In Situ Investigation of the Medieval Copper Alloy Door in Troia (Southern Italy)

Round 1

Reviewer 1 Report

The work deals with the analysis of the elemental composition of a copper alloy door in Troia, the motivation for the work is quite interesting, especially regarding the reduced amount of this kind of door still present today. The authors selected door regions to be analyzed and presented the results of elemental composition by XRF.

I think the article is adequate and in line with Heritage journal, and I describe below my suggestions for its improvement.

General comments:

1. The authors discuss at the beginning of the abstract that they will describe the chemical composition of the patina on the alloy and inlays, however in the rest of the manuscript they treat it as an analysis of the metallic alloy and no longer of the patina. In my opinion this can be a problem, as they discuss the composition of the metal alloy from which the door is made, however, the patina layer on the surface does not necessarily represent the same proportion between the alloy elements of the substrate, as certain elements can be attacked differently by corrosion and give rise to patinas with different compositions from the original base. Even if you do not consider this aspect, a layer of corrosion products formed by dense elements, such as copper and lead, will inevitably attenuate the characteristic X-ray signal of the substrate, which will be reduced in comparison with the signal of the patina elements. Thus, it is difficult to directly relate the surface analysis with the substrate composition.

2. The introduction is well grounded and adequately addresses the motivation behind the work.

3. The first part of methods section is quite compelling, and I think the description of the door and its parts is important to better understanding of the work. The second part, however, could be improved especially regarding the sampling and characteristics of the XRF equipment.

4. I think the results and discussion section is lacking a bit of discussion around the elemental composition values obtained, I have discussed some points in the specific comments.

5. In my opinion, some of the conclusions are too restrictive and are not necessarily fully supported by the results. This need to be revised and rediscussed by the authors after addressing the specific comments suggested bellow.

6. English is clear and understandable, but needs to be homogenized, some parts are in American English and others in British English. For example: moulded (UK) analized (US)

7. I suggest the use of CE instead of AD to disassociate the scientific language from the religious perspective, making the text more inclusive.

8. Sometimes the authors call the master foundryman Oderisius, sometimes Oderisio, I understand they are interchangeable, from Latin and Italian forms of the name, but I suggest using only one of them to let the text clearer.

 

Specific comments:

 

9. L. 50: the hyphen in lost-wax is missing and it is not necessary put it in quotes, because it is the correct term for this process, even industrially nowadays. The same for sand or molded casting in line 58. The same in line 59.

10. L. 101: it is the only “translation” from Latin to English that was put in italic and quoted, the authors must always use the same writing features for every “translation”.

11. Figure 4: The notations a) and b) in figure 4 are not explained in the caption, information of which protome corresponds to a) and b) would be interesting.

12. Figure 6: the authors could inform in which part of the door the picture of complementary engravings was taken; I was curious, but did not manage to find in Figure 1.

13. L. 204: The presented resolution was at the energy of 5.9 keV, close to the peaks of copper, but very far from the characteristic energies of lead and, particularly, of tin. The authors could present the resolution for higher energy ranges. Furthermore, nothing is mentioned about the quantification method. Do the authors obtain the spectrum and then perform the integration or is the composition read directly from the equipment reader? The XRF spectra of each analyzed area could be included as supplementary material.

14. L. 211: The choice of standards is not clear. What are the 5 compositions of the standards? How do they were chosen? why did they use carbonates and oxides if they didn't analyze carbon and oxygen during the experiments? I think the authors must better explain this section of their methods.

15. The authors showed the analyzed points in Figure 1, but there is no explanation around their criteria to choose these points. I think this need to be mentioned and well explained in the methods section. For instance, and perhaps it is just a personal difficult to see in the picture, but lots of points are on dark regions, few on the green patina, and none in the reddish “clean” bronze on the top panels.

16. L. 228: Why the nails would be made of iron or copper? The composition of points 27 and 28 as called “nails” presented 70 to 73 % of copper, around 2% of lead, 7 to 13% of tin and 2 to 4,3 % of iron, as the table indicates, and the authors described in L. 261 and L.262. They seem to be made of brass, alloyed with iron perhaps, less amount of lead, which is understandable, because lead reduces at some extend the mechanical properties of the alloy, which is important for nails, but I don’t understand how they could be iron or copper only. The authors need to clarify this point.

