Anal. Chem. 2006, 78, 8319-8323
Chemical and Morphological Changes of Historical Lead Objects as a Result of the Use of Electrolytic Reduction as a Stabilization Treatment B. Schotte,† A. Adriaens,*,† F. Dhooghe,† D. Depla,‡ M. Dierick,§ M. Dowsett,| E. Lehmann,⊥ and P. Vontobel⊥
Department of Analytical Chemistry, Department of Solid State Sciences, and Department of Subatomic and Radiation Physics, Ghent University, BE-9000 Ghent, Belgium, Department of Physics, University of Warwick, Coventry, CV4 7AL, UK, and Spallation Neutron Source Division, Paul Scherrer Institute, CH-5232 Villigen PSI, Switzerland
This paper focuses on the evaluation of the treatment related to chemical and morphological changes of corroded lead artifacts when using electrolytic reduction as a stabilization method. Synchrotron radiation X-ray diffraction and X-ray photoelectron spectroscopy were used to study the chemical changes of the corrosion layer and on the top nanometer of surface, respectively. Neutron tomography and scanning electron microscopy were used to visualize potential morphological changes on millimeter and micrometer level, respectively. The results of this study have shown that electrolytic reduction is a reliable way to stabilize and conserve active corroded lead artifacts. The corrosion products are actually converted into metallic lead, while the morphological changes due to the treatment are limited. Lead objects generally do not corrode severely when buried or exposed to the atmosphere.1 The formation of a protective corrosion film, mainly consisting of lead carbonates, drastically inhibits the rate of corrosion speed. Therefore, ancient lead objects are usually in a very good condition when found on archaeological sites. However, lead does corrode seriously under conditions of high humidity where organic acids are present.2-4 This accelerated degradation is referred to as active corrosion. Typical conditions of humidity and organic acid vapor can sometimes be found in display cases at museums. The materials that contain these cabinets (wood, glue) degrade and emit organic compounds. Since these display cases are closed and the air exchange is very limited, * To whom correspondence should be addressed. Tel: +32 9 264 4826. Fax: +32 9 264 4960. E-mail:
[email protected]. † Department of Analytical Chemistry, Ghent University. ‡ Department of Solid State Sciences, Ghent University. § Department of Subatomic and Radiation Physics, Ghent University. | University of Warwick. ⊥ Paul Scherrer Institute. (1) Graedel, T. J. Electrochem. Soc. 1994, 141 (4), 922-927. (2) Te´tre´ault, J.; Sirois, J.; Stamatopoulou, E. Stud. Conserv. 1998, 43, 17-32. (3) Te´tre´ault, J.; Cano, E.; van Bommel, M.; Scott, D.; Dennis, M.; Barthe´sLambrousse, M.; Minel, L.; Robbiola, L. Stud. Conserv. 2003, 48, 237250. (4) Degrigny, C.; Le Gall, R.; Guilminot, R. In ICOM-CC 11th Triennial ICOM Meeting Edinburgh 1-6; James & James: London, 1996; Vol. 2, pp 865869. 10.1021/ac061381n CCC: $33.50 Published on Web 11/09/2006
© 2006 American Chemical Society
relatively high concentrations of the organic compounds can be reached. Many precious lead objects have been attacked in this way, thereby losing much surface detail from the original artifact. If no action is taken, the lead objects degrade further until they are transformed into a heap of dust. A possible treatment to stabilize and conserve active corroded lead is electrochemical reduction at constant potential.5-8 The corroded object is set up to represent the working electrode in an electrochemical cell, and the corrosion products are reduced. The various steps in the reduction curve were studied by Degrigny and Le Gall7 and were linked to the changes into the corrosion layer by Schotte et al.8 The method has become generally accepted, and many conservators use it to treat active corroded lead artifacts for their museum collections. Using various analytical techniques, the aim of this work was to provide better insight into potential chemical and morphological changes occurring as a result of the treatment. Synchrotron radiation X-ray diffraction (SR-XRD) and X-ray photoelectron spectroscopy (XPS) were used to study the chemical changes of the surface, including the corrosion layer. Neutron tomography (NT) and scanning electron microscopy (SEM) were used to visualize morphological changes on millimeter and micrometer level, respectively. EXPERIMENTAL SECTION Lead samples obtained from the Muse´e des Arts et Me´tiers (CNAM) were components in a model of a weaving loom (nο. 8710), whose function was to keep tension on the twinning threads. They corroded very badly during storage in their wooden display case. NT, SEM, XPS, and SR-XRD measurements were performed on two samples before and after the reduction treatment. Prior to these analyses, the samples were slightly brushed to remove weakly adhering corrosion particles on the surface. An insulated crocodile clip was fixed to one end of the object, so that the electrical connection would not disfigure it. In addition, for the (5) Organ, R. In ICOM committee for conservation, 2nd Meeting, Amsterdam, The International Council of Museums: Paris, 1969; pp 1-15. (6) Carradice, I.; Campbell, S. Stud. Conserv. 1994, 39, 100-106. (7) Degrigny, C.; Le Gall, R. Stud. Conserv. 1999, 44, 157-169. (8) Schotte, B.; Adriaens, A.; Vandenabeele, P.; Pantos, E. Unpublished work.
