Lasers in the Conservation of Artworks: LACONA VI Proceedings, Vienna, Austria, Sept. 21--25, 2005 (Springer Proceedings in Physics, 116) 3540721290, 9783540721291

Citation preview

richter mrichter utek autek einer hreiner ds.Eds.



J. J.Nimmrichter Nimmrichter W.W.Kautek Kautek M.M.Schreiner Schreiner Editors Editors SPRINGER SPRINGER PROCEEDINGS PROCEEDINGS IN IN PHYSICS PHYSICS116116

Lasers in the Conservation Lasers in the Conservation

Lasers Lasers ininthe theConservation Conservation ofofArtworks Artworks LACONA LACONAVIVIProceedings, Proceedings, Vienna, Vienna,Austria, Austria,Sept. Sept.21–25, 21–25,2005 2005

 

springer proceedings in physics 116

springer proceedings in physics 96 Electromagnetics in a Complex World Editors: I.M. Pinto, V. Galdi, and L.B. Felsen 97 Fields, Networks, Computational Methods and Systems in Modern Electrodynamics A Tribute to Leopold B. Felsen Editors: P. Russer and M. Mongiardo 98 Particle Physics and the Universe Proceedings of the 9th Adriatic Meeting, Sept. 2003, Dubrovnik Editors: J. Trampeti´c and J. Wess 99 Cosmic Explosions On the 10th Anniversary of SN1993J (IAU Colloquium 192) Editors: J. M. Marcaide and K. W. Weiler 100 Lasers in the Conservation of Artworks LACONA V Proceedings, Osnabr¨uck, Germany, Sept. 15–18, 2003 Editors: K. Dickmann, C. Fotakis, and J.F. Asmus 101 Progress in Turbulence Editors: J. Peinke, A. Kittel, S. Barth, and M. Oberlack 102 Adaptive Optics for Industry and Medicine Proceedings of the 4th International Workshop Editor: U. Wittrock 103 Computer Simulation Studies in Condensed-Matter Physics XVII Editors: D.P. Landau, S.P. Lewis, and H.-B. Sch¨uttler 104 Complex Computing-Networks Brain-like and Wave-oriented Electrodynamic Algorithms Editors: I.C. G¨oknar and L. Sevgi 105 Computer Simulation Studies in Condensed-Matter Physics XVIII Editors: D.P. Landau, S.P. Lewis, and H.-B. Sch¨uttler 106 Modern Trends in Geomechanics Editors: W. Wu and H.S. Yu

108 Hadron Collider Physics 2005 Proceedings of the 1st Hadron Collider Physics Symposium, Les Diablerets, Switzerland, July 4–9, 2005 Editors: M. Campanelli, A. Clark, and X. Wu 109 Progress in Turbulence II Proceedings of the iTi Conference in Turbulence 2005 Editors: M. Oberlack, G. Khujadze, S. Guenther, T. Weller, M. Frewer, J. Peinke, S. Barth 110 Nonequilibrium Carrier Dynamics in Semiconductors Proceedings of the 14th International Conference, July 25–29, 2005, Chicago, USA Editors: M. Saraniti, U. Ravaioli 111 Vibration Problems ICOVP 2005 Editors: E. Inan, A. Kiris 112 Experimental Unsaturated Soil Mechanics Editor: T. Schanz 113 Theoretical and Numerical Unsaturated Soil Mechanics Editor: T. Schanz 114 Advances in Medical Engineering Editor: T.M. Buzug 115 X-Ray Lasers 2006 Proceedings of the 10th International Conference, August 20–25, 2006, Berlin, Germany Editors: P.V. Nickles, K.A. Janulewicz 116 Lasers in the Conservation of Artworks LACONA VI Proceedings, Vienna, Austria, Sept. 21–25, 2005 Editors: J. Nimmrichter, W. Kautek, M. Schreiner 117 Advances in Turbulence XI Proceedings of the 11th EUROMECH European Turbulence Conference, June 25–28, 2007, Porto, Portugal Editors: J.M.L.M. Palma and A. Silva Lopes

107 Microscopy of Semiconducting Materials Proceedings of the 14th Conference, April 11–14, 2005, Oxford, UK Editors: A.G. Cullis and J.L. Hutchison

Volumes 70–95 are listed at the end of the book.

J. Nimmrichter W. Kautek M. Schreiner (Eds.)

Lasers in the Conservation of Artworks LACONA VI Proceedings, Vienna, Austria, Sept. 21–25, 2005

With 419 Figures

123

Mag. Johann Nimmrichter Federal Office for Care and Protection of Monuments Arsenal 15/4, 1030 Vienna, Austria E-mail: offi[email protected]

Professor Dr. Wolfgang Kautek University of Vienna, Department of Physical Chemistry W¨ahringer Str. 42, 1090 Vienna, Austria E-mail: [email protected]

Professor Dr. Manfred Schreiner Academy of Fine Arts Vienna, Institute of Science and Technology in Art Schillerplatz 3, 1010 Vienna, Austria E-mail: [email protected]

ISSN 0930-8989 ISBN 978-3-540-72129-1 Springer Berlin Heidelberg New York Library of Congress Control Number: 2007928748 This work is subject to copyright. All rights are reserved, whether the whole or part of the material is concerned, specif ically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microf ilm or in any other way, and storage in data banks. Duplication of this publication or parts thereof is permitted only under the provisions of the German Copyright Law of September 9, 1965, in its current version, and permission for use must always be obtained from Springer-Verlag. Violations are liable to prosecution under the German Copyright Law. Springer is a part of Springer Science+Business Media. springer.com © Springer-Verlag Berlin Heidelberg 2007 The use of general descriptive names, registered names, trademarks, etc. in this publication does not imply, even in the absence of a specif ic statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. Prodcution: SPI Publisher Services Cover design: eStudio Calamar Steinen Printed on acid-free paper

SPIN: 12041855

57/3180/SPi

-543210

Preface

Conservation and protection of works of art as well as of rare remnants of natural history has turned more and more into a race against time. Environments all over the world have become increasingly aggressive causing damage or at least deterioration to surfaces meant to be created for eternity. Conventional techniques do a lot against most of these dangers, but new approaches of high technology have to be explored to preserve the heritage of human civilization as well as the precious specimens of former life such as the feathers’ of birds which died out generations ago. Mechanical and chemical methods are involved in traditional conservation treatments. Contactless cleaning by lasers, on the other hand, is a new and prospering field of laser materials processing. It allows avoiding mechanical disruption and the disadvantage of cleaning fluids – may they be toxic or just water – which could cause potentially long-term degradation of the substrate or health hazards. Moreover, laser cleaning may have the potential to accelerate conservatory work with high quality and moderate costs, and, thus, may help archives’, museums’ and collections’ strained budgets. Laser cleaning in semiconductor, automotive and aerospace industries has already been motivated by cost-savings, yield enhancement, and environmental concerns so that substantial literature about laser processing and cleaning of technical surfaces has accumulated in scientific and technological journals in recent years. This wealth of knowledge and experience, however, is usually not accessible to the conservation, museum, and archiving community. Therefore, the series of the “International Conferences on Lasers in the Conservation of Artworks” – LACONA – was initiated by Costas Fotakis organizing LACONA I 1995 in Heraklion, Greece. This was followed by LACONA II 1997 in Liverpool, Great Britain, LACONA III 1999 in Florence, Italy, LACONA IV 2001 in Paris, France, and LACONA V 2003 in Osnabrück, Germany. The success of these unique conferences motivated the LACONA Permanent Scientific Committee to organize a LACONA VI – this time in the very heart of Europe, in Vienna, Austria.

VI

Preface

The general development in laser conservation has led to the observation that scientific approaches and diagnostics have been introduced in an extent as never before in conservation. The key issues of the state of the art and future developments of laser cleaning of artefacts turned out to be as sketched in the following. Paradigm Change of Conservation. Laser cleaning applies highly localized deposition of heat by a laser beam in contrast to traditional conservation involving both room-temperature mechanical and chemical methods. Advanced Chemical Analysis and Diagnostics. In addition to the inspection by the conservator’s eye, micromorphological and spectroscopic methods are increasingly employed. Inhomogeneity and Precision. The high-precision deliverance of laser radiation to morphologically and chemically inhomogeneous artefact surfaces allows an unprecedented treatment quality. Integration. Merging laser cleaning with complementary conventional restoration steps may provide unrivalled solutions. Automation. Laser precision processing can be highly automated allowing better precision, safety and cost-effectiveness in the future. The 6th International Conference on Lasers in the Conservation of Artworks (LACONA VI) took place in Vienna, Austria, 21–25 September 2005. It represented the above listed new developments which entered the present proceedings volume. Moreover, LACONA VI ran under the auspices of the United Nations endorsed “World Year of Physics 2005” initiative which started by the European Physical Society to demonstrate that natural sciences provide a significant basis for the development of understanding nature, and that scientific research and its applications are a major driving force to scientific and technological development, and remain a vital factor in addressing the challenges of the 21st century. The “World Year of Physics 2005” highlighted the vitality of natural science and its importance in the coming millennium, and will commemorate the pioneering contributions of Albert Einstein in 1905. I want to thank Johann Nimmrichter, Chairman, and Manfred Schreiner, Co-Chairman of LACONA VI, for there unmatched enthusiasm and dedication to make LACONA VI a success. Further there has to be mentioned the invaluable support by the LACONA Permanent Scientific Committee, the LACONA Local Organizers (public institutions in Vienna), the LACONA Local Congress Committee, and last not least the LACONA Sponsors. Finally, I would like to thank Robert Linke and Ed Teppo for their careful and generous support during the preparation of the proceedings of LACONA VI. Vienna, May 2007

Wolfgang Kautek

Contents

List of Committees . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . XVII List of Sponsors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . XXI List of Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .XXIII 1 Serendipity, Punctuated J.F. Asmus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1

Part I Metal 2 Laser Cleaning of Corroded Steel Surfaces: A Comparison with Mechanical Cleaning Methods Y.S. Koh, J. Powell, A. Kaplan, and J. Carlevi . . . . . . . . . . . . . . . . . . . .

13

3 Laser Cleaning of Gildings M. Panzner, G. Wiedemann, M. Meier, W. Conrad, A. Kempe, and T. Hutsch . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

21

4 Current Work in Laser Cleaning of the Porta del Paradiso S. Agnoletti, A. Brini, and L. Nicolai . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

29

5 Cleaning Historical Metals: Performance of Laser Technology in Monument Preservation A. Gervais, M. Meier, P. Mottner, G. Wiedemann, W. Conrad, and G. Haber . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

37

6 Laser Cleaning the Abergavenny Hoard: Silver Coins from the Time of William the Conqueror M. Davis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

45

VIII

Contents

Part II Stone 7 The Application of Laser Cleaning in the Conservation of Twelve Limestone Relief Panels on St. George’s Hall M. Cooper and S. Sportun . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

55

8 The Potential Use of Laser Ablation for Selective Cleaning of Indiana Limestone K.C. Normandin, L. Powers, D. Slaton, and M.J. Scheffler . . . . . . . . .

65

9 Laser Cleaning of a Renaissance Epitaph with Traces of Azurite J. Nimmrichter and R. Linke . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

75

10 Laser Cleaning of Peristyle in Diocletian Palace in Split (HR) D. Almesberger, A. Rizzo, A. Zanini, and R. Geometrante . . . . . . . . . .

83

11 Phenomenological Characterisation of Stone Cleaning by Different Laser Pulse Duration and Wavelength S. Siano, M. Giamello, L. Bartoli, A. Mencaglia, V. Parfenov, and R. Salimbeni . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

87

12 The Cleaning of the Parthenon West Frieze by Means of Combined IR- and UV-Radiation K. Frantzikinaki, G. Marakis, A. Panou, C. Vasiliadis, E. Papakonstantinou, P. Pouli, T. Ditsa, V. Zafiropulos, and C. Fotakis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

97

13 A Comprehensive Study of the Coloration Effect Associated with Laser Cleaning of Pollution Encrustations from Stonework P. Pouli, G. Totou, V. Zafiropulos, C. Fotakis, M. Oujja, E. Rebollar, M. Castillejo, C. Domingo, and A. Laborde . . . . . . . . . . . . . . . . . . . . . . .

105

14 Poultices as a Way to Eliminate the Yellowing Effect Linked to Limestone Laser Cleaning V. Vergès-Belmin and M. Labouré . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

115

15 Experimental Investigations and Removal of Encrustations from Interior Stone Decorations of King Sigismund’s Chapel at Wawel Castle in Cracow A. Koss, J. Marczak, and M. Strzelec . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

125

16 Nd:YAG Laser Cleaning of Red Stone Materials: Evaluation of the Damage C. Colombo, E. Martoni, M. Realini, A. Sansonetti, and G. Valentini

133

Contents

IX

17 Exists a Demand for Nd:YAG Laser Technology in the Restoration of Stone Artworks and Architectural Surfaces? E. Pummer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

143

18 The SALUT Project: Study of Advanced Laser Techniques for the Uncovering of Polychromed Works of Art G. Van der Snickt, A. De Boeck, K. Keutgens, and D. Anthierens . . . .

151

Part III Inorganic Materials 19 Comparison of Wet and Dry Laser Cleaning of Artworks A. Sarzyński, K. Jach, and J. Marczak . . . . . . . . . . . . . . . . . . . . . . . . . . .

161

20 Laser Cleaning of Avian Eggshell L. Cornish, A. Ball, and D. Russell . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

169

21 Removal of Strong Sinter Layers on Archaeological Artworks with Nd:YAG Laser J. Hildenhagen, K. Dickmann, and H.-G. Hartke . . . . . . . . . . . . . . . . . . .

177

22 From the Lab to the Scaffold: Laser Cleaning of Polychromed Architectonic Elements and Sculptures M. Castillejo, C. Domingo, F. Guerra-Librero, M. Jadraque, M. Martín, M. Oujja, E. Rebollar, and R. Torres . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

185

23 Integration of Laser Ablation Techniques for Cleaning the Wall Paintings of the Sagrestia Vecchia and Cappella del Manto in Santa Maria della Scala, Siena S. Siano, A. Brunetto, A. Mencaglia, G. Guasparri, A. Scala, F. Droghini, and A. Bagnoli . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

191

24 Preliminary Results of the Er:YAG Laser Cleaning of Mural Paintings A. Andreotti, M.P. Colombini, A. Felici, A. deCruz, G. Lanterna, M. Lanfranchi, K. Nakahara, and F. Penaglia . . . . . . . . . . . . . . . . . . . . .

203

Part IV Organic Materials 25 Preliminary Results of the Er:YAG Laser Cleaning of Textiles, Paper and Parchment A. Andreotti, M.P. Colombini, S. Conti, A. deCruz, G. Lanterna, L. Nussio, K. Nakahara, and F. Penaglia . . . . . . . . . . . . . . . . . . . . . . . . .

213

X

Contents

26 Simultaneous UV-IR Nd:YAG Laser Cleaning of Leather Artifacts S. Batishche, A. Kouzmouk, H. Tatur, T. Gorovets, U. Pilipenka, V. Ukhau, and W. Kautek . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

221

27 An Evaluation of Nd:YAG Laser-Cleaned Basketry in Comparison with Commonly Used Methods A. Elliott, A. Bezúr, and J. Thornton . . . . . . . . . . . . . . . . . . . . . . . . . . . .

229

28 Novel Applications of the Er:YAG Laser Cleaning of Old Paintings A. Andreotti, P. Bracco, M.P. Colombini, A. deCruz, G. Lanterna, K. Nakahara, and F. Penaglia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

239

29 A Final Report on the Oxidation and Composition Gradients of Aged Painting Varnishes Studied with Pulsed UV Laser Ablation C. Theodorakopoulos, V. Zafiropulos, and J.J. Boon . . . . . . . . . . . . . . . .

249

30 A New Solution for the Painting Artwork Rear Cleaning and Restoration: The Laser Cleaning S.E. Andriani, I.M. Catalano, A. Brunetto, G. Daurelio, and F. Vona . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

257

31 Removal of Simulated Dust from Water-Based Acrylic Emulsion Paints by Laser Irradiation at IR, VIS and UV Wavelengths M. Westergaard, P. Pouli, C. Theodorakopoulos, V. Zafiropulos, J. Bredal-Jørgensen, and U. Staal Dinesen . . . . . . . . . . . . . . . . . . . . . . . .

269

32 Traditional and Laser Cleaning Methods of Historic Picture Post Cards M. Mäder, H. Holle, M. Schreiner, S. Pentzien, J. Krüger, and W. Kautek . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

281

33 Femtosecond Laser Cleaning of Painted Artefacts; Is this the Way Forward? P. Pouli, G. Bounos, S. Georgiou, and C. Fotakis . . . . . . . . . . . . . . . . . .

287

34 Laser Cleaning of Polyurethane Foam: An Investigation using Three Variants of Commercial PU Products U. Staal Dinesen and M. Westergaard . . . . . . . . . . . . . . . . . . . . . . . . . . . .

295

35 Excimer Laser Ablation of Egg Tempera Paints and Varnishes P.J. Morais, R. Bordalo, L. dos Santos, S.F. Marques, E. Salgueiredo, and H. Gouveia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

303

Contents

XI

36 Laser Cleaning of Undyed Silk: Indications of Chemical Change K. von Lerber, M. Strlic, J. Kolar, J. Krüger, S. Pentzien, C. Kennedy, T. Wess, M. Sokhan, and W. Kautek . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

313

37 Determination of a Working Range for the Laser Cleaning of Soiled Silk J. Krüger, S. Pentzien, and K. von Lerber . . . . . . . . . . . . . . . . . . . . . . . .

321

38 Laser Versus Conventional Cleaning Methods: Do the Costs Outweigh the Benefits? P. van Dalen, R. Broere, and H.A. Aziz . . . . . . . . . . . . . . . . . . . . . . . . . .

329

Part V Analytical Techniques 39 Raman Spectroscopy: New Light on Ancient Artefacts P. Vandenabeele and L. Moens . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

341

40 Pigment Identification on “The Ecstasy of St. Theresa” Painting by Raman Microscopy D. Marano, M. Marmontelli, G.E. De Benedetto, I.M. Catalano, L. Sabbatini, and F. Vona . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

349

41 Colorimetry, LIBS and Raman Experiments on Renaissance Green Sandstone Decoration During Laser Cleaning of King Sigismund’s Chapel in Wawel Castle, Cracow, Poland A. Sarzynski, W. Skrzeczanowski, and J. Marczak . . . . . . . . . . . . . . . . . .

355

42 Non-Destructive Observation of the Laser Treatment Effect on Historical Paper via the Laser-Induced Fluorescence Spectra K. Komar and G. Śliwiński . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

361

43 Effects of LIBS Measurement Parameters on Wall Paintings Pigments Alteration and Detection R. Bruder, D. Menut, and V. Detalle . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

367

44 A Parametric Linear Correlation Method for the Analysis of LIBS Spectral Data E. Tzamali and D. Anglos . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

377

45 Investigation on Painting Materials in “Madonna col Bambino e S. Giovannino” by Botticelli D. Bersani, P.P. Lottici, A. Casoli, M. Ferrari, S. Lottini, and D. Cauzzi . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

383

XII

Contents

46 Laser-Induced Plasma Spectroscopy for the Analysis of Roman Ceramics Terra Sigillata A.J. López, G. Nicolás, M.P. Mateo, V. Piñón, and A. Ramil . . . . . . .

391

47 Laser-Induced Fluorescence Analysis of Protein-Based Binding Media A. Nevin, S. Cather, D. Anglos, and C. Fotakis . . . . . . . . . . . . . . . . . . . .

399

48 Applications of a Compact Portable Raman Spectrometer for the Field Analysis of Pigments in Works of Art S. Bruni and V. Guglielmi . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

407

49 Classification of Patinas Found on Surfaces of Historical Buildings by Means of Laser-Induced Breakdown Spectroscopy C. Vázquez-Calvo, A. Giakoumaki, D. Anglos, M. Álvarez de Buergo, and R. Fort . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

415

50 Laser-Induced Breakdown Spectroscopy of Cinematographic Film M. Oujja, C. Abrusci, S. Gaspard, E. Rebollar, A. del Amo, F. Catalina, and M. Castillejo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

421

51 Online Monitoring of the Laser Cleaning of Marbles by LIBS Sulphur Detection V. Lazic, F. Colao, R. Fantoni, V. Spizzichino, and E. Teppo . . . . . . .

429

52 Low Resolution LIBS for Online-Monitoring During Laser Cleaning Based on Correlation with Reference Spectra M. Lentjes, K. Dickmann, and J. Meijer . . . . . . . . . . . . . . . . . . . . . . . . . .

437

53 Pigment Identification on a XIV/XV c. Wooden Crucifix Using Raman and LIBS Techniques M. Sawczak, G. Śliwiński, A. Kaminska, M. Oujja, M. Castillejo, C. Domingo, and M. Klossowska . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

445

54 MOLAB, a Mobile Laboratory for In Situ Non-Invasive Studies in Arts and Archaeology B.G. Brunetti, M. Matteini, C. Miliani, L. Pezzati, and D. Pinna . . . .

453

Part VI Scanning Techniques 55 From 3D Scanning to Analytical Heritage Documentation M. Schaich . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

463

Contents

XIII

56 Cleaning of Painted Surfaces and Examination of Cleaning by 3D-Measurement Technology at the August Deusser Museum, Zurzach P.-B. Eipper and G. Frankowski . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

473

57 Applicability of Optical Coherence Tomography at 1.55 µm to the Examination of Oil Paintings A. Szkulmowska, M. Góra, M. Targowska, B. Rouba, D. Stifter, E. Breuer, and P. Targowski . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

487

58 Varnish Thickness Determination by Spectral Optical Coherence Tomography I. Gorczyńska, M. Wojtkowski, M. Szkulmowski, T. Bajraszewski, B. Rouba, A. Kowalczyk, and P. Targowski . . . . . . . . . . . . . . . . . . . . . . . .

493

59 Multidimensional Data Analysis of Scanning Laser Doppler Vibrometer Measurements: An Application to the Diagnostics of Frescos at the US Capitol J. Vignola, J. Bucaro, J. Tressler, D. Ellingston, A. Kurdila, G. Adams, B. Marchetti, A. Agnani, E. Esposito, and E.P. Tomasini .

499

60 Spectral Domain Optical Coherence Tomography as the Profilometric Tool for Examination of the Environmental Influence on Paintings on Canvas T. Bajraszewski, I. Gorczyńska, B. Rouba, and P. Targowski . . . . . . . .

507

61 Polish Experience with Advanced Digital Heritage Recording Methodology, including 3D Laser Scanning, CAD, and GIS Application, as the Most Accurate and Flexible Response for Archaeology and Conservation Needs at Jan III Sobieski’s Residence in Wilanów P. Baranowski, K. Czajkowski, M. Gładki, T. Morysiński, R. Szambelan, and A. Rzonca . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

513

62 Evaluation by Laser Micro-Profilometry of Morphological Changes Induced on Stone Materials by Laser Cleaning C. Colombo, C. Daffara, R. Fontana, M.Ch. Gambino, M. Mastroianni, E. Pampaloni, M. Realini, and A. Sansonetti . . . . . . .

523

63 A Mobile True Colour Topometric Sensor for Documentation and Analysis in Art Conservation Z. Böröcz, D. Dirksen, G. Bischoff, and G. von Bally . . . . . . . . . . . . . . .

527

64 Reconstruction of the Pegasus Statue on Top of the State Opera House in Vienna using Photogrammetry and Terrestrial and Close-Range Laser Scanning C. Ressl . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

535

XIV

Contents

65 Some Experiences in 3D Laser Scanning for Assisting Restoration and Evaluating Damage in Cultural Heritage L.M. Fuentes, J. Finat, J.J. Fernández-Martin, J. Martínez, and J.I. SanJose . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

543

66 Monitoring of Deformations Induced by Crystal Growth of Salts in Porous Systems Using Microscopic Speckle Pattern Interferometry G. Gülker, A. El Jarad, K.D. Hinsch, H. Juling, K. Linnow, M. Steiger, St. Brüggerhoff, and D. Kirchner . . . . . . . . . . . . . . . . . . . . . .

553

67 Cultural Heritage Documentation by Combining Near-Range Photogrammetry and Terrestrial Laser Scanning: St. Stephen’s Cathedral, Vienna F. Zehetner and N. Studnicka . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

561

68 Laser Engraving Gulf Pearl Shell – Aiding the Reconstruction of the Lyre of Ur C. Rawcliffe, M. Aston, A. Lowings, M.C. Sharp, and K.G. Watkins .

573

69 Fluorescence Lidar Multispectral Imaging for Diagnosis of Historical Monuments, Övedskloster: A Swedish Case Study R. Grönlund, J. Hällström, S. Svanberg, and K. Barup . . . . . . . . . . . . . .

583

R Measurement Technology for Use on Surfaces 70 OptoSurf
of Historic Buildings and Monuments Cleaned by Laser W.P. Weinhold, A. Wortmann, C. Diegelmann, E. Pummer, N. Pascua, Th. Brennan, R. Burkhardt, and L. Goretzki . . . . . . . . . . . .

593

71 Multi-Tasking Non-Destructive Laser Technology in Conservation Diagnostic Procedures V. Tornari, E. Tsiranidou, Y. Orphanos, C. Falldorf, R. Klattenhof, E. Esposito, A. Agnani, R. Dabu, A. Stratan, A. Anastassopoulos, D. Schipper, J. Hasperhoven, M. Stefanaggi, H. Bonnici, and D. Ursu . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

601

72 Time-Dependent Defect Detection by Combination of Holographic Tools E. Tsiranidou, V. Tornari, Y. Orphanos, C. Kalpouzos, and M. Stefanaggi . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

611

Part VII Safety and Miscellaneous 73 Health Risks Caused by Particulate Emission During Laser Cleaning R. Ostrowski, St. Barcikowski, J. Marczak, A. Ostendorf, M. Strzelec, and J. Walter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

623

Contents

XV

74 Generation of Nano-Particles During Laser Ablation: Risk Assessment of Non-beam Hazards During Laser Cleaning St. Barcikowski, N. Bärsch, and A. Ostendorf . . . . . . . . . . . . . . . . . . . . .

