from the secondary electron signals. The image extracted from the secondary electron signals was very important, because the topographical profile of the sample was essential for coupling the chemical information with surface details. The line scan was made across the centre of the image. The radial direction was followed from the middle of the stalagmite outwards. The secondary electron image of the sample was stored to ensure that possibly the same location would be used in the next measurements on the same sample. This image was also used in our analysis in another way. A grey-scale level distribution along the line of the scan was isolated. The grey-scale morphology diagram in its raw form was difficult to analyse but, after smoothing of data points, the internal structure of the scanned area was revealed. All the samples prepared for analysing with the scanning electron microprobe were thick in the spectrometric sense. The 20 keV electrons used in this kind of microprobe were able to penetrate into calcite to a depth of about 5 µm, which was a much shallower layer than any stable calcite strip that could be safely prepared.

A tabletop X-ray capillary microprobe was the second device used. The probe was constructed in the Department of Chemistry of the University of Antwerp (UIA) and has been described elsewhere several times.^26,27 A Siemens rotating anode Mo tube was used for excitation. The inlet of the conical capillary intercepted only a small part of the tube output and this part was applied to the microprobe analysis. The X-ray beam was squeezed to the size of the outlet capillary diameter, i.e., 15 µm. Since the distance from the outlet to the sample was relatively long (about 1.5 mm), the effective beam size on the sample was at least doubled in relation to the above value. The sample was driven in front of the capillary outlet by a set of four computer-controlled step motors. A linear scan was again the preferred option; thus only one motor was effectively working, except for the moment of the initial positioning of the sample. An Si(Li) detector with an energy resolution of 160 eV as measured for the Mn Kα line was applied for the detection of the fluorescent and scattered radiation. The samples, although thinned in comparison with the original material, still had a thickness of about 4 mm and were in the limits of medium thickness category for transmission and infinite for reflection type measurements; thinner samples became brittle, whereas thicker samples led to excessive scattered radiation. The analysed surface was polished.

Synchrotron measurements were performed in Beamline L at the 4.45 GeV positron synchrotron in HASYLAB, Hamburg, and represented the third kind of microprobe. The primary beam was attenuated by the introduction of an 8 mm thick Al sheet and was finally derived by a 10 cm long straight capillary of 30 µm id. The energy range of the primary polychromatic beam was between 1 and 100 keV. The secondary radiation was registered with an HPGe detector. The fluorescent lines of several elements present in the samples and also the selected channel in the scattered white synchrotron radiation (60–70 keV) were chosen for the collection of the data. Our data belong to the first results obtained on such kinds of samples by the use of a synchrotron, although we can cite some other work performed in this facility.²⁸

All the scans with the application of the mentioned methods were carried out on the same samples and at locations as close to one another as possible in order to obtain comparable results. Still, since the measurements were independent and made at different times and places, it was difficult to follow the same positions on the sample with an accuracy better than 100 µm. The runs of the scans covered the whole radial distance from the middle to the edge of the stalactite cross-section and were comparable with the exception of the synchrotron-based measurements. The latter were shortened to about one third of the whole distance owing to the short measuring time available.

The necessity for using three kinds of microprobes for the analysis of the speleothems results from the different characteristics of the described devices. We can consider an electron microprobe as a device with excellent spatial resolution, of the order of micronmetres in our experiments, hardly penetrating the samples below the surface (up to 5 µm) and with a good capability for detecting the low-Z elements. The weakest point of this apparatus is its high detection limit for all elements, restricted by the huge bremsstrahlung background. On the other hand, one can apply part of the bremsstrahlung radiation for the estimation of the matrix features, e.g., the density. For both X-ray microprobes the spatial resolution is limited by the diameter of the capillary outlet and additionally increased by at least the order of millimetres distance between the outlet and the sample. The beam widens significantly within that space.²⁹ In general, the beam size on the sample is within the limits 20–30 µm. The penetration depth of the beam is incomparably greater than for the electron microprobe and can be estimated in centimetres for transmission measurements. Our samples were even slightly transparent for the beam. Nevertheless, the information depth in reflection type measurements is governed by the escape depth of the characteristic radiation and equals 18 µm for Si, 56 for Fe and 97 for Ca. Both X-ray capillary microprobes are very sensitive when determining heavy elements and can operate without a vacuum. Hence the electron microprobe brings results that are strictly surface-oriented, whereas the X-ray microprobes give a much deeper insight into the sample. The low-Z element results from EPMA measurements are supplemented by high-Z element determinations from X-ray microprobes.

