Sulfur accumulation in the timbers of King Henry VIII's warship Mary Rose: A pathway in the sulfur cycle of conservation concern

Sandström et al. 10.1073/pnas.0504490102.

Supporting Information

Files in this Data Supplement:

Supporting Text
Supporting Figure 7
Supporting Figure 8
Supporting Figure 9
Supporting Figure 10
Supporting Figure 11
Supporting Figure 12

Supporting Figure 7

Fig. 7. Normalized sulfur K-edge x-ray absorption near-edge structure (XANES) spectra of powdered segments from core 3 [black-brown oak, from polyethylene glycol (PEG)-treated hull rider, see Fig. 1]. (a) A series of spectra vs. depth from the surface and inwards. The major peak at 2,473 eV originates from reduced sulfur species (elemental sulfur S₈, thiols R-SH, and disulfides R-S-S-R). Oxidized forms, sulfates, and sulfonates (R-SO₃^–), appear in significant amount only at the surface (0–3 mm), as shown by a distinct double peak at »2,482 eV. For total analyses of sulfur, iron, and calcium, see Fig. 12b. (b) Model fitting of standards (mostly in aqueous solution) to the surface spectrum (0–3 mm). Standard XANES spectra used for the fitting are 1 + 1', for disulfides R-S-S-R: (standard: cystine with peaks at 2,472.7 and 2,474.4 eV; 1' is solid) 45%; 2, thiols R-SH (standard: cysteine in aqueous solution, pH 7, 2,473.6 eV) 23%; 3, elemental sulfur (S₈ in xylene 2,473.0 eV) 10%; 4, sulfoxide R(SO)R' (standard: methionine sulfoxide 2,476.4 eV) 5%; 5, sulfonate R-SO₃^– (standard: sodium methylsulfonate solution 2,481.2 eV) 10%; and 6, sulfate SO₄^2– (standard: sodium sulfate in aqueous solution, pH 6, 2,482.6 eV) 7%.

Supporting Figure 8

Fig. 8. The three scanning x-ray microscopy (SXM) images to the right show freshly salvaged (in 2004) oak wood from the Mary Rose after 459 years on the sea floor. On the far left is an ordinary microscopy picture (with light) of a thin wood slice perpendicular to the cell walls, showing lumina (smaller holes) and vessels (larger holes). The S_red image, at 2,473.7 eV, displays two layers of thiols in high concentration in the lignin-rich cell walls of a vessel (lighter color = higher concentration). On the right, bright spots in the image at 2,482.5 eV correspond to sulfate particles, on a general background of sulfate from seawater and some residual absorption from the accumulated reduced forms.

Supporting Figure 9

Fig. 9. Core 5 (Pine): X-ray photoelectron spectra for S_2p electrons at 3 mm and 21 mm depth along core 5 (MR83 T37 from pine wood in air-tight storage), show that sulfur in reduced (S_red) forms dominate over oxidized (sulfates S_ox). At the depth 112 mm (data not shown), the XPS analysis (cf. Fig. 10) gave the relative amounts 92% as reduced sulfur, 5% sulfoxide, and 3 % sulfate. For total analyses of sulfur, iron and calcium, see Fig. 12d.

Supporting Figure 10

Fig. 10. Analyses for reduced and oxidized sulfur was performed by curve fitting of x-ray photoelectron spectra of core 1b at 92.5-mm depth and of the gunshield (cf. Fig. 4d), with three types of sulfur species: 1, sulfate SO₄^2–; 2, sulfoxide R₂SO; and 3, reduced sulfur. The two peaks required for each component correspond to excitation of sulfur 2p electrons from the 2p_1/2 and 2p_3/2 states in the ratio 1:2. The mean S 2p_3/2 binding energies were 1, 168.8 eV; 2, 166.3 eV; and 3, 163.7 eV for the three components, corresponding to sulfate SO₄^2–, sulfoxide, and reduced sulfur species, respectively, in agreement with earlier reported S 2p_3/2 binding energies for these species, and with the relative energy shifts in the corresponding XANES spectra. Note that negative charging by the floodgun shifted the energy scale in Fig. 10 by »–1.6 eV.