Analytical Chemistry, Vol. 78, No. 24, December 15, 2006 8319
Figure 1. Photograph of (left) a corroded sample and (right) the cross section of a similar sample, with the indication of pitting corrosion (A) and uniform corrosion (B).
neutron tomography analyses, it is advisable that the sample is positioned in exactly the same manner before and after treatment so that images may be compared. Therefore, the sample was mounted onto a holder (Figure 1, left). The reduction of the lead coupon was performed in 0.1 M Na2SO4 (Carlo Erba, 99.95%) at -1.3 V using a potentiostat (PGSTAT20, Eco Chem) coupled to a computer running the software Autolab (version 4.9, Eco Chem). The reference electrode used was a Hg/HgSO4 electrode (XR210, Radiometer Analytical) with a potential of 0.658 V versus the standard hydrogen electrode (T ) 295 K). A platinum electrode was used as a counter electrode. All collected data have been referred to the Hg/HgSO4 electrode. After the reduction, the object was rinsed and dried quickly using purged air. If the object were to dry slowly, a superficial layer of lead oxides would be formed.9 SR-XRD data were acquired at the synchrotron radiation station XMaS (Station BM28, European Synchrotron Radiation Facility, Grenoble, France).10 On this station, a beam with a wavelength of 1.5498 Å and with dimensions of 1 mm at 200 µm was used. These dimensions were selected in order to obtain a good signal without saturating the detector. A 2D Mar CCD165 detector (Mar USA Inc.) was used to record the diffraction patterns.11 The angle of the camera to the beam was 35° in order to be able to acquire the signals between 2Θ values 15° and 75°. Under these conditions, the diffraction center is outside the field of view of the camera, and the camera plane intersects the diffraction cones at an angle, to produce elliptical “rings”. The images were processed using a new software program, esaProject (2006 Mark Dowsett, EVA Surface Analysis), which was developed for this purpose. Samples of the corrosion products were taken in several places covering the entire thicknesses of the treated and untreated corrosion layers. The preparation of the samples included grinding the material to a fine powder and fixing it between Kapton tape. The XPS measurements were recorded with a Perkin-Elmer Phi ESCA 5500 system equipped with a monochromated 450 W (9) Degrigny, C. unpublished work. (10) http://www.esrf.fr/UsersAndScience/Experiments/CRG/BM28 (last visited in June 2006). (11) http://www.mar-usa.com/products/marccd (last visited in June 2006).
8320
Analytical Chemistry, Vol. 78, No. 24, December 15, 2006
Al K source. Experiments were recorded with 220-W source power and an angular acceptance of (7°. The analyzer axis made an angle of 45° with the specimen surface. Wide-scan spectra were measured over a binding energy range of 0-1400 eV and a pass energy of 187.85 eV. The C 1s, O 1s, and Pb 4f core levels were recorded with a step of 0.05 eV and a pass energy of 11.75 eV. The C 1s core level taken at 285 eV for adventitious carbon is used as binding energy reference. NT experiments were performed using thermal neutrons at the thermal beam line 32 of the spallation neutron source SINQ, which is known as the NEUTRA facility.12 The neutrons have a polychromatic Maxwellian energy spectrum with the most probable energy at ∼25 meV. The neutron flux at the sample position was 9.4 × 106 n cm-2 s-1. The collimation ratio L/D describing beam divergence was 350, inducing a geometrical unsharpness of ∼0.15 mm. Radiographies were recorded with a Peltier-cooled (-45 °C) slow-scan CCD featuring a 1024 × 1024 pixel chip (camera type DV 434, Andor Technology). The neutron flux was converted into light with a Li-6 doped ZnS screen having a thickness of 46 µm for run 1 (prior to reduction) and 300 µm for run 2 (after reduction). The field of view captured by the 50-mm Pentax F 1.2 normal lens was 97 × 97 mm, representing a nominal pixel size of 95 × 95 µm2. Exposure time was 30 (run 1) and 8 s (run 2). A total of 240 projections were taken while turning the object around 180°. The projections were reconstructed to 3D images by the program OCTOPUS13 and were compared to each other with the software packet VGStudio.14 This software packet is a provider of software technologies for analysis, processing, and visualization of volumetric data. SEM analyses were performed on a Quanta 200F (FEI) at 12.5kV acceleration voltage. RESULTS AND DISCUSSION Figure 1 (left) shows a photograph of one of the corroded samples. The surface is covered with white corrosion products. Some regions are attacked more severely than others and have, as a result, deformed the original shape of the coupon. The latter corrosion structures are referred to as pitting corrosion (indicated as A in Figure 1). A cross section of the embedded coupon shown in Figure 1 (right) gives information about the internal structure of the corrosion layer and shows that the deformations are clearly part of corroded areas. This is reasonable as the corrosion products are more voluminous than the metallic lead. Moreover, the formation of cracks makes these areas very porous. Other corrosion products still follow the contours of the bare metal very well and seem to be very compact. They are referred to as uniform corrosion (indicated as B in Figure 1). In conclusion, two different kinds of corrosion are observed: a thin compact corrosion layer (20 µm) and thick porous patches (
