Medieval Gilding Technology of Historical Metal Threads Revealed by

Sep 19, 2017 - For a more comprehensive list of citations to this article, users are ... Archaeological and Anthropological Sciences 2019 11 (4), 1599...
1 downloads 0 Views 3MB Size
Subscriber access provided by University of Sussex Library

Article

Medieval gilding technology of historical metal threads revealed by electron optical and micro-Raman spectroscopic study of focused ion beam-milled cross sections Tamás Gábor Weiszburg, Katalin Gherdán, Kitti Ratter, Norbert Zajzon, Zsolt Bend#, György Z. Radnoczi, Ágnes Takács, Tamás Váczi, Gábor Varga, and György Szakmány Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.7b01917 • Publication Date (Web): 19 Sep 2017 Downloaded from http://pubs.acs.org on September 20, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Analytical Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 9

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

84x60mm (300 x 300 DPI)

ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 9

1

Medieval gilding technology of historical metal threads revealed by electron optical and micro-Raman spectroscopic study of focused ion beam-milled cross sections Tamás G. Weiszburg◊, Katalin Gherdán*◊†, Kitti Ratter◊, Norbert Zajzon○, Zsolt Bendő◊, György Radnóczi▲, Ágnes Takács◊, Tamás Váczi◊, Gábor Varga◊, György Szakmány◊. ◊

: Eötvös L. University, H-1117, Budapest, Pázmány Péter sétány 1/C, ○: University of Miskolc, H-3515, Miskolc, Egyetemváros, ▲: Institute for Technical Physics and Materials Science RCNS HAS, H-1121, Budapest, Konkoly Thege út 29–33

ABSTRACT: Although gilt silver threads were widely used for decorating historical textiles, their manufacturing techniques have been elusive for centuries. Contemporary written sources give only limited, sometimes ambiguous information, and detailed crosssectional study of the microscale soft noble metal objects has been hindered by sample preparation. In this work, to give a thorough characterization of historical gilt silver threads, nano- and microscale textural, chemical and structural data on cross sections, prepared by focused ion beam milling, were collected, using various electron-optical methods (high-resolution scanning electron microscopy (SEM), wavelength-dispersive electron probe microanalysis (EPMA), electron back-scattered diffraction (EBSD) combined with energy-dispersive electron probe microanalysis (EDX), transmission electron microscopy (TEM) combined with EDX, and micro-Raman spectroscopy. The thickness of the gold coating varied between 70–400 nm. Data reveal nano- and microscale metallurgy-related, gilding-related and corrosion-related inhomogeneities in the silver base. These inhomogeneities account for the limitations of surface analysis when tracking gilding methods of historical metal threads, and explain why chemical information has to be connected to 3D texture on submicrometre scale. The geometry and chemical composition (lack of mercury, copper) of the gold/silver interface prove that the ancient gilding technology was diffusion bonding. The observed differences in the copper content of the silver base of the different thread types suggest intentional technological choice. Among the examined textiles of different ages (13th–17th centuries) and provenances narrow technological variation has been found.

Introduction Man has woven golden threads into textiles and used them for embroidering his clothes for thousands of years1. The first written record on the use of gold metal threads – known by us – comes from the Bible2. Ancient textiles, however, are not merely attractive (Figure 1 and Figure S-1 of the Supporting Information), but are of historical and socio-cultural significance. Their age and provenance is partly coded in the textile decorating metal threads. A key to that code is the manufacturing technique of the threads, not known in detail from historical sources3–7. Golden metal threads are either strips or wires. They can be used „bare” or wound around a fibrous core, such as silk, linen or cotton yarn. The first threads were of pure gold or gold alloys, but later gilt silver and other metal combinations were also used. The threads are thin (strips: 7–40 µm; wires: 70–500 µm; typically), whereas the thickness of their gold coating is supposed to be on the nanometre scale8. In the last four decades intensive research was carried out on the manufacturing techniques of historical gilt silver threads. Chemical changes were traced at the interface between the gold layer and the silver base, as some of the ancient gilding methods identified in the archaeological record or described by historical sources result in the enrichment of certain chemical elements at that interface.

Figure 1. Historical metal strips. Close up of the chasuble shown in Figure S-1 of the Supporting Information: part of a pomegranate motif embroidered with three different silver strip types wound around fibrous cores. 1: one-side-gilt, microscale; 2: nongilt, microscale; 3. non-gilt, macroscale.