631

75 A Novel Portable Multi-Wavelength Laser System A. Charlton and B. Dickinson . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

641

List of Committees

Permanent Scientific Committee Prof. Dr. Wolfgang Kautek (President) University of Vienna Department of Physical Chemistry Waehringer Strasse 42 1090 Vienna, Austria E-mail: [email protected] Prof. Dr. John F. Asmus (Honorary President) IPAPS University of California, San Diego UCSD Physics Dept 9500 Gilman Drive La Jolla, CA 92093, USA E-mail: [email protected] Margaret Abraham Los Angeles County Museum of Art 5905 Wilshire Blvd Los Angeles, CA 99036, USA E-mail: [email protected] Prof. Dr. Giorgio Bonsanti Opificio Delle Pietre Dure di Firenze Via Alafani 78 50121 Firenze, Italy E-mail: [email protected] Dr. Marta Castillejo Consejo Superior de Investigaciones Cientificas Instituto de Química Física Rocasolano Serrano 119 28006 Madrid, Spain E-mail: [email protected]

XVIII

List of Committees

Dr. Martin Cooper The Conservation Centre Whitechapel Liverpool L1 6HZ, UK E-mail: [email protected] Prof. Dr. Klaus Dickmann Fachhochschule Münster Laserzentrum Stegerwaldstr. 39 48565 Steinfurt, Germany E-mail: [email protected] Prof. Dr. Costas Fotakis Foundation for Research and Technology – Hellas (FO.R.T.H.) Institute of Electronic Structure & Laser Vassilika Vouton, P.O. Box 1527 Heraklion 71110, Crete, Greece E-mail: [email protected] Prof. Dr. Eberhard Koenig Freie Universitaet Berlin Kunsthistorisches Institut Koserstrasse 20 14195 Berlin, Germany E-mail: [email protected] Dr. Mauro Matteini Opificio delle Pietre Dure di Firenze Laboratorio Scientifico Viale Strozzi 1 50100 Firenze, Italy E-mail: [email protected] Mag. Johann Nimmrichter Federal Office for Care and Protection of Monuments (Bundesdenkmalamt) Department for Restoration and Conservation (Abteilung für Restaurierung und Konservierung) Arsenal, Objekt 15, Tor 4 1030 Wien, Austria E-mail: [email protected] Dr. Roxana Rãdvan National Institute of Research and Development for Optoelectronics (INOE) Centre for Restoration by Optoelectronical Techniques (CERTO) Platforma Magurele, 1 Atomistilor Str. 76900 Bucharest, Romania E-mail: [email protected]

List of Committees

Dr. Renzo Salimbeni Consiglio Nazionale delle Ricerche Istituto di Elettronica Quantistica Via Panciatichi 56/30 50127 Firenze, Italy E-mail: [email protected] Véronique Vergès-Belmin Laboratoire de Recherche des Monuments Historiques 29 rue de Paris 77420 Champs sur Marne, France E-mail: [email protected] Prof. h.c. Dr. Gert von Bally University of Münster Laboratory of Biophysics, Institute of Experimental Audiology Robert-Koch-Str. 45 48129 Münster, Germany E-mail: [email protected] Prof. Dr. Kenneth Watkins The University of Liverpool Department of Mechanical Engineering Liverpool, L69 3BX, UK E-mail: [email protected] Prof. Dr. Vassilis Zafiropulos Superior Technical Educational Institute of Crete Department of Human Nutrition & Dietetics Ioannou Kondylaki 46, 723 00 Sitia, Crete, Greece E-mail: zafi[email protected]

Local Congress Committee Johann Nimmrichter Chairman, Bundesdenkmalamt, Vienna, Austria Manfred Schreiner Co-Chairman, Academy of Fine Arts, Vienna, Austria Wolfgang Kautek Co-Chairman, Dept. of Phys. Chem., Univ. of Vienna, Austria Wolfgang Baatz Academy of Fine Arts, Vienna, Austria Andrea Böhm Bundesdenkmalamt, Vienna, Austria Dimitrios Boulasikis Conservator-Archaelogist, Mödling, Austria

XIX

XX

List of Committees

Giancarlo Calcagno Conservator-Restorer, Bassano del Grappa, Italy Gabriele Gürtler Bundesdenkmalamt, Vienna, Austria Eva Maria Höhle Bundesdenkmalamt, Vienna, Austria Manfred Koller Bundesdenkmalamt, IIC-Austria, Vienna, Austria Gabriele Krist University of Applied Arts, Vienna, Austria Robert Linke Bundesdenkmalamt, Vienna, Austria Erich Pummer Conservator-Restorer, Rossatz, Austria Johannes Riegl RIEGL Laser Measurement Systems GmbH, Horn, Austria Dieter Schuöcker Vienna University of Technology, Vienna, Austria Christopher Weeks Conservator-Restorer, Tring, UK Robert Wimmer Behindscreen, Vienna, Austria Wolfgang Zehetner Dombaumeister, Architect of St. Stephens Cathedral, Vienna, Austria

Local Organizing Institutions Federal Office for Care and Protection of Monuments Austria (Bundesdenkmalamt) Academy of Fine Arts Vienna (Akademie der bildenden Künste) University of Vienna (Universität Wien) Vienna University of Technology (Technische Universität Wien) Cathedral Masons Lodge of St. Stephens, Vienna (Dombauhütte St. Stephan) International Institute for Conservation (IIC), Austrian Group Austrian Conservator-Restorer Association (Österreichischer Restauratorenverband)

List of Sponsors

The financial support of all organisations is gratefully acknowledged. Riegl Laser Measurement Systems GmbH, www.riegl.com Bundesministerium für Bildung, Wissenschaft und Kultur, www.bmbwk.gv.at Bundesdenkmalamt, www.bda.at Akademie der bildenden Künste, www.akbild.ac.at Dr. Michael Häupl, Mayor of Vienna, www.wien.gv.at Casinos Austria, www.casinos.at COST G7 Artwork conservation by laser, http://alpha1.infim.ro/cost Bundeskammer der Architekten und Ingenieurskonsulenten, www.arching.at Linsinger Kulturgutvermessung, Photogrammetrie, 3D Scanning, www.linsinger.at ofi – Technologie & Innovation GmbH, Abteilung Bauwesen, www.ofi.co.at ELEN GROUP hightech laser, www.elengroup.com Quanta Systems S.p.A. Lasers & Lasersystems, www.quantasystem.com Rest. Felix Mackowitz, [email protected] Rest. Mag. Klaus Wedenig, info@denkmalpflegegmbh.at Rest. Mag. Ralph Kerschbaumer, [email protected] Rest. Erich Pummer, www.lasertech-artcons.at Steinmetzfirma Wolfgang Ecker, [email protected] Rest. Otto Blassnig, [email protected] Steinmetzfirma Rada, www.rada.at Rest. Johann Lindtner, [email protected]

XXII

List of Sponsors

Steinmetzfirma Johann Schaden, www.marmorbau-schaden.at Rest. Johannes Schlögl, [email protected] Rest. Mag. Josef Weninger, [email protected] Landesinnung Wien der Steinmetzmeister, [email protected] Rest. Gerhard Zottmann, www.zottmann.at Rest. Werner Campidell, [email protected] Arctron Ausgrabungen & Computerdokumentationen GmbH, www.arctron.de

List of Contributors

Abrusci, C., 421 Adams, G., 499 Agnani, A., 499, 601 Agnoletti, S., 29 Almesberger, D., 83 Álvarez de Buergo, M., 415 Anastassopoulos, A., 601 Andreotti, A., 203, 213, 239 Andriani, S.E., 257 Anglos, D., 377, 399, 415 Anthierens, D., 151 Asmus, J.F., 1 Aston, M., 573 Aziz, H.A., 329 Bagnoli, A., 191 Bajraszewski, T., 493, 507 Ball, A., 169 Bärsch, N., 631 Baranowski, P., 513 Barcikowski, St., 623, 631 Bartoli, L., 87 Barup, K., 583 Batishche, S., 221 Bersani, D., 383 Bezúr, A., 229 Bischoff, G., 527 Böröcz, Z., 527 Bonnici, H., 601 Boon, J.J., 249 Bordalo, R., 303 Bounos, G., 287 Brüggerhoff, St., 553

Bracco, P., 239 Bredal-Jørgensen, J., 269 Brennan, Th., 593 Breuer, E., 487 Brini, A., 29 Broere, R., 329 Bruder, R., 367 Brunetti, B.G., 453 Brunetto, A., 191, 257 Bruni, S., 407 Bucaro, J., 499 Burkhardt, R., 593 Carlevi, J., 13 Casoli, A., 383 Castillejo, M., 105, 185, 421, 445 Catalano, I.M., 257, 349 Catalina, F., 421 Cather, S., 399 Cauzzi, D., 383 Charlton, A., 641 Colao, F., 429 Colombini, M.P., 203, 213, 239 Colombo, C., 133, 523 Conrad, W., 21, 37 Conti, S., 213 Cooper, M., 55 Cornish, L., 169 Czajkowski, K., 513 Dabu, R., 601 Daffara. C., 523 Daurelio, G., 257 Davis, M., 45

XXIV

List of Contributors

De Benedetto, G.E., 349 De Boeck, A., 151 deCruz, A., 203, 213, 239 del Amo, A., 421 Detalle, V., 367 Dickinson, B., 641 Dickmann, K., 177, 437 Diegelmann, C., 593 Dirksen, D., 527 Ditsa, T., 97 Domingo, C., 105, 185, 445 dos Santos, L., 303 Droghini, F., 191 Eipper, P.-B., 473 El Jarad, A., 553 Ellingston, D., 499 Elliott, A., 229 Esposito, E., 499, 601 Falldorf, C., 601 Fantoni, R., 429 Felici, A., 203 Fernández-Martin, J.J., 543 Ferrari, M., 383 Finat, J., 543 Fontana, R., 523 Fort, R., 415 Fotakis, C., 97, 105, 287, 399 Frankowski, G., 473 Frantzikinaki, K., 97 Fuentes, L.M., 543 Gambino, M. Ch., 523 Gaspard, S., 421 Geometrante, R., 83 Georgiou, S., 287 Gervais, A., 37 Giakoumaki, A., 415 Giamello, M., 87 Gładki, M., 513 Góra, M., 487 Gorczyńska, I., 493, 507 Goretzki, L., 593 Gorovets, T., 221 Gouveia, H., 303 Grönlund, R., 583 Gülker, G., 553 Guasparri, G., 191

Guerra-Librero, F., 185 Guglielmi, V., 407 Haber, G., 37 Hällström, J., 583 Hartke, H.-G., 177 Hasperhoven, J., 601 Hildenhagen, J., 177 Hinsch, K.D., 553 Holle, H., 281 Hutsch, T., 21 Jach, K., 161 Jadraque, M., 185 Juling, H., 553 Kalpouzos, C., 611 Kaminska, A., 445 Kaplan, A., 13 Kautek, W., 221, 281, 313 Kempe, A., 21 Kennedy, C., 313 Keutgens, K., 151 Kirchner, D., 553 Klattenhof, R., 601 Klossowska, M., 445 Koh, Y.S., 13 Kolar, J., 313 Komar, K., 361 Koss, A., 125 Kouzmouk, A., 221 Kowalczyk, A., 493 Krüger, J., 313, 321 Krüger, J/, 281 Kurdila, A., 499 Laborde, A., 105 Labouré, M., 115 Lanfranchi, M., 203 Lanterna, G., 203, 213, 239 Lazic, V., 429 Lentjes, M., 437 Linke, R., 75 Linnow, K., 553 López, A.J., 391 Lottici, P. P., 383 Lottini, S., 383 Lowings, A., 573 Mäder, M., 281 Marakis, G., 97

List of Contributors Marano, D., 349 Marchetti, B., 499 Marczak, J., 125, 161, 355, 623 Marmontelli, M., 349 Marques, S.F., 303 Martínez, J., 543 Martín, M., 185 Martoni, E., 133 Mastroianni, M., 523 Mateo, M. P., 391 Matteini, M., 453 Meier, M., 21, 37 Meijer, J., 437 Mencaglia, A., 87, 191 Menut, D., 367 Miliani, C., 453 Moens, L., 341 Morais, P.J., 303 Morysiński, T., 513 Mottner, P., 37 Nakahara, K., 203, 213, 239 Nevin, A., 399 Nicolás, G., 391 Nicolai, L., 29 Nimmrichter, J., 75 Normandin, K.C., 65 Nussio, L., 213 Orphanos, Y., 601, 611 Ostendorf, A., 623, 631 Ostrowski, R., 623 Oujja, M., 105, 185, 421, 445 Pampaloni, E., 523 Panou, A., 97 Panzner, M., 21 Papakonstantinou, E., 97 Parfenov, V., 87 Pascua, N., 593 Penaglia, F., 203, 213, 239 Pentzien, S., 281, 313, 321 Pezzati, L., 453 Piñón, V., 391 Pilipenka, U., 221 Pinna, D., 453 Pouli, P., 97, 105, 269, 287 Powell, J., 13

Powers, L., 65 Pummer, E., 143, 593 Ramil, A., 391 Rawcliffe, C., 573 Realini, M., 133, 523 Rebollar, E., 105, 185, 421 Ressl, C., 535 Rizzo, A., 83 Rouba, B., 487, 493, 507 Russell, D., 169 Rzonca, A., 513 Śliwiński, G., 361, 445 Sabbatini, L., 349 Salgueiredo, E., 303 Salimbeni, R., 87 SanJose, J.I., 543 Sansonetti, A., 133, 523 Sarzyński, A., 161 Sarzynski, A., 355 Sawczak, M., 445 Scala, A., 191 Schaich, M., 463 Scheffler, M.J., 65 Schipper, D., 601 Schreiner, M., 281 Sharp, M.C., 573 Siano, S., 87 Skrzeczanowski, W., 355 Slaton, D., 65 Sokhan, M., 313 Spizzichino, V., 429 Sportun, S., 55 Staal Dinesen, U., 269, 295 Stefanaggi, M., 601, 611 Steiger, M., 553 Stifter, D., 487 Stratan, A., 601 Strlic, M., 313 Strzelec, M., 125, 623 Studnicka, N., 561 Svanberg, S., 583 Szambelan, R., 513 Szkulmowska, A., 487 Szkulmowski, M., 493 Targowska, M., 487 Targowski, P., 487, 493, 507

XXV

XXVI

List of Contributors

Tatur, H., 221 Teppo, E., 429 Theodorakopoulos, C., 249, 269 Thornton, J., 229 Tomasini, E.P., 499 Tornari, V., 601, 611 Torres, R., 185 Totou, G., 105 Tressler, J., 499 Tsiranidou, E., 601, 611 Tzamali, E., 377 Ukhau, V., 221 Ursu, D., 601 Valentini, G., 133 van Dalen, P., 329 Van der Snickt, G., 151 Vandenabeele, P., 341 Vasiliadis, C., 97

Vázquez-Calvo, C., 415 Vergès-Belmin, V., 115 Vignola, J., 499 von Bally, G., 527 von Lerber, K., 313, 321 Vona, F., 257, 349 Walter, J., 623 Watkins, K.G., 573 Weinhold, W.P., 593 Wess, T., 313 Westergaard, M., 269, 295 Wiedemann, G., 21, 37 Wojtkowski, M., 493 Wortmann, A., 593 Zafiropulos, V., 97, 105, 249, 269 Zanini, A., 83 Zehetner, F., 561

1 Serendipity, Punctuated J.F. Asmus Institute for Pure and Applied Physical Sciences University of California San Diego 9500 Gilman Dr., La Jolla, CA 92093-0360, USA [email protected] Summary. Laser divestment entered the field of art conservation through a nonlinear sequence of positive accidental events (serendipity) that involved the cinema industry, the invention of spread-spectrum and frequency-hopping communications, nuclear space propulsion, and oceanography. The unlikely chain of events began with the invention of a secure military communications system by a Viennese motion picture actress (1942). A first evaluation of the novel communications concept took place during a high-altitude nuclear test (TEAK) over the Pacific Ocean in 1958. The secure radio link proved to be a failure; however, analyses of the backscattered electromagnetic radiation contributed to the realization that nuclear-explosion plasmas need not be spherically symmetrical. Nobel Laureate Freeman Dyson exploited this nuclear option to guide in the design and prototype development of the ORION spaceship that was to rendezvous with the planet Saturn in 1970. For this space vehicle the high-specific-impulse nuclear propulsion was generated by means of superradiant X-ray-beam ablation of the spaceship’s rear surface by the remote detonation of a sequence of asymmetrical bombs projected rearward from the ORION. In the wake of the Nuclear Test Ban Treaty (1962) ORION was canceled. Through a Scripps Institution of Oceanography project in Venice (involving ORION scientists and holographic technology) the nondestructive radiation-ablation process found a resurrection in the field of stone conservation (1972). Ironically, the first major art-conservation project to employ laser ablation (Porta della Carta of the Palazzo Ducale) was paid for in part by Warner Brothers Motion Picture Studios (1980). Finally, the “Venice Laser Statue Cleaner” followed the Viennese actress (Hedy Lamarr/Hedwig Eva Maria Kiesler) to Hollywood where it was employed to treat the granite veneer of the Warner Center (1981).

1.1 Introduction The fields of art conservation and laser science merged, formally and fittingly, in the land of Polyclitus and Democritus with a 1995 event now called LACONA I (held at FORTH). However, appropriate that symbolic recognition of the sources of Western cultural heritage may seem, LACONA VI has, in Vienna, returned to the direct technological genesis of lasers in the service

2

J.F. Asmus

of art. The implausible trajectory of “unintended consequences” that led to the introduction of laser technology into art conservation was triggered in 1941– 1942 when Viennese cinema actress Hedy Lamarr invented a novel (jamming proof) concept for the radio transmission of guidance information to naval torpedoes. Subsequent decades witnessed initial evaluations of the Lamarr modulation schemes that helped uncover new avenues in nuclear weapons design as well as in the invention of the nuclear-propelled spaceship (ORION). Subsequently, the holographic plasma diagnostics developed for the engineering design of the spaceship were applied to the in situ archival recording of deteriorating Venetian statuary. This, in turn, led to the improbable realization that the radiation-propulsion mechanism of ORION could provide a means of selflimiting divestment (and conservation) of crumbling marble statues. The series of “connections” and happy accidents that helped in bringing about LACONA VI are summarized in the paragraphs that follow.

1.2 Hedy Lamarr and Her Communications Patent In 1942 Viennese motion picture actress Hedy Lamarr (Figs. 1.1 and 1.2) (Hedwig Eva Maria Kiesler) of MGM was granted US Patent #2,292,387 for a “Secret Communication System” based on her invention of spread-spectrum (Figs. 1.3 and 1.4) and frequency-hopping concepts. Evidently, the idea was a merging of art and science in that it sprang from her knowledge of the military business of her husband, Fritz Mandl, and her understanding of the player piano (gained from her friendship with artist George Antheil). As her discovery formed the basis of cell phone technology, Wi-Fi protocols, and the wireless Internet, she won a US$1/4M infringement claim against Corel Corporation and received the 1997 Electronic Frontier Award. (Upon receiving the award, 55 years after the fact, her response, “It’s about time,” received almost as much notice as her “au naturel” appearance in the 1933 Czech film, “Ecstasy.”)

Fig. 1.1. MGM motion picture star Hedy Lamarr

1 Serendipity, Punctuated

3

Fig. 1.2. The first page of Hedy Lamarr’s 1942 patent, “A Secret Communication System,” that introduced the frequency-hopping and spread-spectrum concepts to the communications field

Fig. 1.3. Spread-spectrum communication link of Project ARGUS during the highaltitude nuclear detonation, TEAK (inset)

Fig. 1.4. The receiver site on the island of Niihau, Hawaii

The first evaluation of Hedy Lamar’s approach to secure communications was carried out between Hawaiian Pacific Islands in 1958 during the Johnston Island high-altitude nuclear explosion TEAK (3.8 MT at 77 km altitude). Disappointingly, the experimental radio-wave transmission link was completely blacked out by the bomb’s gamma-ray-induced aurora. However, spectral

4

J.F. Asmus

analyses of the backscattered electromagnetic signal revealed that the H-bomb had, through a performance asymmetry, ejected a plasma jet.

1.3 Orion: Nuclear Spaceship The ARGUS backscatter data together with other theoretical and experimental results predicted that nuclear explosive devices possessed the potential for being redesigned into directed-energy radiation sources. Upon this realization, members of the TEAK team joined with theoretical physicist Freeman Dyson and virtuoso minibomb designer Theodore Taylor to exploit and optimize this phenomenon in order to develop a nuclear-propelled spaceship, ORION, for a mission to the planet Saturn (scheduled for 1970). Following a first ORION test flight (1962), the adoption of the Nuclear Test Ban Treaty led to the demise of the program. Figures 1.5–1.7 display a few of the test results of laboratory proof-of-principle ORION technology demonstrations that reveal the impulse delivered by laser ablation.

Fig. 1.5. Deformation of a restrained metallic coin through the impulse delivered by laser ablation pressure at a multigigawatt and kilojoule level

Fig. 1.6. A streak camera record of the laser propulsion of an unrestrained metallic disk to V = 20 km s−1

1 Serendipity, Punctuated

5

Fig. 1.7. Hypervelocity impact crater (and its cross section) produced by the energy released by a laser-propelled projectile

Fig. 1.8. A conceptual portrayal of a nuclear-driven ORION spaceship

Figure 1.8 presents a conceptual rendering of the ORION space vehicle near Mars.

1.4 Laser Divestment in Venice The real-time holographic diagnostics developed for ORION were resurrected for art-conservation purposes in Venice in January 1972. This was a consequence of a collaboration of Scripps Institution of Oceanography geophysicists and ORION project alumni in research directed toward the alleviation of the “acqua alta” problem being experienced by the lagoon. By March 1972 the ruby holographic laser was being employed to clean marble sculpture by means of radiation-induced ablation in accordance with results from the radiationhydrodynamic modeling of the earlier X-ray-beam nuclear-propulsion system (Fig. 1.9). This came about at the suggestion of Lorenzo Lazzarini and Giulia Musumeci of the Venetian Soprintendenza in response to the unacceptable cleaning results on friable stone with conventional air-abrasive and chemical approaches (Fig. 1.10).

6

J.F. Asmus

Fig. 1.9. Arch. Calcagno’s first in situ ruby-laser cleaning demonstration on one of the “Ruskin Capitals” of the Palazzo Ducale in Venice)

Fig. 1.10. Unsuccessful conventional cleaning test areas on a Piazza San Marco facade

This preamble to the formation of LACONA (more than a decade later by scientists at FORTH) came full circle with the entrance of an MGM competitor, Warner Brothers Films. At a 1977 Venice Film Festival event (a showing of “Clockwork Orange” in Asolo), Jack Warner, Jr. offered to divert a portion of his profits from “Clockwork Orange” to pay for a laser restoration feasibility project. He, together with his friends and associates, raised US$ 5,000 toward this end. Arch. Giancarlo Calcagno of the Soprintendenza selected the artwork to be the subject of the first laser cleaning demonstration. The piece he selected in the Porta della Carta was a marble relief depicting “The Last Supper.” It was approximately 60 cm high and 180 cm wide. After a protracted sequence of laser validation tests in the laboratory, the actual laser demonstration took place in 1980 when an Nd:YAG laser was used to clean the marble relief in support of the overall Porta della Carta conservation effort. As the relief had been laid horizontally on its back for treatment, the laser was mounted on a beam above the artwork. The laser beam was directed vertically downward to impinge on the marble surface. The laser head was attached to the supporting beam with a swivel joint so that the laser beam

1 Serendipity, Punctuated

7

Fig. 1.11. Porta della Carta of the Palazzo Ducale in Venice and a vertically mounted laser cleaning the marble relief of “The Last Supper”

Fig. 1.12. The central, cleaned area of the marble piece

could be manually scanned across the surface. The laser functioned in the normal mode (400 µs) at 1 J per pulse. In most areas a spot size of 3 mm was employed. Figure 1.11 shows the Porta della Carta and the laser pointing downward onto the relief. The initial, centrally cleaned area is shown in Fig. 1.12.

1.5 Return to Hollywood and the Cinema A new Warner Brothers corporate office complex was constructed in the Los Angeles area while the laser work proceeded in Venice. The following year (1981), with the completion of the central corporate tower of the Warner Center, the general contractor found that rubber cushioning used during shipping had discolored the tower’s South African granite veneer slabs. Chemical treatments that removed the stains also etched the stone and left it with a frosted appearance. Figure 1.13 shows the central tower of the Warner Center complex. The dark vertical stripes, as well as the horizontal bands at the top and bottom, are the South African granite veneers.

8

J.F. Asmus

Fig. 1.13. Warner center tower with the prominent granite veneer bands and vertical stripes

Fig. 1.14. A detail of laser treatment (bleaching) of the stained granite

As a last resort, the “Venetian” laser was sent to the Warner Tower for a cleaning trial. At low fluxes and high fluences, the laser-induced optical damage in the mineral grains of the stone. The resulting cleavages within the mineral grains resembled the normal heterogeneity of granite, yet masked the in-depth chemical blemish. This approach was selected as the most suitable treatment. Consequently, good fortune made one further appearance when the “Venetian” laser repaid a debt to the cinema industry by removing the blemishes from the exterior granite of the corporate center after the failure of chemical cleaning techniques. Figure 1.14 shows the results of the laser irradiation of the blemished granite veneer.

1.6 Conclusions Histories of developments in science and technology are replete with instances of unintended and/or unanticipated consequences. Sometimes such surprises are favorable. Often they bode disaster. All of the earliest pioneers of the laser have expressed bemusement at the laser’s entry into the field of art conservation practice (as well as its ubiquitous role in the worlds of the audio

1 Serendipity, Punctuated

9

CD and the video DVD). Most certainly, that occurrence is an “untended consequence” of investigations into the very diverse fields of spread-spectrum communications, deep-space nuclear propulsion, holographic plasma diagnostics, and archival holographic recording. In retrospect it is clear that laser surface divestment would have found its way into the field of art conservation at some point. However, the route that did lead initially to the laser in the arts is a testimonial to the tenacious punctuality with which serendipity invades the circuitry of technological progress. This individual route to innovation, beginning, and then returning to Vienna, is one more example (in a myriad of examples) demonstrating that discovery seldom proceeds in a linear and predictable fashion. Acknowledgments Herbert York, Theodore Taylor, Morris Scharff, Edward Creutz, Keith Boyer, Donald Adrian, and Robert Willis are thanked for their oral histories of the ORION and ARGUS Projects. Charles Townes, Ali Javan, Gordon Gould, William Bridges, and Arthur Schawlow are thanked for their oral histories of the invention of the laser. Lorenzo Lazzarini and Giulia Musumeci are thanked for first suggesting the laser cleaning of stone sculpture. The author’s science teacher at Chaffey High School (Mr. Sweihardt) is thanked for arranging summer employment with Project ARGUS at the US Naval Ordnance Laboratory. Much of the material in the above discourse was not published in a timely manner as an earlier manuscript (1978) was rejected (without review) by the Editor of Studies in Conservation as he deemed laser-divestment cleaning “too hypothetical to be taken seriously.” Many of the technical, political, and financial barriers to progress in the activities outlined above were surmounted through the intervention of the Viennese/American, Walter Munk (Director Emeritus, Institute for Geophysics and Planetary Physics, Scripps Institution of Oceanography at the University of California San Diego).