Supporting Figure 11

Fig. 11. X-ray photoelectron spectra reveal for the carbon and oxygen concentrations along core 1b (MR80 T108) an inverse correlation (high C corresponds to low O, cf. Table 2), indicated by the relative sizes of the two C_1s peaks at 285.0 and 286.5 eV, which mainly correspond to CH₂ (lignin) and C-O (cellulose) groups, respectively. The lignin to cellulose ratio increases when erosion bacteria degrade the cellulose of wood, and the variations indicate that the degradation extends throughout the Mary Rose timbers, consistent with the relatively uniform sulfur concentration profile (see Fig. 12a).

Supporting Figure 12

Fig. 12. Analyses of total sulfur, iron, and calcium in mass % along cores from the Mary Rose timbers. (a) Oak core 1a; hull timber of the Mary Rose, magazine stored in airtight wrap. (Upper) X-ray fluorescence line scan along the core 1a (Cox Analytical Systems; see Supporting Text). The gaps in the lines are due to missing pieces of the core lost in destructive analyses. (Lower) Elemental analyses of sulfur (Mikrokemi; see Supporting Text) and analyses of sulfur and iron from x-ray photoelectron spectra. (b) Oak core 3. (Upper) X-ray fluorescence line scan of total sulfur, iron and calcium in mass% along the core 3, from hull rider under PEG treatment (see Fig. 1). Breaks in the lines are due to missing pieces of the core lost in destructive analyses. (Lower) Elemental analyses of sulfur. (c) Oak core 4. (Upper) X-ray fluorescence line scan of sulfur, iron, and calcium in mass % along oak core 4 (MR1A), from hull rider under PEG treatment (see Fig. 1). T he gaps are due to missing pieces of the core lost in destructive analyses. (Lower) Elemental analyses of sulfur. (d) Core 5 (pine wood). X-ray fluorescence line scan of total sulfur, iron, and calcium in mass % along core 5 (MR83 T37) from pine wood in the Block Mills storage.

Supporting Text

Methods and Results of Analyses

Sulfur K-edge X-Ray Absorption Near-Edge Structure (XANES) Measurements. Segments of the wooden cores, a few millimeters each, were filed in inert atmosphere to fine particles, mounted on sulfur-free tape and covered by a 4-mm sulfur-free polypropylene film. Sulfur K-edge XANES spectra were collected of these samples at beamline 6-2, Stanford Synchrotron Radiation Laboratory (SSRL), in fluorescence mode [1 atm (1 atm = 101.3 kPa) helium atmosphere], as described in refs. 1 and 2. The sulfur K-edge, i.e., the 1s electron binding energy, shifts »13 eV from sulfides(-II) to sulfate(VI) and is approximately correlated with the formal oxidation number (3, 4). The main preedge features of the XANES spectra often contain several electronic transitions, which correspond to excitation of 1s core electrons to unoccupied valence states with some 3p character. The sulfur bonding arrangement affects the symmetry-dependent transition probabilities and influences shape, intensity, and position of the peaks in the spectra (5), making it possible to identify characteristic types of sulfur groups (3, 6, 7). The high number of 3p vacancies in sulfates can make the spectral intensity up to 5–6 times higher than that of sulfides.

Note that several schemes are in use for calibrating the absolute energies in the spectra (3, 5, 6). In the current work, we have chosen to calibrate our XANES spectra by assigning the first peak position of solid Na₂S₂O_5×5H₂O, measured before and after the sample, to 2,472.02 eV (6). The instrumental resolution is 0.5 eV (3).

