Pith. sign in

REVIEW 4 major objections 6 minor 34 references

Matching Trace Element Distribution to Mineralogical Phases in Ancient Biotechnology-Derived Metallic Salts: a Multimodal Analysis

T0 review · 4 major / 6 minor · reviewed 2026-07-31 · grok-4.5

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read Spatially mapped X-ray fluorescence and diffraction can assign trace elements in ancient lead carbonate cosmetics to individual mineral phases that bulk averages leave unlinked.

desk verdict Solid incremental archaeometry paper: co-registered 2D-XRF/XRD plus SNR-gated regressions give usable phase-resolved trace maps on psimythion, but “confident assignment” overstates what R^{2}≥0.1 co-location actually proves. read the letter →

arxiv 2607.28492 v1 pith:Y7NQHJM7 submitted 2026-07-30 physics.app-ph

Sumera Rehman , Spyridon Panagiotopoulos , Georgios E Christidis , Athena Athanassiadou , Alessandro Olivo , Effie Photos-Jones , Silvia Cipiccia This is my paper · ORCID
classification physics.app-ph
keywords psimythioncerussitetrace-elementmappingsynchrotronXRF-XRDX-rayptychographyancientcosmeticsleadcarbonatemultimodalimaging
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

Bulk XRF and XRD only give average compositions of ancient metal salts used as pigments or medicines; they cannot show which trace elements sit inside which crystals. This paper shows that combining spatially resolved XRF, XRD and nanoscale X-ray imaging on the same powder grains lets researchers assign traces such as silver, zinc or chromium to cerussite (or rule them out) and infer non-crystalline hosts for iron. The method is demonstrated on fourth-century-BCE Greek psimythion and on a modern sample made by the ancient recipe. Because the chemical–mineralogical profile of an artefact changes from manufacture through burial to conservation, knowing where each element lives turns the composition into a readable life history rather than a single static recipe.

What carries the argument

Pixel-by-pixel linear (or multiple) regression of elemental counts against phase counts, gated by map signal-to-noise ratio, performed on co-registered 10 µm synchrotron XRF and XRD raster maps of the same powder area.

What would settle it

Re-scan the same archaeological powder after deliberate grinding to a uniform sub-beam grain size, or after controlled doping with a known crystalline host of iron or copper; if the R²–SNR assignments reverse or disappear, the confidence thresholds fail.

Watch

Extended reading notes

Core claim

A multimodal scan that co-registers 2-D XRF elemental maps with 2-D XRD phase maps (and optional ptychography) allows confident assignment of trace elements to named crystalline phases—or inference of non-crystalline hosts—inside archaeological and experimental lead carbonate powders, something bulk 1-D XRF plus XRD cannot do.

Load-bearing premise

A regression coefficient of 0.1, chosen by eye and accepted only when the elemental map’s signal-to-noise exceeds 2, is enough to claim that a trace element truly belongs to a named crystal rather than merely sitting nearby or in an overlapping mixture.

Editorial extensions

If this is right

  • Silver that co-locates with cerussite can be read as an impurity inherited from the original lead metal plate.
  • Iron that never co-locates with any detected crystalline phase points to an amorphous hydroxide or organic binder introduced during scraping or formulation.
  • The same pipeline can track how burial or conservation alters the element-to-phase map of any ancient metal-salt pigment or medicament.
  • Particle-size distributions from ptychography supply an independent check that the powder behaved as a true powder under the beam.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Once element-to-phase maps are routine, museum conservation protocols could be rewritten to avoid reagents that preferentially dissolve or re-precipitate the non-crystalline hosts that carry diagnostic traces.
  • The same co-registration logic should transfer directly to other ancient synthetic salts (laurionite, phosgenite, copper acetates) whose bulk analyses currently leave the same origin ambiguities.
  • If amorphous iron phases prove common, routine soft-X-ray or Raman mapping of organics becomes a necessary third channel rather than an optional add-on.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 6 minor

