Professional Documents
Culture Documents
Gas Chromatography-Mass Spectrometry Characterization of The Varnish and Glue of An Ancient 18th Century Double Bass
Gas Chromatography-Mass Spectrometry Characterization of The Varnish and Glue of An Ancient 18th Century Double Bass
Gas Chromatography-Mass Spectrometry Characterization of The Varnish and Glue of An Ancient 18th Century Double Bass
Received 30 August 2006; received in revised form 8 February 2007; accepted 14 February 2007
Available online 21 February 2007
Abstract
A GC–MS investigation is conducted on the double bass “Panormus”, property of Conservatorio di Musica “Vincenzo Bellini” in Palermo. The
most important components of the varnish (fatty acids) and of the glue (proteinaceous amino acids), with which the musical instrument was treated
in the past, are determined. The analyses are carried out by prior derivatization of fatty acids by acidic methanol and of amino acids by acidic
methanol and trifluoroacetic anhydride (TFAA). Analytes identification is achieved by direct comparison with several reference materials and the
use of a digitized library.
© 2007 Elsevier B.V. All rights reserved.
Keywords: Double bass; Art analysis; Derivatization; GC; Binding media; Varnish; Fatty acids; Amino acids; Resins
1. Introduction Usually, the restorer has to deal with two aspects of the var-
nish and glue problem [4]: what kind of varnish and glue were
A functional or conservative restoration work, well-designed originally used and which ones among the present products are
and -executed, should slacken the deterioration process of an suited for the restoration work.
art object. Wooden artworks, although well preserved, suffer It is almost certain that a direct relationship exists between
from several kinds of deterioration attacks which are mainly varnish composition and quality of sound. Therefore, the makers
from biological origin. Physical and chemical agents which con- of stringed instruments, while trying to replicate lightness and
tribute are moisture, temperature fluctuations, light, atmospheric transparency effects and brightness of sounds, wanted to imitate
pollutants, etc. ancient masters and discover their secrets.
Wood restoration techniques vary from work to work and A comparison between ancient stringed instruments from
particular attention is needed for the handling of painted or Cremona (Italy) and modern ones confirmed the existence of
varnished material. differences in the aforementioned quality of sound and in the
After a prior eventual disinfestation from insects or parasites, timbre. The same difference was established between some
preferably in a non-toxic gas chamber [1,2], and a surface clean- Stradivari’s undamaged instruments and other ones from which
ing, the pure restoration consists in the substitution of parts of the varnish was entirely removed (by scalpel or washing) and
the ancient object trying to use the same kind of wood (ancient re-varnished with non-original different products [5]. The var-
or artificially aged, if possible) and the same kind of varnishes nish contribution to the quality of sound plays such an important
and glues [3]. role that is able, in some cases, to compensate possible struc-
tural defects (i.e. the instruments made from Guarneri del Gesù
compared to Stradivari’s) [5].
∗ Corresponding author. Tel.: +39 091 6809366; fax: +39 091 6809399. Coloured varnishes used by the makers of stringed instru-
E-mail address: caruso@pa.ismn.cnr.it (F. Caruso). ments from the 17th to the 19th century can be roughly classified
0021-9673/$ – see front matter © 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.chroma.2007.02.048
F. Caruso et al. / J. Chromatogr. A 1147 (2007) 206–212 207
A novel GC–MS technique, combining different kinds of ysis showed that in methanol solutions more quantity of every
derivatization, for the identification of lipids, proteins and other material was dissolved (excluding Kios mastic tears and Manila
analytes from painted artworks has been proposed by Colombini copal). Results of GC–MS analyses confirmed what stated
and coworkers [25]. above.