For series of spectra, Principal Component Analysis is helpful in establishing the number of different types of functional sulfur groups that can be distinguished. The datfit program in the exafspak program package has been used for that purpose (8). To evaluate the relative amounts of sulfur in different types of sulfur functional groups present in the wood, linear combinations of normalized XANES spectra of appropriate standard sulfur compounds were fitted to the experimental spectra (cf. Table 1). Standard spectra of sulfur species in dilute solution often display sharper features than for solids because of self-absorption of x-rays, as e.g., for particles of elemental sulfur (3, 9), or by transition energies in a "band" structure for interacting species in a crystal structure. The dilute and amorphous organosulfur components in the wood are normally described better by spectra of standards in dilute solution than of solid compounds, as exemplified in Fig. 7b. The fit including a standard XANES spectrum of cystine in solution (1) improved only slightly when introducing solid cystine (1’) as an additional standard spectrum. Also, in the region 2,475-2,476 eV, the fit in Fig. 7b is not perfect with methionine sulfoxide as standard. Dimethyl sulfoxide, which gives a similar spectrum, has been investigated separately, in dilute solution and as a ligand (5). It has been found that covalent metal-oxygen bonding to the sulfoxide group will shift and split the 2,476-eV sulfoxide peak considerably. Hence, the slight shift of the sulfoxide peak in the current spectra may indicate some interaction involving the sulfoxide group of the compounds in the wood.

X-Ray Microscopy. The scanning x-ray microscope (SXM) of beamline ID21 (www.esrf.fr/UsersAndScience/Experiments/Imaging/ID21) of the European Synchrotron Radiation Facility, Grenoble, France, is equipped with a fixed-exit double crystal Si(111) monochromator providing a spectral resolution of 0.5 eV in the x-ray beam, with two parallel silicon mirrors rejecting harmonics, and a tantalum Fresnel zone plate lens generating a submicron probe. Low-grade vacuum was maintained in the sample compartment for the sulfur microspectroscopy measurements to avoid the air absorption at energies around 2.5 keV (10, 11). Wood slices (a few micrometers, cut with razor blades) were raster scanned in the focused beam at energies of characteristic sulfur XANES resonances, »2,473 and 2,483 eV. Emitted fluorescence photons were collected by an energy-dispersive high purity germanium detector to map the distribution of reduced and oxidized sulfur species. Spatially resolved micro-XANES sulfur spectra were acquired without significant charge reduction at precise locations selected on the sample.

X-Ray Photoelectron Spectroscopy (XPS) or Electron Spectroscopy for Chemical Analysis (ESCA). ESCA was performed on oak slices cut at various depths along cores, mounted and transferred in vacuum (<1 ´ 10^–6 Pa) into the sample chamber of a Scienta ESCA-300 instrument for multi-element ESCA analysis (12). The kinetic energy of core shell photoelectrons, excited from all elements in the outermost surface layer (<100 Å) by means of high intensity monochromatic Al Ka x-ray radiation (1,487 eV), was measured to obtain their binding energy. The well resolved C_1s(CH₂) line, assumed at 285.0 eV, was used for charge-calibration. This value resulted in 192.5 eV for the B_1s peak from boric acid, B(OH)₃, in agreement with previously reported values (NIST XPS Database: http://srdata.nist.gov/xps). The photoelectron intensity is proportional to the atomic concentration of each element in the surface layer (<100 Å). The surface potential of the electrically insulating samples was kept constant by an excess current of low-energy electrons (1 eV) from the Scienta flood gun. Long exposure was required before a small reduction of the signal for the oxidized sulfur could be noticed. This observation indicates that the radiation damage during the XPS measurements was insignificant.

ESCA Fittings. The amounts of oxidized and reduced sulfur were determined by considering three different sulfur components in a least-squares fitting procedure, representing each sulfur component by two slightly asymmetric Gaussian peaks with relative intensity 1:2, spinorbit split by 1.18 eV, and with the same full width at half maximum (Fig. 10).