Summary. The manuscript presents a multimodal synchrotron approach (spatially resolved 2D-XRF, 2D-XRD, and X-ray ptychography) applied to archaeological and experimental samples of synthetic lead carbonate (psimythion). Conventional bulk 1D XRF/XRD cannot link minor/trace elements to specific phases; the authors argue that pixel-wise maps plus linear/multiple regression (with SNR gating) allow confident assignment of traces to crystalline phases or inference of non-crystalline hosts. Results on experimental 32X and archaeological 4A, 9A, 13A (plus prior 961/964) show consistent patterns: Fe never correlates with cerussite; Ag, when detected, tracks cerussite; Cu is variable; multiphase 9A shows intimate mixing of cerussite/calcite/quartz. The discussion frames these associations as constraints on manufacture, processing, and burial contributions.

Significance. If the element–phase assignments hold under tighter scrutiny, the work supplies a practical route beyond bulk averages for archaeometric metallic salts and related pigments/therapeutics, and it usefully integrates an experimental control prepared from the Theophrastus recipe. The maps, summed spectra/patterns, SNR values, phase wt% estimates, and ptychographic particle-size data are concrete deliverables. Strengths include the experimental–archaeological pairing, explicit acknowledgment that location does not equal mechanism, and the multiphase multiple-regression treatment of sample 9A. The methodological claim is of genuine interest to applied physics and cultural-heritage materials analysis; the statistical bar for “confident assignment,” however, currently limits how far the central claim can be taken.

major comments (4)
  1. [Methods; Results (Tables 1–6)] Methods and Results (R² threshold and SNR gate): The claim that the multimodal approach can “confidently assign” trace elements to named crystalline phases rests on pixel-wise linear (or multiple) regression with R²≥0.1 as the association cutoff and SNR>2 as the confidence gate, both set by visual inspection of the maps. R²=0.1 is a very weak linear association; many reported “associations” sit near this floor (e.g. Tables 1–6). Without a null model, permutation baseline, spatial-autocorrelation correction, or independent micro-analytical confirmation (e.g. SEM-EDS/μXANES on the same grains), co-location and mixture overlap cannot be rigorously separated from phase occupancy. The language of “confident assignment” should be tempered to “spatial co-variation under stated gates,” or the gates should be justified and stress-tested.
  2. [Results, Experimental sample 32X; Tables 1–2; Figure 4] Results, sample 32X (Tables 1–2, Figure 4): Whole-map Pb–cerussite R² collapses to ~0.02 because large grains fail the Bragg condition; only after post-hoc exclusion of the top-right region does R² rise to 0.23, and most trace-element R² values remain ≤0.03 with SNR≤2. The paper correctly flags the powder-grain-size limitation of XRD mapping, but the partial-map reanalysis is load-bearing for the experimental control yet is not accompanied by a pre-specified exclusion rule or sensitivity check. Either re-prepare/grind 32X to the recipe’s processing stage so that the powder condition is met, or treat 32X associations as inconclusive rather than as the paradigm for manufacture-only traces.
  3. [Results sample 9A; Table 6, Table 9; Discussion] Results/Discussion, sample 9A (Tables 6–7, S1; Discussion): Multiple regression attributes Pb/Cu/Ag/Cr to calcite+cerussite while Ca itself is essentially uncorrelated with calcite (R²~0.015). The authors note this can reflect intimate mixing or undetected/non-crystalline Ca hosts—correctly—but then still tabulate YCa/YCe-style associations in the summary Table 9. For multiphase samples the distinction between lattice incorporation, surface sorption, and spatial mixture overlap must be stated more sharply; otherwise the manufacture/processing/burial separation claimed in the abstract and conclusions overreaches what the regressions show.
  4. [Discussion; Summary and Conclusions; Table 9] Discussion and Summary/Conclusions: The leap from “Fe never with cerussite / Ag with cerussite when present” to excluding Fe from the metal plate and assigning Ag/Zn/Cr to the Pb metal is the paper’s applied payoff, but it inherits the soft statistical gates above and the incomplete processing of 32X. Fe’s host is inferred as amorphous hydroxide/oxyhydroxide or surface adsorption because no XRD peak is seen; that is a reasonable hypothesis, not a positive identification. Conclusions should separate (i) reproducible spatial patterns across samples from (ii) mechanistic origin claims that remain under-determined.
minor comments (6)
  1. [Method / Figure 1–2 captions] Figure numbering/caption inconsistency early on: the experimental-setup narrative refers to “Figures 1B and 1C” for the lead-plate powder while Figure 1 is reserved for the archaeological vessels; Figure 2 is the experimental plate. Align labels.
  2. [Table 9] Table 9 header has duplicated/typo labels (“NCa= not in Calcite; NCa= not in Calcite; YQz= yes in Quartz; YQz= yes in Quartz”). Clean the legend.
  3. [References] Duplicate reference entries for Yuan et al. 2016 appear as both [31] and [33].
  4. [Results, Experimental sample] Ptychography FOV for 32X is limited to 100×100 μm² “because of experimental issues” and yields no usable particle-size statistics; state this limitation once in Results and avoid implying structural parity with the archaeological samples in the comparative discussion.
  5. [Methods, Data processing; Table 7] Semi-quantitative XRD phase wt% precision is given as ±10 wt% at the 10 wt% level; propagate that uncertainty when comparing 9A (69/22/9) to the near-100% cerussite samples.
  6. [Abstract; Introduction] Abstract and Introduction claim the approach is “particularly suited to following changes… from manufacture, to use and burial, to excavation and conservation,” but conservation-stage effects are not actually measured here; soften or flag as prospective.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: observational multimodal correlations on new scans, not quantities forced by definition or self-citation identities.