When describing metal threads micro art objects have to be characterized on the micro- and nanoscale. In the study, preservation and restoration of cultural heritage artefacts the

ACS Paragon Plus Environment

Page 3 of 9

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

2 first and – as in lot of cases not even micro destructive methods are allowed to be used – often the only step is the description of their surface. Theoretically an ideal layer/sandwich structure is possible to be characterized from the surface in submicrometre resolution by depth-profiling methods9, 10. However, the surface of historical art objects is often too rough to perform depth profiling11. Additionally, when studying these objects we should calculate with various, non-layer type subsurface chemical inhomogeneites, reflecting either the original internal 3D texture of the metals, or the imprints of post-production corrosion. Thus, by applying merely surface methods, the nanotexture of the threads cannot be studied at all, as observed heterogeneities in chemical composition might be misinterpreted as technological fingerprints. Another approach is cross section analysis. The traditional method for preparing cross sections available for further microanalysis is mechanical cutting and polishing. However, this destroys the nanotexture of the soft noble metals. Most previous studies analyzed the surface of the threads using various microanalytical techniques: electron probe microanalysis (EPMA)12–19, Auger Electron Spectroscopy (AES)20, Particle Induced X-ray Emission combined with Rutherford Backscattering Spectrometry (PIXE+RBS)21–23, Secondary Ion Mass Spectrometry (SIMS)24. In addition, a few cross sections of the threads were prepared by mechanical methods and studied by EPMA25–29. In this work gilt silver threads (strips and wires) and contemporary silver and gold strips were studied from Medieval and technologically related post-Medieval textiles (13th–17th centuries) of various European provenances (for sample description see Table S-1 of the Supporting Information). Cross sections of undamaged layers were prepared by focused ion beam (FIB) sputtering for imaging and analysis. FIB milling allowed us to create cross sections available for high resolution imaging and microanalysis without damaging the layer structure. Both cross sections and surfaces were studied by a combination of techniques such as high-resolution scanning electron microscopy (SEM), wavelength-dispersive electron probe microanalysis (EPMA), electron back-scattered diffraction (EBSD) combined with energy-dispersive electron probe microanalysis (EDX), micro-Raman spectroscopy, transmission electron microscopy (TEM) combined with EDX. Experimental section The studied samples were taken from the textiles by restorers, following the instructions of art historians. The samples were gently untwisted without any intensive mechanical interaction and, by using carbon paste, were mounted on tilted (45°) sample holders in a way to provide access to both the original edge of the thread (created by the historical thread maker) and the inner parts of the threads as well (Figure 2). Then, in order to protect the gold layer from the ion beam, platinum was deposited on the surface. Deposition was induced by the electron beam (2 kV, 4 nA). Then FIB milling (Ga ion beam, 30 kV, 15 nA) was used to make cross sections perpendicular to the surface to reveal the internal texture of the threads. In the case of a few strips ion milling was also used to create surfaces parallel to the original surface of the thread in order to obtain a 3D reconstruction of the grain texture of the silver base (Figure S-2 of the Supporting Information). The

FIB/SEM equipment used was a FEI Quanta 3D type dual beam system, with a Schottky FEG electron gun equipped with SED, BSED, VcD, silicon-drift X-ray and EBSD detectors.

Figure 2. Sample mounting and cross section preparation. (A) SEM-BSED image of a one-side-gilt silver strip from a 15th century chasuble (sample U13, Table S-1 of the Supporting Information) mounted on a tilted sample holder. O: original edge cut by the historical thread maker, N: newly created edge, result of sample preparation, dashed line: trace of the planned cross section; (B, C) Steps of creating FIB-cut cross sections on strips. (B) Pt deposition. (C) Cross-section preparation by ion milling. Pt: platinum (deposited), F: FIB-cut cross section.