X-Ray Fluorescence Line Scan Analysis. An Itrax wood scanner from Cox Analytical Systems (Gothenburg, Sweden) was used to obtain high resolution concentration profiles of total sulfur and iron by automatic line scans along cores (13). X-ray fluorescence was excited by means of a focused Cu Ka x-ray beam, and an energy-dispersive solid state x-ray detector provided »0.5-mm resolution and an analytical depth into the wood of »0.1 mm. Pellets prepared of oak wood powder, PEG 4000, known concentrations of iron(III) sulfate and gypsum, were scanned to achieve quantitative calibration.

Elemental Analysis. Mikrokemi (Uppsala, Sweden) performed total elemental sulfur analyses. The sample (a few milligrams) is weighed in a tin capsule. Oxygen is injected, the tin metal oxidizes, the temperature momentarily rises to »1,800°C, and the sample is combusted. The combustion gases, CO₂, H₂O, NO_x, SO₂, and SO₃, are led into a reduction chamber where SO₃ is reduced to SO₂. The amount of SO₂ is determined by gas chromatography. This type of analysis requires total sulfur concentrations of at least 0.1 mass % for good accuracy; it is destructive and can only provide sulfur concentration profiles with low resolution (see e.g., Figs. 12 a–c).

1. Sandström, M., Jalilehvand, F., Persson, I., Gelius, U., Frank, P. & Hall-Roth, I. (2002) Nature 415, 893–897.

2. Sandstorm, M., Jalilehvand, F., Persson, I., Fors, Y., Damian, E., Gelius, U., Hall-Roth, I., Dal. L., Richards, V. L. & Godfrey, I. (2003) in Conservation Science 2002, eds. Townsend, J. H., Eremin, K. & Adriaens, A. (Archetype Books, London), pp. 79–87.

3. Pickering, I. J., Prince, R. C., Divers, T. & George, G. N. (1998) FEBS Lett. 441, 11–14.

4. Vairavamurthy, M. A., Maletic, D., Wang, S., Manowitz, B., Eglinton, T. & Lyons, T. (1997) Energy & Fuels 11, 546–553.

5. Damian, E., Jalilehvand, F., Abbasi, A., Pettersson, L. G. M. & Sandström, M. (2005) Phys. Scripta T115, 1077–1079.

6. Hedman, B., Frank, P., Penner-Hahn, J. E., Roe, A. L., Hodgson, K. O., Carlson, R. M. K., Brown, G., Cerino, J., Hettel, R., Troxel, T., et al. (1986) Nucl. Instr. Methods A246, 797–800.

7. Sarret, G., Connan, J., Kasrai, M., Bancroft, G. M., Charrié-Duhaut, A., Lemoine, S., Adam, P., Albrecht, P. & Eybert-Bérard, L. (1999) Geochim. Cosmochim. Acta 63, 3767–3779.

8. George, G. N., George, S. J. & Pickering, I. J. (2001) exafspak Computer Program Package (Stanford Synchrotron Radiation Laboratory, Menlo Park, CA).

9. Pickering, I. J., George, G. N., Yu, E. Y., Brune, D. C., Tuschak, C., Overmann, J., Beatty, J. T. & Prince, R. C. (2001) Biochemistry 40, 8138–8145.

10. Susini, J., Salomé, M., Fayard, B., Ortega, R. & Kaulich, B. (2002) Surf. Rev. Lett. 9, 203–211.

11. David, C., Kaulich, B., Barrett, R., Salomé, M. & Susini, J. (2000) Appl. Phys. Lett. 77, 3851–3853.

12. Gelius, U., Wannberg, B., Baltzer, P., Fellner-Feldegg, H., Carlsson, G., Johansson, C.-G., Larsson, J., Münger, P. & Vegerfors, G. (1990) J. Electron Spectrosc. Relat. Phenom. 52, 747–785.

13. Sandström, M., Fors, Y., Jalilehvand, F., Damian, E. & Gelius, U. (2005) in Proceedings of the Ninth ICOM Group on Wet Organic Archaeological Materials Conference, Copenhagen, 2004, ed. P. Hoffmann (Deutsches Schiffahrtsmuseum, Bremerhaven, Germany), in press.