full rationale

The paper’s load-bearing claims are measured pixel-wise associations (linear/multiple regression R² between 2D-XRF elemental maps and 2D-XRD phase maps, gated by map SNR) on newly scanned archaeological samples 4A/9A/13A and experimental 32X, plus particle-size stats from ptychography. Those R²/SNR numbers and the resulting element–phase table are empirical outputs of the scans, not algebraic rearrangements of fitted inputs or of quantities defined in terms of the claimed associations. Prior self-citations [6] and [7] supply the manufacturing recipe/context and earlier bulk/multimodal results on different samples (961, 964) that are merely summarized alongside the new data in Table 9; they are not invoked as uniqueness theorems, ansatzes, or identities that manufacture the present maps. Threshold choices (R²≥0.1, SNR>2, post-hoc exclusion of large Bragg-silent grains) are methodological judgment calls that affect confidence in the interpretation, but they do not make the reported correlations true by construction. No self-definitional loop, fitted-input-as-prediction, or renaming of a known result is present. Score 0 is therefore appropriate.

Assumptions & free parameters 4 free parameters · 5 assumptions · 0 invented entities

Load-bearing structure is experimental: synchrotron multimodal maps plus author-chosen statistical gates for ‘association’. No new physical entities. The claim rests on standard diffraction/fluorescence physics, the recipe/archaeological provenances, and ad hoc correlation thresholds and grain exclusions that directly control which element–phase links are called confident.

free parameters (4)
  • R² association threshold = 0.1
    Authors set R²=0.1 by visual inspection of maps as the cut between ‘some correlation’ and ‘low correlation’; this gate decides which trace elements are reported as phase-associated.
  • SNR confidence gate = SNR < 2 → low confidence
    SNR<2 is declared low confidence so that many weak R² values are discarded; chosen for this work without external calibration.
  • XRD phase wt% semi-quant precision = ±10 wt% at 10 wt%
    Diffrac.EVA integrated-peak normalization reported with estimated ±10 wt% at the 10 wt% level; used when stating cerussite/quartz/calcite fractions.
  • Partial-map exclusion of large grains (sample 32X) = exclude top-right large-grain region
    Top-right large Pb particles lacking XRD signal were discarded post hoc, raising Pb–cerussite R² from ~0.02 to 0.23; the exclusion region is hand-selected.
assumptions (5)
  • domain assumption Pixel-wise spatial correlation (linear or multiple regression) between XRF and XRD maps implies the trace element is hosted by or mixed with that crystalline phase at the 10 µm scale.
    Core interpretive step in Methods/Results; co-location is treated as phase assignment, with acknowledged ambiguity for overlapping mixtures (e.g. Pb with calcite+cerussite in 9A).
  • domain assumption XRD mapping is reliable only when crystallites are smaller than the illuminating beam so the sample behaves as a powder (Bragg condition statistically satisfied).
    Stated in Introduction with citation [13]; motivates exclusion of large grains in 32X and the grinding caveat in Discussion.
  • ad hoc to paper Absence of an XRD peak for a host of Fe (or excess Ca) implies a non-crystalline or out-of-angular-range phase rather than a failed detection of a major crystalline host.
    Used in Discussion to infer amorphous Fe hydroxide/oxyhydroxide or adsorbed/organic Fe when Fe is NCe across all samples.
  • domain assumption Theophrastus’s 4th c. BCE recipe and the stated burial/container contexts correctly describe manufacture and post-depositional pathways for the studied powders.
    Frames stages a–d and interpretation of Ag vs Fe vs quartz-hosted metals; Supplementary Note 1 provenances.
  • standard math Standard synchrotron XRF quantification from integrated peak areas normalized to 100% and XRD phase ID via Diffrac.EVA are adequate for relative trace/phase comparisons at the reported precision.
    Methods data-processing section; no full fundamental-parameters uncertainty budget.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Matching Trace Element Distribution to Mineralogical Phases in Ancient Biotechnology-Derived Metallic Salts: a Multimodal Analysis." pith.science (2026). https://pith.science/paper/Y7NQHJM7

@misc{pith2026260728492,
  author       = {Pith},
  title        = {Pith review of: Matching Trace Element Distribution to Mineralogical Phases in Ancient Biotechnology-Derived Metallic Salts: a Multimodal Analysis},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/Y7NQHJM7}},
  note         = {Machine review of arXiv:2607.28492}
}
read the original abstract

Conventional X-ray fluorescence (XRF) and X-ray diffraction (XRD) analysis applied to the investigation of ancient metal salts used as pigments and/or therapeutics provide bulk average compositions in two stand-alone data sets; however, major elements aside, these two sets cannot inform on the spatial distribution of one with respect to the other. To address this issue, we present here a multimodal approach incorporating spatially resolved XRF, XRD and nanoscale X-ray imaging applied to the analysis of archaeological and experimental samples of synthetic lead carbonate (PbCO3 - Greek psimythion); psimythion was used in antiquity as a cosmetic and/or a therapeutic for external applications. The experimental sample was produced according to a well-documented recipe dated to the 4th century BCE. In this paper we demonstrate that by using a multimodal approach we can confidently assign trace elements to individual crystalline or to infer the existence of non-crystalline phases. Although the assignment of an element to a phase (i.e. the location) is now possible, the origin underlying it (i.e. the mechanism) is not always clear. Trace elements do not 'control' chemical/mineralogical composition, but they can influence it. Our approach is particularly suited to following changes in the artefact's chemical/mineralogical profile, from its manufacture, to use and burial, to excavation and conservation.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

34 extracted references

  1. [1]

    Nature, 1999

    Walter, P ., et al., Making make-up in Ancient Egypt. Nature, 1999. 397(6719): p. 483-484

    1999

  2. [2]

    Analytical Chemistry, 2010

    Tapsoba, I., et al., Finding out Egyptian Gods’ secret using analytical chemistry: biomedical properties of Egyptian black makeup revealed by amperometry at single cells. Analytical Chemistry, 2010. 82(2): p. 457-460. 19

    2010

  3. [3]

    Communications Chemistry, 2018

    Beck, L., et al., Absolute dating of lead carbonates in ancient cosmetics by radiocarbon. Communications Chemistry, 2018. 1(1): p. 34

    2018

  4. [4]

    Caley, E.R. and J.F . Richards, Theophrastus on Stones: Introduction, Greek text, English translation, and commentary. 1956

    1956

  5. [5]

    2020, University of Glasgow

    Jocks, I.T., Scribonius Largus' Compounding of Drugs (Compositiones medicamentorum): introduction, translation, and medico-historical comments. 2020, University of Glasgow

    2020

  6. [6]

    Archaeol Anthropol Sci, 2020

    Photos-Jones, E., et al., On metal and 'spoiled' wine: analysing psimythion (synthetic cerussite) pellets (5th-3rd centuries BCE) and hypothesising gas-metal reactions over a fermenting liquid within a Greek pot. Archaeol Anthropol Sci, 2020. 12(10): p. 243

    2020

  7. [7]

    Open Access Journal of Archaeology and Anthropology, 2024

    Cipiccia, S., et al., Tracing impurities to minerals: overlapping XRF and XRD data sets derived from synchrotron techniques. Open Access Journal of Archaeology and Anthropology, 2024. 5(4): p. OAJAA. MS. ID. 000616

    2024

  8. [8]

    Keune, K. and J.J. Boon, Analytical imaging studies of cross-sections of paintings affected by lead soap aggregate formation. Studies in conservation, 2007. 52(3): p. 161- 176

    2007

Show all 34 references
  1. [9]

    Accounts of chemical research, 2010

    Cotte, M., et al., Synchrotron-based X-ray absorption spectroscopy for art conservation: looking back and looking forward. Accounts of chemical research, 2010. 43(6): p. 705- 714

    2010

  2. [10]

    Heritage, 2024

    Gianoncelli, A., et al., X-ray synchrotron radiation to look at pigments in antiquities: overview and examples. Heritage, 2024. 7(4): p. 2118-2137

    2024

  3. [11]

    Physics Reports, 2012

    Bertrand, L., et al., Development and trends in synchrotron studies of ancient and historical materials. Physics Reports, 2012. 519(2): p. 51-96

    2012

  4. [12]

    Rodenburg, J. and A. Maiden, Ptychography. Springer Handbook of Microscopy, 2019: p. 819-904

    2019

  5. [13]

    Nanoscale, 2015

    Birkbak, M.E., et al., Diffraction scattering computed tomography: a window into the structures of complex nanomaterials. Nanoscale, 2015. 7(44): p. 18402-10

    2015

  6. [14]

    Kallithrata et al, in preparation

  7. [15]

    Microscopy and Microanalysis, 2018

    Batey, D., et al., Coherence Branch at I13, DLS: The Multiscale, Multimodal, Ptycho- tomographic End Station. Microscopy and Microanalysis, 2018. 24(S2): p. 40-41

    2018

  8. [16]

    The ExcaliburRX-3M X-Ray Photon Counting Area Detector for Coherent Diffraction Imaging at the I13 Beamline at Diamond Light Source

    Williams, S.J., et al. The ExcaliburRX-3M X-Ray Photon Counting Area Detector for Coherent Diffraction Imaging at the I13 Beamline at Diamond Light Source. in 2017 IEEE Nuclear Science Symposium and Medical Imaging Conference (NSS/MIC). 2017

    2017

  9. [17]

    Hurst, and A

    Rodenburg, J., P . Hurst, and A. Cullis, Transmission microscopy without lenses for objects of unlimited size. Ultramicroscopy, 2007. 107(2-3): p. 227-231

    2007

  10. [18]

    Nature Communications, 2018

    Yu, Y .-S., et al., Three-dimensional localization of nanoscale battery reactions using soft X-ray tomography. Nature Communications, 2018. 9(1): p. 921

    2018

  11. [19]

    Scientific Reports, 2017

    Deng, J., et al., X-ray ptychographic and fluorescence microscopy of frozen-hydrated cells using continuous scanning. Scientific Reports, 2017. 7(1): p. 445

    2017

  12. [20]

    Maiden, A.M. and J.M. Rodenburg, An improved ptychographical phase retrieval algorithm for diffractive imaging. Ultramicroscopy, 2009. 109(10): p. 1256-1262

    2009

  13. [21]

    Enders, B. and P . Thibault, A computational framework for ptychographic reconstructions. Proceedings of the Royal Society A: Mathematical, Physical and Engineering Sciences, 2016. 472(2196): p. 20160640

    2016

  14. [22]

    Journal of Applied Crystallography, 2016

    Marchesini, S., et al., SHARP: a distributed GPU-based ptychographic solver. Journal of Applied Crystallography, 2016. 49(4): p. 1245-1252

    2016

  15. [23]

    Guizar-Sicairos, M. and J.R. Fienup, Phase retrieval with transverse translation diversity: a nonlinear optimization approach. Optics Express, 2008. 16(10): p. 7264-7278

    2008

  16. [24]

    Journal of Applied Crystallography, 2020

    Wakonig, K., et al., PtychoShelves, a versatile high-level framework for high- performance analysis of ptychographic data. Journal of Applied Crystallography, 2020. 53(2): p. 574-586. 20

    2020

  17. [25]

    Rau, and S

    Batey, D., C. Rau, and S. Cipiccia, High-speed X-ray ptychographic tomography. Scientific Reports, 2022. 12(1): p. 7846

    2022

  18. [26]

    2014, University of Sheffield: Sheffield

    Batey, D., Ptychographic Imaging of Mixed States. 2014, University of Sheffield: Sheffield

    2014

  19. [27]

    Rasband, and K.W

    Schneider, C.A., W.S. Rasband, and K.W. Eliceiri, NIH Image to ImageJ: 25 years of image analysis. Nat Methods, 2012. 9(7): p. 671-5

    2012

  20. [28]

    O'Day, and M

    Carroll, S.A., P .A. O'Day, and M. Piechowski, Rock−Water Interactions Controlling Zinc, Cadmium, and Lead Concentrations in Surface Waters and Sediments, U.S. Tri-State Mining District. 2. Geochemical Interpretation. Environmental Science & Technology,

  21. [29]

    Inorganic Chemistry, 2012

    Smith, S.J., et al., Novel Synthesis and Structural Analysis of Ferrihydrite. Inorganic Chemistry, 2012. 51(11): p. 6421-6424

    2012

  22. [30]

    Environmental Science & Technology, 2014

    Callagon, E., et al., Incorporation of Pb at the Calcite (104)–Water Interface. Environmental Science & Technology, 2014. 48(16): p. 9263-9269

    2014

  23. [31]

    Environmental Science & Technology, 2016

    Yuan, K., et al., Replacement of Calcite (CaCO3) by Cerussite (PbCO3). Environmental Science & Technology, 2016. 50(23): p. 12984-12991

    2016

  24. [33]

    Matching Trace Element Distribution to Mineralogical Phases in Ancient Biotechnology -Derived Metallic Salts: a Multimodal Analysis

    Yuan, K., et al., Replacement of Calcite (CaCO3) by Cerussite (PbCO3). Environmental Science & Technology, 2016. 50(23): p. 12984-12991. 21 Supplementary Material for: “Matching Trace Element Distribution to Mineralogical Phases in Ancient Biotechnology -Derived Metallic Salts...

    2016

  25. [1998]

    Description This is a pit grave oriented along an east -west axis

    Excavators M Bessios and Athena Athanassiadou. Description This is a pit grave oriented along an east -west axis. It contained the burial of a woman; her skull laid at the eastern end. The grave had been looted; the tomb robbers had disturbed only the area around the skull, wi...

    1924

  26. [2009]

    A pit grave oriented along the N -S axis

    Excavators M Bessios and Athena Athanassiadou. A pit grave oriented along the N -S axis. It was cut into the fill of an earlier grave, so its boundaries could not be precisely determined, and part of the skeleton was disturbed during excavation. The deceased was a young indivi...

Pith tools

Reviewed July 31, 2026 · model on record in the stance chip above.