In the present work, we report on a procedure for the anal- The double bass varnish samples were treated with 0.50 mL
ysis of fatty and amino acids’ GC–MS profile identification, of methanol and left in a screwed cap tube in a dark and dry
based: (i) on the extraction of fatty acids with methanol and place for about two weeks at room temperature.
derivatization with acidic methanol; (ii) on proteins hydrolysis
with aqueous HCl and a double step amino acids’ derivatization 2.2.2. Treatment
with acidic methanol and TFAA. Our experimental procedures Every suspension of reference material was filtered through a
are based on the choice of common and low-cost derivatiz- small column of anhydrous sodium sulphate to eliminate water
ing reagents and solvents. Furthermore, our procedures can be and solid residues. The column was previously packed with the
adopted by restoration and conservation laboratories to charac- relative solvent. The filtered solutions were dried in a water
terize lipid binders and proteinaceous glues in artworks. bath at 50–60 ◦ C under a gentle nitrogen stream. The resulting
The whole procedures have been validated by trials on several residue was treated with 2.00 mL of hexane containing hexade-
reference materials and by the direct comparison with proto- cane (internal standard) at 20 L/L. Subsequently, 2.00 mL of
cols available in literature regarding other kinds of artefacts about 1:5 (v/v) HCl in methanol were added to each tube to
[12,16–18,20,26–32]. produce a biphasic system which was heated in a 50–60 ◦ C
water bath for approximately 2 h. After cooling to room tem-
2. Experimental perature, the hydrocarbon phase was separated and filtered on
a small column of anhydrous sodium sulphate – again to elimi-
2.1. Sampling nate water and possible solid residues – previously packed with
hexane with hexadecane. Solutions were then injected into the
Seven samples of varnish and two samples of glue were taken GC–MS.
from the ancient artefact. Samples were taken by gently scraping Regarding the three oils, 50 L of each were diluted with
the surface of the instrument with a scalpel at nine points, thought 2.00 mL of hexane with hexadecane, treated with 2.00 mL of
to be the most representative ones. the aforementioned methylating reagent, esterified (or transes-
Samples collection is described in Table 1. terified) and cooled. The hydrocarbon phase was then separated
and filtered.
2.2. Experimental procedure for the analysis of fatty acids The double bass samples were treated like the solid reference
materials.
Trials of fatty acids analyses and subsequent rigorous anal-
yses were carried out on several reference materials: three 2.3. Experimental procedure for the analysis of amino acids
siccative oils (linseed, boiled linseed and walnut) and seven dif-
ferent kinds of resins (Manila copal, Kios mastic tears, sandarac, Two GC–MS analyses were carried on each glue sample
orange shellac, lemon shellac, white shellac and Kusmi shellac). taken as references: rabbit glue and bone glue (20 mg and
Afterwards, the seven varnish samples from the double bass were 28 mg, respectively). Eight pure amino acids were used as
analysed. standards: aspartic acid, glutamic acid, phenylalanine, glycine,
4-hydroxyproline, proline, serine, valine, to allow identification
2.2.1. Extraction of these in test samples.
For every solid reference material, a solubility test with
three different solvents (acetone (polar aprotic solvent), chlo- 2.3.1. Hydrolysis
roform (apolar solvent) and methanol (polar protic solvent)) Following indications from the literature [33,34], each sam-
was conducted for one week. Initial approximate visual anal- ple of modern and double bass glue was transferred into a 50 mL
Table 1
Sampling description and weight of the analyzed samples
Sample Sampling description Sample ID Mass of analyzed sample (mg)
Sample 1 Varnish sample from the intern left f-hole. VCB 1 0.7
Sample 2 Taken from intern right f-hole. VCB 2 0.4
Sample 3 Taken from intern upper-right f-hole. VCB 3 0.7
Sample 4 Taken from intern left f-hole VCB 4 5.1
Sample 5 Taken from the bottom of the instrument. VCB 5 2.8
Sample 6 Taken from the right edge of the f-hole. VCB 6 1.8
Sample 7 Taken from the right edge of the harmonic case and from the bottom. VCB 7 2.3
Sample 8 Glue taken from the intern left side of the harmonic case. CCB 8 12.1
Sample 9 Glue taken from the intern bottom of the harmonic case. CCB 9 49.7
F. Caruso et al. / J. Chromatogr. A 1147 (2007) 206–212 209
round bottom flask, treated with about 5 mL of HCl 1:1 (v/v) in provided by CRPR. Reference materials were chosen on the
water and refluxed for 24 h into a hot oil bath (100–110 ◦ C). Pure basis of historical sources of old violins varnish composition
amino acids were only treated with HCl 1:1. [5].
