the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Short communication: Synchrotron-based elemental mapping of single grains to investigate variable infrared-radiofluorescence emissions for luminescence dating
Mariana Sontag-González
Raju Kumar
Jean-Luc Schwenninger
Juergen Thieme
Sebastian Kreutzer
Marine Frouin
Abstract. During ionising irradiation, potassium (K)-rich feldspar grains emit infrared (IR) light, which is used for infrared-radiofluorescence (IR-RF) dating. The late-saturating IR-RF emission centred at ~880 nm represents a promising tool to date Quaternary sediments. However, in the present work, we report the presence of individual grains of K-feldspar displaying an aberrant IR-RF signal shape, whose combined intensity contaminates the sum signal of an aliquot composed of dozens of grains. Our experiments were carried out at the National Synchrotron Light Source (NSLS-II) on coarse (> 90 µm) K-feldspar grains of five samples of different ages, nature and origin in order to characterise the composition of grains yielding the desired or contaminated IR-RF emission. Using micro-X-ray-fluorescence (µXRF), we successfully acquired element distribution maps of fifteen elements (<1 µm resolution) of the surface of grains previously used for luminescence dating. In keeping with current theories of IR-RF signal production, we observed a correlation between the relative proportions of Pb and Fe and the shape of the luminescence signal: most grains with the desired IR-RF signal shape had high Pb and low Fe contents. Interestingly, these grains were also defined by high Ba and low Ca contents. Additionally, this study represents a proof-of-concept for mapping the oxidation states of Fe-ions using micro-X-ray absorption near-edge structure spectroscopy (µXANES) on individual grains. The high spatial resolution enabled by synchrotron spectroscopy makes it a powerful tool for future experiments to elucidate long-standing issues concerning the nature and type of defect(s) associated with the main dosimetric trap in feldspar.
- Preprint
(1166 KB) - Metadata XML
-
Supplement
(1251 KB) - BibTeX
- EndNote
Mariana Sontag-González et al.
Status: final response (author comments only)
-
RC1: 'Comment on gchron-2023-14', Svenja Riedesel, 24 Jul 2023
Sontag-Gonzáles et al. Short communication: Synchrotron-based elemental mapping of single grains to investigated variable infrared-radiofluorescence emissions for luminescence dating
General comments
The manuscript entitled Short communication: Synchrotron-based elemental mapping of single grains to investigated variable infrared-radiofluorescence emissions for luminescence dating presents exciting new results on the use of synchrotron-based x-ray fluorescence and x-ray absorption near-edge structure spectroscopy on feldspar single grains to link spatially resolved element and oxide concentrations to infrared radiofluorescence (IR-RF) emissions. Generally, I think that this is a really interesting study and will provide some helpful insights into understanding the sources of undesired IR-RF emissions. However, I have two general comments regarding the content and focus of this study.
General comment 1. The paper has been submitted as a short communication and on multiple occasions throughout the manuscript it is stated that the study should be regarded as a proof of concept or preliminary study. Nevertheless, I am wondering, if the others do not wish to hold on to their data until some gaps have been filled, prior to then re-submit the study as full research article. In my opinion this would enable the authors to perform additional measurements, which are already briefly commented on in this manuscript, and to also further shape the focus of this study, which leads me to my second general comment.
General comment 2. To me the focus and purpose of the study does not become entirely clear. I think that some changes to the structure and the text could help in tightening the focus and in the following I list some suggestions the authors might like to consider when revising their manuscript:
- The introduction, especially the explanation of luminescence dating and some parts on IR-RF (lines 34-48, lines 59-67) are too long and they distract from the actual study. At some point early on in the introduction a brief statement outlining the importance of this study is needed. The last paragraph of the introduction (lines 68-74) could be expanded and linked to the brief statement of the necessity of this study.
- In my opinion the multi-grain IR-RF results (section 4.1) are not needed, and I would suggest removing all sections related to the multi-grain experiment, so that the paper is solely focussed on single grain analyses.
- Section 2 might benefit from some restructuring and revising of some of the paragraphs. Maybe starting with the problems (e.g. overlap of signals, unwanted growth signals etc.) and then explaining the current state of the art would be an option?