Pieces of the stalactite were initially examined and selected for other experiments by the use of an Eclipse E400 optical microscope with a Coolpix 950 digital camera, manufactured by Nikon Europe (The Netherlands). Our sample was prepared in one only available manner; we cut transversely the piece that we had. The longitudinal cross-section was not available as the amount of the stalagmite was not sufficient. A laminar structure was clearly observed on the surface of the sample and could be analysed in the radial direction. The image-processing program Screen Measurement G was applied for the analysis of surfaces and for the studies of surface morphology. The results of the morphological linear profile option, presented as the grey-scale levels versus the distance, could be used as a substitute for the optical scanning analysis³⁰ and compared with the results from other kinds of microprobes. This was an interesting option, since it allowed the comparison of the auxiliary signals from different microprobes: selected bremsstrahlung channel and secondary electron signal in the electron microprobe; Rayleigh and Compton scattered signals in the tabletop X-ray microprobe; and selected white radiation channel in the synchrotron-based X-ray microprobe with the optical signals. The last could even be observed with the naked eye in the optimal situation. Moreover, in this indirect way, the optical signals could be coupled with a fluorescent signal of the most important constituent, Ca. Some efforts to couple the different scattered signals from different analytical methods are sometimes made.³¹

The UV fluorescence method, although similar in its assumptions and advocated by some researchers,^32,33 was not applied in this study.

Results

Each of the microprobes detected the signal of Ca, the most important element of the matrix. There were different auxiliary signals taken from each microprobe, which essentially should describe the topographic details and density variations of the samples. The detection of other elements was dependent on the abundance of the element in the matrix and the specific analytical abilities of the microprobes. One should remember that the electron microprobe allows one to detect lighter elements much better than other devices. We registered the calcium and iron signals very easily by the use of this microprobe. They were by far the main and the very common elements. Less common were Mg, Si, P, S, Cl, Br and I. The Ar signal was very specific—it was an artefact, resulting from the situation that air was present in some openings, splits or holes in the boundary regions of the stalagmite. There were very characteristic patterns of distribution of the mentioned elements. The calcium content showed the boundaries of the laminar structure of the sample that could be observed in parallel in the linear grey-scale profile extracted from the secondary electron image of the sample (Fig. 1). These boundaries correspond to the very sharp minima in the calcium signal. Owing to the statistical fluctuations, there are many minima in the Ca spectrum and one would have difficulty recognising which of them are the boundaries in the structure and which are the fluctuations. In that case it is helpful to have a second type of distribution. We can observe some elements as inclusions. Si inclusions corresponded very well to the boundaries in the Ca (Fig. 2) and secondary electron profiles, as did part of the iron inclusions also. The Ar signals also matched the above-mentioned minima. Concerning the last case, we can claim that the material at the minima has a loosened structure, partially filled with air. Another part of the iron inclusions corresponded well with the bromine inclusions, suggesting that both elements were associated in one compound (FeBr₃?; speciation studies were not carried out). The distributions of magnesium and sulfur were slightly different from that of Ca; however, they had also some zone structures and both elements matched each other fairly well. The profiles of phosphorus and chlorine were very similar to each other, but slightly dissimilar to the previously mentioned ones. Summarising the electron microprobe investigation, we can claim that the distribution pattern of calcium carbonate determined the laminar structure of stalagmite. Moreover, the boundaries were emphasised by the presence of the silicon (as silica?) and iron inclusions and there were some pores or openings in the boundary zones, as testified by the Ar signal. Some elements were associated, e.g., iron with bromine (but not especially well with iodine) in inclusions and magnesium with sulfur in a shape of the continuous zoned distribution, not corresponding to that of calcite. What is interesting is that the ratio of magnesium to calcium (Fig. 3) has its maximum in the boundary regions of the sample that are more or less strictly indicated by the elevated levels of the iron signals.