17. From L. 237 to L. 241: all the restored areas must be identified in Table 1, not only those from 14 to 17. It was rather complicated to keep track of which parts were and were not restored when just looking the table.

18. From L. 247 to L. 256: I think the authors correctly did not quantify the silver in the inlays, but in my opinion, they should have not quantified the other elements either. For example, they claim that inlays 18, 19 and 20 are made of copper with a high content of lead and traces of silver, but measure 19 has only 6% lead, less than zinc, which is not even mentioned until the end of paragraph.

Additionally, if the silver foils were applied onto the base alloy, then silver is not an alloying element, but a coating. And this brings a problem with the capability of XRF, as the authors mentioned, all samples had infinity thickness, which is not true in the case of coated metals, for instance this group of analysis. As silver is highly dense and have a high mass attenuation coefficient in the range of some keV’s, even a foil with 10 microns will attenuates most of the signal. The authors must address this issue in the paper and discuss this part in a more qualitative way.

19. From L. 267 to L. 271: The authors claim the chlorine concentration could not be from atmospheric corrosion, but instead from chemicals used in the 1951 restoration. I would not be this categoric. Firstly, because few greenish areas, which are typical of copper chloride patinas, seem to be analysed. Secondly, in the bottom frames, the inscriptions are greener, and I assume that these parts are decaled, which generates electrolyte confinement and consequently concentration of the solution in whatever ion is present, for example chlorides.

Besides this, rains inevitably contain some chloride concentration, this article (doi:10.1016/j.corsci.2008.10.008) collected rain from north Italy and shows appreciable amounts of Cl^-. During wet and dry cycles, chlorides could concentrate in the surface causing chlorine-rich patinas.

Of course, this also could be related to the restoration chemicals, but for this argument to be more solid, a better description of the patination process applied in 1951 must be included. It is a shame that this reference is not easy to be found, at least it is not available online in the archivi dell’Istituto Centrale del Restauro.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments for author File: Comments.pdf

Author Response

We thank the Reviewer for his valuable comment. We are truly sorry for his words.

Our aim was to investigate the patinas of these precious medieval doors using the portable instrumentation we have available.

The method used in this research is certainly not innovative and it is similar to that used in publication [1], but the case study is different.

In particular, the door of Troy is part of a limited series of medieval artifacts on which there is not much scientific literature. In our opinion, any additional information we can discover will certainly help future research in the field. Also for this reason, it is important to compare the results obtained in the various investigated doors.

We have edited much of the manuscript and we hope that this revised version will be appreciated for eventual publication.

Reviewer 3 Report

The presented work is a great contribution to understanding and preserving this great work of art.

I would accept this work without any changes.

Author Response

We thank the Reviewer for his comment. We are honored to read his words.

Round 2

Reviewer 1 Report

I think the authors have improved several points of the manuscript. But some aspects could still be better discussed. In particular with regard to self-absorption, not necessarily what it is, but which is the implications in this research. For example, one of my questions in the first round was about how the authors used "alloy composition", which is the base material when testing only the surface.

The authors could have exploited this fact in their discussion. But instead, they only changed the term to "alloy patina" all-around the manuscript. In some parts this makes sense and the text was more correct, however, in other parts, the two things were not interchangeable or needed some extra clarification. For example in the conclusions:

"The experimental results obtained have shown that the two wings and the frame of the medieval door in Troia have the same patina’s chemical composition". Here they could correlate the patina composition with the substrate because the same patina indicates de same substrate.

However here: "A binary copper-lead alloy patina was used in their production in a ratio of about 5:1. The alloy patina of the studied door is however different from eight medieval copper doors analyzed in Italy."

the alloy patina was not at all used in their production. In this case, was the substrate. The authors could then link with the relationship between patina composition and alloy composition.

Regarding the methods, nothing was discussed about the integration method used to calculate the composition from the spectra.

Author Response

Please see attached document.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments for author File: Comments.pdf

Author Response

Please see attached document.

Author Response File: Author Response.pdf