Specific comments
On the role of Pb and Fe for feldspar luminescence (mostly section 2):
a) Lines 91-92: I would suggest phrasing this more carefully, since Erfurt (2003) compared the emission spectra of a feldspar and KCl:Pb and inferred a relationship of the emissions due to the presence of Pb. Additionally, the two main oxidation states of Pb are Pb2+ and Pb4+ and a transition to Pb+ via a highly reactive radical is rather unlikely. This could be a particular reason to wait for the XANES measurements for Pb oxides before further considering this hypothesis.
b) Lines 97-98: Original references should be given here for the debate on Fe oxides in feldspars. A reference to include here could be Kirsh and Townsend (1988), especially because these authors also suggest further options for the role Fe might play for the red emission. The reference for the reaction of Fe4+ to Fe3+ due to the capture of an electron would be Jain et al. (2015), although the presence of Fe4+ is rather unlikely in a natural mineral.
Line 111: Are these methods really the “best” suited methods to tackle the described problem? Either here or later in the paper, some advantages and disadvantages of using these methods should be highlighted. For instance, the penetration depth of the x-rays would be useful to know, especially, since in the conclusion a hypothesis of the effect of surface coating on luminescence is mentioned.
Section 4.4: The paragraph is rather short and contains a contradiction. In line 237 it is suggested that Fe exists on the surface of the grain, but in line 238 an inclusion within the crystal is mentioned. Knowing the penetration depth of the x-rays and expanding this section, might help in clarifying this issue.
Technical comments
Line 26: A correlation is not statistically validated in this study; I would suggest exchanging the word “correlation” with “trend”.
Line 76: It should read: “of twenty-one to…”
Line 192-193: No references are given for the statement on the link of luminescence behaviour and volcanic origin of the samples. May I please ask the authors to delete this statement or to add a reference?
Lines 199-202: Since the microXRF measurement results have not been described yet, I would suggest removing the statement of potential trends of K content and IR-Rf signal at this point.
Lines 213-222: Maybe a boxplot or pie chart or some other type of illustration could be used to display the relative contributions of each element for each group of grains?
Fig. 5: Would it be possible to provide the reader with some numerical values for the axis of these ternary diagrams? Relative intensities are plotted, but on what basis? It would also be helpful, if the grains, which are displayed in Figure 6 could be somehow highlighted in Fig. 5.
Fig. 7: Please indicate where this inset relates to the whole figure.
Fig. 8: Are the areas displayed in Figure 8 the same as those in Figure 6? If yes, may I please ask the authors to highlight this in the figure caption. Alternatively, some overview images of the grains (similar to Fig. 6) could perhaps be given?
Lines 250-252: Since no oxidation states of Pb were measured, it is unclear why this sentence is included in the conclusion of the paper.
Line 254: It should be “element and oxidation state maps” and not mineralogical maps.
Lines 258-260: Apologies if I missed it, but I think this is the first time that the hypothesis of elements on the surface of the grains influencing the luminescence signal is mentioned in the manuscript, except for the brief mentioning in section 4.4.
Citation: https://doi.org/10.5194/gchron-2023-14-RC1 -
AC1: 'Reply on RC1', Mariana Sontag-González, 25 Aug 2023
Thank you for the comments and feedback. Below we are responding to each of your comments.
General comment #1: We agree that more data would automatically make our observations more robust. However, getting granted access to the National Synchrotron Light Source II for an experiment is highly competitive. All beam time is allocated based on a peer-reviewed proposal process. We are continuously submitting proposals to the Brookhaven National Laboratory with the hope of being awarded beam time in the near future.
In our opinion, the sample preparation and methodology we have implemented during this beam time has never been used in the luminescence dating discipline. Because of its pioneering aspect, it should be worth publishing as a short communication. In addition, our current dataset and results are advancing our understanding of oxidation states in feldspar, which was suggested to exist but have never been directly obtained or measured at this resolution.
Thank you also for the suggestions in your comment #2. We will streamline the introduction and more strongly highlight the need for this study, and restructure section 2. However, we feel that the results from the multi-grain experiment (section 4.1) are an important argument in favor of the need for the present work, as it demonstrates the potential danger in measuring IR-RF on multi-grains. The vast majority of IR-RF studies have been on multi-grain aliquots, and the influence of the overlap of different signals from different grains has not received much attention in the literature so far. In section 4.1, we attempted to show through one example that the multi-grain measurements can be unreliable.