Fig. 1 Main element (Ca) profile (solid line), determined with the electron microprobe, compared with the linear grey-scale profile (dashed line), extracted from the secondary electron image of the sample.

Fig. 2 Comparison of the Ca distribution (dashed line) with the Si profile (solid line) from electron microprobe measurements; note the correspondence between silicon peaks and some of calcium minima, determining the laminar structure to some extent.

Fig. 3 Mg/Ca ratio (dashed line) extracted from electron microprobe experiments versus iron profile (solid line). Note the correlation of magnesium maxima with iron peaks, which means with the laminar boundaries.

The X-ray capillary tabletop microprobe measured the signals of several additional elements, mainly Sr, Fe and Ti. Two additional auxiliary signals, coherently and incoherently scattered radiation of the Mo Kα line, gave an insight into the density variations in different locations of the samples. Just the last two signals together with the calcium signal create the basis for the differentiation of the annular structure of the stalagmite. Still, there is a problem, i.e., both scattered signals should be inverted by, e.g., subtraction from the arbitrary selected level. This results from the fact that material in the boundary region either has a greater electron density or is much more chaotically ordered and it scatters X-rays more intensely. After the inversion, both profiles look as in Fig. 4. Both spectra look very similar, except the region between step numbers 45 and 80. The original Rayleigh signal is much higher in this location, whereas the Compton signal is undisturbed, as is the Ca signal. If there is no elemental reason for such behaviour, the orientation of the monocrystals may be a reason. Monocrystals can be observed in this region of a stalagmite even with the naked eye. Fig. 5 shows a comparison of the calcium profile with iron, dispersed into inclusions. Many of these inclusions match the boundary regions of stalagmites, as was observed with the previous microprobe. One can even conclude that all boundary zones include pronouncedly elevated levels of iron. At the same time, these zones have elevated levels of Ti and Ce (the last is an artefact, as was explained). Sr was another element detected; its distribution is strange and some very clear maxima occur regularly. The first minimum in the Sr content coincides well with the mentioned pronounced deviation in the Rayleigh spectrum, but later this dependence vanishes. The regularities in the Sr profile demand a deeper investigation if one wished to couple them to the geochemical and thermal history of stalagmites.

Fig. 4 Comparison of two auxiliary signals (Compton, dashed line; Rayleigh, solid line), obtained with the X-ray tabletop microprobe. Note the good correlation of those two scans, except for the structure in the Rayleigh profile between steps 45 and 80. Both scans are inverted.

Fig. 5 Ca profile (dashed line) from X-ray tabletop microprobe in relation to Fe scan (solid line). The iron peaks delimit the zone boundaries.

The X-ray synchrotron microprobe detected easily some heavier elements, namely Sr and Ba, in addition to Fe and I. As artefacts, the presence of La and Ce was registered. Both elements were introduced during polishing the samples, since they are components of the standard optical polishing powder. Their signals were not useless, since they could be correlated with the laminar structure of the stalagmite. In Fig. 6(a), the coincidence of the Ce signals with the minima of the calcium signals (boundary regions of the laminar zones) is clearly shown. In a similar way, the iron inclusions match the same places [Fig. 6(b)]. The distributions of Sr and Ba were very similar, but they differed from the calcium profile, and also the magnesium profile, extracted from the electron microprobe measurements and presented here together with heavy alkaline earth elements after re-calibration of the axes [Fig. 7(b)]. Trying to exploit the parallel results from all microprobes, we added Fig. 7(a), with the calcium profile, and Fig. 7(c), with the profiles of anionic substances, all tailored to the run of the synchrotron microprobe. All the data in Fig. 7 were smoothed to make the comparisons easier. The alkaline earth metal cations differ significantly in their patterns. As we observed earlier (Fig. 3), the magnesium maxima coincide well with the annular boundaries. The patterns of Ba and Sr follow each other but they are dissimilar in relation to both Ca and Mg. We can observe very clear maxima in the distributions of Ba and Sr; as was shown by the X-ray tabletop microprobe measurements, they were regularly located every 7.2 mm on our sample and this corresponds fairly well with the measurements by Wogelius et al.¹⁰ on similar calcite formations. This pattern did not match the laminar structure as drawn by the Ca scan. It should be examined by microdiffraction studies, if the calcite structure is not disturbed by the influence of the strontianite and witherite structures. Such a situation can be expected if the surrounding solution is sufficiently enriched in strontium.¹⁴ Both of the last two minerals mentioned belong to the aragonite group, not calcite. Probably there are some amounts of calcium sulfate, chloride and phosphate dissolved in solid calcite, fairly uniformly.

