Ibid- Caltech Library System

Electronic Delivery Cover Sheet

WARNING CONCERNING COPYRIGHT RESTRICTIONS

The copyright law of the United States (Title 17, United States Code) governs the making of photocopies or other reproductions of copyrighted materials. Under certain conditions specified in the law, libraries and archives are authorized to furnish a photocopy or other reproduction. One of these specified conditions is that the photocopy or reproduction is not to be "used for any purpose other than private study, scholarship, or research". If a user makes a request for, or later uses,a photocopy or reproduction for purposes in excess of "fair use", that user may be liable for copyright infringement. This institution reserves the right to refuse to accept a copying order if, in its judgement, fulfillment of the order would involve violation of copyright law.

Spectrochimica Acta Part B 59 (2004) 1663 1667 www.elsevier.com/locate/sab

In situ multi-element analyses by energy-dispersive X-ray fluorescence on varnishes of historical violinsB

Jean-Philippe Echard*
Laboratoire de recherche et de restauration, Muse e de la musique, Cite de la musique, 221, avenue Jean-Jaure ` s, 75019 Paris, France Received 1 November 2003; accepted 10 May 2004 Available online 3 September 2004

Abstract Varnishes of Italian violins and other historical stringed instruments have been analyzed by energy-dispersive X-ray fluorescence (EDXRF). The instruments whose varnishes were to be analyzed were chosen from the collection kept in Muse e de la Musique in Paris. Direct analyses were performed on instrument varnishes, without any sampling and non-destructively, showing inorganic elements such as lead, mercury and iron that could be related to siccatives or pigments. Analytical results and their comparison with old formulae or traditional recipes of violin varnishes, as with the few previous analytical results, will be discussed. D 2004 Elsevier B.V. All rights reserved.
Keywords: Energy-dispersive X-ray fluorescence; Violin varnish

1. Introduction Violins made by old Italian masters, especially those working in Cremona, are even today the most revered of instruments by musicians and violin makers. The varnishes of these historical instruments are particularly distinguished for their beauty and resistance. Modern violin makers aspire to reproduce these varnishes, which seem till today to be unequalled though. The varnishes of historical instruments are mixtures of natural organic substances (such as resins, oils, spirit, waxes, organic colorants) and, in lesser proportions, of inorganic substances, added mainly to influence the color, the transparency, other physical properties, and the drying time for oilbased varnishes. These ingredients are quoted in many varnish recipes published during the 17th and 18th centuries [1,2] but none of these bibliographic sources is by name related to the Cremonese makers like Andrea and Nicolo
This paper was presented at the International Congress on X-Ray Optics and Microanalysis (ICXOM XVII), held in Chamonix, Mont Blanc, France, 22-26 September 2003, and is published in the special issue of Spectrochimica Acta Part B , dedicated to that conference. * Tel.: +33 1 44 84 46 82; fax: +33 1 44 84 46 41. E-mail address: jpechard@cite-musique.fr (J.-P. Echard). 0584-8547/$ - see front matter D 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.sab.2004.05.026
B

Amati, Antonio Stradivari and Giuseppe Guarneri ddel Gesu ` T. For these reasons, the varnishes of these instruments have been the object of numerous observations, suppositions [3,4], and analytical studies of the chemical composition [59]. The inorganic substances have been characterized by means of total X-ray fluorescence (TXRF) [10,11] on small flakes of varnish sampled on instruments. In this paper, a non-sampling and non-destructive energydispersive X-ray fluorescence (EDXRF) technique was applied to the characterization of inorganic additives to understand better the varnishing techniques of the Italian violinmakers, especially in Cremona. A great number of analyses on each instrument were then made possible to improve the statistic parameters on heterogeneous materials such as nonindustrial varnishes, simultaneously preserving the integrity of these violins of high cultural and historical value. 2. Experimental 2.1. Selection of musical instruments to be studied Among the musical instruments kept in the Muse e de la Musique in Paris are a representative number of instruments