Specific comments (authors’ responses are in bold-type after each reviewer’s comment):
On the role of Pb and Fe for feldspar luminescence (mostly section 2):
- Lines 91-92: I would suggest phrasing this more carefully, since Erfurt (2003) compared the emission spectra of a feldspar and KCl:Pb and inferred a relationship of the emissions due to the presence of Pb. Additionally, the two main oxidation states of Pb are Pb2+ and Pb4+ and a transition to Pb+ via a highly reactive radical is rather unlikely. This could be a particular reason to wait for the XANES measurements for Pb oxides before further considering this hypothesis.
- Thank you for your comment. Here we just intended to state hypotheses based on previous work, but we will rephrase this sentence to avoid any misunderstanding.
- b) Lines 97-98: Original references should be given here for the debate on Fe oxides in feldspars. A reference to include here could be Kirsh and Townsend (1988), especially because these authors also suggest further options for the role Fe might play for the red emission. The reference for the reaction of Fe4+ to Fe3+ due to the capture of an electron would be Jain et al. (2015), although the presence of Fe4+ is rather unlikely in a natural mineral.
- Thank you, we will add the original references.
Line 111: Are these methods really the “best” suited methods to tackle the described problem? Either here or later in the paper, some advantages and disadvantages of using these methods should be highlighted. For instance, the penetration depth of the x-rays would be useful to know, especially, since in the conclusion a hypothesis of the effect of surface coating on luminescence is mentioned.
- We will add a paragraph referring to the X-ray penetration depth. For our samples, this is not entirely straightforward since we measured whole grains (unpolished), whose topography can change the depth from which the resulting XRF can still escape (attenuation length). Additionally, this depth depends on material density, which would be different for grains of different compositions as present in our dataset. Different elements will also have different attenuation lengths, e.g., for Fe, we consider 30–40 µm are probed. As the reviewer rightfully pointed out, this is a disadvantage of the method.
Section 4.4: The paragraph is rather short and contains a contradiction. In line 237 it is suggested that Fe exists on the surface of the grain, but in line 238 an inclusion within the crystal is mentioned. Knowing the penetration depth of the x-rays and expanding this section, might help in clarifying this issue.
- We apologise for this misunderstanding. We wrote ‘the surface of the grain’ to highlight that our measurements are of the whole grain and we have no information about the composition of the total volume, but we are in fact, not measuring the surface, but a rim of the grain (see comment above). We will rephrase for clarity.
Technical comments (authors’ responses are in bold-type after each reviewer’s comment):
Line 26: A correlation is not statistically validated in this study; I would suggest exchanging the word “correlation” with “trend”.
- Yes, ‘trend’ would be more appropriate in this case, thank you.
Line 76: It should read: “of twenty-one to…”
- We will correct this.
Line 192-193: No references are given for the statement on the link of luminescence behaviour and volcanic origin of the samples. May I please ask the authors to delete this statement or to add a reference?
- We will rephrase this statement to clarify the link.
Lines 199-202: Since the microXRF measurement results have not been described yet, I would suggest removing the statement of potential trends of K content and IR-Rf signal at this point.
- We will rephrase this paragraph to state that only a sub-population of grains displays the expected IR-RF signal and that our hypothesis is that only these are K-rich feldspar grains.
Lines 213-222: Maybe a boxplot or pie chart or some other type of illustration could be used to display the relative contributions of each element for each group of grains?
- Unfortunately, the detector is not calibrated to convert the XRF intensity to elemental contribution (e.g., in weight) of the different elements. We can only plot relative XRF intensities of each element, as we used in Fig. 5, where we focussed on the five most relevant elements. We can add a boxplot to Fig. 4 with the relative XRF intensities of each element for the three groups.
Fig. 5: Would it be possible to provide the reader with some numerical values for the axis of these ternary diagrams? Relative intensities are plotted, but on what basis? It would also be helpful, if the grains, which are displayed in Figure 6 could be somehow highlighted in Fig. 5.