Chemical composition was determined using the FEI Quanta 3D FIB/SEM equipment, coupled with a silicon drift energy-dispersive X-ray spectrometer (20 kV, 1.7 nA), an AMRAY 1830i SEM instrument equipped with an energydispersive X-ray spectrometer (20 kV, 1 nA) and a JEOL JXA-8600 Superprobe having four wavelength-dispersive X-

ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 9

3 ray spectrometers (20 kV, 20 nA). Average copper concentrations of the silver base were estimated from the largest accessible areas (60–350 µm2) of the cross sections that, based on BSE images, could be considered to be homogeneous. As the copper-silver binary system is phase-sensitive to relatively small changes in concentration at low copper values, concentration data is given not only in wt%, but also in at%. EBSD analysis was carried out using the FEI Quanta 3D FIB/SEM equipment, coupled with a Hikari detector (20 kV, 4 nA). Patterns were processed by the EDAX-TSL software package. EBSD was intended to be used to study the crystal phases, and the submicrometre texture – especially grainsize and grain boundaries – of the base metal, on the same surface where chemical and texture analysis was carried out, without the need of etching the surface. EBSD revealed that for complete characterization of the silver base, especially for the identification of all the crystal phases present, the use of TEM is necessary. Crystal phase characterization of the some ten nm thin copper lamellae in between the silver grains was carried out by conventional TEM using a Philips CM-20 microscope operated at 200 kV. The sample was prepared in cross section by embedding it between two Si wafers into a special Ti supporting disk (3 mm in diameter) to protect the sample from heat during thinning30.The sample was then mechanically thinned to about 50 µm in thickness and ion-milled by 10 keV Ar+ ions at grazing incidence (1–4°) to the surface.

Image analysis was used to estimate the volume percents of copper-rich inclusions and pores. A selection of inclusions of the silver base were analyzed by micro-Raman spectroscopy. Spectra were recorded on a HORIBA JobinYvon LabRAM HR800UV edge filter-based Raman microspectrometer. The instrument is coupled to an Olympus BXFM microscope, a 100× objective (N.A. 0.9) was used to focus laser light onto the sample and to collect the scattered light. The confocal hole that acted also as the spectrometer entrance slit was 100 µm. Raman spectra were excited with a 785 nm diode laser (approx. 15 mW laser power on the sample). Signal collecting times were set between 10 s×30 and 30 s×60, according to signal and background intensities. Results and discussion We found that gilt silver threads can be described in cross section by characterizing their following four aspects: 1) nanotexture and nanoscale chemical behavior of the silver base, 2) morphology and chemistry of the surface gold coating, 3) geometry and chemistry of the gold–silver interface, and 4) shapes of the edges (strips only). Nanotexture and chemical composition of the silver base The nanotexture of the silver base is different in wires and strips. In cross section the silver grains in wires are equiaxed (diameter ~100–200 nm), whereas in strips they are diskshaped, flattened parallel to the surface of the strips. The grain size distribution of the strips has two maxima (at 200–300 and 600–800 nm) (Figure S-3 of the Supporting Information).

ACS Paragon Plus Environment

Page 5 of 9

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

4 Figure 3. Nanotexture of the silver base. SEM-SED (A, B) and TEM (C, D) images of the micro- and nanotextures of the silver base of a strip from a 15th century codice (sample U8, Table S-1). (A) FIB-milled cross section with an I2-type feature. (B) Detail of (A) with E1 and E2 type features. (C) Fine texture of silver grains and E1 type (arrows) features. (D) Selected-area electron diffraction (SAED) pattern of the area (C) showing the separate copper and silver phases.

In terms of copper content, two types of silver base can be distinguished. Low-copper (0,5– 2,5 wt%, 1–4 at%) is typical for drawn wires and strips gilt on both sides, whereas highcopper (4–10 wt%, 7–16 at%) is characteristic for one-sidegilt strips (Figure S-4 of the Supporting Information). In wires, silver includes all copper in solid solution (Figure S-5 of the Supporting Information). In one-side-gilt strips, excess copper is heterogeneously distributed in two generations of copper crystals. Additionally, in both types of threads angular, copper-dominant inclusion particles are dispersed in subordinate amounts. Intensive corrosion may also be present in one-sidegilt strips, resulting in heterogeneous copper depletion. Special features causing inhomogeneity of the silver base We distinguished inclusion type (I-type), exsolution type (E-type), gilding technology type (GT-type) and corrosion type (C-type) features in the threads.

I-type features (300 nm–1 µm sized relict inclusions) are not related to the gilding technology but to the imperfectness of metallurgy of the silver raw material, preceding the manufacture of the threads (Figure 3, 4). Because of their minor concentration (