Fig. 6 (a) Ca scan (dashed line) from the synchrotron microprobe in parallel with the Ce profile (solid line). Note the correlation of the minima in the calcium profile with the peaks in the cerium profile. (b) Similar comparison for Ca (dashed line) and Fe (solid line) with the same microprobe.

Fig. 7 (a) Smoothed Ca profile obtained from the electron microprobe. (b) Smoothed profiles of Mg (solid line with crosses) from the electron microprobe and Sr (solid line with open squares) and Ba (dashed line with solid circles) from the X-ray synchrotron microprobe, after re-scaling to show the locations on the sample, analogous to locations from (a). (c) Smoothed profiles of anionic species (S, solid line with open squares; Cl, dashed line with solid circles; P, dashed line with crosses) from the electron microprobe, re-scaled as previously.

The profile obtained by the use of the selected energy channel (60–70 keV) in white scattered radiation was obscure (not shown here). Someone very experienced in image analysis would certainly recognise the boundaries of the laminar structure, but rather invoking at the same time the Ca and Ce profiles. It was interesting that we found a very clear profile of iodine, but it was dissimilar to the other elemental profiles. This discovery may suggest that the cave from which the stalagmite was taken was under some influence of seawater or contact with it.

All the results compliment each other to the degree allowed by the nature of the particular microprobe and by the possibility of making measurements by different microprobes strictly at the same locations.

Conclusions

Our studies confirmed that calcite was an essential building material for the stalagmite examined. However, the situation is not simple. When analysed with the tabletop X-ray microprobe, unexpected differences occurred in the coherently and incoherently scattered spectra. Since there was no significant change in the chemical composition of the locations with the undisturbed Compton and disturbed Rayleigh spectra, we decided to ascribe the reason to the crystallographic orientation of the calcite monocrystal against the direction of the exciting beam. Indeed, one could observe in the middle of the stalagmite some large, clear crystals, glued together. This structure vanished in the outward direction and the new structure was much more homogeneous.

Another concern was the chemical composition. The results from the three different microprobes were complementary. We found that the calcite structure was everywhere enriched. This is best illustrated in Fig. 7, where one can observe the main pattern as given by the Ca profile [Fig. 7(a)], superimposed scans of other alkaline earth metals [Fig. 7(b)] and three anionic substances, relatively evenly distributed. The same does not apply to the iodine that can be clearly observed, in the best way with the synchrotron X-ray microprobe, and is distributed according to some pattern, dissimilar to any other among those studied. The presence of iodine indicates some contact of the sample with seawater but, on the other hand, we did not find clear traces of seawater elements such as Na and K, both of which usually are associated with iodine. When they are really present in the samples, as is the case with petrified wood, our microprobes have no problems with their detection.²²

The second type of general pattern, characteristic of the stalagmite structure, includes the concentration of some elements in the boundary regions of the laminar structure. This concerns mainly iron, that was detected by all the microprobes, silicon and titanium—they all exist in the form of sharp inclusions—magnesium, that is only enriched in the boundary regions. Such additions testify that the structure in the boundary region is fairly loose. Some workers have noted that the porosity reaches 10% in a stalagmite structure, but this was studied only on the macroscale.^25,34 The boundaries, otherwise indicated by the minima in the Ca profiles, are relatively very narrow, of the order of 150–200 µm, in comparison with the size of a typical ring in the range 1500–3500 µm. Owing to the irregularity of the scans, resulting from the statistics and from the nature of the object, it is useful to compare the positions of the calcium minima with the inclusions of heavier elements to be sure which boundaries are real and not random.