- We can mark the grains from Fig. 6 in the ternary diagram. The diagrams are plotted according to the XRF intensity (not the chemical composition), and thus only serve to compare grains to each other (e.g., grain A has more Fe than grain B), not to discern, e.g., which grains have more Fe than Pb in their compositions. We can add axes labels to the diagrams, they would read 0–100%, meaning that a hypothetical point on the centre of the axis between, e.g., Fe and Pb would have the same XRF intensities (integrating counts in the deconvoluted peaks) for these two elements and no peak for the third element in the ternary diagram.
Fig. 7: Please indicate where this inset relates to the whole figure.
- The inset highlights the shift in the absorption edges for the three oxidation states. The units of the inset relate directly to the units of the main figure (the same underlying curves are shown, only the axis limits differ). We will clarify this in the caption.
Fig. 8: Are the areas displayed in Figure 8 the same as those in Figure 6? If yes, may I please ask the authors to highlight this in the figure caption. Alternatively, some overview images of the grains (similar to Fig. 6) could perhaps be given?
- It is the same grain in both figures, but the locations are slightly different. We will add an overview image, as suggested.
Lines 250-252: Since no oxidation states of Pb were measured, it is unclear why this sentence is included in the conclusion of the paper.
- XANES measurements are possible for any element with XRF emissions within the measured energy limits. We highlighted the possibility of Pb oxidation states maps in the section ‘Conclusions and future work’ to suggest that they would also be informative to investigate the origin of the IR-RF, given the work by Erfurt et al., (2003). We can remove this suggestion from the sentence.
Line 254: It should be “element and oxidation state maps” and not mineralogical maps.
- We will correct this.
Lines 258-260: Apologies if I missed it, but I think this is the first time that the hypothesis of elements on the surface of the grains influencing the luminescence signal is mentioned in the manuscript, except for the brief mentioning in section 4.4.
- We mention this possibility in section 4.1 and will expand on this hypothesis in that section.
Citation: https://doi.org/10.5194/gchron-2023-14-AC1
-
AC1: 'Reply on RC1', Mariana Sontag-González, 25 Aug 2023
-
RC2: 'Comment on gchron-2023-14', Anonymous Referee #2, 25 Jul 2023
Review of “Synchrotron-based elemental mapping of single grains to investigate variable infrared-radiofluorescence emissions for luminescence dating”
This is an interesting work in which authors tried to identify the causes in variation in shapes of the IR-RF curves and attributed them to variations in elemental concentrations inside mineral grains. The authors also conclude such variations can result in wrong estimation of doses. Although methodology adopted seems reasonable, there are still several scientific aspects, which are not clear and need to be addressed.
Comments:
- Page 2 line 38, Introduction: “rely on the capacity of”. Its not capacity, its property of defects. Change appropriately.
- Page 2 line 40, 41, Introduction: “In the laboratory, the total amount of energy stored” and “energy absorption rate (dose rate, Gy a-1)”. Energy per unit mass is dose.
- Page 2 line 43, Introduction: “quartz because of its high abundance and resistance to weathering”. Besides these two, the fast to bleach OSL signal makes it most appropriate for geological dating.
- Page 2 line 45, Introduction: “considering a low dose rate of 1 Gy ka-1”. Why to mention ‘low’ here? Please delete low
- Page 2 line 64, Introduction: “saturation cap at around 1500 Gy, reducing” cap can be deleted
- Page 3 line 80, Method rationale: “study, between one and three grains (~…….. curves”. Statement not clear, consider revising. What is the reason for bad match? Bad match is often observed due to sensitivity changes. Pls refer (Varma, V., Biswas, R.H., Singhvi, A.K., 2013. Aspects of Infrared Radioluminescence dosimetry in K-feldspar. Geochronometria 40, 266-273.)
- Page 3 line 83, Method rationale: “grains that emitted a detectable signal displayed the expected decay shape.” What is meaning of expected shape here? Exponential decay?
- Page 3 line 89, Method rationale: “With dose exposure, the 955 nm emission increases and overlaps with the 880 nm peak.” Why is it so? How dose increases the 955 nm emission, does it mean more 955nm recombination centres are being regenerated? Does it reflect multiple trap system?
- Page 4 line 101, Method rationale: “The thermal stability of the ~710 nm emission has been, however…. the measured IR-RF” The red TL emission in feldspar is generally considered more stable than conventional IRSL method, why the red IR-RF is unstable?