Fig. 2. TIC of methanol extracted and derivatized solution of VCB 6 sample (a); TIC of methanol extracted and derivatized solution of VCB 5 sample (the one from
the restored bottom of the instrument) (b). The internal standard (hexadecane) is labelled as “Rif. C16H34”.
In order to further confirm this evidence, the ratio between unsaturated fatty acids in the same siccative oil and, as a conse-
the areas of peaks relative to azelaic acid dimethyl ester and quence, the contemporaneity of different samples [16].
to palmitic acid methyl ester (Azel/P) is considered. This Azel/P ratios for double bass varnish samples are also
ratio allows to establish the oxidative degradation degree of reported in Table 2.
Table 2
Experimental P/S and Azel/P ratios of the double bass varnish microsamples 3.2. Proteinaceous amino acids determination in glue
and arithmetic means of the values samples
Sample Experimental P/S ratio Experimental Azel/P ratio
Amino acids derivatization involves, apart from the esterifica-
VCB 1 2.76 ± 0.01 0.31 ± 0.01
VCB 2 3.34 ± 0.01 0.31 ± 0.01
tion of the carboxylic groups by acid methanol, the quantitative
VCB 3 3.43 ± 0.01 0.35 ± 0.01 acetylation of the amino groups with TFAA.
VCB 4 2.88 ± 0.01 0.34 ± 0.01 The two TICs of the glue double bass samples are shown in
VCB 6 3.40 ± 0.01 0.32 ± 0.01 Fig. 3a and b.
VCB 7 2.81 ± 0.01 0.36 ± 0.01 The identification of amino acids derivates peaks is achieved
Mean 3.1 ± 0.3 0.33 ± 0.02
VCB 5 1.52 ± 0.01 0.08 ± 0.01
by the direct comparison of retention times of the chro-
matograms peaks and by mass spectra of the eight amino acids
F. Caruso et al. / J. Chromatogr. A 1147 (2007) 206–212 211
Fig. 3. TIC of hydrolyzed and derivatized solution of CCB 8 sample (a); TIC of hydrolyzed and derivatized solution of CCB 9 sample (b); TIC of hydrolyzed and
derivatized solution of bone glue sample (c); TIC of hydrolyzed and derivatized solution of rabbit glue sample (d). The internal standard (hexadecane) is labelled as
“Rif. C16H34”.
used as references. The amino acids found in the analyzed sam- Both double bass glue samples (CCB 8 and CCB 9) show
ples are listed in Table 3. high relative concentration in glycine (Gly), proline (Pro) and
A mass fragment at 69 m/z is present in each mass spec- 4-hydroxyproline (Pro-OH). It is well known that an abundance
trum of the considered derivatized amino acids. This fragment is of these three amino acids reveals the presence of collagen-like
assigned to the formation of the CF3 + cation in agreement with proteins amounting an animal glue [41]. In fact, the presence
literature [40]. of hydroxyproline, an amino acid very rarely found in other
The relative amino acids percentage composition of the two proteins, prevents collagens from an early folding during biosyn-
double bass glue samples is obtained by integration of the peaks thetic process [42].
areas (Table 3). The results of the analysis of the two available samples of
modern animal glue used in wooden artworks restoration (bone
glue and rabbit glue) confirms the presence of collagen deriving
Table 3
amino acids (Fig. 3c and d and Table 3).