Fortunately, there is a fair number of artefacts associated with the measurements and such artefact signals can serve as an indication of the boundary locations. Using the electron microprobe, one can easily detect the argon, included in air, entrapped in the porous material of the boundary zone. This testifies once again to the loosened structure of the boundaries. Some additional materials can also be included in openings, as were La and Ce from the polishing powders used during the surface preparation. Whereas the Ar signal is proof of the existence of closed pores, La and Ce testify to the open holes in the boundary regions; probably, both structures co-exist. In the literature, cases where the air included in pores was analysed, are well-known.^3,35

The behaviour of Mg is unique. It was found that this element was enriched in the boundary positions. Otherwise, it is known that an increase in Mg content is associated with a warmer temperature of the water during the stalagmite deposition. In turn, it would be equivalent to the claim that the boundary regions were created in the environment with higher temperature than for the creation of all the other stalagmites. This is in conflict with our knowledge, since the deposition process stops during the winter and this coincides with the boundaries in stalagmites. This aspect demands further studies.

Some auxiliary signals were very useful for establishing the equivalence between the morphological structure of the sample and its local chemical composition. Probably, the linear profile, extracted from the secondary electron image (Fig. 1) of samples and Compton and Rayleigh signals (Fig. 4) from the tabletop X-ray microprobe, can be readily adopted for this aim, as was found earlier in the case of petrified wood.²³ On the other hand, the application of the selected channel from the white synchrotron was doubtful, which is strange since this radiation is composed mainly of many Compton contributions, even of higher order,³⁶ and it was used previously in an efficient way for the analysis of other periodic structures.

The electron and X-ray microprobes proved their effectiveness for the delivery of useful signals for tracing the stalagmite structure and composition and to couple the data obtained with different chemical, climatic and hydrological factors. In this sense these methods can be useful to the same extent as isotope-searching instruments for this kind of research. Unfortunately, our sample was only an exemplary piece, without any correction with mentioned databases. Moreover, we had only a fragment of the stalagmite, allowing us to reconstruct the radial but not the longitudinal profile of the stalagmite. The second profile is preferred by investigators. To illustrate that the two are not always identical and should be treated carefully, we present here a comparison of the radial and longitudinal profiles as extracted by us from the data from an optical microscope (Fig. 8) concerning a similar sample.⁵ The problem of having a representative sample is probably more important in this field than the problem of the selection a method for studying objects.

Fig. 8 Comparison of the single stalagmite optical profiles: radial (dashed line) with longitudinal (solid line). Extracted from ref. 5.

ack

One of the authors (A.K.) is deeply grateful to the Flemish Research Fund for a fellowship that enabled this study to be carried out. Part of the study was financed by a grant from the Hallym Academy of Sciences, Hallym University, Korea.