- Page 4 line 111, Method rationale: “The μ‐XRF and μ‐XANES techniques are best suited for this purpose by producing high-resolution maps of elements and their oxidation states”. It will be good to provide some details about the mentioned techniques and their usefulness for present work.
- Page 3 line 83, Method rationale: “The use of synchrotron μXRF allows us to improve the spatial resolution compared with previous uses of μXRF (e.g., Buylaert et al., 2018) by reducing the beam spot size from ~25 μm to 1 or 0.5 μm”. It is indeed impressive that spot size is smaller and we can work at higher resolutions, but how will it effect S/N ratio and thus elemental concentration estimation? In addition, since spot size is smaller, only few grains analysis may be possible. In such cases, how can we get the statistical representation of entire grain population just based on few grains studies?
- Page 4 line 128, Material and instrumentation: Normally Tuff samples are expected to contain Fe rich species. Is this a deliberate choice to look the effect of Fe in the samples as 2 out of 5 are tuff samples?
- Page 5 line 134, Material and instrumentation: Why only the sample H22550 was etched with HF. Why not same is performed for other samples?
- Page 5 line 134, Material and instrumentation: “Multi-grain and single-grains … National Laboratory” How correspondence between IRRF and XRF signals is established?
- Page 5 line 143, Material and instrumentation: “2016). Grains were fixed on a polymer microscope slide…………” What are spectral and luminescence characteristics of the base material used?
- Page 5 line 144, Material and instrumentation: “XRF maps were obtained by scanning across pre-selected regions on the grains 90 x 90 μm maps,” What is the basis of ROIs selection?
- Page 5 line 149, Material and instrumentation: “resolution of 0.67 μm was achieved by focusing the beam with” Is it Xray beam focussing or luminescence focussing, please specify.
- Page 5 line 150, Material and instrumentation: “ An incident beam energy of 13.5 keV was” Why this energy chosen any specific reason?
- Page 5 line 150, Material and instrumentation: “fluorescence was detected through the sum of 4 silicon drift detectors” Why these four detectors were used? Can we use PMT instead? whats the advantage we get with use of these detectors. Can you provide geometry of measurements and experimental setup?
- Page 5 line 152, Material and instrumentation: “All XRF measurements were normalised to the corresponding incident X-ray flux ” X-ray sources are normally found inhomogenous spatially and temporally. Does this can effect your measurements?
- Page 5 line 150, Material and instrumentation: “The XANES maps had a resolution of 0.5 μm (60 x 60μm).” Are units correct? How does 60 um X 60 um translate to 0.5um? not clear.
- Page 5 line 150, Material and instrumentation: “we varied the incident beam energy according to the absorption edge values obtained from the μXANES measurements of Fe standards (Fe foil, pyrite, hematite).” How the specific absorption edge values were estimated?
- Page 5 line 150, Material and instrumentation: “(i) the total Fe (at 7.275 keV),” The energies mentioned here are quite precise. How much is normally the resolution. Since electronic energy levels of specific elements are quite low in energy (~few eVs) compared to what is being provided, then why this much precision is needed?
- Page 6 line 165, Results: “500 Gy succeeded by an increase, roughly following a saturating exponential shape” What is reason behind increase to saturating exponential behaviour? Why should there be an increase at all considering physics aspect? What is the nature of sample X7343, Is it similar to volcanic tuff?
- Page 6 line 169, Results: “contamination, potentially coming from coating around the grains, we” Why is it assumed that coating could be responsible?
- Page 6 line 171, Results: “Despite using density … high Fe content”. Does that mean it is Na or Ca feldspar grains? Have you performed XRD analysis on bulk to see the mineralogy of samples?
- Page 6 line 176, Results: “their visual appearance under a microscope” What were the visual indicators considered for choosing K-Feldspar?
- Page 6 line 178, Results: “grains. The regenerated IR-RF signals showed a clear distinction between the two aliquots (Fig. 1), proving it is possible to separate the two observed IR-RF shapes.” This is quite a qualitative way. I am not sure how to progress using only visually inspected grains. The visual appearance and selection can vary depending upon geological settings of grains as well as person observing them. Is there any other rigorous way of making such selection?