1664

J.-P. Echard / Spectrochimica Acta Part B 59 (2004) 16631667

made in Cremona and Brescia, cities reputed to be important instrument making centers in Italy, between the 16th and the 18th century. Among them (see Table 1) are five violins, a kit (or pochette ) and a head of a viola damore made by Antonio Stradivari, a viola made by Andrea Amati circa 1575, a violin made by Nicolo Amati in 1639 and one made by Giuseppe Guarneri ddel Gesu ` T in 1742. Another Guarneri ddel Gesu `T dated 1733 from a private collection, two citterns (plucked stringed instruments) made earlier in Brescia (one of them having maybe been restored and revarnished by Stradivari [12]) and a violin made by Goffredo Cappa, a pupil of Nicolo Amati, were studied too. These instruments are the evidences of a continuous tradition of instrument making, with strong father-to-son or master-to-pupil transmission of know-how. In addition, samples of non-varnished woods, raw materials and ingredients for varnishes were analyzed. 2.2. Selection of areas to be analyzed Because violins might have had varnish retouches after their making, due to wear and tear or accidents, each instrument has been observed and photographed under normal light and ultraviolet (UV) light prior to EDXRF analyses. This method made it possible to map different strates or states of the varnish, such as boriginal varnishQ, bretouchingQ, bground layerQ, bwoodQ or bdirty areasQ. However, only the results concerning the making of the boriginalQ varnishes will be described in this paper. In violin varnishes, inorganic additives are dispersed in a solidified organic matrix and, consequently, their concentration is subject to change from one point to another. Each of these bstatesQ has been analyzed in several places to increase the reliability of the analytical results statistically. The instruments have been observed and photographed under optical microscope with reflected light, to describe the mineral particles (average size, shape, color, etc.) included in the varnishes.

2.3. EDXRF analyses The EDXRF device ARTTAX (INTAX, Berlin, Germany) was equipped with molybdenum anticathode. A 0.6 mm collimator was used. It was operated at 45 kV, 600 AA. Direct analyses on violin varnishes were carried out at a distance between 5 and 6 mm of the instrument in 400 s blive-timeQ period. The ARTTAX device includes a Silicon Drift Detector X-Flash (RONTEC), an analyzer and a software used for spectra acquisition, elements recognition, background and peak areas calculation. The bcollimator to sampleQ distance was adjusted using a digital video micro-camera attached to the measuring head of the system, close to the X-ray collimator. It has been measured that small variations of this distance (less than 1 mm), would not affect significantly the net peak areas (less than 2%). Because only a qualitative study was performed, these variations were considered as negligible. Neither relative mass fraction nor elemental content in the varnishes have been calculated.On every instrument, several analyses were carried out on the soundboard, the back, the ribs and the head. Depending on the size of each instrument and the degree of heterogeneity of the surface varnish (boriginal varnishQ, bretouchingQ, bground layerQ, bwoodQ or bdirty areasQ), between 8 and 44 analyses were necessary to have a good statistical approach of the elemental content of each bstateQ of the varnish, as observed under UV light.

3. Results and discussion 3.1. Wood preparation The inorganic elements present in wood have to be considered, because they are visible in the EDXRF spectra of the varnishes. Violins are essentially made of two species of

Table 1 Detection of lead, mercury and arsenic with EDXRF in the varnishes of the studied instruments Maker G. Virchi G. Virchi, attributed to A. Amati N. Amati A. Stradivari A. Stradivari A. Stradivari A. Stradivari A. Stradivari A. Stradivari A. Stradivari G. Guarneri, ddel Gesu `T G. Guarneri, ddel Gesu `T G. Cappa B. Calcagni, attributed to Instrument cittern cittern viola violin violin, bLonguetQ violin, bTuaQ violin, bDavidoffQ violin, bProvignyQ Kit/(or pochette ) violin, bSarasateQ head of a viola damore violin Violin, bAlardQ violin violin Date 16th c., 2nd half circa 1570 circa 1580 1639 1692 1708 1708 1716 1717 1724 1733 1742 circa 1710 1740 Place Brescia Brescia Cremona Cremona Cremona Cremona Cremona Cremona Cremona Cremona Cremona Cremona Cremona Saluzzo Genova Pb X n.d. n.d. X X X X X X n.d. X X X X n.d. Hg n.d. n.d. n.d. n.d. X n.d. n.d. X n.d. X n.d. n.d. n.d. n.d. n.d. As n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. X

The bXQ symbol means that the element has been detected in the instrument varnish. bn.d.Q stands for bnot detectedQ.

J.-P. Echard / Spectrochimica Acta Part B 59 (2004) 16631667

1665

wood: spruce for the soundboard and maple for the back, the ribs, the neck and the head. The EDXRF spectra were carried out on modern samples of non-varnished spruce and maple show noticeable differences, especially concerning the potassium content: Maple seems to have a higher potassium content than spruce and it has to be taken in consideration for further spectra interpretation. It is however quite uneasy to compare such results with the spectra of worn areas on historical instruments. In these areas, not only a possible wood preparation, but also sweat and dust might have considerably modified the elemental composition (see Section 3.4). 3.2. Nature of the varnish The varnishes used for musical instruments from the 16th century up to the middle of the 19th century in Europe are of three different types: drying oil-based varnishes, essential oil-based varnishes, and spirit varnishes. Drying oil-based varnishes (based on linseed oil or walnut oil, for example) have a very long drying time. The most common siccative for these varnishes was litharge (PbO) with other lead compounds, and more rarely white couperose , a zinc sulfate [2]. The presence of lead (Pb) in historical varnishes can be a reliable indicator for drying oil-based varnishes. Lead is detected by EDXRF in 11 of the 15 instruments studied (see Table 1). In particular, all the Stradivari and Guarneri ddel Gesu ` T contain lead in their varnishes, except the Stradivari bSarasateQ violin, dated 1724. The lead detected can be attributed to a siccative, and thus indicates the varnishes are drying oil based. 3.3. Coloring the varnish On the instruments analyzed in this research, the detection of elements such as Hg, As, Fe and Mn can show evidences

of mineral pigments incorporated to the varnish for coloring purposes [13]. Arsenic (As) is contained in orpiment (As2S3), a yellow mineral pigment. This element has been detected on a violin attributed to Bernardo Calcagni. The microscopic observations on this instrument showing yellow particles approx. 20 Am in diameter confirmed the use of orpiment in the elaboration of the varnish. Mercury (Hg) is contained in cinnabar and vermilion (HgS). This element has been detected on three Stradivari violins, especially as a major constituent on the bSarasateQ violin, as shown in Fig. 1. On this instrument, microscopic observations of red particles approx. 1040 Am in diameter confirmed the use of cinnabar or vermilion in the elaboration of this varnish (see Fig. 2). Iron (Fe) and sometimes manganese (Mn) are contained in their oxidized forms in ochre, umber and other earths, giving a wide range of colors from yellow to dark brown. When detected, these elements indicate probable presence of earths in the varnish. However, these elements could have accidentally been incorporated by contamination from dust or tools and recipients used for the preparation of the varnish. Microscopic observations however do not allow positive identification of particles of such pigments. 3.4. Influence of playing and handling instruments On every violin, several analyses were run in areas where varnish has been worn or removed by repeated contact with the skin of musicians. These areas are the bottom part of the soundboard, that the musician applies to his chin to hold the instrument when playing and the upper right part of the soundboard, over which the musician often places his left hand to play the highest notes. Even with the introduction of the chin rest during the 19th century, to minimize the contact of the chin with the soundboard and careful wiping after

Fig. 1. EDXRF spectrum of the varnish on the Stradivari bSarasateQ violin.

1666

J.-P. Echard / Spectrochimica Acta Part B 59 (2004) 16631667

Fig. 2. Microscopic photography showing red particles in the varnish of the Stradivari bSarasateQ violin.

playing, sweat is liable to degrade the varnish or at least to dry over it. Characteristic elements for sweat were detected specifically in these areas with EDXRF. The main electrolytic elements excreted by sweat are sodium, chloride, potassium, calcium, zinc and copper [14,15]. With the exception of sodium, which is not detected with the EDXRF system used for this research, all these elements were detected in these areas with significantly bigger peaks than in areas less or not exposed to sweat (see Fig. 3).

4. Conclusion Historical violin varnishes were investigated using in situ EDXRF spectrometry. For the first time, inorganic elements were identified in the varnishes of instruments made by Stradivari, Amati, Guarneri ddel Gesu ` T and others using a non-sampling and non-destructive technique. The presence of siccative was deduced from the lead detected in the majority of the instruments investigated. This means the varnishes of these instruments could be oil-based. Orpiment

Fig. 3. Detail of the EDXRF spectra of areas not exposed to sweat (bottom) and exposed to sweat (top), showing signals of typical inorganic elements excreted by human sweat.

J.-P. Echard / Spectrochimica Acta Part B 59 (2004) 16631667

1667

and vermilion or cinnabar were identified as pigment incorporated to varnishes as coloring agents.

Acknowledgments The author is much obliged to Jean-Jacques Rampal, violin-maker and expert and to his staff for allowing the analyses on the 1733 Guarneri ddel Gesu ` T and to Christine Beno t, Centre de Recherche et de Restauration des Muse es de France, for valuable discussions.

References
[1] F. Meyer, Oil, spirit, amber and wine, The Strad 103 (1223) (1992 (March)) 250 251. [2] Watin, Lart du peintre, doreur, vernisseur, Paris, 1755, pp. 88 90. [3] W.H. Hill, et al., Antonio Stradivari, His Life and Work, Dover Publications, New York, 1963. [4] S.F. Sacconi, in: Casteli et Bourhis (Ed.), Les bsecretsQ de Stradivarius, 1989, pp. 159 191. [5] R. White, An examination of varnish from three 18th century musical instruments, ICOM Committee for Conservation Report, 78/16/1, 5th Triennal Meeting, Zagreb, 1978.

[6] F. Meyer, Surface coatings of stringed musical instruments, Dartington Violin Conference Proceedings, 1995, pp. 29 42. [7] C.Y. Barlow, J. Woodhouse, Firm ground? A detailed analysis of ground layers under the microscope, The Strad 100 (1187) (1989) 195 197; C.Y. Barlow, J. Woodhouse, Firm ground? A detailed analysis of ground layers under the microscope, The Strad 100 (1188) (1989) 275 278. [8] C.Y. Barlow, J. Woodhouse, Of old wood and varnish: peering into the can of worms, J. Catgut Acoust. Soc. 1 (4) (1989) 2 9 (series II). [9] J. Nagyvary, The chemistry of a Stradivarius, Chem. Eng. News 66 (21) (1988) 24 31. [10] A. Von Bohlen, F. Meyer, Microanalysis of old violin varnishes by total reflection X-ray fluorescence, Spectrochim. Acta, Part B: Atom. Spectrosc. 52 (1997) 1053 1056. [11] A. Von Bohlen, Multielement microanalysis and pattern characterization of historical violin varnishes, J. Trace Microprobe Tech. 17 (2) (1999) 177 187. [12] J. Dugot, personal communication, Paris, 2003. [13] Artists pigments, vol. 1 (1986), 2 (1993) and 3 (1997), National Gallery of Art, Washington, Oxford University Press. [14] S. Robinson, A.H. Robinson, Chemical composition of sweat, Physiol. Rev. 34 (1954) 202 220. [15] J.R. Cohn, E.A. Emmett, The excretion of trace metals in human sweat, Ann. Clin. Lab. Sci. 8 (4) (1978) 270 275.