References

1. M. Gascoyne, Q. Sci. Rev., 1992, 11, 609.
2. H. P. Schwarcz, Archaeometry, 1980, 22, 3.
3. A. Matthews, A. Ayalon and M. Bar-Matthews, Chem. Geol., 2000, 166, 183.
4. D. Genty, M. Massault, M. Gilmour, A. Baker, S. Verheyden and E. Keppens, Radiocarbon, 1999, 41, 251.
5. G. M. Thompson, D. N. Lumsden, R. L. Walker and J. A. Carter, Geochim. Cosmochim. Acta, 1975, 39, 1211.
6. P. Thompson, D. C. Ford and H. P. Schwarcz, Geochim. Cosmochim. Acta, 1975, 39, 661.
7. I. J. Vinograd, B. J. Szabo, T. B. Coplen and A. C. Riggs, Science, 1988, 242, 1275.
8. K. R. Ludwig, K. R. Simmons, B. J. Szabo, I. J. Winograd, J. M. Landwehr, A. C. Riggs and R. J. Hoffman, Science, 1992, 258, 284.
9. J. M. Vadillo, I. Vadillo, F. Carrasco and J. J. Laserna, Fresenius' J. Anal. Chem., 1998, 361, 119.
10. R. A. Wogelius, D. G. Fraser, G. R. T. Wall and G. W. Grime, Geochim. Cosmochim. Acta, 1997, 61, 2037.
11. J. Paquette and R. J. Reeder, Geochim. Cosmochim. Acta, 1995, 59, 735.
12. D. J. J. Kinsman, J. Sediment. Petrol., 1969, 39, 486.
13. S. L. Brantley, J. T. Chesley and L. L. Stillings, Geochim. Cosmochim. Acta, 1998, 62, 1493.
14. R. H. Parkman, J. M. Charnock, R. F. Livens and D. J. Vaughan, Geochim. Cosmochim. Acta, 1998, 62, 1481.
15. N. E. Pingitore Jr. and M. P. Eastman, Geochim. Cosmochim. Acta, 1986, 50, 2195.
16. N. E. Pingitore Jr., F. W. Lytle, B. M. Davies, M. P. Eastman, P. G. Eller and E. M. Larson, Geochim. Cosmochim. Acta, 1992, 56, 1531.
17. J. D. Rimstidt, A. Balog and J. Webb, Geochim. Cosmochim. Acta, 1998, 62, 1851.
18. A. Katz, Geochim. Cosmochim. Acta, 1973, 37, 1563.
19. R. H. Parkman, J. M. Charnock, F. R. Livens and D. J. Vaughan, Geochim. Cosmochim. Acta, 1998, 62, 1481.
20. A. Goede and J. C. Vogel, Palaeogeogr. Palaeoclimat. Palaeoecol., 1991, 88, 121.
21. W. Dreybrodt, Chem. Geol., 1980, 29, 80.
22. A. Kuczumow, B. Vekemans, O. Schalm, W. Dorriné, P. Chevallier, P. Dillmann, C.-U. Ro, K. Janssens and R. Van Grieken, J. Anal. At. Spectrom., 1999, 14, 435.
23. A. Kuczumow, B. Vekemans, O. Schalm, L. Vincze, W. Dorriné, K. Gysels and R. Van Grieken, X-Ray Spectrom., 1999, 28, 282.
24. A. Kuczumow, P. Chevallier, P. Dillmann, P. Wajnberg and M. Rudaś, Spectrochim. Acta, Part B, 2000, 55, 1623.
25. A. Baker, D. Genty, W. Dreybrodt, W. L. Barnes, N. J. Mockler and J. Grapes, Geochim. Cosmochim. Acta, 1998, 62, 393.
26. K. Janssens, L. Vincze, J. Rubio and F. Adams, J. Anal. At. Spectrom., 1994, 9, 151.
27. K. Janssens, B. Vekemans, L. Vincze, F. Adams and A. Rindby, Spectrochim. Acta, Part B, 1996, 51, 1661.
28. N. C. Sturchio, M. R. Antonio, L. Soderholm, S. R. Sutton and J. C. Brannon, Science, 1998, 281, 971.
29. A. Kuczumow and S. Larsson, Appl. Optics, 1994, 33, 7928.
30. D. Genty and Y. Quinif, J. Sediment. Res., 1996, 66, 275.
31. Neutron, X-ray and Light Scattering, ed. P. Lindner and Th. Zemb, North-Holland, Amsterdam 1991..
32. A. Baker, D. Genty and P. L. Smart, Geology, 1998, 26, 903.
33. A. Baker and D. Genty, J. Hydrol., 1999, 217, 19.
34. A. C. Kendall and P. L. Broughton, J. Sediment. Petrol., 1978, 48, 519.
35. H. P. Schwarcz, R. S. Harmon, P. Thompson and D. C. Ford, Geochim. Cosmochim. Acta, 1976, 40, 657.
36. L. Vincze, K. Janssens, B. Vekemans and F. Adams, J. Anal. At. Spectrom., 1999, 14, 529.

---