Part I

Metal

2 Laser Cleaning of Corroded Steel Surfaces: A Comparison with Mechanical Cleaning Methods Y.S. Koh1 , J. Powell2 , A. Kaplan2 , and J. Carlevi2 1

2

Kiruna Center for Conservation of Cultural Property, Arent Grapegatan 20, 98132 Kiruna, Sweden, [email protected] Luleå University of Technology, SE-971 87 Luleå, Sweden

Summary. Conservation often requires the removal of oxide layers from metal artifacts and new cleaning methods are being developed all the time. This paper provides a quantitative comparison of eight cleaning methods, three of which are mechanical (brushing or micro-blasting with Al2 O3 or glass beads) and five of which are laser dependent (TEA CO2 or Nd:YAG laser, with or without surface water). Surface profilometry and scanning electron microscopy have been used to compare the cleaned surfaces with the original, known, surface geometries.

2.1 Introduction For the purposes of the conservation of metal artifacts it is often necessary to remove surface oxide layers without damaging the metal below. Various mechanical and chemical surface treatments are available for the removal of surface corrosion and other contaminants [1], but these can damage the underlying metal. Consequently, there is a great deal of interest in developing a new cleaning technology which is less aggressive to the metal surface below the oxide layer. Over the past few years the traditional cleaning methods employed by conservators have been augmented by treatments involving the use of lasers [2, 3]. Laser cleaning is a non-toxic, environmentally friendly and non-contact process which can be carefully controlled to minimise damage to the metal beneath a corroded surface. For this reason this process has potential advantages over traditional chemical or mechanical cleaning methods. This paper directly compares the relative effectiveness of eight methods of cleaning a corroded steel surface (listed in Table 2.1).

2.2 Experimental Work To avoid use of actual historic artefacts and also to quantify the relative performance of the cleaning techniques, it was decided to produce a large

14

Y.S. Koh et al. Table 2.1. A list of the cleaning techniques compared in this study Mechanical methods

Laser-based methods

Micro-blasting with glass TEA CO2 laser (10,600 nm) Micro-blasting with Al2 O3 Nd:YAG laser (1,064 nm) on a dry surface Rotating steel brush Nd:YAG laser (1,064 nm) on a wet surface Nd:YAG laser (532 nm) on a dry surface Nd:YAG laser (532 nm) on a wet surface

Fig. 2.1. A typical cross section of a grooved sample

number of similar samples with known surface topology and corrosion levels. The cleaned surfaces could then be compared with each other and with a set of machined but uncorroded reference samples. The material used to produce the samples was a carbon steel (SS 1672) with the chemical formulation Fe 98.7%, C 0.47%, Si 0.25% and Mn 0.60%. The samples used in this work were machined to have a surface covered in parallel grooves with a cross section of the type shown in Fig. 2.1. The depth of the grooves was 0.25, 0.5 or 2.0 mm. The sample size was 30 × 30 × 13 mm. The samples were degreased with acetone before being systematically corroded in a corrosion-chamber. The samples were corroded for 3, 5 or 7 weeks by exposure to 0.1 M NaCl solution, which was sprayed on the samples twice a day.

2.3 Results Figure 2.2 shows the results of the cleaning trials for lightly corroded 0.5 mm grooves (i.e. peak-to-peak distance of 1 mm, peak height of 0.5 mm). The results have been arranged with the most effective cleaning method towards the top of the figure and the least effective towards the bottom (the uncorroded reference sample is shown first and the uncleaned, corroded reference sample is shown at the bottom of the figure). After reviewing a number of statistical roughness comparisons, it was found that a simple comparison of the range of groove amplitude was the most effective. Although corrosion and cleaning had little effect on the maximum groove amplitude, it had a considerable effect on the minimum amplitude measured where the original groove peak or trough would have been located. This location is easy to identify as the pitch of the grooves is known. This

2 Laser Cleaning of Corroded Steel Surfaces Pictures

Optical profilometry

Reference no corrosion

2mm Microblast, Al2O3 Microblast, glass

Amplitude Range (mm) 0.47 – 0.46 = 0.01 0.45 – 0.42 = 0.03 0.45 – 0.42 = 0.03

Nd:YAG laser 532 nm, wet

0.45 – 0.42 = 0.03

Nd:YAG laser 1064 nm, dry

0.42 – 0.37 = 0.05

Nd:YAG laser 1064 nm, wet

0.45 – 0.42 = 0.03

Rotating stainless steel brush

0.45 – 0.42 = 0.03

TEA CO2 laser 10600 nm

0.45 – 0.35 = 0.10

Nd:YAG laser 532 nm, dry

0.41 – 0.24 = 0.17

Reference corrosion

15

0.46 – 0.22 = 0.24

Fig. 2.2. A comparison of cleaning results for the 0.5 mm deep grooves and light corrosion

reduction in groove amplitude is either due to the corrosion products filling the bottom of the grooves or the corrosion process removing the peaks of the grooves. The groove amplitude measurements give only a rough guide to the effectiveness of the cleaning method and this is made clear by comparison of the micro-blast Al2 O3 and the stainless steel brush results in Fig. 2.2. Although

16

Y.S. Koh et al.

the groove amplitude measurements are identical for both samples, it is obvious that the original groove profile has been retained only in the case of the micro-blast sample. The steel brushing method has, by its nature, eroded the upper surface of the grooves giving the grooves a sharper peak than the original machined profile. Using this combination of groove amplitude measurements and visual assessment of the profiles, it was possible to rank the effectiveness of the eight cleaning methods for the 0.5 and 0.25 mm deep grooves for light, moderate and severe corrosion. The results of this grading procedure are presented in Table 2.2. By allocating points to each process in Table 2.2 depending on their performance in each case, it is possible to give an overall performance ranking and this is presented in Table 2.3. In the case of the moderately corroded samples there is not much difference between the profile for the two least effective cleaning methods and the uncleaned reference sample. Only the two micro-blast methods and the Nd:YAG – wet surface techniques reveal a pattern of regular sharp points. In the case of the heavily corroded samples the situation is even worse. Here, only the micro-blasted specimens give us any useful information about the original surface topology. An appropriate analogy here is that of a signal to noise ratio: The regular cycle of the original, uncorroded surface can be considered a signal and the corrosion process can be assumed to be obscuring this signal with random surface features or ‘noise’. The cleaning process attempts to

Table 2.2. A comparison of effectiveness of the eight cleaning methods for different grooves and corrosion conditions Light corrosion 0.5 mm 0.25 mm Groove Groove Best




Worst

A B C D E F G H

A B C D E G F H

Moderate corrosion 0.5 mm 0.25 mm Groove Groove A B E C D H G F

A B C E H G F D

Heavy corrosion 0.5 mm 0.25 mm Groove Groove A B C E F D H G

A B E C D H G F

A: Micro-blast Al2 O3 , (2 bar, Ø 0.050–0.075 mm), B: Micro-blast glass beads (6 bar, Ø 0.075–0.15 mm), C: Nd:YAG laser 532 nm per wet (pulse duration 10 nS, repetition rate 2.5 Hz, maximum pulse energy 300 mJ), D: Nd:YAG laser 1,064 nm per dry (pulse duration 10 nS, repetition rate 2.5 Hz, maximum pulse energy 600 mJ), E: Nd:YAG laser 1,064 nm per wet, F: Stainless steel brush (5,000 revolutions per min), G: TEA CO2 laser 10,600 nm (pulse duration 100 nS, repetition rate of 20 Hz, maximum pulse energy 4 J), H: Nd:YAG laser 532 nm per dry

2 Laser Cleaning of Corroded Steel Surfaces

17

Table 2.3. An overall ranking of the eight cleaning methods Ranking Best


Worst

1st 2nd 3rd 4th 5th 6th 7th 8th

Method Micro-blast – Al2 O3 Micro-blast – Glass Nd:YAG 532 nm – Wet Nd:YAG 1064 nm – Wet Nd:YAG 1064 nm – Dry Nd:YAG 532 nm – Dry Stainless steel brush TEA CO2 laser

remove the noise and re-identify the signal. Figure 2.3 shows that the signal was more rapidly corrupted by the corrosion process in the case of the 0.25 mm deep grooves. The results shown here are for the moderate corrosion samples but it is clear that even the micro-blasting techniques are less effective than they were in the case of the deeper grooves. The reason for this decrease in the signal to noise ratio is simply that the shallower grooves represent a ‘weaker signal’ than the deeper ones. This ‘weaker signal’ is more easily lost under the action of the corrosion process.

2.4 Discussion Reviewing all the results presented so far it is clear that the best cleaning results are obtained by the micro-blasting processes and the next most effective techniques involve the use of an Nd:YAG laser on a wet surface. It is clear from this result that the effective removal of corrosion products from an iron (or steel) surface must involve some mechanical action. In the case of the micro-blasting processes this mechanical action is provided by the momentum transfer from the high velocity Al2 O3 or glass particles. For the Nd:YAG/wet surface technique the mechanical action is provided by the rapid expansion and partial vapourisation of the infiltrated water in oxide layer which is heated by the laser energy. Only when the oxide layer is very shallow it is possible to remove it effectively by simple laser irradiation (at the energy densities considered here). In this case the mechanical action is simply one of local, thermally induced expansion leading to fracture and removal of the brittle oxide surface. Figure 2.4 provides some clues to the difference between the cleaning mechanisms for the dry and wet samples. The electron micro-graphs presented in Fig. 2.4 compare the cleaned surfaces after wet and dry laser cleaning with the 1,064 nm Nd:YAG laser. The samples presented here are those with 2 mm deep grooves (which were intrinsically too rough for examination by the optical profilometry used on the 0.25 and 0.5 mm deep grooves). The dry ‘cleaned’ surface is covered in a rather thick oxide layer which is clearly cracked and porous. It

18

Y.S. Koh et al. Pictures

Optical profilometry

Reference, no corrosion

2mm Microblast, Al2O3 Microblast, glass

Amplitude Range (mm) 0.23 – 0.21 = 0.02 0.18 – 0.15 = 0.03 0.19 – 0.15 = 0.04

Nd:YAG laser 532 nm, wet

0.27 – 0.17 = 0.10

Nd:YAG laser 1064 nm, wet

0.19 – 0.10 = 0.09

Nd:YAG laser 532 nm, dry

0.16 – 0.08 = 0.08

TEA CO2 laser 10600 nm

0.09 – 0.06 = 0.03

Rotating stainless steel brush

0.14 – 0.09 = 0.05

Nd:YAG laser 1064 nm, dry

0.08 – 0.03 = 0.05

Reference, corrosion

0.16 – 0.10 = 0.06

Fig. 2.3. A comparison of cleaning results for 0.25 mm grooves and moderate corrosion

seems likely that, in the case of the wet cleaned surface, the water infiltrated these cracks and pores. The incident laser has a wavelength (1,064 nm) which would pass though any surface water without being absorbed by it. The laser energy would be absorbed by the solid oxide layer which would rapidly heat up the water in the cracks and pores. The transfer of the heat from the solid to the surrounding water has been noted by Zapka et al. [4] and Grigoropoulos and Kim [5]. However, these studies do not entirely explain the results of this

2 Laser Cleaning of Corroded Steel Surfaces Nd:YAG laser 1064 nm, wet

× 25

SEM

× 100

SEM

× 500

× 25

SEM

× 100

SEM

× 500

19

4mm

Nd:YAG laser 1064 nm, dry

Fig. 2.4. Scanning electron micro-graphs demonstrating the difference in the cleaned surface for wet and dry laser cleaning

present investigation because here, we are dealing with an adherent coating of oxide tens or even hundreds of microns thick. A probable explanation for the removal of thick, wet oxide layers is the action of the sudden expansion of the vapourising water on the cracks and pores within the layer. The forces generated by vaporisation would rapidly open the cracks and pores to shatter the brittle oxide coating.

2.5 Conclusions Within the scope of this study, micro-blasting techniques cleaned iron oxides from steel surfaces more effectively than laser methods. It was also demonstrated that wet surface Nd:YAG laser techniques were more effective than dry surface techniques. This increase in effectiveness is probably the result of the break up of the oxide layer by the sudden expansion and vaporisation of trapped liquid in cracks and pores. Steel brushing cleaned the oxide from the surface in some cases but this was accompanied by substrate erosion. Finally, it was shown that the TEA CO2 laser was less effective than the Nd:YAG laser in removing oxide layers. Acknowledgements The authors gratefully acknowledge the Swedish National Heritage Board, who sponsored this study. Also thanks goes to Johnny Grahn and Tore Silver for the technical help, Luleå University of Technology, Sweden.

20

Y.S. Koh et al.

References 1. J. M. Cronyn, The Elements of Archaeological Conservation, Routledge, London, 1992. 2. J. F. Asmus, in Laser Techniques and Systems in Art Conservation, Edited by R. Salimbeni, SPIE Vol. 4402, 1–7, 2001. 3. M. Cooper, Laser Cleaning in Conservation: An Introduction, Oxford, 1998. 4. W. Zapka, W. Ziemlich, and A. C. Tam, in Applied Physics Letters Vol. 58, 2217, 1991. 5. C. P. Grigoropoulos and D. Kim, in Laser Cleaning, Edited by B. Luk’yanchuk, 229, Singapore, 2002.

3 Laser Cleaning of Gildings ∗

M. Panzner1 , G. Wiedemann1 , M. Meier3 , W. Conrad2 , and A. Kempe1 , and T. Hutsch1 1

∗ 2 3

Fraunhofer Institute Material and Beam Technology Winterbergstr. 28, 01277 Dresden, Germany [email protected] Freelance Restorer, Obere Parkstr. 10, 06295 Lutherstadt Eisleben, Germany Lower Saxony Department of Preservation of Ancient Monuments Scharnhorststr. 1, 30175 Hannover, Germany

Summary. Results of laser cleaning experiments on different gilding types like leaf gilding and fire gilding are presented in this contribution by means of three tested art objects. The reflectivity of gold is advantageously high for the typical laser cleaning wavelength of 1,064 nm. Additionally, to avoid damage like gold loss, the transfer of the absorbed laser pulse energy into the art object by thermal conduction is considered. Fire gilded surfaces are most easily cleaned because of the good heat transfer conditions which imply a high threshold intensity with respect to damage. This is different for leaf gilded surfaces but suitable laser cleaning parameters have also been found for this case. The results of laser cleaning experiments are presented by photography, microscopy, SEM and EDX analysis.

3.1 Introduction The surfaces of many objects of art and architecture are gilded by various techniques like fire, leaf or electrochemical gilding. Especially in the case of outdoor objects exposed to rain, dust and environmental pollution usually accumulate as dirt layers on the gilded surface. In many cases, the gold layer below is damaged because of corrosion or other aging processes like the development of cracks (craquelure) in the ground layer of leaf gildings. Mechanical cleaning of such surfaces would cause further loss of gold. So a contactless cleaning method like laser cleaning could be advantageous. The pressure wave induced by the ablation of dirt cause much lower mechanical stress than mechanical cleaning techniques would [1]. Laser cleaning makes use of the fact that, in virtually any particular case, one can find a range of exposure parameters where selective removal of dirt is feasible so that damage to the object is precluded. Finding suitable parameters is not a matter of trial and error but can be greatly eased by the knowledge of the optical and thermal properties of the materials involved [2].

22

M. Panzner et al.

Fig. 3.1. Reflectivity of gold, copper and silver vs. wavelength [3], with Nd:YAG laser and its harmonics shown

Fig. 3.2. Scheme of the energy deposition flow of heat during and after absorption of laser pulses

Above 700 nm, the reflectivity of gold exceeds 95% for the Nd:YAG laser wavelength of 1,064 nm (Fig. 3.1). Obviously it would not make sense to apply the harmonics of the Nd:YAG laser to clean gildings. Thus the selective removal of surface pollution from solid gold by laser cleaning with Nd:YAG laser should be possible without damage. The deposited laser power of 0

a) 15.00

20.00

ala gly val 4.1.A 10.7 24.8 2.8 4.1.B 6.3 14.5 2.7

leu 4.0 3.6

25.00

ile 1.9 1.8

30.00

met 0.0 0.0

ser 3.4 5.3

35.00

40.00

pro phe asp glu lys 5.1 1.2 12.6 23.0 4.5 11.6 2.6 14.1 32.8 0.7

hyp tyr 6.1 0.0 3.9 0.0

b)

c)

3 2

colla −10

−8

uovo database

1

−6

−4

−2

0 0

2

4

4.1.A

−1 −2 −3

casein a

4.1.B

−4

Fig. 25.5. Chromatogram of sample 4.1.A and amino acids percentage content of two samples from a stain on the ancient paper (a); and principal component analysis loading plot (b) and score plot (c)

220

A. Andreotti et al.

score plot towards the casein cluster. The laser ablation of the material from the paper caused a decrease of the proline content. The amino acids profile of parchment was similar to that of animal glue, but in a few samples after laser ablation, a decrease of glutamic acid, and proline was observed. Materials deposited on the paper surface are easier to ablate than materials which penetrated into the structure. This is not due to the laser energy absorption, but to the separation between laser spots and inner paper structure. When the materials impregnate the fibres it is more difficult to control the laser action avoiding secondary effects on the substrate. Future experimentation will take into account the Er:YAG laser combination with traditional restoration procedures to verify if many of these effects can be reduced.

25.3 Conclusions The cleaning of paper is more difficult than cleaning of other materials because of the very broad parameters encountered: Different stains on the same substrate can achieve an excellent result or a bad one. The same stain on a different substrate could in one case be difficult while, on another, an easy and complete stain removal is possible. The thin layer of the superficial materials, the strong adhesion among these materials, the fibrous structure of substrates, the wide variety of paper, and their state of conservation can be complex with many variables which complicate the approach to the problems with use of the Er:YAG laser and require further studies and applications. The laser cleaning is an effective procedure when helped by a wet cleaning. Acknowledgement Ente Cassa di Risparmio of Florence is gratefully acknowledged for the large contribution given for this research.

References 1. E. Adamkiewitz, P. Bracco, M. P. Colombini, A. De Cruz, G. Lanterna, M. Matteini, K. Nakahara, O. Sartiani, and M. L. Wolbarsht, in Journal of Cultural Heritage, Vol. 4(1001), 202, 2003. 2. A. Andreotti. M. P. Colombini, G. Lanterna, and M. Rizzi, in Journal of Cultural Heritage, Vol. 4(1001), 355, 2003.

26 Simultaneous UV–IR Nd:YAG Laser Cleaning of Leather Artifacts ∗

S. Batishche1 , A. Kouzmouk1 , H. Tatur1 , T. Gorovets2 , U. Pilipenka3 , V. Ukhau3 , and W. Kautek4 1

∗ 2 3

4

National Academy of Sciences of Republic of Belarus, Institute of Physics, F. Scorina Ave. 68, 220012 Minsk, Belarus [email protected] National Art Museum of Belarus, Minsk, Belarus Research Technological Enterprise “Belmicrosystems” of “Integral” Amalgamation, Minsk, Belarus University of Vienna, Department of Physical Chemistry, Waehringer Str. 42 1090 Vienna, Austria

Summary. Ancient leather samples from original upholstered furniture were treated with nanosecond Nd:YAG laser radiation with wavelengths of 1,064, 532, and 266 nm. The novel approach was the simultaneous application of these wavelengths. It opened new approaches for laser cleaning leather. Extensive diagnostics such as absorbance of different layers of leather, chemical composition and microscopic inspection studies before and after cleaning were conducted. Advantageous results with simultaneous UV–IR (266 nm + 1, 064 nm) radiation are presented and discussed.

26.1 Introduction Laser cleaning of biogenetic artifacts has concentrated mainly on painting varnish, paper, and parchment in recent years [1–5]. In this context, best results for the removal of dark contaminants from, for example, cellulose and collagen materials were observed with visible laser radiation (532 nm) allowing maximum contrast between the light absorption of foreign materials and the substrate [4, 5]. Leather processing such as laser engraving and marking, in contrast to cleaning, has been developed to an industrial process relying on the in-depth ablation of the substrate. In this chapter, results optimized multiwavelength conditions of laser cleaning dark contaminants on a leather surface with a Nd:YAG laser are presented.

222

S. Batishche et al.

26.2 Experimental Methods The prototype Nd:YAG laser cleaning system [6, 7] employed in this study allows the spatial and temporal overlap of the fundamental beam (1,064 nm, up to 300–500 mJ) with its second (532 nm, up to 800 mJ), and fourth (266 nm, up to 300 mJ) harmonic frequencies. The absorbance spectra measurements of leather samples were performed by a Varian Cary 500 spectrometer. The cleaning results were evaluated with an optical microscope (Stemi 2000-C) and a scanning electron microscope (JEOL 840, Stereoscan-360 with EDXSpectrometer AN-10000). The colorimetric investigations were conducted following the standards indicated by a microspectrophotometer (MPV-SP, Leitz). In these investigations three light sources (A – incandescent lamp, C – daylight lamp, D65 – luminescent lamp) were used. The study of the elemental composition of different layers of leather samples before and after cleaning was conducted by the methods of local X-ray spectral analysis on a raster-type electronic microscope (S-360 with an AN-10000 analyzer) and by secondary-ion mass spectroscopy (IMS-4F, CAMECA). The following parameters were employed: primary ions O2 + , Cs+ ; element range from H to U; depth resolution of 5–30 nm.

26.3 Results and Discussion Ancient leather samples from original upholstered furniture were laser treated (Fig. 26.1). A schematic structure of leather samples is shown in Fig. 26.2. The

Fig. 26.1. Original leather samples from an upholstered furniture. Austrian Museum of Applied Arts/Contemporary Art (MAK), Vienna

26 Simultaneous UV–IR Nd:YAG Laser Cleaning of Leather Artifacts

223

Fig. 26.2. Schematic layer structure of a contaminated leather surface

surface was covered by a thin dark layer of pollutants located mostly in cracks and grooves. A number of areas may be chosen corresponding to particular conditions of the contaminant layer (1) an area on the right side of the sample that had been subjected only to minimal light and minimal contacts, referred to as “original,” (2) an area on the right side of the sample that was subjected to intense light but minimal contacts, “faded,” (3) an area on the right side of the sample that was subjected to intense light and strong contacts, “polluted,” (4) a transition layer between the leather itself and the dye layer, “intermediate,” (5) an area on the right side of the sample from which the layers of pollutants, varnish, dye, and intermediate were scraped off, “base,” and (6) an area on the backside of the sample, “backside.” Figure 26.3 shows approximate (including absorbance and scattering of light) absorbance spectra of various parts of the sample. The spectra were taken from KBr pellets (20 mm in diameter, 1 mm thickness, mass 1 g) with powdered material from various parts of the sample surface (5 mg, thin layers were scraped off the surface). The absorbance spectra of pure KBr was subtracted. The densities of contaminant, leather, and intermediate phases were found to be 0.706, 1.276, and 0.987 g cm−3 , respectively. From Fig. 26.3, it is seen that the absorbance for all samples show localized peaks in the ranges of 500, 360, and 260 nm. For 266-nm radiation, however, a maximum is observed around 260 nm. Use of 266-nm radiation obviously allows removing pollutants layer by layer to a certain extent. Irradiation of the backside with 35 and 120 mJ cm−2 gives analogous results – a minor rise in absorbance in the vicinity of 355 nm in comparison with the nonirradiated backside. That suggests that 266-nm irradiation causes little photochemical changes of a collagen structure of leather.

224

S. Batishche et al.

4000

266 nm

3000

K, cm−1

355 nm

2000

KB01 KB11 KB21 KB31 KB41 KB51 KB71 KB91 KB61 KB81

532 nm 1000 1064 nm

0 300

400

500

1100

λ, nm Fig. 26.3. Typical absorbance spectra of various parts of a contaminated leather (see Fig. 26.1): Original – clean part kept in the dark, Faded – clean part that was exposed to sunlight, Polluted – polluted part, Intermediate – a layer between the top side and the base, Base – a layer of the base just below the Intermediate, Cleaned – polluted parts of the top layer cleaned by 266-nm laser radiation at 35 mJ cm−2 , Cleaned01 – faded parts of the top layer cleaned by 266-nm laser radiation at 35 mJ cm−2 , Backside− Ini – polluted part of the verso side (the untreated base), Backside− Irr and Backside− Irr− 01 – polluted parts of the verso side cleaned by 266-nm laser radiation at 35 and 120 mJ cm−2 , respectively

It should be noted, that the mechanical properties of the backside changed substantially upon treatment. This was obvious due to the fact that the particles scraped off from the irradiated backside showed a fluffy structure. The irradiation of the top side gives a more substantial rise in absorption in the vicinity of 355 nm as compared to the nonirradiated one. These spectral changes may involve photochemical changes just in the varnish layers due to the low penetration depth of the UV light. The inspection of the surface of leather samples before and after laser cleaning with optical microscopy, scanning electron microscopy, secondary ion mass spectroscopy, colorimetry and EDX allowed several conclusions: 1. The density of carbon is substantially higher on the surface than in the bulk. A higher density of Cr and Fe is also observed on the surface. In the bulk, Na, Al, Ca, K are increased. The Ti density is practically the same on the surface and in the bulk. These differences are connected, most

26 Simultaneous UV–IR Nd:YAG Laser Cleaning of Leather Artifacts

225

Table 26.1. Lab color coefficients

1 2 3

L∗

A a∗

b∗

26.0 25.4 21.9

10.0 9.8 3.0

10.3 11.0 1.3

light source C L∗ a∗ 24.6 24.0 21.7

7.1 7.0 1.4

b∗

L∗

D65 a∗

b∗

8.2 8.9 1.0

24.6 23.9 21.7

7.9 7.7 1.7

8.0 8.7 1.0

Fig. 26.4. The surface of the leather sample after (at the left) and before (at the right) cleaning by nanosecond laser radiation simultaneously with the wavelengths of 266 and 1,064 nm at fluence F266 = 35 mJ cm−2 and F1,064 = 80 mJ cm−2 , respectively

likely, with peculiarities of original leather manufacture, or with airborne contaminants, handling, intrinsic impurities from soil, etc. 2. The analysis of the chemical composition of the contaminated surface before and after laser treatment showed the presence of a large number of chemical elements (C, O, Na, S, Si, K, Ca). There were also some traces of Cl, Al, Mg, Fe. After laser treatment, concentrations of such elements as S, Si, K, Ca, Cl, Al, Mg, Fe decreased, while peaks of O, Al, Na increased. 3. Color coefficients for xyz, XYZ, Lab, and LCH color calculation systems for different angles of incidence were measured. Table 26.1 presents Lab

226

S. Batishche et al.

color coefficients at angle of incidence of 10◦ , where color differences are most distinct, for the original area (1), the polluted area before (3), and after (2) laser treatments at 266 nm. The initial difference between L∗ a∗ b∗ coefficients of the original area and the polluted area before laser treatment for the light sources A and D65 amounts to 1.5–9 times. The final difference between the color coefficients of the original area and the polluted area after laser treatment at 266 nm becomes small for all of the light sources and angles of incidence. These investigations show that the best laser cleaning results are produced with 266 nm at F = 35–100 mJ cm−2 and with the simultaneous combination of 266 nm and 1,064 nm with F = 35–100 mJ cm−2 and 80–250 mJ cm−2 , respectively. After laser cleaning under optimum conditions the surface gets a dull luster, which can be explained by the partial removal of the varnish layer. Expert evaluation shows that results with 266+1, 064 nm in combination look more attractive (Fig. 26.4). Laser radiation at 532 or 1,064 nm resulted in the damage of the surface in all fluence ranges.

26.4 Conclusions The data obtained suggest that the laser treatment of an ancient polluted leather surface by UV + IR radiation of nanosecond duration at 266 nm and 266 + 1, 064 nm may result in cleaning of the surface, recovering the original color and preserving most of the superficial varnish structure. Expert evaluation shows that the 266 + 1, 064 nm wavelength combination yields the most attractive cleaning result in respect to other wavelengths. Acknowledgments We acknowledge partial financial support by the ISTC project B-373-2 “Laser cleaning of art works of metals, paper, parchment, fabric, leather, and painting: research of possibilities, development of technologies and laser equipment”. W.K. thanks P. Noever and M. Trummer of the Austrian Museum of Applied Arts/Contemporary Art (MAK), Vienna, for providing the original samples in the context of the EUREKA project “Laser Cleaning of Paper and Parchment (LACLEPA)” N 1681.

References 1. W. Kautek and E. König (Eds.), Lasers in the Conservation of Artworks I, Restauratorenblätter, Vienna 1997. 2. M. Cooper, Laser Cleaning in Conservation, Butterworth-Heinemann, 1998. 3. W. Kautek, S. Pentzien, P. Rudolph, J. Krüger, and E. König, in Applied. Surface Science Vol. 127–129, 746, 1998.

26 Simultaneous UV–IR Nd:YAG Laser Cleaning of Leather Artifacts

227

4. P. Rudolph, F. J. Ligterink, J. L. Pedersoli Jr., M. van Bommel, J. Bos, H. A. Aziz, J. B. G. A. Havermans, H. Scholten, D. Schipper, and W. Kautek, in Appl. Phys. A, Vol. 9, 181, 2004. 5. J. Kolar, M. Strlic, S. Pentzien and W. Kautek, in Appl. Phys. A 71, 87, 2000. 6. A. Anisimov, S. Batishche, A. Egglezis, C. Fotakis, A. Kouzmouk, P. Pouli, H. Tatur, and V. Zafiropulos, in 3-rd International Workshop on New Trends in Laser Cleaning, October 3–4, 2003, Crete, Greece. 7. S. Batishche, A. Englezis, T. Gorovets, A. Kouzmouk, U. Pilipenka, P. Pouli, H. Tatur, G. Totou, and V. Ukhau, in Applied Surface Science, Vol. 248 (1–4), 264, 2005.

27 An Evaluation of Nd:YAG Laser-Cleaned Basketry in Comparison with Commonly Used Methods A. Elliott1 , A. Bezúr2 , and J. Thornton3 1

2

3

Walters Art Museum, Department of Conservation and Technical Research, 600 North Charles Street, Baltimore MD 21201-5185, USA [email protected] Department of Conservation, The Art Institute of Chicago, 111 South Michigan, Chicago IL 60603-6110, USA Art Conservation Department, Buffalo State College, 1300 Elmwood Avenue, Rockwell Hall 230, Buffalo NY 14222-1095, USA

Summary. While in storage and on exhibition, baskets can accumulate dirt that is aesthetically undesirable and even harmful. The nature of the woven structure, as well as the porosity of organic materials, causes difficulty in the removal of accumulated dirt. This chapter presents results from a study of basket-cleaning methods focusing on how Nd:YAG laser-cleaned samples are compare with those cleaned by more commonly used methods. Cleaning tests were performed on stem, bark, and root sample materials in order to examine the effects of cleaning on a variety of plant materials that are commonly encountered with basketry. Photography, optical microscopy, and scanning electron microscopy were used to document and compare the effectiveness and drawbacks of these methods. The results indicated that plant materials with protective cuticle layers can be effectively cleaned using lowtech methods and such fibers would not greatly benefit from laser cleaning. Materials without protective cuticle layers are more sensitive to mechanical cleaning and could possibly be more safely cleaned using lasers.

27.1 Introduction The problem of surface dirt on basketry has been approached from many different angles due to the difficulty of cleaning. The nature of the woven structure and the fibrous quality of plant materials complicate the cleaning process. Dirt easily becomes imbedded in the rough, uneven surfaces. A study to compare cleaning methods was undertaken to address the scarcity of published information on the cleaning of basketry [1]. This study examined several commonly used cleaning methods including a brush and vacuum, cotton swabs lightly dampened with deionized water, and groomstick [2]. Lasers were included in the study as a possible new cleaning tool because they have been

230

A. Elliott et al.

Fig. 27.1. Epidermis (based on Florian 1990: 8)

used successfully for cleaning other organic materials. In this chapter, we will present the findings of the laser component of this study in comparison with the most effective commonly used methods. It is important to remember that the irreversible nature of cleaning necessitates a careful approach to treatment, particularly with basketry materials as they often have evidence of ethnographic use. Any residue that could be associated with use should not be disturbed during the removal of post-collection dirt and grime. The ability of cleaning methods to leave residue undisturbed was not investigated. Criteria for evaluating the appropriateness of a cleaning method include (1) the effectiveness of dirt removal, (2) damage to fibers and the weave structure, and (3) the retention/deposition of residues from cleaning materials. As our study confirms, the first two criteria are related to the characteristics of the basket, including the type of fiber used, the structure of the weave, and condition of the artifact. The influence of a basket material’s morphology on the success of cleaning cannot be understated. Basketry materials usually fall into one of four categories – roots, stems, leaves, or bark. Stems, roots, and leaves all have an outer layer of epidermal cells which protect their internal structure. These layers have openings called stomata that regulate air and vapor transmission. In addition, the epidermal layers on stems and leaves produce a waxy cuticle layer that covers the structure (Fig. 27.1). The cuticle is not comprised of one homogeneous layer but several layers of differing chemical compositions [3]. The cuticle layers help to reflect ultraviolet and infrared radiation, as well as waterproof the plant. If present after processing, these protective layers can help prevent damage to the fiber during cleaning and handling. Inner bark on the other hand is found within a woody stem. Its location within the stem eliminates the need for protective layers. When the root and bark fibers are pulled apart during processing, the cells are split longitudinally, leaving a vulnerable structure with no outer protective cells.

27 An Evaluation of Nd:YAG Laser-Cleaned Basketry

231

Fig. 27.2. Stem sample

27.2 Experimental Methods 27.2.1 Sample Materials and Preparation This chapter discusses controlled irradiation with lasers, along with the three most commonly used cleaning methods: brushing and vacuuming, swabbing with cotton lightly dampened with deionized water, and swabbing with groomstick. Three different plant materials were sampled for cleaning tests with six different methods [4]. An unidentified stem material, cedar bark, and spruce root were chosen to represent the variety of plant structures encountered on artifacts. The first sample set was taken from an Italian basket constructed of an unknown stem material (Fig. 27.2). The fibers appeared to be in excellent condition, with the exception of some areas that appeared more fibrous in texture possibly indicating fiber damage. These areas were avoided during sampling. The other samples were taken from artifacts belonging to the Buffalo State College Art Conservation Department’s study collection. Both of these artifacts originate from tribes from the northwest coast of North America. One object was a woven mat constructed of the black, orange-red, and tan-colored inner bark from a cedar tree (Fig. 27.3). The other basket consisted of twined spruce root (Fig. 27.4). Both the spruce root and cedar bark artifacts showed some wear and brittleness. The types and degree of soiling varied between the three types of sample materials. The stem material had a heavy layer of gray particulate. The spruce root had a darkened appearance that appeared to be imbedded soiling, along with some light gray particulate soiling on the surface. The cedar bark had very little surface soiling, particularly for the purposes of this study. It was artificially soiled by mixing dirt collected from artifact storage areas and carbon black pigment. This mixture was dusted onto the lightly dampened surface of the sample set. Three samples measuring approximately 3.18 × 3.18 cm were used for each cleaning method for each material to allow for the observation of variations. In

232

A. Elliott et al.

Fig. 27.3. Cedar bark sample

Fig. 27.4. Spruce root sample

addition, a control sample was used for each sample set. An effort was made to choose samples with similar degrees of soiling. An additional control was used for the stem material. The two-layer construction of the basket maintained an unsoiled woven surface on the inner sides of each layer. A sample was taken from this area for comparison with the uncleaned sample. 27.2.2 Cleaning Procedures The brush and vacuum technique involved lightly brushing the samples with a fan brush while directing particulate into a vacuum nozzle. This technique was performed for 10 s on each of the three sample types. Swabs were dampened with deionized water and blotted to remove the excess. The swabs were rolled across the surface or individual elements using the swab tip to reach into crevices when needed. This technique was performed for 35 s on the cedar bark and stem samples and 25 s on the spruce root samples. A 2.54×0.48 cm section

27 An Evaluation of Nd:YAG Laser-Cleaned Basketry

233

Fig. 27.5. Stem control sample

Fig. 27.6. Water-cleaned stem

of groomstick was applied to the end of a bamboo skewer. The groomstick was rolled across the surface for 25 s on each of the three materials. The laser cleaning was performed using a Lynton Lasers Phoenix Q-switched Nd:YAG laser at the infrared wavelength of 1,064 nm. Fluence was determined by dividing pulse energy by the area of the beam (estimated using carbon paper). A pulse rate of 2 Hz was maintained for all cleaning tests. All samples were cleaned using fluences of 0.39, 0.45, and 0.56 J cm−2 . Additional testing was performed at the fluences of 0.20, 0.29, and 0.35 J cm−2 for the stem material and 0.17, 0.35, and 0.42 J cm−2 for the spruce root material. No further testing was done on the cedar bark material due to a limited supply of sample material. 27.2.3 Documentation and Analysis Samples were documented before and after cleaning using a Nikon D-100 digital camera. Optical microscopy and scanning electron microscopy were used on the control samples and on one of the three samples from each cleaning method and material.

234

A. Elliott et al.

Fig. 27.7. Laser-cleaned stem (0.56 J cm−2 )

Fig. 27.8. Laser-cleaned stem (0.35 J cm−2 )

27.3 Cleaning Results 27.3.1 Stem All four cleaning methods were visually effective in reducing soiling (Fig. 27.5). Brushing and vacuuming produced the most even cleaning because the brush was able to reach the deep interstices of the weave structure (Fig. 27.6). Waterdampened swabs were unable to achieve the same effect but revealed more of the original sheen than other methods. Groomstick removed a moderate amount of dirt. The laser-cleaned samples significantly reduced soiling over the entire surface, including the deep interstices, but left an overall dull appearance on some of the samples. Photomacrographs taken before and after treatment show only minor fiber disturbances with the commonly used cleaning methods and no change with the laser-cleaned samples. Scanning electron microscopy showed the brush and vacuum and the water-dampened swabs to be slightly abrasive to the cuticle surface, while there was little damage with the groomstick-cleaned sample. Damage was very apparent with laser cleaning at the higher fluences of 0.39, 0.45, and 0.56 J cm−2 . At the highest fluence, the cuticle was largely ablated leaving the underlying cell walls exposed (Fig. 27.7). This damage was also apparent with the 0.45 J cm−2 sample. The cuticle layer of the 0.39 J cm−2 sample appeared to have been partially

27 An Evaluation of Nd:YAG Laser-Cleaned Basketry

235

Fig. 27.9. Cedar bark control

Fig. 27.10. Brush and vacuum cleaned bark

reduced. The samples cleaned with a fluence below 0.35 J cm−2 had larger amounts of dirt remaining but also have intact cuticle layers (Fig. 27.8). 27.3.2 Cedar Bark The three low-tech methods for dirt removal were only marginally effective. Photomacrographs showed damage to the sample surfaces in the form of removed or damaged fibers (Fig. 27.9). Groomstick was clearly damaging and should not be used with this type of material. As with the stem material, brushing and vacuuming removed more of the dirt trapped between the woven elements, while the water-dampened swabs revealed more fiber sheen. Lasers were clearly seen as less damaging on this type of material. Loose and damaged fibers were left undisturbed and there was an overall reduction in surface dirt. SEM analysis of the cleaned samples revealed the brush and vacuum method to be the least damaging of the more common cleaning methods (Fig. 27.10), while water-dampened swabs were clearly the most damaging due to the removal of the delicate cell walls (Fig. 27.11). The laser-cleaned samples at the highest fluence of 0.56 J cm−2 showed some disturbance and removal of the cell walls (Fig. 27.12). Damage, if any, at the lower fluences of 0.39 and 0.45 J cm−2 was difficult to distinguish from the original condition of the

236

A. Elliott et al.

Fig. 27.11. Water-cleaned bark

Fig. 27.12. Laser-cleaned bark (0.56 J cm−2 )

fibers. It is likely that no laser-induced damage occurred at the 0.39 J cm−2 cleaned sample. The lack of extra sample material prevented further testing. 27.3.3 Spruce Root Of the low-tech methods, brushing and vacuuming was the only one to produce visually acceptable results. Although swabbing with water-dampened cotton swabs or groomstick removed some light surface dirt, the pressure produced by these methods was too great. The laser-cleaned samples were the only ones to show any significant reduction in dirt, likely due to the imbedded nature of the soiling. SEM analysis was difficult to interpret with this material due to the complex structure and apparent damage on the control sample. The results were inconclusive.

27.4 Conclusions This study suggests that the appropriate cleaning method for a basket depends on the type of fiber used for construction. Materials that have protective

27 An Evaluation of Nd:YAG Laser-Cleaned Basketry

237

cuticle layers can be cleaned with a wider variety of techniques than materials without cuticle layers. Commonly used methods such as a brushing and vacuuming and swabbing with water-dampened cotton, or a combination of both are sufficiently effective on materials with protective cuticle layers. This study established a damage threshold for lasers on materials with cuticle layers and those without. Lasers could be beneficial on materials without protective cuticle layers such as spruce root and cedar bark materials. These more fragile materials are easily damaged during cleaning with more traditional materials. The loosely bound fibers of cedar bark are easily lifted and disturbed using traditional methods. Dirt also becomes easily embedded in fibrous materials such as cedar bark and spruce root making them difficult to clean. In addition, the weave structure and condition of many baskets constructed of these types of fragile materials make the pressure of swab cleaning impractical, while lasers do not involve physical contact or pressure. In conclusion, while lasers can provide visibly effective cleaning, cellular damage that is not visible to the unaided eye could occur. Assuming that appropriate fluences are chosen, laser cleaning may be useful for cleaning problematic basket materials and structures, especially since low-tech methods appear to cause comparatively more damage. Acknowledgments We would like to thank the following people for their generous time and support of this project: Dr. Peter Bush, Tony Sigel, Pamela Hatchfield, Ruth Norton, and Claire Munzenrider. Angela Elliott would like to acknowledge funding of her graduate studies and projects by the following entities: the Leo and Karen Gutmann Foundation, Buffalo State College, the Getty Grant Program, the Andrew W. Mellon Foundation, the Samuel H. Kress Foundation, the National Endowment for the Arts, the Kenzie Art Conservation Fellowship, and the Museum of New Mexico Foundation.

References 1. Ruth Norton’s section on cleaning in The Conservation of Artifacts from Plant Materials is an extremely useful guide on the topic. M.E. Florian, D.P. Kronkright, and R.E. Norton, J. Paul Getty Trust, Los Angeles, (1990). 2. Groomstick is vulcanized cis-1,4-polyisoprene with titanium dioxide used as a filler. It is described as a molecular trap by the manufacturer and has a very tacky quality. Available from Talas, New York; http://www.talasonline.com. 3. P.J. Holloway, in The Plant Cuticle, Ed. by D.F. Cutler, K.L. Alvin, and C.E. Price, Linnean Society Symposium Series, No. 10, (1980). 4. The two less effective methods not discussed in this paper were the use of a Magic Rub polyvinyl chloride eraser in combination with vacuuming and Smoke Off sponges, a polyisoprene sponge marketed for the removal of soot. Available from Talas, New York; http://www.talasonline.com.

28 Novel Applications of the Er:YAG Laser Cleaning of Old Paintings A. Andreotti1 , P. Bracco2 , M.P. Colombini1 , A. deCruz2 , G. Lanterna2 , K. Nakahara2 , and F. Penaglia1 1

2

3

Dipartimento di Chimica e Chimica Industriale, Università di Pisa, via Risorgimento 35, 56126 Pisa, Italy Opificio delle Pietre Dure, Fortezza da Basso, Viale Filippo Strozzi 1, Firenze, Italy Duke University, Department of Chemistry, Durham, NC, USA

Summary. This chapter focuses on the use of Er:YAG laser cleaning technique for the removal of unwanted and/or degraded materials both from a large series of reference standards (overpainting, varnishes, patinas, and restoration materials) which simulate the layering of old paintings, and also examples from old paintings. A series of diagnostic controls (optical microscopy, SEM, FT-IR, GC–MS, and topographic techniques) were designed to study the effects of the laser radiation on the surface components, including morphological, optical, and chemical examination. The most significant results show that an effective thin-layer-removal of about 90% is obtained by submitting the painted surfaces to the laser exposure, while the rest of cleaning is rapidly accomplished in safety by applying mild solvents or aqueous methods. Consequently, possible interference with the original substrate can be noticeably minimized. No degradation compound induced by laser energy was formed. The laser cleaning procedure applied on an oil painting canvas “Morte di Adone” (seventeenth century), and on a panel tempera painting “San Nicola e San Giusto” of Domenico di Michelino (fifteenth century) shows that the surfaces cleaned by this system exhibit a morphology quite similar to that obtained by traditional cleaning methods.

28.1 Introduction The cleaning of painted surface is one of the most critical operations in conservation. Contemporary criteria for the cleaning procedures require a selective and progressive removal of materials, which can enable an expert conservator to preserve any external thin original glaze or even old varnishes or patinas. Moreover, for environmental needs and safety reasons it is required to abandon toxic solvents cleaning and to use alternative methods, such as aqueous methods or less-toxic solvent cleaning. Er:YAG laser ablation, in fact, meets these requirements. Actually, laser techniques have demonstrated very promising

240

A. Andreotti et al.

applications for diagnostic and conservation purposes in art conservation [1]. As reported in the literature [2], Er:YAG laser exposure at 2.94 µm on a surface dampened with a liquid containing –OH groups effectively removes old varnish and other encrustations without inducing unwanted chemical or physical changes on the original painted surfaces on canvas and wood panels. On the basis of the previous results [3,4], this chapter presents an exhaustive protocol for the use of Er:YAG laser technique in the removal of superficial layers from paintings to obtain a comprehensive knowledge on laser efficiency, fluence, threshold limits, and its practical cleaning approach combined with other ancillary methods. These parameters are tested first on the laboratory models and afterwards on old paintings. The study and control of the polychrome surfaces before and after removal of the coatings by Er:YAG laser were performed by a series of diagnostic analyses which provide fundamental information on the morphology and chemical composition of the treated surfaces.

28.2 Experimental Methods Experiments of Er:YAG laser cleaning were conducted firstly on the laboratory tempera or oil paint models with top layers varying from natural resin varnishes, oil-resin varnishes, synthetic varnishes/fixatives/adhesives, artificial patinas to overpaintings, which were previously prepared in 1999 [3]. The top layers examined by laser ablation were shellac (natural resin varnish); ketone and vinylic resin (synthetic varnish); mastic and walnut oil mixture varnish, boiled linseed oil varnish (oil based varnish); burnt umbercasein overpainting, Naples yellow-linseed oil overpainting, Naples yellowcasein overpainting, burnt umber-linseed oil overpainting on gypsum/rabbit glue (dark-colour/light-colour overpainting and overpainting on new ground); and EVA-based resin and acrylic resin (BMA and EA/MMA). In order to study the behaviour of the paint layer (tempera/oil) exposed directly to laser, uncoated surfaces were also examined. On the paint model simulating multiple layers (burnt umber-linseed oil overpainting, stucco with gypsum/rabbit glue, mastic varnish, yellow ochre/lead white in egg tempera, ground of gypsum/rabbit glue), laser ablation was tested to prove its gradual thinning action. All the laser tests have been executed under the control of stereo microscope by painting conservators in order to study the action of laser ablation and to determine threshold for each material examined. The laser ablation testing has been proceeded with application of progressively increasing energy to the surface, firstly starting with dry methods and secondly with wet methods. To study the overall effects of laser, the energy levels beyond the threshold limit were also examined. The auxiliary wetting liquids, applied with small cotton swab, were O–H containing substances (distilled water, ethanol, water– ethanol mixture (1:1, v/v), distilled water with 1% surfactant [Tween20 ]);

28 Novel Applications of the Er:YAG Laser Cleaning of Old Paintings

241

non-O–H containing liquid (light aliphatic hydrocarbons [Ligroine]); non-O– H containing liquid added with O–H containing substances [White Spirit with 15% diethylene glycol]). The range of energies used was between 3 and 200 mJ, with a laser beam diameter of 1 mm at 10 and 15 Hz macropulse frequencies. c of Mona The Er:YAG lasers used were “CrystaLase 2940” and Light Scalpel Laser Inc., USA. A previously published procedure [3, 4] was employed for the operative laser conditions and the collection of ablated materials. These materials were analysed by PY–GC–MS (Hewlett Packard, Palo Alto, CA, USA) and GC–MS (Thermo Electron Corporation, USA) methods for the analysis of synthetic polymers, amino acids, fatty acids, and terpenoids [5, 6]. OM with VIS and UV sources (Zeiss Axioplan) were used to observe the painted specimen surfaces and their modification after the laser pulses. SEM (Leica Cambridge) and EDS (Link-Oxford) were used to study the complex morphology changes of the surfaces both in flat samples and in cross sections. FTIR (Thermo-Quest) was used to compare the materials left on the surfaces and the ablated materials collect on the cover-slip glass. µ-Profilometry INOA prototype (Florence, Italy) was used to measure micrometre differences in depth after the different laser pulses.