Relative amino acids weight percentage for double bass glue samples CCB 8
and CCB 9 and for modern bone glue (BG) and rabbit glue (RG) samples
4. Conclusions
Amino acid Relative weight percentage
CCB 8 CCB 9 BG RG The use of the little destructive (in terms of sample mass)
Gly 21 32 15 17 procedure permits the identification of fatty and amino acids
Val – 5 2 1 in samples of varnish and glue of the 18th century double bass
Ser 7 – 3 3 Panormus by Vincenzo Trusiano and gives useful information
Asp 3 6 6 4
for restorer, historians of art and luthiers, in good agreement with
Pro 35 23 19 20
Pro-OH 19 15 21 22 historical references about the construction of ancient musical
Glu 4 10 18 18 instruments. Furthermore, data analyses shed some relevant light
Phe 11 9 17 14 on the ancient formulations (the workshop secrets) of varnishes
In sample CCB 8, valine derivate peak area was not detected while, in CCB 9, and glues of the makers of stringed instruments. In particular, the
serine derivate peak area was not detected. Amino acids are ordered according use of walnut oil as a component of the binding medium for this
to their elution times. ancient double bass varnish is detected; a previous restoration
212 F. Caruso et al. / J. Chromatogr. A 1147 (2007) 206–212
treatment on the bottom of the ancient instrument is confirmed [19] Y.J. Yang, M.H. Choi, M.-J. Paik, H.-R. Yoon, B.C. Chung, J. Chromatogr.
by the experimental data; the use of animal glue as adhesive in B 742 (2000) 37.
the construction of the instrument is detected. [20] J. Dron, R. Linke, E. Rosenberg, M. Schreiner, J. Chromatogr. A 1047
(2004) 111.
[21] L.D. Metcalfe, A.A. Schmitz, Anal. Chem. 33 (1961) 363.
Acknowledgements [22] P. Campo, G.A. Sorial, M.T. Suidan, A.D. Venosa, Talenta 68 (2006) 888.
[23] W.W. Christie, J. Lipid Res. 23 (1982) 1072.
The authors wish to thank the Conservatorio di Musica “Vin- [24] C. Tokarski, E. Martin, C. Rolando, C. Cren-Olivé, Anal. Chem. 78 (2006)
cenzo Bellini” (Palermo, Italy) for providing the ancient musical 1494.
[25] A. Andreotti, I. Bonaduce, M.P. Colombini, G. Gautier, F. Modugno, E.
instrument and the whole staff of CRPR (in particular, Mr. Gio- Ribechini, Anal. Chem. 78 (2006) 4490.
vanni Garofalo) for the technical support in the sampling and [26] M.J. Casas-Catalán, M.T. Doménech-Carbó, R. Mateo-Castro, J.V.
working operations on the musical instrument. Gimeno-Adelantado, F. Bosch-Reig, J. Chromatogr. A 1025 (2004)
269.
References [27] A. Casoli, P.C. Musini, G. Palla, J. Chromatogr. A 731 (1996) 237.
[28] R.P. Evershed, S.N. Dudd, M.S. Copley, R. Berstan, A.W. Stott, H. Mot-
tram, S.A. Buckley, Z. Crossman, Acc. Chem. Res. 35 (2002) 660.
[1] A. Gambetta, E.L. De Capua, M. Mercatini, P. Cicolani, Bollettino ICR –
[29] J.V. Gimeno-Adelantado, R. Mateo-Castro, M.T. Doménech-Carbó, F.
Nuova Serie, no. 2, Nardini, Firenze, 2001, p. 104.
Bosch-Reig, A. Doménech-Carbó, M.J. Casas-Catalán, L. Osete-Cortina,
[2] R. Not, Atti della Giornata di Studio, la Biblioteca Comunale di S. Agostino
J. Chromatogr. A 922 (2001) 385.
di Taormina, Interventi per la salvaguardia e la conservazione del patrimo-
[30] E. Manzano, A.G. Bueno, A. Gonzalez-Casado, M. del Olmo, J. Cult. Herit.
nio librario. Regione Siciliana Assessorato BB. CC. AA. e P. I., Centro
1 (2000) 19.
Regionale Progettazione e Restauro, Palermo, 2003, p. 105.
[31] C. Marinach, M.-C. Papillon, C. Pepe, J. Cult. Herit. 5 (2004) 231.