- Page 6 line 184, Results: “presumed to be the low-K, Fe rich minerals identified via SEM-EDS” Low K means possibly high Na or Ca, why only Fe is considered. Fe if present should be in form of defects, which should be in ppm level. Can uXRF measure to such low concentration levels? If Fe is appearing as major element in feldspar separates, it means it is present in stoichiometric formula and in that case, mineralogy of sample would be different. Please suggest if I am missing something.
- Page 6 line 186, SG IR-RF characterisation: “signal of twenty-two individual grains coming” Can you specify mineralogy of each grains, which are picked for such measurements?
- Page 6 line 165, SG IR-RF characterisation: “2): Category #1 for grains with a decreasing IR-RF signal, category #2 for grains with an increasing IR190 RF signal, and category #3 for grains with a flat signal”. How many grains falls in each category and is there any link to the provenance.
- Page 6 line 165, SG IR-RF characterisation: “we also observed the unwanted decreasing IR-RF signal for one of the four grains for sample H22550, which is from a coastal sand deposit.” We expect IR-RF signal to decrease with irradiation, so why it is said unwanted ?
- Page 6 line 193, SG IR-RF characterisation: “When the total signal of the theoretical aliquot was composed of more than 50% of signal from the category #2 grain, we observed the same decay shape as in figure 1 for a multi-grain aliquot sample X7343” Obviously, since the two different category of grains having two different IR-RF characteristics are being added, so the result will depend on the proportion of the individual populations in the mixture. More importantly, it is important to know, how these two grains are different with respect to crystallography or stoichiometry or defects concentration. Is the nature of curve repeatable over repeated bleaching and irradiation cycles?
- Page 7 line 201, SG IR-RF characterisation: As mentioned by authors, long-term signal stability may not be there for bad traps, is there a way to prove it? How do we know it without experiment?
- Page 7 line 204, SG IR-RF characterisation: “Further, our results demonstrate…” I agree with this statement, but it is still not clear how can we segregate K-Feldspar and other minerals. Manually it will not be possible on routine basis.
- Page 7 line 207, Subgrain μXRF elemental maps: “We then fitted each of the 18 225 spectra for….” this statement is not clear
- Page 7 line 212, Subgrain μXRF elemental maps: “characterise visible inclusions (see Table 1).” The number of grains analysed per samples are quite small to represent the statistics of system. Can we consider them as representative of whole samples? It is difficult to conclude unless sufficient data points exists.
- Page 7 line 214, Subgrain μXRF elemental maps: “all contain K, Pb, Fe and Ba, among other elements (Fig. 4).” What is typical concentration of these elements? Considering K is a major element present in stoichiometric formula, how much is relative concentration of the other elements?
- Page 7 line 223, Subgrain μXRF elemental maps: “grains from category #3 cluster relatively close to those from category #2, suggesting” How and why does this effect the IR-RF properties? These are observations, but what is the reason for IR-RF signal due to such clusters is not clear.
- Page 7 line 237, Mapping oxidation states with μXANES: “suggest that Fe exists on the surface of this feldspar grain in its Fe3+ and Fe2+ states.” It is great that using uXANES, we could map the presence of Fe on the surface of feldspar grain, but luminescence or IR-RF is a volumetric phenomenon. How this observation is helpful in explaining the IR-RF signal.
-
AC2: 'Reply on RC2', Mariana Sontag-González, 25 Aug 2023
We thank the reviewer for their positive assessment of our work. We have attempted to clarify and/or correct the aspects pointed out by the reviewer, as set out in the attached document. (Note that the numbering follows the html version of the reviewer’s comments, which is slightly different from the pdf version due to an extra line in comment 4.)
Mariana Sontag-González et al.
Data sets
Short communication: Synchrotron-based elemental mapping of single grains to investigate variable infrared-radiofluorescence emissions [Dataset] (v.1.0.0) M. Sontag-González, R. Kumar, J.-L. Schwenninger, J. Thieme, S. Kreutzer, and M. Frouin https://doi.org/10.5281/zenodo.7971805
Mariana Sontag-González et al.
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 221 | 70 | 21 | 312 | 35 | 9 | 9 |
- HTML: 221
- PDF: 70
- XML: 21
- Total: 312
- Supplement: 35
- BibTeX: 9
- EndNote: 9
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1