28.3 Results and Discussion In general, the optimal energy thresholds for thin top layers (thickness ≈ 15 µm), of natural or synthetic resin, is between 8 and 13 mJ with auxiliary liquid containing O–H bond. In particular, synthetic polymers such as Plexisol (n-butyl methacrylate), or Plextol (methyl methacrylate/ethyl acrylate) were successfully removed, while BEVA (ethylene vinyl acetate) was not ablated: however, a deep surface modification was provoked by the laser exposure which allowed the cleaning by a simple swabbing technique. Oil-based varnish models resulted quite resistant to laser ablation and showed a morphology (Fig. 28.1) with an increased roughness of the surface after laser ablation. This increase, together with the introduction of chemical agents, permits to achieve a cleaned surface in a shorter time using a low concentration of chemicals. Thick oil-based overpaintings (about 30–50 µm) can be ablated at an energy level between 30 and 100 mJ with auxiliary hydroxyl liquids. Repeated exposure may be performed according to the thickness and nature of the superficial layers, without exceeding the established energy threshold. This allows a gradual removal of the overpainting without causing damage or discoloration of the substrate. Experiments on the multilayer model demonstrated a gradual and very thin laser ablation as shown in Fig. 28.2, where the energy of each passage and the wetting agent is reported. The thick overpainting (burnt umber in boiled linseed oil) was gradually thinned by five laser applications (100 mJ, 80 mJ, and 50 mJ) using distilled water/ethanol mixture (1:1) except for the last treatment, where white spirit

242

A. Andreotti et al.

Fig. 28.1. Microscope images of the sample YOET 1-1-3 (shellac varnish on lead white/yellow ochre dispersed in egg yolk tempera; gypsum/glue ground). All the images are referred to the same sample observed under optical microscopy and SEM (gold sputtering was used). The fragment was then embedded in resin and observed again in OM and SEM. All the images show the unexposed surface on the left-hand side

Fig. 28.2. The gradual action of Er:YAG laser is shown on a laboratory multilayers sample: burnt umber-linseed oil overpainting on stucco (gypsum/rabbit glue) which covers mastic varnish on yellow ochre/lead white egg tempera

added with 15% diethylene glycol was used. Then the thick stucco with gypsum and rabbit glue was thinned by five laser applications (150, 80, 50, and 20) using distilled water/ethanol mixture (1:1). The thin layer of mastic varnish as well as the egg tempera with yellow ochre/lead white underneath was well preserved. The laser cleaning was then applied on a fifteenth-century tempera painting on panel, “S. Nicola e S. Giusto”, attributed to Domenico di Michelino, from S. Giusto in Piazzanese church in Prato (Italy), under restoration at the

28 Novel Applications of the Er:YAG Laser Cleaning of Old Paintings

243

4–5 mJ, white spirit + solvent

laser

1. fatty emulsion added with artificial saliva and coccocollagene 2. fatty emulsion pH7 3. ligroin 5–6 mJ, white spirit + solvent 1. fatty emulsion added of resin Soap made with DCA-TEA 2. fatty emulsion pH7 3. ligroin 4. fatty emulsion added of citric acid 5. fatty emulsion ph7 6. ligroin

Old tested area with traditional solvent method

Fig. 28.3. Fifteenth-century tempera painting on panel, “S. Nicola e S. Giusto”, attributed to Domenico di Michelino. Cleaning tests with combined Er:YAG laser and emulsions (top and middle) compared with an old test area (bottom) cleaned with mechanical tools and traditional solvents

Opificio delle Pietre Dure (Laboratories of restoration in Florence, Italy). The laser ablation was executed on the degraded greyish thin and compact film tightly bound to the original layer (a very fragile white tempera). This layer locally contains discontinuous old brown varnish and some residues of greycoloured overpainting in the area of the mitre of Saint Nicolas (Fig. 28.3). The chemical identification of the grey patina was performed by analysing the organic material with the GC–MS analysis: the amino acids’ percentage data showed that the patina was mainly constituted by egg with small traces of animal glue. Moreover, the FTIR analysis demonstrated the presence of calcium oxalate on sample surfaces. This area was considered to be suitable for laser test, as it was extremely difficult to execute a selective and safe cleaning even by using sophisticated aqueous methods based on emulsions containing surfactants, resin soaps, or artificial saliva [7, 8]. After preliminary tests using the energy between 3 and 9 mJ at the frequency of 15 Hz, and wetting agents, white spirit or distilled water/ethanol mixture (1:1, v/v), a combined method with laser ablation and aqueous and mechanical method was considered to be appropriate. Initially the energy level of 8 mJ with white spirit as wetting agent and a clearing agent as water/ethanol mixture (1:1, v/v) after laser ablation seemed to offer a good result as far as removal is concerned. The combined method with laser at 5–6 mJ with white spirit as wetting agent and aqueous methods using an emulsion added with resin soap, followed by a

244

A. Andreotti et al.

milder pH 7 emulsion and by Ligroin one, seemed to be also rather efficient. However, the first method did not seem to guarantee a selective and gradual cleaning, judging from observation under stereoscopic microscope especially due to extreme sensibility of the original colour to aqueous solution; the second combined method either was considered inappropriate, as it removed also the antique inner varnish which was considered to be preserved. Finally, the appropriate method consists of: – A laser ablation at 5–6 mJ with wetting agent (white spirit), which permits homogeneous superficial desegregation of the grey-coloured layer – A following cleaning with a fatty emulsion [7] (Brij 35 2 g, artificial saliva (mucin 0.25 g, tribasic ammonium citrate 0.25 g, deionized water 100 ml) 20 ml, coccocollagen 2.5 ml, ligroin 80 ml) – A treatment with a clearing agent, a pH 7 emulsion (Brij 35 2 g, deionized water 10 ml, Tween 20 (nonionic surfactant) 2 ml, ligroin 90 ml) [8] – A final treatment with ligroin alone By this way, it was possible to achieve a real selective removal of the degraded materials without removing the old varnish underneath. This approach is in agreement with the latest criteria of contemporary conservation, oriented towards the minimum intervention and the reduction of toxic chemicals. The laser cleaning approach was also used for removing the insoluble and hard overpaintings on the canvas painting “Morte di Adone” (seventeenth century). Using various cleaning approaches based on polar or basic solvents, acid solution, enzymes, enzyme soaps, chelating agents, and surfactants, any efficient and safe removal of the unwanted materials was observed; nor the scalpel was successful due to the hardness of the overpaintings and their strong adhesion on the original substrate. The investigation on the grey overpainting in the area of the Venere’s shoulder by the GC–MS analysis, highlighted the presence of egg as proteinaceous material and a small amount of beeswax. The analysis of the organic material, laser sampled in different areas from a tick overpainting on the Venere’s face, showed that it was mainly constituted by animal glue (Fig. 28.4) with a small trace of a pinacae resin. Laser pulse was repeated several times on the surface, in decreasing energy levels (18–20, 15, and 10 mJ) as shown in Fig. 28.5. This procedure permits the removal of the most of the overpainting and allows a final cleaning with application of a mild solvent mixture and a soft employment of the scalpel without removing the old varnish. A noninvasive analysis, µ-profilometry, executed before, during, and after laser cleaning allows to quantify the depth of laser ablation as 3–5 µm (Fig. 28.6). This avoids the need to take samples for cross sections. The operative conditions adopted for the removal of the overpainting were 15–18 mJ energy, wetting liquid isopropanol/ligroine 50%, and isopropanol 27%, followed by a clearance with the same liquid, repeated several times. Results show a very homogeneous and superficial removal of

28 Novel Applications of the Er:YAG Laser Cleaning of Old Paintings

245

3 Database 1 Pos. 3, 2° passage, 15 mJ

2 egg

2 Pos. 3,1° passage, 19 mJ

1 3 Pos. 2,1° passage, 18 mJ 4 Pos. 1,1° passage, 19 mJ

0 −8

−6

−4 animal glue

−2

0

2

−1

−2

−3

4

5 Pos. 1,1° passage, 15 mJ 6 Pos. 2,2° passage, 17 mJ

casein

7 Pos. 3–4,° passage, 10 mJ

isopropylic alcohol + ligroine (1:1) as wetting agent

Fig. 28.4. Principal component analysis of samples from seventeenth century oil painting on canvas “Morte di Adone”: the relative amino acid percentage contents of the ablate material from different area of Venere’s face reveal the presence of animal glue

Fig. 28.5. Seventeenth-century oil painting on canvas “Morte di Adone”: four overpainting removal tests realised by combined Er:YAG laser and emulsions actions. Left: tested area; right: close details

Fig. 28.6. Laser micro-profilometry results of first and second exposure of Er:YAG laser and emulsion on area No. 1. Left: level difference between unexposed surface (orange-red) and first passage ablation level (yellow-green); centre: difference between surface and second passage level (cyan); right: relative level difference between first and second passage (5 µm on average)

246

A. Andreotti et al.

irregular top layers after each laser application. However, small changes in laser exposure (about 10 mJ) and in the composition of the solvent mixture (isopropanol/ligroine 1:1), may also give successful results as shown for the different areas in Fig. 28.5.

28.4 Conclusions The Er:YAG laser ablation offers a gradual, homogeneous, and OH-selective cleaning with its thinning action of a few micron’s depth. The general mechanism of ablation consists in modifying the surface by breaking up of a few micron’s depth, which enables the conservator to complete the cleaning using less invasive methods which are not efficient if used by themselves. The preliminary tests under microscope are indispensable to obtain the optimal threshold for each specific material, and to avoid possible surface modification or chromatic alteration of the original substrate which may not be directly observed. The macropulse frequency of 15 Hz was confirmed as the optimum to obtain a more homogeneous ablation. No chromatic alteration was observed below the energy threshold, except for certain colours such as yellow ochre, some mineral iron based blue, or Naples yellow in casein. The use of hydroxylwetting agents is appropriate, as it increases the efficiency of laser ablation as well as safety of the original substrate. It is important to note that chemical analyses show that ablation under the threshold limits provokes no significant variation in chemical composition of both ablated materials and substrate. Analyses also show that the energy levels used below the safety threshold, combined with the conservators’ skills, permit gradual and progressive cleaning. Particularly, repeated laser applications at lower fluencies in respect to the optimal energy thresholds allow the cleaning of degraded. To conclude, we can assert that the combination of Er:YAG laser, used within the energy thresholds of each material, together with the ultimate chemical and biochemical systems, allows the cleaning of a broad variety of unwanted layers. With the suggested cleaning procedure, conservators/restorers have a further and alternative chance to properly solve difficult restorations.

Acknowledgements Ente Cassa di Risparmio of Florence is gratefully acknowledged for the large contribute given for these researches. The authors are very gratefully to Raffaella Fontana, Enrico Pampaloni, and Maria Chiara Gambino (INOA, UNIFI) for the µ-profilometry measurements and interpretations, to Fabrizio Cinotti for the macroimages on old paintings, and to Annette Keller for the graphic elaborations.

28 Novel Applications of the Er:YAG Laser Cleaning of Old Paintings

247

References 1. A. deCruz, S. Hauger, and M. L. Wolbarsht, in Opt. Photon. News, Vol. 10, 36, 1999. 2. A. deCruz, S. Hauger, and M. L. Wolbarsht, in Journal of Cultural Heritage, Vol. 1, 173, 2000. 3. E. Adamkiewitz, P. Bracco, M. P. Colombini, A. De Cruz, G. Lanterna, M. Matteini, K. Nakahara, O. Sartiani, and M. L. Wolbarsht, in Journal of Cultural Heritage, Vol. 4, 202, 2003. 4. A. Andreotti. M. P. Colombini, G. Lanterna, and M. Rizzi, in Journal of Cultural Heritage, Vol. 4, 355, 2003. 5. I. Bonaduce and M. P. Colombini, in Journal of Chromatography, Vol. 1028, 297, 2004. 6. I. Bonaduce and M. P. Colombini, in Rapid Communication in Mass Spectrometry, Vol. 17, 2523, 2003. 7. P. Cremonesi, L’uso degli enzimi nella pulitura di opera policrome, Padova 1999. 8. P. Cremonesi, L’uso di tensioattivi e chelante nella pulitura di opere policrome, Padova 2001.

29 A Final Report on the Oxidation and Composition Gradients of Aged Painting Varnishes Studied with Pulsed UV Laser Ablation C. Theodorakopoulos1 , V. Zafiropulos2 , and J.J. Boon3 1

2

3

University of London, School of Biological and Chemical Sciences, Birkbeck College, Malet Street, London WC1E 7HX, United Kingdom [email protected] Laboratory of Applied Physics, Department of Human Nutrition & Dietetics, Technological Educational Institute of Crete, I. Kondylaki 46, 723 00 Sitia, Crete, Greece FOM Institute for Atomic and Molecular Physics, (AMOLF), Kruislaan 407, Amsterdam, The Netherlands

Summary. This paper discusses findings that establish the ageing-induced compositional and crosslinking gradients across the depth-profiles of two accelerated aged natural resin varnishes: dammar and mastic, which are commonly applied to paintings. Profile measurements of laser-processed films using a KrF excimer laser, as well as online measurements of the C2 emission by laser-induced breakdown spectroscopy (LIBS), showed a significant reduction of the ablation step and ablation yield with depth, respectively. Direct temperature mass spectrometry (DTMS) showed that the oxidation products formed upon ageing were gradually eliminated across depth, which affected the depth-wise optical properties of the films studied by UV/VIS spectrophotometry. The total ion currents of the DTMS in the electron ionisation mode (EI, 16 eV) demonstrated also a gradual reduction of the pyrolysis yield which corresponded to a gradual depth-wise elimination of the high MW fractions that was confirmed by size exclusion chromatography (SEC). The gradients were established also by surface analyses, such as matrix-assisted laser desorption/ionisation-time-offlight-MS (MALDI-TOF-MS) and attenuated total reflection-Fourier transformed infrared spectroscopy (ATR-FTIR), which indicated that the action of the KrF excimer laser is non-destructive to the varnish when optimal fluences are used for the laser cleaning process. The extracted data enabled the quantification of the compositional gradients and unraveled a significant feature of natural resins, which would not have been possible without the use of KrF excimer lasers.

29.1 Introduction The laser cleaning of old master paintings and other works of art coated with aged and deteriorated varnishes has been proposed since the early 1990s [1–3].

250

C. Theodorakopoulos et al.

The proposed method is based on excimer laser ablation [4–6] and provides superior selectivity owing to a stepwise varnish removal on the micron scale. Since these advances came to light, the main objective has always been the ‘safe’ removal of aged varnish coatings from photosensitive paint substrates [7]. Technically this is difficult, as UV laser irradiation results in paint discoloration once the laser photons reach the paint [8, 9]. Recently, it was shown that whilst the paint medium (oil or egg binding) is not visually affected, discoloration occurs in the presence of inorganic pigments only [10]. Regardless of the minor scale that discoloration occurs, that is a few nm deep in the surface of the exposed pigment particles [11], the damage caused on the surface is irreversible. So far, two ways have been proposed to prevent paint discoloration (a) the use of online control of the laser ablation process by employing techniques that either monitor the plume emission or the treated surface [12–15] and (b) the retention of a thin film of varnish on the surface to prevent irradiation of the underlying, photosensitive paint layers [7, 13]. Online control based on analysis has had remarkable results that assisted the improvement of the laser cleaning methods, but in practice it is not always possible to detect damage before its occurrence. We found that several types of analysis can be used at several stages of the cleaning procedure to assess the progress of the treatment and to validate the integrity of the treated surface. Regardless of the use of any complementary analytical assistance, the retention of a thin film of the varnish has been proven to be essential and very effective [16], provided that the thickness of the remaining varnish is larger than the optical absorption length of the material at the laser wavelength. Therefore, since the excimer laser cleaning of a coated surface must be terminated before the total ablation of the varnish, the most significant factor for the safety of the underlying surface is the chemistry of the varnish. An exceptionally controlled removal of aged varnish with the minimum possible transmission towards the paint is guaranteed once the ‘optimum’ fluence that maximizes the ablation yield per laser photon is being used throughout the cleaning procedure [7]. However, the depth-wise chemical change of aged natural varnishes results (a) in the change of the laser ablation rate [17, p. 164] and (b) in the transition of the ‘optimal’ fluences to higher energy densities as the cleaning proceeds deeper within the varnish [17, p. 174]. This explains the change of the interaction of excimer laser pulses with successive depths of the same varnish films using the same fluence [18]. Thus, the optical, oxidation, crosslinking and compositional gradients, which have recently been established across the depth profiles of aged natural resin films [17], play a dominant role in the laser cleaning of natural varnishes. Herein, we summarize the highlights of a comprehensive investigation of the compositional gradients of accelerated aged dammar and mastic resin films, which would not have been possible without the aid of KrF excimer laser ablation.

29 A Final Report on the Oxidation and Composition Gradients

251

29.2 Experimental Methodology Mastic and dammar in xylene 40% (w/v) were spin coated to produce 55-µm thick films. After drying, ageing was accelerated in a Sunset CPS, Heraeus R xenon-arc fadeometer. Wavelengths longer than 295 nm were Instruments employed to imitate sunlight exposure [19]. The films received a light dosage of 160 Mlux h, then exposed to open air for 45 days, followed by 30 days storage in the dark. R , COMPex series, KrF excimer laser (λ = 248 nm) was A Lambda Physik used with a pulse duration of 25 ns and source energy of 380 mJ. Standard laser ablation rate studies [7] were performed on both films. The etching depths R S5P. Laser cleaning was based upon were determined with a Perthometer scanning adjacent areas of the films with a pre-calculated number of laser pulses across the Gaussian profile of the laser beam, as described in [7, 16] and [17, p. 159]. Zones of cumulative ∼3 µm steps were carried out as in [18]. Laser-induced breakdown spectroscopy (LIBS) [20] was carried out to determine the carbon dimer emission, which is associated with the degree of crosslinking [7] and condensation [17, p. 87] as a function of depth. Online transmission studies [7] were carried out to quantify the amount of laser light transmitted in the rear of the films during the ablation process. Mass spectrometric, chromatographic, and spectroscopic analyses were separated into two groups to determine the depth-wise properties of the aged varnishes (a) those which determined the chemical state of the ablated surfaces, such as attenuated total reflection-Fourier transformed infrared spectroscopy (ATR-FTIR) and a specifically planned methodology for matrixassisted laser desorption/ionisation-time-of-flight-MS (MALDI-TOF-MS) [17, p. 280] and (b) those which determined the mean chemical properties of the remaining films across depth, such as UV/VIS spectrophotometry, direct temperature-resolved mass spectrometry (DTMS) [21] – both the summation DTMS mass spectra and the total ion currents (TICs) in EI (16 eV) and NH3 /CI (250 eV) ionisation modes were studied – the multivariant factor discriminant analysis (MFDA) [22, 23] of the DTMS data, as well as high performance-size exclusion chromatography (HP-SEC).

29.3 Results and Discussion A first indication of the existence of the gradients across the films was the change of the ablation rate against fluence with depth [17, p. 164]. Using constantly ‘optimal’ fluences, which were determined on the surface of the films, it was observed that the ablation step was minimized particularly after the removal of the 15 µm surface layers in both films [18]. Oxidation gradients were illustrated by the DTMS both in the summation MS and the TICs [17, p. 268, 18]. A good representation of the oxidation gradients detected by DTMS was demonstrated by MFDA. Figure 29.1 shows the molecular modifications across the MFDA-discriminant function (DF1) vs. the depth of the accelerated aged

C. Theodorakopoulos et al. DTMS discrimination function 1

252

1.2

Depth steps

1.0 0.8 Thresholds of highly aged and less deteriorated material

0.6 0.4 0.2 0.0

fresh dammar: independent signal from film thickness

−0.2 −0.4 0

5

10 15 20 25 30 35 Depth from surface (mm)

40

45

Fig. 29.1. MS projections of MFDA-DF1 in the 15-µm thick accelerated aged dammar film. The corresponding DF1 vs. depth plot of mastic has a similar trend [17, p. 279]

Intensity (a.u.)

Dammar

0

10 20 30 40 Mean depth from surface (mm)

50

Fig. 29.2. Intensity of the C2 emission in the laser plume at ∼ 515 nm (filled circle) and ∼ 470 nm (circle) vs. the mean depth for an accelerated aged dammar varnish observed by LIBS (KrF excimer laser at 0.9 J cm−2 ). A few pulses were required to optimise the laser-coating interaction as described in [6]. The maximum intensity at the 10–15 µm uppermost layers corresponds to a high degree of crosslinking (and condensation) at the surface [7], which eventually decreases with depth. Mastic showed a similar trend [17, p. 180]

dammar film. The corresponding MFDA-DF1 plot of mastic had a similar trend. The oxidation gradients have also been detected by micro-FTIR [7] and ATR-FTIR [17, p. 228] both showing a gradual reduction of the carbonyl absorption (∼ 1, 710 cm−1 ) with depth in accelerated aged natural resin films. MALDI-TOF-MS indicated that the oxidative change with depth is not only quantitative but also qualitative, since the uppermost 10–15 µm layers had

29 A Final Report on the Oxidation and Composition Gradients

% DTMS-TIC Pyrolysis

no of pulses crosslinking

90

50 40

80 30 70 20 60 10 50

0 0

5

10 15 20 Depth from surface (mm)

No of KrF Excimer Laser Pulses

60

100

253

25

Fig. 29.3. The leap of the ablation step at ∼ 15 µm from the surface of the aged dammar film using successive 248-nm laser pulses [18] corresponds to the evident jump of the intensity of pyrolysis during the DTMS investigation of the successive depth-steps of dammar. The latter is a marker for the reduction of the MW [25] with depth. The same observation has been made for the accelerated mastic film [17, p. 246]

oxidation secondary products produced by exposure to the ambient UV wavelengths, whereas below the highly oxidised surface there was a zone in the bulk down to 25–30 µm oxidized by the absorption of visible light [17, p. 280]. Both films were almost unaffected from ageing at depths greater than ∼ 25–30 µm from surface [17, p. 273, 18]. Analysis of the ablation plume with LIBS showed that the carbon dimer emission observed by the C2 Swan band system, the intensity of which has been associated with the degree of crosslinking [7], decreases abruptly after the removal of the 15 µm surface layers (Fig. 29.2) [17, p. 177]. Crosslinking gradients of aged dammar and mastic films have been verified by (a) the depth-wise reduction of the methylene-to-methyl ratios monitored by microFTIR [24] and ATR-FTIR [17, p. 228], (b) the reduction of the intensity of pyrolysis during the DTMS runs of the remaining films [17, p. 264, 18] and (c) the gradual decrease of the MW determined by HP-SEC [17, p. 284]. These findings show that the highly oxidative fraction of the aged varnishes, both in terms of quality and quantity, as well as the high MW crosslinks are generated at the uppermost zone of the films. During the laser cleaning, the ablation step per laser pulse had a linear response to the laser-induced etching depth (for a certain fluence), as long as the process was limited at the highly degraded surface layer [18]. Once this layer was removed the ablation step changed, as shown in Fig. 29.3. All data reflect theories on the oxygen ‘starvation’ with depth [26, 27], verify Beer’s Law and indicate a faster reduction of the UV compared with the visible wavelengths as the ambient light is transmitted across the depthprofile of the films. Indeed, UV/VIS spectrophotometric measurements on

C. Theodorakopoulos et al. Optical Absorption Length (mm)

254

50 40 30

Mastic Dammar Threshold of highly aged, UV oxidised material

20 10

non UV oxidized less degraded material

0 250 275 300 325 350 375 400 425 450 Wavelength (nm)

Fig. 29.4. Optical absorption lengths against wavelength of ambient light (250– 450 nm) for dammar and mastic varnish films. The corresponding optical absorption lengths of the 248-nm laser are one order of magnitude shorter [7]. Wavelengths shorter than 325–350 nm were completely absorbed by the 10–15 µm surface layers of the films [17, p. 227]

the successive laser-ablated zones of the films showed that the ambient radiation up to 350 nm is completely absorbed by the 15 µm dammar and mastic varnishes [17, p. 227] (Fig. 29.4). At the beginning of these studies it was shown that yellowing, that is the most readily monitored side effect of resin ageing, was gradually reducing across depth of aged varnishes uncovered by means of nanosecond KrF excimer laser ablation [16]. It could have been suggested that this observation was due to laser-induced bleaching of the yellow chromophores, which are unsaturated quinones that absorb in the blue (400–490 nm) [28]. The presented findings, however, demonstrated that these chromophores are not generated in the bulk of the aged resin films showing that, under controlled circumstances, the possibility of photochemical damage in the photoablated films is negligible. According to the UV/VIS data and considering a ‘fresh’ natural varnish suddenly exposed to sunlight (λ > 250 nm), it could be suggested that autooxidative processes, i.e. free radical formation, oxygen uptake and crosslinking, which contribute to the eventual chemical and physical degradation of such a film are eliminating with depth. This was effectively reflected in the 248-nm ablation of the films. During the laser irradiation, the 248-nm photons are first absorbed causing (a) electronic excitations across the optical absorption length of the film, which in turn causes (b) shock waves, that is photochemical and photomechanical action in the irradiated area, leading to (c) bondbreakage and desorption of excited species into the gas phase [6, 29]. Because of the gradient in absorption and the distribution of different amounts and types of chromophores across depth [17, p. 258], the mechanisms that initiate the ablation of natural resin films are gradually altered as the thickness of the films is eventually reduced. Shock waves are dependent on the number of cova-

29 A Final Report on the Oxidation and Composition Gradients

255

lent bonds that are involved in the preliminary stages of the bond-breakage process [30]. Since both the degrees of absorption, which is indicative of the degree of oxidation, and crosslinking decrease with depth, the intensity of the photochemical and photomechanical actions following the release of the laser pulse also decreases with depth, as is indicated by the change of the ablation step vs. depth (Fig. 29.3).

29.4 Conclusions This paper establishes the ageing-induced compositional and crosslinking gradients formed across the depth-profiles of two natural resin varnishes: dammar and mastic, which was possible with the aid of nanosecond KrF excimer laser ablation. The significant depth-wise reduction of the degree of oxidation corresponded to a decreasing absorption, which in addition to the reduction of crosslinking with depth, alters the ablation rate across depth of such films. The determination of the gradients, especially with surface analysis, showed that laser cleaning with excimer lasers and optimal fluences does not deteriorate further aged coatings made of natural resins. Upon ageing, the 10–15 µm surface layers of the dammar and mastic varnishes were utterly deteriorated absorbing the UV wavelengths of ambient light. The free transmission of the blue light (λ > 400 nm) after having removed the top 25–30 µm layers of the aged films indicates that, upon ageing, yellow chromophores are not produced in the bulk of such films. Thus, chemical deterioration leading to visual degradation of natural resin films is a surface phenomenon. Hence, cleaning of aged (natural) varnish-paint systems can be accomplished by removing surface layers only. Acknowledgements This work would not have been possible without the EU Large Installations Plan DGXII (HCM) program ERBCHGECT920007 at the Ultraviolet Laser Facility at IESL/FORTH, the FOM Program 49 funded by FOM and NWO and the Foundation of the State Scholarships, Greece (IKY), which funded the PhD research of C. Theodorakopoulos. Professor Dr. Wolfgang Kautek, University of Vienna, Austria, and Dr. Klaas Jan van den Berg, ICN, The Netherlands, are acknowledged for their participation in the PhD examination board.

References 1. E.I. Hontzopoulos, C. Fotakis, and M. Doulgeridis, in SPIE: 9th International Symposium on Gas Flow and Chemical Lasers Vol. 1810, Bellingham, Washington, 748, 1993. 2. N. Morgan, Opto and Laser Europe 7, 36 (1993). 3. C. Fotakis, Optics and Photonics News 6, 30 (1995).