[3] C. Quaglierini, L. Amorosi, Chimica e Tecnologia dei Materiali per l’arte,
[32] L. Osete-Cortina, M.T. Doménech-Carbó, R. Mateo-Castro, J.V. Gimeno-
Zanichelli, Bologna, 1991, p. 496.
Adelantado, F. Bosch-Reig, J. Chromatogr. A 1024 (2004) 187.
[4] V. Massa, G. Scicolone, Le Vernici per il Restauro – I Leganti, Nardini,
[33] R.H. Garrett, C.M. Grisham, Biochemistry, second ed., Brooks/Cole,
Firenze, 1991.
Pacific Grove, CA, 1999, p. 112.
[5] S.F. Sacconi, I Segreti di Stradivari, Libreria del Convegno, Cremona, 1979.
[34] R.K. Murray, D.K. Granner, P.A. Mayes, V.W. Rodwell, Harper’s Illustrated
[6] J.-P. Echard, Spectrochim. Acta, Part B 59 (2004) 1663.
Biochemistry, 26th ed., Lange Medical Books/McGraw-Hill, New York,
[7] A. Van Bohlen, F. Meyer, Spectrochim. Acta, Part B 52 (1997) 1053.
NY, 2003, p. 20.
[8] L. Burgio, R.J.H. Clark, L. Sheldon, G.D. Smith, Anal. Chem. 77 (2005)
[35] J.S. Mills, R. White, The Organic Chemistry of Museum Objects,
1261.
Butterworths–Heinemann, Oxford, 1994.
[9] B.J. Marquardt, S.R. Goode, S.M. Angel, Anal. Chem. 68 (1996) 977.
[36] M. Lazzari, O. Chiantore, Polym. Degrad. Stab. 65 (1999) 303.
[10] J.M. Rosenfeld, Anal. Chim. Acta 465 (2002) 93.
[37] P. Vandenabeele, B. Wehling, L. Moens, H. Edwards, M. De Reu, G. Van
[11] P. Hušek, J. Chromatogr. B 717 (1998) 57.
Hooydonk, Anal. Chim. Acta 407 (2000) 261.
[12] R. Mateo Castro, M.T. Doménech Carbó, V. Peris Martı́nez, J.V. Gimeno
[38] J.S. Mills, Stud. Conserv. 11 (1966) 92.
Adelantado, F. Bosch Reig, J. Chromatogr. A 778 (1997) 373. [39] F. Cappitelli, T. Learner, O. Chiantore, J. Anal. Appl. Pyrol. 63 (2002) 339.
[13] J. Peris-Vicente, J.V. Gimeno Adelantado, M.T. Doménech Carbó, R.
[40] E. Pretsch, J.-T. Clerc, J. Seibl, W. Simon, Tables of Spectral Data for Struc-
Mateo Castro, F. Bosch Reig, J. Chromatogr. A 1101 (2006) 254.
ture Determination of Organic Compounds, second ed., Springer-Verlag,
[14] M.-J. Paik, K.-O. Lee, H.-S. Shin, J. Chromatogr. B 721 (1999) 3.
Berlin, 1989.
[15] K.-L. Woo, J.-I. Kim, J. Chromatogr. A 862 (1999) 199.
[41] R.K. Murray, D.K. Granner, P.A. Mayes, V.W. Rodwell, Harper’s Illustrated
[16] M.P. Colombini, F. Modugno, M. Giacomelli, S. Francesconi, J. Chro-
Biochemistry, 26th ed., Lange Medical Books/McGraw-Hill, New York,
matogr. A 846 (1999) 113.
NY, 2003, p. 38.
[17] M.P. Colombini, F. Modugno, S. Giannarelli, R. Fuoco, M. Matteini,
[42] D.J. Prockop in W.J. Lennarz, M.D. Lane, Encyclopedia of Biological
Microchem. J. 67 (2000) 385.
Chemistry, Academic Press, Elsevier Science, San Diego, CA, 2004, p.
[18] V. Pitthard, S. Stanek, M. Griesser, T. Muxeneder, Chromatographia 62
482.
(2005) 175.