*3.4. XRF Imaging*

method (−1.2 V to 1.2 V).

Cl-

A closer look at Figure 2 shows the optical images before and after cleaning, as well as the Hg L<sup>α</sup> image. It is apparent from Figure 2C that the XRF image has revealed a finely detailed portrait of the little girl and the pretty purse with a very clean background, as if all the tarnish were removed. The most noticeable difference is that, while the various cleaning methods applied to the plate show nonuniform cleaning in the optical image depending on the cleaning condition and the region of interest, removal of tarnish from the daguerreotype is complete in the XRF image, which reveals fine details everywhere across the portrait.

The XRF image can be fitted using the software package PyMCA [19,20], in which the X-ray fluorescence peaks are fitted and the area under the curve is the intensity contributing to the image. This procedure removes contributions from the overlapping peaks, such as Au L<sup>α</sup> in the case of Hg Lα. An XRF image can also be obtained using Hg Lβ, Au Lα, and Au Lβ. The images from the Hg L PyMCA fit, and Hg L<sup>β</sup> and Au L<sup>α</sup> without fitting, are shown in Figure 6. It is interesting to note that the image is also finely revealed in the Hg L<sup>β</sup> map and is noticeable in the Au map, albeit thinly veiled. The presence of a veiled Au image indicates that a gilding step using a gold chloride solution was applied after the image particles were formed and fixed when the plate was made, so that Au was found all over the plate, tracking how the gliding was done at the time. An image can still be observed from the Au fluorescence, suggesting that Au tracks the density and the distribution of the image particles, as well as the featureless region of the Ag plate.

*Heritage* **2021**, *4* FOR PEER REVIEW 9

across the portrait.

depending on the cleaning condition and the region of interest, removal of tarnish from the daguerreotype is complete in the XRF image, which reveals fine details everywhere

bution of the image particles, as well as the featureless region of the Ag plate.

The XRF image can be fitted using the software package PyMCA [19,20], in which the X-ray fluorescence peaks are fitted and the area under the curve is the intensity contributing to the image. This procedure removes contributions from the overlapping peaks, such as Au Lα in the case of Hg Lα. An XRF image can also be obtained using Hg Lβ, Au Lα, and Au Lβ. The images from the Hg L PyMCA fit, and Hg Lβ and Au Lα without fitting, are shown in Figure 6. It is interesting to note that the image is also finely revealed in the Hg L<sup>β</sup> map and is noticeable in the Au map, albeit thinly veiled. The presence of a veiled Au image indicates that a gilding step using a gold chloride solution was applied after the image particles were formed and fixed when the plate was made, so that Au was found all over the plate, tracking how the gliding was done at the time. An image can still be observed from the Au fluorescence, suggesting that Au tracks the density and the distri-

**Figure 6.** XRF image from left to right using the Hg L PyMCA fit, and Hg Lβ and Au Lα without fitting. **Figure 6.** XRF image from left to right using the Hg L PyMCA fit, and Hg L<sup>β</sup> and Au L<sup>α</sup> without fitting.

When comparing the Hg Lα image in Figure 2C with the Hg L PyMCA and L<sup>β</sup> images (Figure 6), it appears that while the Lα and Lβ images without fitting were of equally good quality, the PyMCA image showed a slightly better spatial resolution and contrast at closer scrutiny. This is because the Hg and Au Lα lines could not be completely resolved with the solid-state detector (SSD) without fitting. We also tracked the Cu Kα line, which did not show any image as the signal came from the Cu plate, and Cu was not involved in the formation of image particles. The Ag L emission was too weak at this excitation energy to be detected, and did not reveal any image either. It will be of interest to track Ag with tender X-ray excitation at just above the Ag L3 edge (3351 eV) [3]. When comparing the Hg L<sup>α</sup> image in Figure 2C with the Hg L PyMCA and L<sup>β</sup> images (Figure 6), it appears that while the L<sup>α</sup> and L<sup>β</sup> images without fitting were of equally good quality, the PyMCA image showed a slightly better spatial resolution and contrast at closer scrutiny. This is because the Hg and Au L<sup>α</sup> lines could not be completely resolved with the solid-state detector (SSD) without fitting. We also tracked the Cu K<sup>α</sup> line, which did not show any image as the signal came from the Cu plate, and Cu was not involved in the formation of image particles. The Ag L emission was too weak at this excitation energy to be detected, and did not reveal any image either. It will be of interest to track Ag with tender X-ray excitation at just above the Ag L<sup>3</sup> edge (3351 eV) [3].

### **4. Conclusions 4. Conclusions**

We conducted chemical and electrochemical cleaning procedures on various sites from a single, partially tarnished daguerreotype. We found that both methods of cleaning were effective at removing the tarnish and restoring the image. While all procedures seemed to improve the image by some degree, the tarnish was never removed uniformly or entirely. The chemical cleaning procedures were sufficient to remove the halides/white haze from the surface. Optical images taken after cleaning still appeared to have areas with a brown/orange tinge on the surface. Electrochemical cleaning was sufficient at removing the sulfides from the surface in addition to the halides, and did it faster. Both the Barger and the Wei electrocleaning methods improved the visual appearance of the image. Again, they did not always remove the orange/brown tarnish colour form the surface. We conducted chemical and electrochemical cleaning procedures on various sites from a single, partially tarnished daguerreotype. We found that both methods of cleaning were effective at removing the tarnish and restoring the image. While all procedures seemed to improve the image by some degree, the tarnish was never removed uniformly or entirely. The chemical cleaning procedures were sufficient to remove the halides/white haze from the surface. Optical images taken after cleaning still appeared to have areas with a brown/orange tinge on the surface. Electrochemical cleaning was sufficient at removing the sulfides from the surface in addition to the halides, and did it faster. Both the Barger and the Wei electrocleaning methods improved the visual appearance of the image. Again, they did not always remove the orange/brown tarnish colour form the surface. As noted in the introduction, daguerreotypes can vary significantly depending on when they were made, the methods and equipment of the artist who made them, and the conditions under which they were stored. By performing these experiments on small regions of the same daguerreotype, we tried to obtain the most consistent possible initial conditions for comparison. The effectiveness of the treatment will depend entirely on the original integrity of the daguerreotype. They are not uniform. This work confirms previous observations that the only sure way to retrieve the complete contents of the daguerreotype is through synchrotron radiation X-ray fluorescence imaging. This technique will ensure that events of historical significance from a tarnished plate can be retrieved. With the XRF method, even if the daguerreotype is severely tarnished, provided there is still sufficient mercury on the image particles on the surface, the image in its entirety can still be reconstructed through digitizing the XRF images. The cleaning methods have been shown to improve the image optically; nevertheless, this should be undertaken with caution. Once the Hg is gone, the image will be lost forever. To refine the daguerreotype cleaning methods even more, further research would need to be performed to determine the detailed chemical

composition of the substrate plate, the image particles, the tarnish, and its interplay with the environment, such as the daguerreotype housing and the protecting glass.

**Supplementary Materials:** The following are available online at https://www.mdpi.com/article/10 .3390/heritage4030089/s1. Figure S1: Experimental setup for electrochemical cleaning and chemical cleaning. (a) Schematic for the three-electrode set up for electrocleaning. (b) Actual set up for electrocleaning; the area of interest is confined by the perimeter of the cell, which leaves an oval mark on the plate after cleaning. The working electrode (daguerreotype), counter electrode (Pt), and reference electrode (Ag/AgCl) are noted. (c) Setup for chemical cleaning with the three-well cell clamped down on the plate. This setup leaves behind three small oval marks on the plate. (see Figure 2, middle panel and text). Figure S2: Experimental arrangement for the XRF imaging. The focussed beam (yellow line from left to right) with a spot size of 30 µm × 20 µm is stationary. The plate is mounted on a three-axis stage that moves the plate across the beam with submicrometre precision, pixel by pixel. The fluorescence X-rays are collected with a four element SSD (VotexME4). The data are stored in a multichannel analyzer (MCA). Desired energy windows are set to collect element-sensitive maps (see Figure S4). Figure S3: A snapshot of the MCA display during a scan (top); the abscissa is photon energy and the ordinate axis is intensity in a semi-log plot. The Hg L intensities fit using PyMCA are shown in the bottom panel (both L<sup>α</sup> and L<sup>β</sup> are used, black dotted curve). Figure S4: EDX maps of Ag, S, Au, Hg, and Cl, and the backscattered (BSE) SEM image (black and white) for the chemical cleaning solutions discussed in Figure 3A,B. A: before cleaning; B: after 3% Na2SO<sup>3</sup> .

**Author Contributions:** Conceptualization, R.R.M., J.J.N., and T.-K.S.; methodology, A.S., F.F., J.J.N., Y.Z.F., and T.-K.S.; lab experiment and analysis, A.S. and F.F.; A.S. is an undergraduate student under the supervision of J.J.N., R.R.M., and T.-K.S.; synchrotron, XRF imaging experiment and analysis, Y.Z.F., S.M.H., and T.-K.S.; writing—original draft, T.-K.S.; writing—review and editing, all of the authors. All of the authors have read and agreed to the published version of the manuscript. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was supported by the Natural Science and Engineering Research Council of Canada (NSERC, RGPIN-2019-05926), the Canada Research Chair Program (CRC), the Canada Foundation for Innovation (CFI), and the University of Western Ontario (UWO). This research used resources of the Advanced Photon Source, an Office of Science User Facility operated for the U.S. Department of Energy (DOE) Office of Science by Argonne National Laboratory, and was supported by the U.S. DOE under contract no. DE-AC02-06CH11357.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Data used in this work are available from the authors upon request.

**Acknowledgments:** CLS@APS where XRF imaging was conducted is an operation of the Canadian Light Source, which is supported by the Canada Foundation for Innovation (CFI), the Natural Sciences and Engineering Research Council (NSERC), the National Research Council (NRC), the Canadian Institutes of Health Research (CIHR), the Government of Saskatchewan, and the University of Saskatchewan.

**Conflicts of Interest:** The authors declare no conflict of interest.