256

C. Theodorakopoulos et al.

4. B. Luk’yanchuk, N. Bityurin, S. Anisimov, and D. Bäuerle, in Excimer Lasers, Ed. by L.D. Laude. Kluwer Academic Publishers, The Netherlands. 59, 1994. 5. R. Srinivasan, in Laser Ablation, Ed. by J.C. Miller, Springer Series of Material Science Vol. 28, Springer, Berlin Heidelberg, 107, 1994. 6. D. Bäuerle, in Laser Processing and Chemistry, Third, rev. and enl. edn., Springer, Berlin Heidelberg New York, 2000. 7. V. Zafiropulos, in Laser Cleaning, Ed. by B. Luk’yanchuk, World Scientific, Singapore, 343, 2002. 8. A. Athanassiou, A.E. Hill, T. Fourrier, L. Burgio, and R.J.H. Clark, J. Cult. Herit 1. s209 (2000). 9. V. Zafiropulos, T. Stratoudaki, A. Manousaki, K. Melesanaki, and G. Orial, Surf. Eng. 17, 249 (2001). 10. M. Castillejo, M. Martin, M. Oujja, D. Silva, R. Torres, A. Manousaki, V. Zafiropulos, O.F. Van den Brink, R.M.A. Heeren, R. Teule, A. Silva, and H. Gouveia, Anal. Chem 74, 4662 (2002). 11. B. Luk’yanchuk, and V. Zafiropulos, in Laser Cleaning, Ed. by B. Luk’yanchuk, World Scientific, Singapore, 393, 2002. 12. I. Zergioti, A. Petrakis, V. Zafiropulos, and C. Fotakis, in Proceedings of LACONA I; Restauratotenblätter Sonderband, Vienna, 57, 1997. 13. V. Zafiropulos, and C. Fotakis, in Laser in Conservation: an Introduction, Ed. by M. Cooper, Butterworth Heineman, Oxford, 79, 1998. 14. J.H. Scholten, J.M. Teule, V. Zafiropulos and R.M.A. Heeren, J. Cult. Herit. 1, 215 (2000). 15. R. Teule, H. Scholten, O.F. van den Brink, R.M.A. Heeren, V. Zafiropulos, R. Hesterman, M. Castillejo, M. Martin, U. Ullenius, I. Larsson, F.A. GuerraLibrero, H. Gouveia, and M.B. Albuquerque, J. Cult. Herit. 4, 209 (2003). 16. C. Theodorakopoulos and V. Zafiropulos, J. Cult. Herit. 4, 216 (2003). 17. C. Theodorakopoulos, PhD Thesis, RCA, London, with FORTH-IESL, Heraklion, and FOM-AMOLF, Amsterdam, 2005; URL: http://www.amolf.nl/publications/theses/theodorakopoulos/theodorakopoulos.html; 18. C. Theodorakopoulos, V. Zafiropulos, C. Fotakis, J.J. Boon, J. van der Horst, K. Dickmann, and D. Knapp, in Springer Proceedings in Physics, Vol. 100, Edited by K. Dickmann, C. Fotakis, and J.F. Asmus. Springer, Berlin Heidelberg, 255, 2005. 19. R.L. Feller, in Accelerated Aging: Photochemical and Thermal Aspects, Ed. by D. Berland. The J. Paul Getty Trust, L.A. 45, 1994. 20. D. Anglos, Appl. Spectrosc. 55, 186A (2001). 21. J.J. Boon, Int. J. Mass Spectrom., Ion Process 118/119, 755 (1992). 22. R. Hoogerbrugge, S.J. Willig, and P.G. Kistemaker, Anal. Chem. 55, 1710 (1983). 23. W. Windig, J. Haverkamp, and P.G. Kistemaker, Anal. Chem. 55, 81 (1983). 24. V. Zafiropulos, A. Manousaki, A. Kaminari, and S. Boyatzis, in ROMOPTO: Sixth Conference on Optics, Edited V.I. Vlad, SPIE Vol. 4430, 181, 2001. 25. G.A. van der Doelen, PhD Thesis University of Amsterdam. 28, 1999. 26. A.V. Cunliffe and A. Davis, Polym. Degrad. Stabil. 4, 17 (1982). 27. T. Fukushima, Durability Build. Mat. 1, 327 (1983). 28. M.W. Formo, in Baile’s Industrial Oil and Fat Products Vol. 1, Ed. by D. Swern, John Wiley & Sons, New York, 722, 1979. 29. B. Luk’yanchuk, N. Bityurin, S. Anisimov, and D. Bäuelre, Appl. Phys. A 57, 367 (1993). 30. R. Srinivasan and B. Braren, Chem. Rev. 89, 1303 (1989).

30 A New Solution for the Painting Artwork Rear Cleaning and Restoration: The Laser Cleaning ∗

S.E. Andriani1 , I.M. Catalano1 , A. Brunetto3 , G. Daurelio2 , and F. Vona4 1

∗ 2

3

4

InterAteneo Physics Department “M. Merlin” of the University of Bari – InterDepartment Center “Search Laboratory for the Cultural Heritage Diagnostic” – Via Amendola 173, 70126, Bari, Italy [email protected] C.N.R.-I. N. F. M. – National Institute for the Physics of Matter – Regional Lab. L.I.T. Laser Innovation Technology Transfer and Training – c/o Physics Department “M. Merlin” of the University and Polytechnic of Bari, Via Amendola 173, 70126 Bari, Italy Private Restorer & Researcher in Art Restoration of “Restauri Brunetto di Brunetto Anna” firm, via Tormeno, 63, 36100 Vicenza, Italy Soprintendenza per il patrimonio Storico, Artistico e Etnoantropologici delle provincie di Bari e Foggia, Italy

Summary. Before restoring a painting, in order to assure a good level of adhesion between the canvas and the preparation layer or to reline the painting, it is often necessary to consolidate the canvas by intervening on the painting rear. Traditional cleaning techniques, chemical combined with mechanical ones, show an important drawback: The cleaning process and technique enfeeble permanently the canvas. The present work performs a comparative study for evaluating both the cleaning process efficiency and the canvas integrity preservation by using various cleaning methods, including Nd:YAG laser systems and traditional techniques. The effects of a short free running mode (λ = 1,064 nm, pulse duration of 40–110 µs), a long Q-switched mode (λ = 1,064 nm, pulse duration of 200 ns) and a Q-switched mode (λ = 1,064 and 532 nm, pulse duration of 6 ns) of Nd:YAG laser irradiation on the hemp canvas of a seventeenth century painting are investigated. The analyses using FTIR spectroscopy and degradation mapping by optical microscope, with photographs taken before, during and after the cleaning process, were carried out. The work is still in progress.

30.1 Introduction Several physical techniques such as microanalytic methods, optical techniques and laser applications can be employed successfully for diagnosis and maintenance of cultural heritage. During the last few years, many laboratories have tested lasers to solve some cleaning problems. The excellent results of laser cleaning to remove unpleasant and harmful pollution layers from

258

S.E. Andriani et al.

many substrates, mainly limestone and marble, stimulated the researchers to try this method on various kinds of supports with different natural or synthetic pigmented surfaces. Many experimental papers have demonstrated the mechanisms operating during laser cleaning and they give a reasonable understanding of the interaction of laser radiation with the complex structure of stone artworks [1, 2]. However, with a polychrome masterpiece, a new variable is introduced that should not be of underestimated and that is the presence of pigments and their related binders [3, 4]. The laser application on painted artworks is more demanding due to the high sensitivity of paint layers (different pigments and binding media) to light. Amongst the many interesting challenges to be tackled is the discoloration of pigments upon laser irradiation [4,5]. Some laser application difficulties are met with restoration of the back surface of a painting support, where often there is an inhomogeneous polychrome layer. In fact, not only the painting needs restoring but also the support (the painting rear). Alteration layers, past consolidation treatments, adhesives applied on the rear surface, spillage of pigments and binders on the back all damage the rear surface of a painting as well as soil the front painting. In order to assure a good level of adhesion between the canvas and the preparation layer or to reline the painting, it is often necessary to consolidate the canvas by intervening on the back of a painting. Traditional cleaning techniques, chemical combined with mechanical ones, show an important drawback: the cleaning process and technique enfeeble irremediably the canvas. Here various Nd:YAG lasers with different pulse duration: Q-switch (τ = 6 ns), long Q-switch (τ = 200 ns), short free running (τ = 40 µs), and different wavelengths, fundamental and second harmonic (λ = 1,064 nm and λ = 532 nm) have been tested. Optical microscopy (OM) and stereomicroscopy with an image storage system (Nikon LUCIA) and Fouriertransform infra-red spectroscopy (FTIR) have been used to characterize and evaluate the composition of the untreated and the laser-irradiated canvas samples, as well as to assess the effects induced by the laser radiation. The aims of this innovative research, still in progress, are: – The evaluation of the true efficiency of the different laser sources employed in cleaning operations compared with traditional ones – The understanding of the principles, the threshold parameters and effectiveness of the laser cleaning method – The research of the optimum working conditions for the materials to be removed, through the establishment of laser parameters (short free running, long Q-switch or Q-switch, wavelength, repetition rate and energy density) without causing damage to the canvas or preparation layer underneath the painting

30 A New Solution for the Painting Artwork

259

30.2 Experimental Methods The rear surface canvas specimens used in the present work derive from the “centre” (150 × 50 cm2 ) of the “S. Sebastiano curato dalle Pie donne” painting. This oil painting on canvas comes from the S. Chiara church, Bari (Puglia-South of Italy) and it dates back to the sixteenth century. The centre was probably an addition of the seventeenth century during the church restoration, in order to increase the painting size. The centre edges, creased and lacerated, showed holes brought by nails that had probably served for fixing the painting on a loom. The back of the canvas sample (Fig. 30.1) shows some cuts, holes and tears. It is covered by a “mash” substance (“beverone”), brown coloured, composed of a mixture of desiccative oils and resins that was brushed on the back of the painting. When this mixture has been applied, it restored the paintings to its original brightness but, subsequently after some time, the same beverone has blackened and defaced the painting. Pigment traces (black and red coloured) and preparation layer (red and brown coloured) that have penetrated crossed the canvas weave of the painting are evident. The preparation and the pigments leakage “spillage” were so thick that they had been leveled, closing the “reading” of the weave in the past (Fig. 30.2).

Fig. 30.1. Centina sample

Fig. 30.2. Beverone and traces of the preparation layer and pigment leakage

260

S.E. Andriani et al.

Paint layer Preparation layer Canvas Mixture: Beverone+Pigment and preparation leakage Fig. 30.3. Typical painting stratigraphy

A schematic cross-section representation of a typical painting is shown in Fig. 30.3, in which a brown oily, resinous layer (beverone) is reported. The usual cleaning process of the rear surface of the canvas is mechanical/chemical and manual, made with the use of emery paper and/or scalpel and, where the alterations are more tenacious, at first dampened with H2 O and immediately after “scraped off” with a scalpel or then dampened with chemical solvents, then scraped with scalpel and buffered with alcohol. 30.2.1 Laser Systems and Experimental Procedures The irradiation experiments were carried out with different laser systems. Three portable Nd:YAG laser sources called SMART CLEAN, VARIO (produced by EL.EN. Group, Florence, Italy) and PALLADIO (produced by Quanta System-EL.EN. Group, Florence, Italy) were employed for the sample laser cleaning operations. The first produces up to 1 J per pulse in a 40-µs pulse duration (SFR), at single shot or 5–10 or 20 Hz repetition rate. The second laser (VARIO) emits up to 350 mJ per pulse in 200 ns (LQS) pulse duration at single shot or 5–10 Hz repetition rate. Both lasers, using only the fundamental wavelength (λ = 1,064 nm), are equipped with an optical fibre delivery system of different lengths and use a 1-mm core guiding fibreglass. The fibres terminate with an optical manipulator which permits adjustment of the spot size and fluence (J cm−2 ) onto the irradiated surface. Due to the beam delivery via an optical fibre, the spatial energy profile has a top-hat shape and poses a lower risk for hot spots, thermal effects and discoloration on the surface. The PALLADIO laser system is a Q-switched source that produces up to 450 mJ per pulse in 6 ns pulse duration (QS), at single shot or 5–10 or 20 Hz repetition rate. It offers two choices for the wavelength (1,064 and 532 nm). The laser system delivers the beam through a seven mirror, multi-articulated arm with a handpiece equipped with a focusing lens in order to adjust the spot size and homogeneity. The applied energy densities were calculated from the laser energy measured with a calorimeter (power meter) and the size of the circular laser spot monitored with a thermal sensitive paper. Optical microscopy and stereomicroscopy (OM) with a storage image system (LUCIA) and Fourier-transform infra-red spectroscopy (FTIR) have

30 A New Solution for the Painting Artwork

261

been used in order to characterize the surface to evaluate the composition of the untreated and laser-irradiated canvas samples, as well as to assess the effects induced by the laser radiation. OM (optical microscopy, Nikon ECLIPSE E80i/RT and stereoscopic NIKON Model SMZ-800) was performed on the canvas sample before and after the laser irradiation. FTIR (Perkin Elmer 1600) was employed to determine the inorganic and organic compounds present in the beverone and to evaluate the effects of laser–beverone interaction. For the FTIR analysis, 1 mg of sample was homogenized with 0.2 mg of KBr, and the disks formed were examined in an FTIR spectrophotometer in a transmittance range between 4,000 and 400 cm−1 . 30.2.2 Samples Twelve different adjacent zones (4 × 10 cm2 , Fig. 30.4) were selected on rear painting in order to compare chemical, mechanical and laser cleaning efficiency as well as drawbacks. FTIR analyses have revealed probably the composition of the beverone: linseed oil and/or animal glue (ν(CH2 ) 2,926 cm−1 , 2,853 cm−1 ), pigment in media and soiling. FTIR analysis was not able to identify the composition of the layer; perhaps the preparation of powder sampling is not the suitable analytical technique in this case. However, the difficulty of sampling a homogenous section of sample within even an extremely heterogeneous mash of pigment, media, glue and oil must not be ignored. The basic condition for laser cleaning to be effective without underlying substrate destruction is that the soiling ablation threshold is significantly

Fig. 30.4. Painting rear details with different test sections: Zones (1, 2, 3, 4, 5, 6, D, C) are laser cleaning areas; Zones (A, B, E and H) employed mechanical–chemical cleaning methods

262

S.E. Andriani et al.

lower in comparison with that of substrate. However, the complexity of the canvas rear alteration layers and the sequence of different alteration layers as shown, step-by-step, with OM observations, represent a serious and difficult problem to the application of the laser cleaning process. In order to prevent the direct interaction of the laser beam with light-sensitive paint materials (including pigments and binders) and the risk of excessive material removal (superficial, or even structural, damage), different laser sources and exposure parameters (wavelength, energy, pulse duration, spot size, dry/wet cleaning) were tested. Samples: Sections A, B2, H, E The sections A, B2 , H, E (Fig. 30.5) of the hemp canvas were cleaned with traditional cleaning techniques (chemical and mechanical ones) such as emery paper, scalpel and water, scalpel and solvent (1:2 Amile:metilformalamide) and scalpel and emery paper. The use of a scalpel or an emery paper gives a complete superficial cleaning such that all alteration layers were removed even if they generate some canvas mechanical damage. In fact, as shown in Fig. 30.5 (Zone E), the scalpel/emery paper erosion mechanism has not discriminated beverone from canvas so it has completely removed the layers but, at the same time, it irremediably damaged the weave and has not cleaned the weave interstices. The scalpel has levelled the surface, making it unsuitable for the painting relining. Even if scalpel and water are applied, in order to take advantage of the water solvent action, the result is almost the same (Fig. 30.5, zone B2 ). Chemical treatments rely on chemical reactions with the alteration layers; the solvent action gives a more cleaning easiness. In fact, after the chemical reaction, the use of a “soft” abrasive mechanism (scalpel) preserves the canvas weave better as shown by the comparison in Fig. 30.5, zone A–H. Anyway, after a chemical treatment, it is recommended to buffer with alcohol in order to interrupt the chemical solvent penetration and relative chemical reaction, preserving the painting layer on the front of the painting. So apply-

B2

A-H

E

Fig. 30.5. Identification of zones. B2: scalpel + water; A: emery paper; H: scalpel + chemical solvent; E: (Left) emery paper; (Right) scalpel

30 A New Solution for the Painting Artwork

263

ing this methodology, it is impossible to get a real control of the chemical action into the painting. Samples: After Laser Treatments, Sections 1–6 Some different tests, executed by laser, are described next. No examples using the SFR laser are in the following descriptions because this laser showed a high thermal effect that blackened the canvas irremediably. In Table 30.1, the different zones with the laser systems employed are correlated. The absorption of the laser radiation and the induced effects depend on the complex stratigraphic composition and inhomogeneity of chromatic characteristics of the beverone. As a result of this complexity, for each individual case studied, some results for the surfaces as well as the laser operating parameters and various experimental conditions are reported on the following. Sections 1 and 2 The working laser parameters are: Vario (τL = 200 ns, φ = 3 mm, ν = 10 Hz, E = 300 mJ, F = 4.2 J cm−2 ) in zones 1a and 2a. The cleaning of the beverone is not complete, in fact, Fig. 30.6a and, in detail, Fig. 30.6b, shows the weave canvas still obstructs so this laser was inefficient on this thick beverone layer. In zone 1b, after the first unsatisfactory Vario laser cleaning result with E = 250 mJ, the remaining area was test cleaned with another laser system: Palladio (λ2 = 532 nm, τL = 6 ns, φ = 3 mm, ν = 10 Hz, E = 200 mJ, F = 2.8 J cm−2 ). The results show a complete surface and interstices cleaning. The following OM observation, Fig. 30.6c, reveals the effects of QS λ2 laser cleaning: a canvas fibre abrasion. Section 3 The working laser parameters are Palladio (λ1 = 1,064 nm, τL = 6 ns, φ = 4 mm, ν = 10 Hz, E = 110 mJ, F = 0.8 J cm−2 ). Figure 30.7 shows the complete laser cleaning of canvas and the absence of interstices residues. This Table 30.1. Different canvas sections corresponding to laser system used: Short free running (SFR), long Q-switch (LQS), Q-switch with λ = 1,064 nm (QS λ1), Q-switch with λ = 532 nm (QS λ2) and xn , where n indicates the number of laser treatments laser/section

D a

SFR LQS QS λ1 QS λ2

1 b

a

b

2 a

x

x1

x

2

x

x x1

3 a

4 b

5 wet

x 2

x

x

x

x

6

a

b

x

x

a

b

x

x1 x2

264

S.E. Andriani et al.

a

b

c

Fig. 30.6. Sections 1 and 2: (a) zones 1 and 2; (b) zone 1a, after (on the left) and before (on the right) laser treatments; (c) zone 1b, particulars of some canvas fibres. Magnification, ×10

a

b

Fig. 30.7. Section 3: (a) canvas during the cleaning; (b) microimage (×10) on some canvas weaves, completely cleaned

section represents the best laser exposure parameters for the painting rear cleaning. Section 4 The three zones of Section 4 are irradiated with the same laser system: Palladio (λ2 = 532 nm, τL = 6 ns, φ = 4 mm, ν = 10 Hz) but with different laser energy (fluence) values in order to resolve some problems that appeared after cleaning. In fact, in zone 4a, after laser cleaning with E = 120 mJ(F = 1 J cm−2 ) the canvas was completely clean (Fig. 30.8) but a discoloration appears. It seems that the laser, by the ablation process, is responsible of the greenish colour effect (Fig. 30.8I). Other laser tests showed that the greenish colour effect occurs only after encrustation/soiling removal and not after the tests laser irradiation of canvas without beverone. A possible explanation is the physicochemical or thermal alteration of the remaining surface due to a high energy laser beam. By other tests using energy E < 120 mJ, no ablation effects are evident; so it is possible to assume E = 120 mJ is the energy threshold for λ2 = 532 nm. FTIR analysis did not support the physicochemical hypothesis since, as is shown in Fig. 30.9, no modification of the layers

30 A New Solution for the Painting Artwork

265

4w 4a I

II

4b

III

Fig. 30.8. (Left) Zone 4 using the Palladio laser, λ2 = 532 nm. (I) Zone a, ×20; (II) Zone b, ×10; (III) Zone w, ×10 magnification

Fig. 30.9. FTIR spectrum of black and red zones before and after the laser cleaning with λ2 = 532 nm and λ1 = 1,064 nm. No chemical modification of the compounds is evident

266

S.E. Andriani et al.

composition are evident. Other hypotheses for the greenish effect are the deposition/redeposition of residues or laser ablation products or residues due to incomplete cleaning. In Zone 4b, if the colour canvas change was due to the incomplete laser cleaning, E = 200 mJ, φ = 3 mm(F = 2.8 J cm−2 ) was used for a few round trips on the zone with the purpose to prevent the greenish effect. This solution did not resolve the problem but a more greenish and abrasive effects on canvas (probably due to a high fluence), were observed (Fig. 30.8II). Another possible solution was tested in Zone 4w: E = 120 mJ, φ = 4 mm(F = 1 J cm−2 ) with some distilled water on the encrustations before irradiating (wet laser cleaning). The water enhanced the cleaning effect, so it is important to wet the surface in those cases where the energy density and number of pulses must be kept to a minimum value to avoid possible substrate damage. At the first time, it seems that the water prevents the greenish effect and the canvas abrasion (Fig. 30.8III), perhaps thanks to the lower energy density (F) combined with the water solvent action. However these exposure parameters were ineffective with thick beverone island layers. It was also difficult to prevent water penetration/adsorption into the painting and its harmful solvent action. Microimages (Fig. 30.8) were not sufficient to find a possible explanation of the greenish effect and the work is still in progress. Section 5 The working laser parameters are Vario (τL = 200 ns, φ = 3 mm, ν = 10 Hz) in Zone 1a E = 200 mJ(F = 2.8 J cm−2 ), E = 250 mJ(F = 3.5 J cm−2 ) in Zone 2a. The cleaning results are the same for Section 1a and 2a; no differences are evident. Section 6 The working laser parameters are Palladio (τL = 6 ns, φ = 3 mm, ν = 10 Hz, λ1 = 1, 064 nm) E = 250 mJ in Zone 6a, and the same in Zone 6b (E = 250 mJ, F = 3.5 J cm−2 ). The cleaning of the beverone is not complete perhaps because of a rapid movement of the optical manipulator due to high fluence employed. Figure 30.10I shows the canvas weave still obstructed and thick beverone red islands remaining. In Zone 6b, after the first unsatisfactory laser cleaning (executed by Palladio laser) with the same parameters of Zone 6a, we tried to remove the remaining islands with another laser type: Palladio (λ2 = 532 nm, τL = 6 ns, φ = 3 mm, ν = 10 Hz, E = 150 mJ, F = 2.1 J cm−2 ). The results show a complete surface and interstices cleaning. However, the OM observations reveal the effects of QS λ2 laser cleaning: clearer areas and darker areas, probably due to microabrasion or chromatic alteration of canvas fibres.

30 A New Solution for the Painting Artwork

I

267

II

Fig. 30.10. (I) Zone 6; (II) zone D

Section D The working parameters in Zones Da and Db, were the Palladio laser source (λ2 = 532 nm, τL = 6 ns, φ = 3 mm, ν = 10 Hz, E = 180 mJ, F = 2.5 J cm−2 ) but only with one or two round trips. After cleaning, the greenish effect appears evident (Fig. 30.10II). In Zone Db, another laser irradiation by Palladio (τL = 6 ns, φ = 4 mm, ν = 10 Hz, λ2 = 1, 064 nm, E = 200 mJ, F = 1.6 J cm−2 ) was performed. It is important to underline that the Zone Db, after the second laser treatment, did not show the green effect and the possible explanations are still under investigation.

30.3 Conclusions The behaviour of the cleaning method based on different Nd:YAG laser sources gives different results depending on the type of laser system and exposure parameters. These preliminary tests show the effectiveness of laser cleaning method comparing with traditional ones and they individualize the optimum working conditions for the substances to be removed. The correct laser parameters by using Palladio source (Q-switch, λ1 = 1, 064 nm, τL = 6 ns, φ = 4 mm, ν = 10 Hz, E = 110 mJ) without causing damage to the canvas and to the preparation layer underneath the paint (front of canvas) are obtained in Section 3. On laboratory tests, some drawbacks were observed due to the incomplete cleaning or to the unknown greenish effect. Further investigations

268

S.E. Andriani et al.

should be done on the cause of greenish effect and on its dependence of wavelength since in Section 6 (Q-switch, λ1 = 1,064 nm) no greenish effect is observed with respect to Section 4 (Q-switch, λ2 = 532 nm). It is important to underline that in all examined experimental laser setups, no alteration or loss of painting fragments are observed on the front of the painting. At least in any case where a laser cleaning process was applied, it is extremely important to find “safety” limits for the application of these cleaning methods to prevent irreversible damage on the artworks. Acknowledgements The authors wish to thank Restorers of the Soprintendenza per il Patrimonio Storico, Artistico ed Antropologico delle Province di Bari e Foggia Laboratory for their precious collaboration and help in carrying out the traditional cleaning techniques.

References 1. M. Cooper, Laser Cleaning in Conservation – An Introduction. 2. M. C. Gaetani and U. Santamaria, in Journal of Cultural Heritage, Vol. 1, 199, 2000. 3. A. Sansonetti and M. Realini, in Journal of Cultural Heritage, Vol. 1, 189, 2000. 4. C. Theodorakopoulos, V. Zafiropulos, et al. in Lacona V Book of abstracts, 72, 2003. 5. A. Brunetto “L’utilizzo della strumentazione laser per la pulitura delle superfici nei manufatti artistici” Edizioni il Prato.

31 Removal of Simulated Dust from Water-Based Acrylic Emulsion Paints by Laser Irradiation at IR, VIS and UV Wavelengths M. Westergaard1,2 , P. Pouli3 , C. Theodorakopoulos3,4 , V. Zafiropulos3,5 , J. Bredal-Jørgensen1 , and U. Staal Dinesen1 1 2

3

4

5

School of Conservation, Royal Academy of Fine Arts, Copenhagen, Denmark Art Conservation Centre, Kronborg Castle, Elsinore, Denmark [email protected] Foundation for Research and Technology-Hellas (FO.R.T.H.), Institute of Electronic Structure and Laser, Heraklion, Crete, Greece RCA/ V&A Conservation, School of Humanities, Royal College of Art, Kensington Gore London SW7 2EU, UK Laboratory of Applied Physics, Human Nutrition & Dietetics, Technological Educational Institute of Crete, Ioannou Kondylaki 46, 723 00 Sitia, Crete, Greece

Summary. This study aims to investigate whether laser cleaning may be a valuable method for the removal of soiling from water-based acrylic emulsion paints in comparison to traditional cleaning methods. Acrylic-grounded canvas was painted with three different paints (yellow ochre, titanium white and red alizarin) in a polybutyl-acrylate and methyl methacrylate binder. An acrylic binder was used as a reference. The samples were covered with carbon, SiO2 and soot. Cleaning process ablation rate studies were carried out with a Q-switched Nd:YAG laser at 1,064, 532 and 355 nm and a KrF Excimer laser at 248 nm. The energy densities varied from 0.03 to 0.69 J cm−2 . The irradiated tests at 248 nm were monitored by LIBS analysis. On the samples irradiated at 1,064 nm, various analytical methods were carried out. A determined alteration of the titanium white paint resulted in a marked decrease in the glass transition temperature (Tg ). Furthermore, discoloration (yellowing) occurred on the binder and the titanium white paint. The ochre darkened slightly but the alizarin was unchanged. When compared with the samples cleaned with water-based solvents, the samples cleaned with laser appeared cleaner. However, SEM/EDX and ATR showed that SiO2 was still present on the surface after laser cleaning at the tested conditions.

31.1 Introduction The aim of these experiments was to evaluate whether laser cleaning of soiled water-based acrylic emulsion paints could turn out to be a better alternative to a traditional cleaning with water-based solvents, known among conservators to be quite complicated.

270

M. Westergaard et al.

Water-based acrylic emulsion paints are fast-drying paints with a low glass transition temperature (Tg ) [1]. The softness of these paints may lead to the surface attracting dust particles that can become embedded, and thus difficult to remove. Furthermore, the surface of the paint may come off using traditional water-based solvents, depending on the content of delicate pigments and the amount of dissolving detergents and pigment dispersing agents [2]. In addition, it may be altered mechanically by the movements of the cotton swab when cleaning or just by the grease from fingerprints. A non-contact cleaning method, like laser cleaning, would be preferable. The ablation of varnish on oil paintings at UV wavelengths is a welldocumented and recognized cleaning method [3, 4]. On the contrary, to the authors’ knowledge, there is no documented comparison between surface cleaning of acrylic emulsion paints executed with laser vs. with traditional solvents. Fourrier et al. [5] reported on some good results in relation to UV laser ablation of particles (SiO2 ) from polymers contrary to Real et al. [6] who did some experiments with UV laser ablation of soot, among other materials, from different acrylic surfaces. At too high energy densities, the surface went off while at lower energy densities, the cleaning lacked in sufficiency. Furthermore, a grey discoloration was observed on the white acrylic primers. A considerable number of previous experiments with laser irradiation, mainly of inorganic pigments, have been executed since the mid 1990s and, in general, these pigments discolour. Some of the pigments, like white lead, red lead and zinc white, will darken only temporarily, that is, from a couple of hours up to a whole week after the laser irradiation. Other pigments like cinnabar decompose and permanently darken [7,8]. According to publications, as by Chappé et al. [9], it is notable that 1,064 nm seems like the wavelength at which an ablation threshold can be found in relation to laser irradiation of inorganic pigments, like zinc white, red lead, ochre, sienna, azurite, malachite, cobalt blue, ultramarine blue, bone black and also the organic alizarin. In this study we intended to get preliminary results on laser ablation of artificial dust on acrylic emulsion paint. The test materials were chosen with colours of a well-known brand on a primed canvas. The artificial soiling consisted of known components in a dust layer executed according to the contents of ordinary museum dust.

31.2 Experimental Methods To determine the ablation threshold and evaluate the self-limiting character of the laser cleaning process, ablation rate studies were carried out with a Q-switched Nd:YAG laser at 1,064, 532 and 355 nm and a KrF Excimer laser at 248 nm. The irradiated tests at 248 nm were monitored by Laser induced breakdown spectroscopy (LIBS) analysis. The number of pulses and application of wet or dry cleaning was evaluated and the energy densities varied from 0.03 to 0.69 J cm−2 . Pilot tests were executed at all four wavelengths

31 Removal of Simulated Dust from Water-Based Acrylic Emulsion Paints

271

Table 31.1. Composition of the binder, acrylic paints and soilings 1. titanium white paint (TiO2 ) 2. yellow ochre paint (FeOOH) 3. alizarin paint (1,2-dihydroxyanthraquinone) 4. binder presumably: polybutyl-acrylate/methylmethacrylate P(nBA/MMA) [11]

1. 3.33 g carbon (lampblack) + 6.66 g SiO2 + 0.01 g soot (from beeswax) 2. carbon (lampblack) 3. SiO2 4. 3.33 g carbon (lampblack) + 6.66 g SiO2

in order to test the cleaning effects within the UV, VIS and IR regions. The most promising results of these pilot/preliminary experiments led to the final and more detailed experiments at the most promising wavelength. Analyses were executed on the irradiated samples after the final experiments: Colour measurements (CIE–L∗ a∗ b∗ ), differential scanning calorimetry (DSC), scanning electron microscopy (SEM), energy dispersive X-ray analysis (EDX) and infrared spectroscopy/attenuated total reflectance (ATR). An evaluation of laser cleaning compared to traditional cleaning with selected solvents on the same samples was also executed. Four primed canvases were each painted with one of the three different acrylic paints and with the acrylic binder only (all of the brand “FINITY” from Winsor & Newton). The primer consisted of washed chalk, titanium white pigment and an acrylic binder. The samples were then soiled with artificial dust (see Table 31.1). The dust was composed according to previous studies [10] on the composition of museum dust concluding that the components are mainly earth (silicates) and carbon/soot and may have a greasy consistency. Laser cleaning tests were carried out on all four reference groups (without dust) and on all three types of paints and binder, each covered by all four types of dust. On the same samples, cleaning with water-based solvents was also carried out. The lasers used in this study were (1) A BMI Q-switched Nd:YAG (series 5022 DNS 10) emitting at 1,064, 532 and 355 or 266 nm. Its pulse duration was in the range of 5–7 ns and the pulse repetition rate could vary from 1 to 10 Hz. (2) A Lambda Physics KrF Excimer (COMPex 110), emitting at 248 nm with a pulse duration of 30 ns and repetition rate up to 100 Hz. The choice of traditional solvents was based on an enquiry among conservators in DK and on an investigation carried out among 29 conservators in Canada in 2002 [12]. It was decided to use two water-based solvents: (1) water (1 dl) added a detergent (“Agepon”) (10 drops) and (2) water added 6% tri-ammonium citrate. Six pilot irradiations were executed on each type of sample at each wavelength and the size of the ablations was 1 × 1.5 cm. The energy densities used with the Nd:YAG at the fundamental wavelength were ranging from 0.1 up

272

M. Westergaard et al.

Table 31.2. Parameters at the further experiments with laser irradiation executed at 1,064 nm Experiment A B a

Energy (mJ)

Spot sizea (cm2 )

81.1 81.1

0.47 0.47

Energy density (J cm−2 ) 0.15 0.15 + H2 O

No. of pulses per square

Hz

40 10–15

2 2

(0.8 cm × 0.75 cm × π/4)

to 0.64 J cm−2 and 1–40 pulses were applied. At 532 nm, the energy densities were in the range of 0.03–0.69 J cm−2 and 1–5 pulses were applied, while energy densities from 0.15 up to 0.26 J cm−2 and 1–10 pulses were used at 355 nm. At 248 nm, 1–7 tests were carried out on each type of paint (two on acrylic binder, four on titanium white paint, three on ochre and seven on alizarin) and the energy densities varied from 0.1 to 0.38 J cm−2 with a beam overlap of 80–90%. Based on the results of these pilot tests, the final irradiations were carried out using the 1,064 nm wavelength. For the final laser cleaning test, the parameters (“A” in Table 31.2) of most satisfying preliminary tests (evaluated under the microscope) were repeated in a larger area. However, the cleaning effect was insufficient and a thin water film was applied prior to the irradiation (“B” in Table 31.2) leading to a much better effect. Apparatus applied for the analyses (1) Optical evaluations. An Axiotech optical light stereomicroscope (Zeiss), fitted with differential interference contrast (DIC) capability. (2) Colour measurements (CIE–L∗ a∗ b∗ ). A Konica Minolta CM 2600d Spectrophotometer, double rays with Xenon flash as a light source. The measuring geometry was D/8, the specular component was excluded and the lighting was D65 . (3) Differential scanning calorimetry (DSC). A thermal analysis (TA) instruments DSC Q 1000, coupled with a liquid nitrogen cooling system (LNCS). (4) and (5) Scanning electron microscopy (SEM) and energy dispersive X-ray analysis (EDX). A JEOL JSM 5310, LV vacuum scanning electron microscope. Data processing was executed with the analysis programme LINK ISIS from Oxford. (6) Infrared spectroscopy/attenuated total reflectance (ATR). A spectrum one, FT-IR spectrometer coupled with a Perkin–Elmer universal ATR sampling accessory with the use of a diamond crystal. Spectra collections and calculations were executed with PE Spectrum software.

31 Removal of Simulated Dust from Water-Based Acrylic Emulsion Paints

273

31.3 Results and Discussion 31.3.1 Visual Observations Even at energy densities as low as 0.1 J cm−2 at 248 nm and 0.15 J cm−2 at 355 nm, discoloration occurred on the binder sample (turned greyish) as well as on titanium white (turned blue-greyish) and ochre (turned brownish). The organic alizarin paint was cleaned without any discoloration at 248 nm (with an energy density of 0.22 J cm−2 applying 1 pulse and 80% overlap for the scanning), and at 532 nm (with an energy density of 0.17 J cm−2 , adding water and 2–3 pulses). Generally, no discoloration occurred on alizarin paint apart from a barely visible bleaching of the reference (no dust) at 248 nm. The tests at 532 nm on the binder, the ochre and the titanium white paint either led to an incomplete cleaning at 0.03 J cm−2 or a more complete cleaning (at an energy density of 0.17 J cm−2 ) but leading to the abovementioned discolorations in the binder, titanium white paint and ochre paint (Fig. 31.1). At 1,064 nm, 40 pulses and an energy density of 0.15 J cm−2 resulted in a much more satisfactory cleaning effect. No discoloration of the binder and the alizarin paint was observed but a slight bluish discoloration of the titanium white paint and a very vague greenish tone on the ochre could be detected. When adding a thin water film to the surface before irradiation at 1,064 nm and using an energy density of 0.15 J cm−2 , the cleaning effect seemed complete with only 10–15 pulses and no discoloration could be observed with the naked eye. However, after about a week in the dark, the surface of the titanium white and the binder became somewhat yellow, mostly on samples from which carbon-containing dirt had been removed. The yellowing was more

Fig. 31.1. Microscope image of titanium white paint after laser ablation of carbon and SiO2 at 532 nm with an energy density of 0.17 J cm−2 using 2–3 pulses per irradiated square on a wetted surface. Discoloration of the paint and remains of dirt in the surface can be observed

274

M. Westergaard et al.

pronounced on the titanium white paint. This yellowing bleached to some degree when exposed to daylight for 24 hs. Visual observations indicated that the best results were obtained by cleaning at 1,064 nm, with an energy density of 0.15 J cm−2 , 10–15 pulses and application of a water film. By means of the above-mentioned equipment, these results from the test run were analysed. 31.3.2 Colour Measurements Colour measurements showed that the ochre and alizarin paint became somewhat darker, diminishing their L-value by 3.36 and 3.28, respectively, when ablating carbon and SiO2 . This may be due to residues of carbon on the surface. The titanium white reference sample (no dust) also became a little darker after the irradiation, reducing its L-value by 0.63. Laser removal of dust however led to higher b-values indicating a yellowing. b-Values were higher on the titanium white paint than on the binder. Furthermore, these b-values were increasingly higher proportional to the carbonic content of the ablated dust layer, e.g. the ablation of pure carbon on the titanium white paint led to the highest increase of the b-value (+4.79) compared to the increase of the b-value (+0.48) measured after ablation of SiO2 . The discoloured (into yellow) binder and titanium white paint did not bleach/recover completely after exposure to daylight. The discoloured binder after ablation of carbon, SiO2 and soot was bleached/recovered showing a shift in b-value from 3.80 to 2.34. On the titanium white paint after ablation of carbon, SiO2 and soot, the bleaching reduced the yellowing from 4.89 to 2.57 in b-value. As the b-values of the references of the binder and the titanium white paint before laser irradiation were 0.34 and 1.09, respectively, the surfaces were still somewhat yellow. 31.3.3 DSC Measurements It show that the Tg of the three acrylic paints and the binder before laser cleaning ranged from 10.58 to 13.64 ◦ C. This was expected as the Tg of related polymers range from 5 to 17 ◦ C [11,13,14]. The Tg of the binder reference was found to be 11.84 ◦ C; while for the titanium white paint reference it was measured to be 10.58 ◦ C. On the ochre and alizarin acrylic paint reference samples, the Tg was found to be higher, 13.64 and 13.22 ◦ C, respectively. After the laser cleaning at 1,064 nm, only the samples with organic materials, that is, the binder and the alizarin paint were found to have slightly higher Tg : 1.08 and 2.06 ◦ C, respectively. However, the Tg of the paints containing inorganic pigments, namely ochre and titanium white, decreased: ochre paint by 3.09 ◦ C and the titanium white remarkably by 18.60 ◦ C, see Fig. 31.2.

31 Removal of Simulated Dust from Water-Based Acrylic Emulsion Paints Sample: Tio2 -1064nm Size: 4.0000 mg Method: Heat/Cool/Heat Comment: Tio2, 1064 nm 1.0

275

File: C:\TA\Data\DSC\kcs\Tio2-1064nm.001 Operator: kcs Run Date: 06-Mar-03 12:14 Instrument: DSC Q1000 V6.19 Build 227

DSC

Heat Flow (mW)

0.5

0.0

−0.5

−9.60
C

−8.02
C(1) 0.1043J/g/
C(1)

−6.98
C −1.0

−1.5 −50

−40

−30

−20

Exo Up

−10

0

10

20

Temperature (
C)

30

40

50

60

70

80

Universal V3.5B TA Instruments

Fig. 31.2. DSC measurements on titanium white acrylic emulsion paint after laser ablation of simulated dust at 1,064 nm, with an energy density of 0.15 J cm−2 , 10–15 pulses and application of water. The curve is indicating a decrease in the Tg to −8.02◦ C from 10.58◦ C

31.3.4 SEM SEM was executed on a thin cross section from the yellowed titanium white paint after laser cleaning carbon, SiO2 and soot (at 1,064 nm with an energy density of 0.15 J cm−2 and application of water), but neither in the pigment grains nor in the binder were any morphological changes found. Likewise, no morphological changes were found on the surface neither in the pigment grains of the samples ablated at 1,064 nm. 31.3.5 EDX Mapping EDX mapping on samples after laser ablation of SiO2 alone or of the mixture of carbon, SiO2 and soot showed that SiO2 residues were found on both. When wet laser ablation was performed at 1,064 nm, with an energy density of 0.15 J cm−2 and 10–15 pulses, less SiO2 was found. After the traditional cleaning with water-based solutions, no morphological changes were observed on the surfaces of the samples. 31.3.6 ATR (FT-IR) ATR was executed after the laser cleaning at 1,064 nm but neither on reference samples nor on samples originally coated with carbon, SiO2 and soot any

276

M. Westergaard et al.

change in the binder was observed. No changes could be detected in the pigments either. As regards the traditional cleaning, with a detergent and with triammonium citrate, no changes in the chemical combination of the binder occurred. After wet laser cleaning, residues of SiO2 could still be detected at 1,114 cm−1 , particularly on the binder and the ochre paint samples. No residues of carbon could be observed. After traditional cleaning of the binder and paints, both SiO2 and carbon were detected at 1,114 cm−1 . However, the SiO2 content was higher after laser cleaning than after the traditional cleaning leading to the observation that, under the applied conditions, the laser was less efficient than the traditional cleaning when ablating SiO2 . The results of the analyses indicate that the binder absorbed the irradiation and turned greyish in all the pilot tests except for dry cleaning at 1,064 nm with an energy density of 0.15 J cm−2 and 40 pulses. This could be due to a discoloration of the underlying primer containing titanium white paint turning blue-greyish above the threshold at all four wavelengths in the pilot tests, and to the transparency of the acrylic binder. Ochre paint turned brownish at 248, 355 and 532 nm, indicating transformation into hematite and black magnetite [15]. However, at 1,064 nm (0.15 J cm−2 and 40 pulses), ochre turned greenish, presumably caused by black magnetite mixed with the yellowish of the ochre. When water was applied before irradiation at 1,064 nm, ochre turned slightly darker as revealed from the colour measurements. The permanent grey-bluish discoloration of the titanium white paint at the pilot test is hardly due to a pollution of the pigment with zinc white as Chappé et al. [9] mention that the discoloration of the zinc white is only temporary. The yellowing of the titanium white and the binder appeared on samples ablated at 1,064 nm after cleaning of dirt containing carbon and when applying water. After irradiation also reference samples were kept in the darkness and exposed to temperatures around 30◦ C and they did not show any yellowing. Consequently, the high temperature and the darkness can be excluded as factors causing the yellow discoloration of these polymer paints. According to Strlic et al. [16], the degree of yellow discoloration of various organic surfaces rose in accordance with the thickness of the layer of carbon. But to the author’s knowledge, the application of water (in combination with ablation of carbon) has not yet been reported to induce discoloration. One theory, which is applied in relation to the remaining yellowing patina after laser cleaning of stonework, is relying on changes in the refractive index due to the creation of voids in the surface [17]. However, the indications up to now show that this theory is not applicable on acrylic paints. Firstly, the yellow discoloration recovered to some extent when exposed to daylight and secondly, no morphological changes were noted with SEM, nor did ATR show any alterations in the composition of the acrylic binder after the laser treatment. A possible explanation to the yellowing is that the titanium white pigment particles (in the paint and primer) as well as the carbon particles act as heat sources to the surrounding organic medium. The temperature in the paint can rise significantly and trigger a quick decomposition of the organic binder. This

31 Removal of Simulated Dust from Water-Based Acrylic Emulsion Paints

277

decomposition might very well lead to a creation of chromophores and thus a yellowing of the material. Water helps the heat transfer from the particles to the organic binder. Chromophores bleach when exposed to daylight, and so did the examined discoloured samples of binder and titanium white paint. The degree and speed of the polymer recovery is dependent on the Tg . The lower the Tg , the quicker the recovery process. An indication of the decomposition triggered by the titanium white was the remarkable decrease of Tg by 18.60◦ C, leading to a pronounced softness of the paint, due to broken chains in the molecular structure. The binder, on the other hand, turned somewhat yellow too, but increased in Tg by 1.08◦ C. This could indicate that the decomposition and the yellow discoloration were not related. Further tests including more paints containing other pigments should be executed at 1,064 and 2,940 nm with an Er:YAG laser too, as the best results were obtained in the IR area. deCruz et al. [18] reported on good results concerning the cleaning of soot from oil paintings with short pulses and application of water. Also, Bracco et al. [19] ablated varnish from oil paintings at this wavelength. There is a possibility that the lower energy at 2,940 nm combined with the maximum absorption of water at this wavelength might lead to “steaming off” of acrylic emulsion paints without further heating and therefore without any discoloration. When comparing the cleaning effect of laser cleaning at 1,064 nm (with an energy density of 0.15 J cm−2 , 10–15 pulses and adding water) with the cleaning effect of the traditional cleaning method (triammonium citrate or a detergent in an aqueous solution) by means of optical observations, it was obvious that the samples cleaned with laser seemed much cleaner than those cleaned traditionally. The detergent (Agepon) seemed more efficient than triammonium citrate. Samples traditionally cleaned from carbon were poorly cleaned, while those cleaned from only SiO2 seemed perfectly cleaned. SEM and FT-IR eventually proved the latter to some extent. The same analysis showed that more SiO2 residues could be detected on the laser-cleaned samples.

31.4 Conclusions With the given parameters, 248 and 355 nm, seemed unsuitable for the laser cleaning of acrylic binder, titanium white paint and ochre paint, due to discoloration and to an insufficient cleaning effect. The alizarin paint seemed laser safe, however, and was cleaned without any discoloration at 248 nm, with an energy density of 0.22 J cm−2 , 1 pulse and 80% overlap. Apparently, 532 nm was not suitable for the laser cleaning of acrylic binder, titanium white paint and ochre paint, due to discoloration and insufficient cleaning effect. Optical evaluations indicated that this wavelength and the applied parameters (at an energy density of 0.17 J cm−2 , 2–3 pulses and application of water) seemed adaptable for the cleaning of alizarin paint. At 1,064 nm and an energy density

278

M. Westergaard et al.

of 0.15 J cm−2 , 10–15 pulses and application of water, the cleaning effect was satisfactory but colour measurements showed a slight darkening of the ochre and, after a week kept in the dark, the binder and titanium white had turned yellowish. The yellowing was most pronounced on the titanium white. The yellowing was partly recovered when exposed to daylight and is presumably due to the creation of chromophores. On the titanium white paint samples, the catalysing TiO2 pigment probably intensified the oxidation process while the transparent binder could be optically influenced by the TiO2 content in the underlying primer. The catalysing effect of the TiO2 probably contributes to the remarkable decrease of Tg by 18.60◦ C, leading to a pronounced softness of the titanium white paint. This is ascribed to broken chains in the molecular structure of the binder. No morphological changes in the surface of any of the samples were noted with SEM after the laser cleaning at 1,064 nm nor did ATR reveal any changes in the chemical composition of the polymer. When comparing with samples cleaned with water-based solvents, samples cleaned with laser gave a much cleaner result. However, SEM/EDX and ATR showed that SiO2 was still present on the surface after laser cleaning and the yellowing of titanium white, as well as changes in the Tg , questions the suitability of laser cleaning on acrylic emulsion paint, at least at the given parameters. Acknowledgements The authors would like to thank Marie Vest from The Royal Library, Denmark, Jan Jørn Hansen and Jetti van Lanschot from the School of Conservation at The Royal Academy of Fine Arts, in Denmark, Yvonne Shashoua at The National Museum, Denmark, Tom Learner at Tate Gallery in London, England and Costas Fotakis from FO.R.T.H., Institute of Electronic Structure and Laser, Heraklion, Crete, Greece.

References 1. J. Crook, and T. Learner, The Impact of Modern Paints. London: Tate Gallery Publishing Ltd. 200, 2000. 2. C. Lamb, The Conservation of Modern Paintings: Introductory Notes on Papers to be Presented. London 1982. London, United Kingdom Institute for Conservation & Tate Gallery. 6, 1982. 3. V. Zafiropulos and C. Fotakis, in Laser Cleaning in Conservation: an Introduction, Edited by M. Cooper, Oxford, 79, 1998. 4. V. Zafiropulos, in: Laser Cleaning, Edited by B. Luk’yanchuk. World Scientific. Singapore, New Jersey, London, Hong Kong, 343–392, 2002. 5. T. Fourrier, G. Schrems, T. Mühlberger, J. Heitz, N. Arnold, D. Bäuerle, M. Mosbacher, J. Boneberg, and P. Leiderer, in Applied Physics A 72, Materials, Science & Processing. 1–6, 2001.

31 Removal of Simulated Dust from Water-Based Acrylic Emulsion Paints

279

6. W. A. Real, I. Zergioti, Y. Spetsidou, and D. Anglos, in The 11th Triennial Meeting: Proceedings. Edinburgh; ICOM Committee for Conservation, 303–308, 1996. 7. P. Pouli, D. C. Emmony, C. E. Madden, and I. Sutherland, in Applied Surface Science Vol. 173, 252, 2001. 8. B. Luk’yanchuk and V. Zafiropulos, Chapter 8.3 in: Laser Cleaning, Edited by B. Luk’yanchuk. World Scientific, Singapore, New Jersey, London, Hong Kong, 393–414, 2002. 9. M. Chappé, J. Hildehagen, K. Dickmann, and K. Bredol, in Journal of Cultural Heritage Vol. 4. suppl. 1., 264, 2003. 10. Y. H. Yoon and P. Brimblecombe, Studies in Conservation, Vol. 45, 127, 2000. 11. T. Learner, O. Chiantore, and D. Scalarone, in The 13th Triennial Meeting Rio de Janeiro: Preprints. Rio de Janeiro 2002.; ICOM Committee for Conservation. 911, 2002. 12. A. Murray, C. Contreras de Berenfeld, P. Y. Chang Sue, E. Jablonski, T. Klein, C. M. Riggs, C. E. Robertson, and A. W. M. Tse, in Materials Research Society Fall Meeting, Materials Issues in Art and Archaeology VI: Proceedings Vol. 712. Boston 2001. Boston: Materials Research Society. 83–90, 2002. 13. M. F. Meckelburg, C. P. Tumosa, and J. D. Erlebacher, Polymer: Preprints. u.p./u.å., Boston: American Chemical Society. Division of Polymer Chemistry, William E. Casp. 297–298, 1994. 14. P. M. Whitmore and V. G. Colaluca, in Studies in Conservation Vol. 40, 51, 1995. 15. A. Athanassiou, A. E. Hill, T. Fourrier, L. Burgio, and R. J. H. Clark, in Journal of Cultural Heritage, Vol. 1, 210, 2000. 16. M. Strlic, J. Kolar, V. P. Selih, and M. Marincek, in Applied Surface Science, Vol. 236, 2003. 17. V. Zafiropulos, C. Balas, A. Manousaki, G. Marakis, P. Maravelaki-Kalaitzaki, K. Melesanaki, P. Pouli, T. Stratoudaki, S. Klein, J. Hildenhagen, K. Dickmann, B. S. Luk’yanchuk, C. Mujat, and A. Dogariu, in Journal of Cultural Heritage Vol. 4, 249, 2003. 18. A. deCruz, M. L. Wolbarsht, and P. A. Hauger, Journal of Cultural Heritage, Vol. 1, 173, 2000. 19. P. Bracco, G. Lanterna, M. Matteini, K. Nakahara, O. Sartiani, A. deCruz, M. L. Wolbarsht, E. Adamkiewicz, and M. S. Colombini, in Journal of Cultural Heritage, Vol. 4. suppl.1., 202, 2003.

32 Traditional and Laser Cleaning Methods of Historic Picture Post Cards ◦

M. Mäder1 , H. Holle2 , M. Schreiner1 , S. Pentzien3 , J. Krüger3 , and W. Kautek3,4 1

◦

2

3

4

Institute of Science and Technology in Art, Academy of Fine Arts, Schillerplatz 3, 1010 Vienna, Austria present address: Freiberger Compound Materials, Am Junger Löwe Schacht 5, 09599 Freiberg, Germany Institute of Conservation and Restoration, Academy of Fine Arts, Schillerplatz 3, 1010 Vienna, Austria, [email protected] Division Surface Technologies, Federal Institute for Materials Research and Testing (BAM), Unter den Eichen 87, 12205 Berlin, Germany Department of Physical Chemistry, University of Vienna, Waehringer Str. 42, 1090 Vienna, Austria

Summary. Traditional paper cleaning techniques are not always sufficient for cleaning artefacts. The pulsed laser cleaning can offer a valuable tool for solving problematic cases in paper conservation. Comparative studies of traditional and laser cleaning were performed on historic picture post cards. The results demonstrate the possibilities of partial laser cleaning using nanosecond laser pulses at a wavelength of 532 nm.

32.1 Introduction Historic post cards from the beginning of the twentieth century belong to the field of applied arts. Today they not only are treasured objects for collectors but also offer an insight in the development of art and culture as well as the printing technology. A couple of such objects show noticeable contaminations on their surfaces, i.e. they are covered with a thin layer of dust and dirt. Therefore, the images and the inscriptions can be hardly recognized. Laser cleaning [1–11] should be applied to remove the impurities without an alteration of the card boards, pigments and dyes as well as the writing media. Before laser cleaning, parts of the objects were cleaned with traditional dry as well as aqueous cleaning methods [12, 13]. Conservation techniques for paper include mechanical scratching and the use of a brush, eraser or draft clean powder (a granulated eraser). Paper may also be cleaned with water, organic solvents, enzymes, etc. Satisfactory results are often

282

M. Mäder et al.

obtained by a treatment using cellulose ethers like carboxymethyl cellulose (CMC), methyl cellulose (MC) or hydroxypropyl cellulose (HPC). Nevertheless, the current aqueous cleaning methods are not always sufficient, in particular if the paper is coated with sensitive printing as well as writing media. Additionally, microscopic investigations and X-ray fluorescence analysis (XRF) were carried out in order to characterize the materials and their possible alterations provoked by cleaning.

32.2 Experimental Two historical picture post cards were used in this case study exemplarily: “Moorlake” and “Gruss aus Bad Ems” (Fig. 32.1). Both cards were produced at the beginning of the twentieth century using high-quality chromolithography. Besides the colourful images, they also show handwritten inscriptions of the senders. Additionally, the surfaces of the cards are covered by extensive contaminations accumulated over the years. 32.2.1 Traditional Paper Cleaning First of all, the loose parts of the dirt were removed using a soft brush. Subsequently, draft clean powder was applied to reduce the thickness of the adherent deposits. In the case of the post card “Moorlake”, the left half of the surface area was treated by the granular eraser, whereas the right half of the post card “Bad Ems” was cleaned in this way. In order to test their cleaning behaviour, three different cellulose ethers were used in a second cleaning step on the post card “Bad Ems”. For these purposes the surface area was subdivided into three horizontal parts and each cleaning solution was applied to one of these stripes (Fig. 32.2). In the third step small vertical stripes on the left and the right side of the card were treated a second time using the most effective CMC solution.

Fig. 32.1. General view of the picture post cards: “Moorlake” (left) and “Gruss aus Bad Ems” (right)

32 Traditional and Laser Cleaning Methods

283

Fig. 32.2. Wet cleaning treatment using different cellulose ethers: after a first cleaning step (left), the CMC treatment was repeated on small vertical stripes (right)

Fig. 32.3. Laser cleaning station (Federal Institute for Materials Research and Testing, Berlin)

32.2.2 Laser Cleaning Selected areas of the objects were treated with the radiation of a Nd:YAG laser running at 532 nm wavelength providing a pulse duration of 8 ns. Figure 32.3 shows the computer-controlled laser cleaning station. The laser system with the scanning device is mounted above the object to be cleaned. The sample can be aligned on a movable holder. Volatile ablation products in the form of gas or dust can be exhausted. The whole laser-cleaning workstation fulfils the safety requirements of the Laser Class 1 condition. In addition to the laser light, the object may be illuminated by UV or visible light. A camera system allows the observation of the object also during the laser treatment.

284

M. Mäder et al.

The focused beam with a (1/e2 ) Gaussian diameter of 230 µm was scanned over the working area at a fixed laser fluence. The fluences were varied between 53 and 154 mJ cm−2 . The number of laser pulses on every point ranged from 25 to 400. The scan pattern, i.e. the lateral distance between the spot centres, was selected in such a manner that the overlap of spots with a Gaussian distribution of the fluence yielded a nearly homogeneous impact of the laser beam on the working area. 32.2.3 Material Analysis Energy dispersive X-ray fluorescence analysis could be applied to characterize the chemical composition of the cardboards as well as the printing and writing materials in a non-destructive manner. The portable µ-XRF instrument [14] consists of a Mo-tube with a polycapillary lens ensuring a beam spot of about 100 µm. The tube voltage and current were set to 35 kV and 0.6 mA, respectively. X-ray spectra were taken by an electronically cooled drift chamber detector at an acquisition time of 200 or 600 s.

32.3 Results and Discussion The traditional cleaning methods proved suitable for a careful removal of dirt layers from the paper surface. Particularly, the dry techniques allowed a gradual proceeding, but the visible improvement was restricted in the case of extensive contaminations. More pronounced effects could be achieved by treatments with the cellulose ether solutions. The best cleaning result was obtained for the post card “Bad Ems” treated with the 2.5% CMC. A repeated processing with the same solution resulted in a further removal of material from the card surface. However, the result was an “over-cleaning” in this special case, i.e. the original surface of the post card and especially the printing materials were also affected by the second treatment. In preliminary experiments, the laser parameters were optimized in order to get visible cleaning effects. In particular, the laser fluence and the number of pulses per spot were varied. Best results were obtained in a relatively low range of laser fluences between 53 and 100 mJ cm−2 utilizing a low pulse number of 25 per spot. An increase in these values resulted in additional destruction, i.e. the printing material and the covering layer of the card boards were already removed. The general views of both post cards (Fig. 32.1) show clearly the areas irradiated with the laser light. The selected areas of laser treatment are distinguished by an explicit cleaning effect, i.e. the contaminations could be reduced significantly. Figure 32.4 depicts a detail of the post card “Moorlake”. The rectangular area was treated with a laser fluence of 53 mJ cm−2 and 25 pulses per spot. The figure of the man, the printed red letters and the hand-writing are more visible than before. Also microscopic investigations demonstrated

32 Traditional and Laser Cleaning Methods

285

Fig. 32.4. Detail of the post card “Moorlake” showing an area after laser treatment (25 pulses per spot with 53 mJ cm−2 )

that no alterations of the cardboard and the colour of the printed images and letters were observable due to the laser treatment in this case. The other case study on the post card “Bad Ems” indicated that an adherent dirt layer, which partially resisted the cellulose ethers, could be removed by a proper laser treatment. The subsequent visual evaluation of the treated areas revealed that the selected laser fluence of 100 mJ cm−2 already reached the range of destruction in which the printed materials were partially affected. Material analysis using XRF allows the determination of the chemical elements which are characteristic for the materials used for covering, printing as well as writing. The colour palette of the identified printing materials contains mainly inorganic pigments like cinnabar, red lead, lead white, Naples yellow, various ferric oxides and chromium green. Organic dyestuffs were used only in a few cases. Additionally, different coatings on the cardboards were detected. Whereas the paper of the post card “Bad Ems” was coated with Permanent White, a chalk containing ground was used for the preparation of the “Moorlake”. An iron gall ink could be identified as writing material used for the hand-written inscriptions.

32.4 Conclusions Traditional cleaning of paper using dry and/or wet methods offers the possibility of large area as well as partial treatments in order to remove contaminations layer-by-layer from the surface. The dry (mechanical) cleaning with brushes and eraser agents often proves to be insufficient in cases of intense dirt. The application of water or aqueous solvents holds the risk of paper degradation as well as damage (loss) of paints or inks. Partial laser cleaning of picture post cards is demonstrated successfully. The results show that the dirt on the surface, which is particularly resistant

286

M. Mäder et al.

to dry cleaning, can be removed and the obscured pictures and writings could be visualized again. For this purpose, laser fluence levels must stay below the ablation and destruction threshold of the paper substrate as well as the coating materials including paints and inks. Further studies are required to expand the understanding of laser interaction with coated papers in order to avoid any risks for the objects. Acknowledgement We acknowledge partial financial support by the European Co-operation in the field of Scientific and Technical Research, “Artwork Conservation by Laser” (COST G7, Short Term Scientific Mission).

References 1. W. Kautek, S. Pentzien, J. Krüger, and E. König, in Lasers in the Conservation of Artworks I, Restauratorenblätter (Special Issue) Edited by W. Kautek and E. König, Mayer & Comp., Wien, 69, 1997. 2. W. Kautek, S. Pentzien, P. Rudolph, J. Krüger, and E. König, in Appl. Surf. Sci., Vol. 127–129, 746, 1998. 3. P. Rudolph, S. Pentzien, J. Krüger, W. Kautek, and E. König, in Restauro, Vol. 104 (6), 396, 1998. 4. J. Kolar, M. Strlic, S. Pentzien, and W. Kautek, in Appl. Phys. A, Vol. 71, 87, 2000 5. J. Kolar, M. Strlic, D. Müller-Hess, K. Troschke, S. Pentzien, W. Kautek, in J. Cultural Heritage, Vol. 1, 221, 2000. 6. W. Kautek, S. Pentzien, D. Müller-Hess, K. Troschke, and R. Teule, in SPIE, Vol. 4402, 130, 2001. 7. D. Müller-Hess, K. K. Troschke, J. Kolar, M. Strlic, S. Pentzien, and W. Kautek, in Restauro, Vol. 8, 604, 2001. 8. P. Rudolph, F. J. Ligterink, J. L. Pedersoli Jr, M. van Bommel, J. Bos, H. A. Aziz, J. B. G. A. Havermans, H. Scholten, D. Schipper, and W. Kautek, in Appl. Phys. A, Vol. 79, 181, 2004. 9. P. Rudolph, F. J. Ligterink, J. L. Pedersoli Jr., H. Scholten, D. Schipper, J. B. G. A. Havermans, H. A. Aziz, V. Quillet, M. Kraan, B. van Beek, S. Corr, H.-Y. Hua-Ströfer, J. Stokmans, P. van Dalen, and W. Kautek, in Appl. Phys. A, Vol. 79, 941, 2004. 10. H. Scholten, D. Schipper, F. J. Ligterink, J. L. Pedersoli Jr., P. Rudolph, W. Kautek, J. B. G. A. Havermans, H. A. Aziz, B. van Beek, M. Kraan, P. van Dalen, V. Quillet, S. Corr, and H.-Y. Hua-Ströfer, in Lasers in the Conservation of Artworks, Springer Proceedings in Physics, Vol. 100, 11, 2005. 11. E. Pilch, S. Pentzien, H. Mädebach, and W. Kautek, in Lasers in the Conservation of Artworks, Springer Proceedings in Physics, Vol. 100, 19, 2005. 12. O. Wächter, in Restaurierung und Erhaltung von Büchern, Archivalien und Graphiken, Hermann Böhlaus Nachf., Wien Köln Graz, 1982. 13. D. van der Reyden, in JAIC, Vol. 31, 117, 1992. 14. µ-XRF “COPRA”: Prototype for X-ray fluorescence analysis of artifacts, developed and built within an EU-Project (SMT4-CT98-2237).

33 Femtosecond Laser Cleaning of Painted Artefacts; Is this the Way Forward? ∗

P. Pouli1 , G. Bounos1,2 , S. Georgiou1 , and C. Fotakis1,2 1

∗ 2

Institute of Electronic Structure and Lasers (IESL), Foundation for Research and Technology-Hellas (FORTH), P.O. Box 1527, Heraklion, Crete, 71110, Greece [email protected] Department of Physics, University of Crete, Heraklion, Greece

Summary. The laser cleaning of painted artefacts relies on the synergy of thermal, photochemical and photomechanical effects, which are involved in laser ablation. A crucial issue, however, for a successful cleaning intervention is the spatial confinement and control of these effects for safeguarding the original surface from potential damage. Extensive studies have shown that in many cases there is an optimum interplay of laser and material parameters, which resulted in successful laser cleaning applications. The laser pulse duration is an important parameter in this context. The scope of this work has been the exploration of any advantages, which may be offered by using ultrafast UV (248 nm) lasers for the cleaning application of sensitive painted artworks. To achieve this goal, comparative study on the ablation rate and threshold of femto- and nanosecond laser pulses of typical varnishes (dammar, mastic, etc.) have been performed. Femtosecond pulses appear to be superior in terms of the spatial resolution and etching resolution and this fact has been demonstrated for both technical samples and original objects. Additionally, possible induced photochemical modifications have been investigated by monitoring the photoproduct laser-induced fluorescence of varnish-systems doped with low concentrations of welldefined photosensitive dopants (e.g. PhenI). It is established that irradiation with f s UV laser pulses results in minimal photochemical modifications. Importantly, the amount of photochemical products is largely independent from the optical properties (i.e. absorptivity) of the varnish. Considering the recent advances in ultrafast laser technology, the use of such lasers appears to provide a viable approach.

33.1 Introduction Femtosecond (fs) laser technology has been demonstrated to offer particular capabilities for material processing overcoming several of the limitations of processing with nanosecond (ns) laser pulses. In particular three major advantages of fs laser processing have been emphasized in previous studies [1]. First, multiphoton processes enable the processing of even nominally transparent substrates [2, 3]. Second, the thermal diffusion is minimal thereby enabling

288

P. Pouli et al.

processing with minimal/negligible thermal degradation [1]. Third, there is no plasma shielding so that there is maximum coupling of the laser energy into the substrate [4]. As a result of these factors, the morphological aspects of the processed substrates are far superior to those attained with ns laser pulses [4]. Although the previous advantages are very important in the cleaning of painted artworks [5, 6], there are a number of additional aspects that become important/vital and may be decisive for the success of the femtosecond laser processing. A most crucial issue concerns the extent of photochemical modifications effected to the substrate. In fact, this question issue is central to all femtosecond laser processing schemes of molecular/organic systems, including biological and medical applications. In relationship with these issues there is almost no work reported in the literature. We report here first/preliminary studies in the use of femtosecond laser technology for the processing/restoration of painted artworks. As demonstrated by extensive previous work with nanosecond laser technology [5–7], this is a highly demanding application in which several issues have to be carefully optimised. In particular, given the high photolability of the substrates, photochemical effects upon fs ablation/irradiation become a crucial factor. Although femtosecond laser ablation may considerably improve the morphology, it is unclear to what extent it affects photochemical effects. To explore any advantages which may be offered by using ultrafast pulse lasers for the cleaning application of painted artworks, comparative studies on typical varnishes (dammar and mastic) have been performed. The cleaning result of fs (500 fs) and ns (30 ns) laser pulses at 248 nm was evaluated on the basis of 1. The spatial resolution, i.e. the etching efficiency and the ablation threshold (FT H ) of the studied materials in both regimes (fs and ns) 2. The extent of the induced photochemical modifications to the remaining material, by monitoring the photoproduct laser-induced fluorescence of varnishes doped with very low concentrations of the photosensitive dopants, as a function of the laser fluence and the number of laser pulses 3. The morphology of the ablation spots in order to investigate the extent of “influence” to the substrate

33.2 Experimental Methods Two types of commercially available varnishes, dammar and mastic of Chios, have been studied. Neat and doped films are cast on quartz plates from a dichloromethane solution and are subsequently dried for 48 h. Typical film thickness is in the 30 ± 10 µm range. Femtosecond laser pulses (500 fs) at 248 nm were generated using a XeCl excimer pumped dye laser system based on the principle of a distributed feedback dye laser (DFDL). The energy output is 10–30 mJ per pulse while the

33 Femtosecond Laser Cleaning of Painted Artefacts

289

average pulse-to-pulse fluctuation is 15%. The beam was focused perpendicularly on the sample by means of a quartz lens (f = +300 mm). Typical fluence values were between 0.05 and 0.75 mJ cm−2 . Nanosecond laser pulses (10 ns) were emitted from a compact excimer laser (Braggstar 200, TUI Laser) operating at 248 nm (KrF). The maximum energy was 16 mJ while the laser beam was focused on the sample by means of a fused silica plano-convex lens (f = +100 mm). Typical fluences were between 4 and 500 mJ cm−2 . All irradiation experiments were performed in air. A standard laboratory arrangement was used for the pump/probe LIF measurements. For consistency reasons, irradiation of the samples with both fs and ns laser systems was performed under the same recording conditions. Photoproduct fluorescence was induced with the ns laser at 248 nm at fluences of 5–10 mJ cm−2 , in surface area slightly smaller than the area irradiated with the “pump” beam to avoid additional photolysis products. An optical fibre placed nearly perpendicularly to the sample and in close proximity to its surface (∼ 1 cm) was employed to collect the emitted signal while cut-off filters were used to minimize detection of laser-scattered light. The emission signal is spectrally analysed in a 0.20 m grating spectrograph equipped with a 300 grooves/mm grating and the spectrum is recorded on an optical multichannel analyzer (OMA III system, EG&G PARC Model 1406) interfaced to a computer. Etch depth measurements, following laser ablation, were performed by a mechanical stylus profilometer (Perthometer S5P, Mahr) and the irradiated areas were examined under an optical microscope (ME600, Nikon).

33.3 Results and Discussion 33.3.1 Spatial Resolution In laser cleaning applications on painted artworks, the major attention is focused on the spatial confinement of the intervention in order to ensure minimal damage risk to the underlying original surfaces [5,6]. This is particularly important in cases where the unwanted surface layers are very thin and thus it is necessary to ensure that the mean removed depth per laser pulse, as well as the thermal load to the substrate, are the minimum possible. It is by now well established [3, 4, 7] that multiphoton processes in the femtosecond laser irradiation result in reduction in the effective optical penetration depth and, as a consequence, in a significant decrease to the actual etching rate. Given that the examined varnishes are a mixture of oligomers of some chemical similarity to polymers, it is expectable that their etching efficiency presents similar dependence. Figure 33.1 illustrates the etching depths of dammar and mastic by 248 nm excimer laser pulses of 500 fs and 30 ns duration. In the ns case, the etch rate rises proportionally to the applied fluences up to about 150 mJ cm−2 , where the etch rate saturates reaching a plateau. On the other hand, the fs etch

290

P. Pouli et al.

Etch Rate (mm/pulse)

2,0 1,6 1,2 0,8 dammar (fs) dammar (ns) mastic (fs) mastic (ns)

0,4

0

200

400

600

Fluence (mJ/cm2)

Fig. 33.1. Etch rate (in µm/pulse) vs. fluence (in mJ cm−2 ) for dammar (black dots) and mastic (red squares) in the femtosecond (full dots/squares) and the nanosecond region (open dots/squares)

curves appear almost linear, with lower slope in comparison to the ns one. The mean removed depth per pulse is much lower with the shorter laser pulse width and ablation initiates at significantly lower fluence values than in the ns regime, suggesting that fs laser ablation may ensure a higher degree of resolution and thus minimal risk to the original artwork surface. Another important observation is that, in the fs case, the etch rate lines of the two studied varnishes are nearly equal. Mastic shows a higher small signal absorption coefficient than dammar at the irradiation wavelength (248 nm) and as a result a smaller optical penetration depth is expected, which justifies a lower etch depth rate in the ns case (Fig. 33.1). The fact that in the fs irradiation these lines appear very close implies that the mean removed material per laser pulse is nearly independent from the optical properties of the material, which may be of great advantage in the specific cleaning applications. 33.3.2 Photochemical Modifications Another very important issue in the laser cleaning of painted artworks lays on the extent of the photochemical modifications that may be effected to the substrate (i.e. the remaining varnish layer) [8]. Based on extensive previous work of our group in polymers [9,10], the induced photochemical modifications was investigated by monitoring the photoproducts formed in doped varnish systems by means of laser induced fluorescence (LIF) in a “pump-probe” scheme. The approach is based on the following key-points: 1. It was found [7] that the dissociation and subsequent radical activity of the dopant is particularly sensitive to the alterations of the polymer environment and thus it is expected that the study of photoproduct activity of the dopant will lead to a representative picture of the actual photochemistry of the varnish layers.

33 Femtosecond Laser Cleaning of Painted Artefacts

291

Fluorescence Intensity (a.u.)

2000 1800 1600 Dammar F(ns) = 270 mJ/cm2

1400 1200 1000 800 600 400

Dammar F(fs) = 65 mJ/cm2

200 0 300

400 wavelength (nm)

500

Fig. 33.2. Comparison of the product LIF spectra from 9-Iodophenanthrene (1%wt) doped dammar irradiated above the ablation threshold (FT H ) with a single fs (black line) and ns (red line) laser pulses (λ = 248 nm). For consistency reasons in both cases the LIF spectra was obtained with the same probing beam (λ = 248 nm, 10 ns) under the same conditions

2. The employed dopants, usually halo-aromatics, are characterised by a simple photodissociation reactivity pattern resulting exclusively into aryl and halogen radicals. The laser-induced modifications of these products can be easily and uniquely characterised and quantified via LIF. For the effective comparison of the extent of photochemical changes induced from the different laser systems, it was chosen to use the same probing beam in both experiments. Furthermore in order to ensure one-photon excitation of product, ns pulses were selected, at very low fluences (in the range of 4– 6 mJ cm−2 ). In Fig. 33.2, the product LIF spectra recorded from 9-Iodophenanthrene (1% wt) doped dammar irradiated above ablation threshold (FT H ) with a single fs and ns laser pulse at 248 nm are comparatively illustrated. The emission band at ∼374 nm, which does not appear either before irradiation with the pump laser or upon irradiation of the neat varnish films, is attributed to emitting photoproducts. This band is ascribed to the 1 B3u →1A1g transition of PhenH [11, 12]. Given that the illustrated LIF spectra were recorded under the same conditions, it is clear that product formation is much reduced in the fs irradiation as compared with ns irradiation for a single pump pulse with fluence well above the FT H . The superiority of the fs pulses as regards the limited photochemical modifications was also supported by quantitative characterisation and comparison of the products of the doped varnishes as a function of the fluence of the “pump” beam and the applied number of pulses. Indeed when comparing the product fluorescence intensity upon fs and ns irradiation as a function of the fluence of the “pump” beam [13], it was clearly shown that photochemical

292

P. Pouli et al.

Fig. 33.3. Optical microscope photographs of craters created on thick (∼50 µm) mastic varnish film by ns (a) and fs (b) laser pulses at 248 nm. Irradiation parameters: 200 pulses at 350 mJ cm−2 in the ns case and 300 pulses and 265 mJ cm−2 in the fs case

modifications are remarkably lower when shorter pulses of the same wavelength are applied. Furthermore we observe that the fs irradiation of the two systems results in nearly equal product yields despite their considerable difference in the absorptivity at 248 nm. In other words the amount of product formation is largely independent of the absorptivity of the varnish. This feature can be particularly important for cases where multiple conservation treatments have been applied, as it may enable highly precise treatment, nearly independently of the optical properties of the substrate. Furthermore studies aiming to compare the accumulation of the PhenH photoproducts (at λ = 374 nm) as a function of successive “pump” laser pulses of ns and fs duration [13] have shown that at low fluences (below the ablation threshold) both irradiation regimes show the same behaviour; increasing product formation with increasing number of pulses and fluences. On the other hand a distinctively different picture was observed when higher laser fluences (above the ablation threshold) were applied, where product formation in the fs case tends to decrease with increasing laser fluences. The above observations indicate that the thermal effected zone is limited in the fs case even at extensive irradiation conditions (fluence and number of pulses). Thus the importance of using shorter laser pulse width of higher fluence values (significantly above the ablation threshold) for an effective and photochemically limited material removal is highlighted. 33.3.3 Morphological Aspects Finally, in laser cleaning of artworks, the morphology of the laser-irradiated surfaces is of major importance as melting phenomena or irregular structures may result into potential damage features on the underlying original surface. In Fig. 33.3, the morphology of the ablation spots created by ns (a) and fs (b) laser pulses on relatively thick varnish layers are comparatively examined under the optical microscope. It is clear that compared to the ns ablation spots, the fs ones result into a sharply defined etch craters with clear edges and no indication of thermal modifications (melting). This is in agreement with the observations reported by Küper et al. [2] in PMMA and Teflon.

33 Femtosecond Laser Cleaning of Painted Artefacts

293

33.4 Conclusions This study aimed to explore any advantages which may be offered by using ultrafast pulsed lasers for the cleaning of sensitive painted artworks. To this end fs and ns laser irradiation of technical varnish samples (dammar and mastic) has been compared on the basis of (a) the spatial resolution, (b) the extent of the induced photochemical modifications and (c) the morphology of the irradiated surfaces. It was shown that fs pulses ensure higher etching resolution and edge precision without any melting phenomena. Product formation was proven to be much reduced in the fs regime, while it was shown that the amount of the product formation is largely independent of the absorptivity of the material to be removed (varnish). Thus ultrashort lasers can enable a highly precise treatment nearly independent of the optical properties of the substrate, opening new perspectives in the cleaning interventions of sensitive and particularly demanding painted artworks, such as modern paintings. Acknowledgement The authors would like to thank Dr. M. Doulgeridis, director of conservation in the National Gallery of Athens, for valuable discussions and suggestions.

References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13.

D. Bäuerle, in Laser Processing and Chemistry; Springer, Berlin, 2000. S. Küper and M. Stuke, Appl. Phys. Lett. 54, 4 (1988). J.K. Frisoli, Y. Hefetz, and T.F. Deutsch, Appl. Phys. B 52, 168 (1991). S. Küper and M. Stuke, Appl. Phys. B 44, 199 (1987). V. Zafiropulos and C. Fotakis, Ch. 6 in Laser Cleaning in Conservation: an Introduction, Ed. by M. Cooper, Butterworth Heinemann, Oxford, 79, 1998. V. Zafiropulos, Ch. 8 in Laser Cleaning, Ed. B. Luk’yanchuk, World Scientific, Singapore, 337, 2002. S. Georgiou, Adv. Polymer Sci. 168, 1 (2004). S. Georgiou, V. Zafiropulos, D. Anglos, C. Balas, V. Tornari, and C. Fotakis, Appl. Surf. Sci. 127, 738 (1998). M. Lassithiotaki, A. Athanassiou, D. Anglos, S. Georgiou, and C. Fotakis, Appl. Phys. A 69, 363 (1999). G. Bounos, A. Athanassiou, D. Anglos, S. Georgiou, and C. Fotakis, J. Phys. Chem. B 108, 7052 (2004). M. Dzvonik, S. Yang, and R. Bersohn, J. Chem. Phys. 61, 4408 (1974). J.B. Birks, in Photophysics of Aromatic Molecules; John Wiley & Sons: London, 232, 1970. P. Pouli, G. Bounos, S. Georgiou, and C. Fotakis, Appl. Phys. A (submitted )

34 Laser Cleaning of Polyurethane Foam: An Investigation Using Three Variants of Commercial PU Products ∗

U. Staal Dinesen1,2 and M. Westergaard2,3 1

2 ∗ 3

National Workshops for Arts and Crafts, Copenhagen, Denmark/Louisiana Museum of Modern Art, Humlebæk, Denmark School of Conservation, Royal Academy of Fine Arts, Copenhagen, Denmark [email protected]/[email protected] Art Conservation Centre, