First Identification of a Gypsum-Based Preparatory Layer on Polychrome Wooden Figurines from the Mawangdui Han Tomb No. 1 (2nd Century BCE), Changsha, China
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsThe paper under review is devoted to the investigation of ancient wooden figurines found in a tomb of Han dynasty times. This is a very interesting multidisciplinary work, which is good as from the historical, as from materials science points of view. Paper reports on studies of mineral composition of preparatory layer and pigments that cover figurines. In general, I believe it should be published after minor revision of several moments listed below.
The first question is about the detection of amorphous material in a studied sample (lines 238, 248, etc.). As for me, I didn’t find any real evidence of it presence. SEM images show rather significant amount of this “amorphous” phase, which should result in a presence of well visible humps on the XRD pattern. In addition, you wrote that “overlap at 29.19° indicates … fine-grained or amorphous phases”, which is not correct as well. If you have well-shaped diffraction peaks, the phase can’t be really amorphous. Of course, it might have low crystallinity in comparison to other studied phases, but not amorphous. There are only two ways of amorphous phase to be detected that I can see here: 1) measure its chemical composition using SEM EDX and show that you don’t have such compounds found by XRD; 2) If you have SEM equipped with EBSD detector, you can use it instead of XRD to show an absence of diffraction. So, your phrase that “SEM analysis further revealed the presence of amorphous film-like or fibrous material on the surface of gypsum crystals” (lines 490-491) can be regarded as correct only in second case. Using SEM only to get images, or in combination with EDX won’t give you information on amorphous or crystalline origin. By the way, you have mentioned Rietveld refinement, but didn’t give an image of profile (powder pattern) convergence – is it not good to be shown? Are there many unaccounted peaks left?
The second moment relates to the discussion of larnite appearance in the studied sample (lines 468-475). Your explanation that non-uniform calcination took place is more philosophical than experimentally confirmed. Moreover, further it’s written that presence of carbonates may be a result of “natural impurities introduced during raw material sourcing”. I think that larnite, which also associates with quartz in some rocks, is also a natural contaminant. And material was prepared at temperatures not more than ~400 C. Of course, some non-uniform calcination very likely took place to form some amounts of anhydrite, but not to form larnite.
Minor comments:
- line 55: “major polity” – may be better is “major state”?
- line 60: “such” – should starts with capital S.
- line 233: “although” – should starts with capital A.
- line 237: “structural arrangement” – may be change for “texture” or “aggregation”?
- line 306: “Spectroscopic Composition” – is not correct title. Think about “Mineralogical composition and spectroscopy studies”.
Author Response
248, etc.). As for me, I didn’t find any real evidence of it presence. SEM images show rather significant amount of this “amorphous” phase, which should result in a presence of well visible humps on the XRD pattern. In addition, you wrote that “overlap at 29.19° indicates … fine-grained or amorphous phases”, which is not correct as well. If you have well-shaped diffraction peaks, the phase can’t be really amorphous. Of course, it might have low crystallinity in comparison to other studied phases, but not amorphous. There are only two ways of amorphous phase to be detected that I can see here: 1) measure its chemical composition using SEM EDX and show that you don’t have such compounds found by XRD; 2) If you have SEM equipped with EBSD detector, you can use it instead of XRD to show an absence of diffraction. So, your phrase that “SEM analysis further revealed the presence of amorphous film-like or fibrous material on the surface of gypsum crystals” (lines 490-491) can be regarded as correct only in second case. Using SEM only to get images, or in combination with EDX won’t give you information on amorphous or crystalline origin.
Response by QL：
We sincerely thank the reviewer for this insightful comment. We acknowledge that the original manuscript did not provide conclusive evidence for the presence of amorphous phases, and this was an oversight on our part. Indeed, without additional structural analysis, such as EBSD, we are not able to definitively determine the amorphous or crystalline nature of the observed features. Unfortunately, our laboratory does not currently have access to EBSD facilities, but we hope to revisit this issue in future studies using preserved samples.
we have revised the manuscript accordingly:
In lines 238–244, the discussion of microstructural features has been extensively rewritten. All mentions of “amorphous” have been removed and replaced with non-specific terms such as “film-like” and “structureless”, which more accurately describe the observed morphology without implying structural disorder.
In line 316, we have deleted the phrase “and potentially the presence of fine-grained or amorphous phases,” which we now recognize as an unsupported assumption based on the available XRD data.
In line 490, we retained only the description “film-like” to characterize surface features observed in SEM, again avoiding any unverified reference to amorphous character.
By the way, you have mentioned Rietveld refinement, but didn’t give an image of profile (powder pattern) convergence – is it not good to be shown? Are there many unaccounted peaks left?
Response by QL：
We appreciate the reviewer’s attention to the completeness of the Rietveld refinement analysis. The profile fitting image has been prepared and will be included as a Supplementary Figure (Figure S1) in the revised submission.
According to our refinement results, the main crystalline phases are clearly identified as gypsum, larnite, quartz, calcite, and anhydrite. The profile shows quite good agreement between the observed and calculated patterns, and nosignificant unexplained peaks observed. The refinement parameters supporting the reliability of the phase identification.
The second moment relates to the discussion of larnite appearance in the studied sample (lines 468-475). Your explanation that non-uniform calcination took place is more philosophical than experimentally confirmed. Moreover, further it’s written that presence of carbonates may be a result of “natural impurities introduced during raw material sourcing”. I think that larnite, which also associates with quartz in some rocks, is also a natural contaminant. And material was prepared at temperatures not more than ~400 C. Of course, some non-uniform calcination very likely took place to form some amounts of anhydrite, but not to form larnite.
Response by QL：
We thank the reviewer for raising this important point. We agree that the current explanation regarding the formation of larnite through non-uniform calcination is hypothetical and not directly supported by experimental confirmation.
As the reviewer correctly points out, larnite can occur as a natural mineral phase, particularly as an associate of quartz in specific geological settings. However, in the context of our study, it is worth noting that gypsum deposits in Hunan Province, which are sedimentary in origin typically do not provide the geo-thermal or metamorphism conditions for larnite formation, however, the exact provenance of the gypsum used in these mawangdui is currently unknown, and we agree that this uncertainty prevents us from drawing any firm conclusions regarding natural contamination. As such, we have decided not to include speculative interpretations in the revised manuscript.
Nevertheless, based on archaeological and historical studies of early pyrotechnology, including ancient metallurgy and proto-ceramics, we believe that temperatures approaching 1000°C could have been locally achieved in small, enclosed kiln environments—even using wood or charcoal as fuel. This temperature range falls within the stability field for larnite formation, especially under conditions involving very small localized hot spots condition. Therefore, we propose to retain the non-uniform calcination hypothesis in the revised text, but with revised wording to clarify its speculative nature.
In response to the reviewer’s valuable suggestion, we will also incorporate an additional statement acknowledging the alternative possibility that larnite may derive from mineralogical associations with quartz in the raw gypsum material
We appreciate the reviewer’s critical insights, which have helped us improve both the clarity and caution of our interpretation.
Minor comments:
- line 55: “major polity” – may be better is “major state”?
- line 60: “such” – should starts with capital S.
- line 233: “although” – should starts with capital A.
- line 237: “structural arrangement” – may be change for “texture” or “aggregation”?
- line 306: “Spectroscopic Composition” – is not correct title. Think about “Mineralogical composition and spectroscopy studies”.
Response by QL：
We thank the reviewer for their careful reading and helpful suggestions. All minor issues have been addressed as follows:
We appreciate the reviewer’s attention to these details, which have helped improve the readability and precision of the manuscript.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for Authorscoatings-3590395
Title: First Identification of a Gypsum-Based Preparatory Layer on Polychrome Wooden Figurines from a Mawangdui Han Tomb No. 1 (2nd Century BCE), Changsha, China
Authors: Ningning Xu, Tingyan Ren, Pan Xiao and Qi Liu
Section: Surface Characterization, Deposition and Modification
Special Issue: Advanced Technology in Surface Characterization and Conservation for Architectural and Archaeological Heritage
The authors present a detailed study of the white preparatory layer and polychrome pigments on painted wooden figurines excavated from Mawangdui Tomb No. 1, dating to the Han dynasty. For me, this is a very positive aspect of applying science to understanding the life of ancient civilisations. The techniques used, such as XRF mapping, SEM, AT-IR / Raman and XRD are typical in such cases. Although, from a chemical point of view, the gypsum and calcite with minor phase such as anhydrite and larnite have been studied, in general the article presents a high level of analysis. The background of the methods used and the measurement conditions are adequately described (with all detailed information such as scans, resolution and etc.). All in all, I have one suggestion. IR and Raman methods are complementary. Have the authors ever heard of the fact that a mode can be IR active, Raman inactive and vice-versa however not at the same time? This fact is called the mutual exclusion rule. For this reason, I suggest that sections 3.3.2 and 3.3.3 be combined.
For molecules with little or no symmetry, the modes are likely to be active in both IR and Raman. In case of SO42 ions: For an isolated SO4 2– ion (Td symmetry), there are four normal modes: ν3(F2) = ca. 1150 cm–1 (triply degenerate νas(S–O)); ν1(A1) = ca. 980 cm–1 (νs(S–O)); ν4(F2) = ca. 600 cm–1 (triply degenerate δas(OSO)) and ν2(E) = ca. 450 cm–1 (doubly degenerate δs(OSO)). All the four vibrations are Raman active, whereas only ν3 and ν4 are infrared active [66, 67]. If the Td symmetry of the SO42– ion is lowered, i.e. by intermolecular hydrogen bonding, the degenerate vibrations are split and the Raman-active modes also become infrared active. I agree with the general assignment of the author. However, if the authors have mentioned about ‘’ A weaker band at 1009 cm⁻¹ corresponds to the symmetric stretching mode (ν1),’’ it needs more comment about the lowering symmetry of SO42- (or just distorted Td symmetry), which activated the ν1 vibrations in IR spectrum.
Some references: - M. Alikhani, M. Hakimi, K. Moeini, V. Eigner, M. Dusek, J. Inorg. Organomet. Polym. Mater. 30, 2907 (2020) / - K. Nakamoto, Infrared and Raman Spectra of Inorganic and Coordination Compounds, part B (New York Inc, J. Wiley & Sons, 2009) / - M.K. Marchewka, J. Chem. Res. 8, 518 (2003) / - E.A. Secco, Can. J. Chem. 66, 329 (1988)
- b) part 3.3.3 Raman Identification – in my opinion it is not necessary to write (lines 370-378)‘’ Compared to ATR-FTIR, which offers a broader overview of composite materials, confocal Raman spectroscopy enables the targeted identification of specific microstructures within heterogeneous samples. This technique allows for more localized interpretation of molecular composition at the microscale. In materials with low crystallinity and limited long-range order, Raman scattering bands tend to broaden, making weak peaks more difficult to resolve. A smaller particle size also reduces Raman scattering efficiency, attenuating low-intensity signals. In addition, factors such as molecular orientation, hydrogen bonding, and intermolecular interactions may enhance or suppress certain vibrational modes, causing dominant peaks to appear while others remain undetected.’’
It is unfortunate that the authors did not include a comparison of the IR and Raman spectra of their samples with the gypsum (CaSO₄·2H₂O) and calcite (CaCO₃).
Due to the above, I consider that the work can be interesting for Coatings readers. I propose to accept after minor revisions.
Comments for author File: Comments.pdf
Author Response
All in all, I have one suggestion. IR and Raman methods are complementary. Have the authors ever heard of the fact that a mode can be IR active, Raman inactive and vice-versa however not at the same time? This fact is called the mutual exclusion rule. For this reason, I suggest that sections 3.3.2 and 3.3.3 be combined.
Response by QL：
We thank the reviewer for this suggestion. In the revised manuscript, Sections 3.3.2 and 3.3.3 have been combined into a single subsection to better reflect the complementary nature of ATR-FTIR and Raman spectroscopy.
For molecules with little or no symmetry, the modes are likely to be active in both IR and Raman. In case of SO42 ions: For an isolated SO4 2– ion (Td symmetry), there are four normal modes: ν3(F2) = ca. 1150 cm–1 (triply degenerate νas(S–O)); ν1(A1) = ca. 980 cm–1 (νs(S–O)); ν4(F2) = ca. 600 cm–1 (triply degenerate δas(OSO)) and ν2(E) = ca. 450 cm–1 (doubly degenerate δs(OSO)). All the four vibrations are Raman active, whereas only ν3 and ν4 are infrared active [66, 67]. If the Td symmetry of the SO42– ion is lowered, i.e. by intermolecular hydrogen bonding, the degenerate vibrations are split and the Raman-active modes also become infrared active. I agree with the general assignment of the author. However, if the authors have mentioned about ‘’ A weaker band at 1009 cm⁻¹ corresponds to the symmetric stretching mode (ν1),’’ it needs more comment about the lowering symmetry of SO42- (or just distorted Td symmetry), which activated the ν1 vibrations in IR spectrum.
Some references: - M. Alikhani, M. Hakimi, K. Moeini, V. Eigner, M. Dusek, J. Inorg. Organomet. Polym. Mater. 30, 2907 (2020) / - K. Nakamoto, Infrared and Raman Spectra of Inorganic and Coordination Compounds, part B (New York Inc, J. Wiley & Sons, 2009) / - M.K. Marchewka, J. Chem. Res. 8, 518 (2003) / - E.A. Secco, Can. J. Chem. 66, 329 (1988)
Response by QL：
We fully agree with the reviewer that the mutual exclusion rule is an important theoretical principle when analyzing the spectroscopic activity of vibrational modes, particularly in symmetric anions such as SO₄²⁻. We appreciate the reviewer’s insightful observation regarding the ν₁ (A₁) mode of SO₄²⁻ appearing in the IR spectrum at 1009 cm⁻¹.
In response to this, we have made the following revisions in the revised manuscript:
in Section 3.3.2, we have added a sentence noting that the appearance of the ν₁ mode in the IR spectrum likely indicates a lowering of the ideal Td symmetry of the SO₄²⁻ ion, potentially due to intermolecular hydrogen bonding or structural distortions in the sulfate matrix. The discussion now acknowledges that such symmetry lowering can result in IR activation of modes that are otherwise Raman-exclusive under ideal symmetry.
We have also cited the references recommended by the reviewer Nakamoto (2009), to support this interpretation.
Furthermore, based on this important addition, we have also indicated the position of 1009 cm⁻¹ in Figure 8(a).
- b) part 3.3.3 Raman Identification – in my opinion it is not necessary to write (lines 370-378)‘’ Compared to ATR-FTIR, which offers a broader overview of composite materials, confocal Raman spectroscopy enables the targeted identification of specific microstructures within heterogeneous samples. This technique allows for more localized interpretation of molecular composition at the microscale. In materials with low crystallinity and limited long-range order, Raman scattering bands tend to broaden, making weak peaks more difficult to resolve. A smaller particle size also reduces Raman scattering efficiency, attenuating low-intensity signals. In addition, factors such as molecular orientation, hydrogen bonding, and intermolecular interactions may enhance or suppress certain vibrational modes, causing dominant peaks to appear while others remain undetected.’’
Response by QL：
We thank the reviewer for this suggestion. In response, we have deleted the paragraph in lines 370–378, which contained general information about Raman scattering mechanisms.
It is unfortunate that the authors did not include a comparison of the IR and Raman spectra of their samples with the gypsum (CaSO₄·2H₂O) and calcite (CaCO₃).
Response by QL：
We appreciate this comment and agree that direct comparison enhances interpretive clarity. To address this, we have added a Supplementary Figure (Figure S2) presenting the ATR-FTIR and Raman spectra of standard gypsum and calcite, alongside the spectra of our sample. We used publicly available reference data from the RRUF Project database for this comparison. Figure S2 demonstrates strong spectral correspondence between the sample and the reference materials, particularly in the SO₄²⁻ and CO₃²⁻ vibrational regions, supporting the accuracy of our spectral assignments.
Due to the above, I consider that the work can be interesting for Coatings readers. I propose to accept after minor revisions.
Author Response File: Author Response.pdf
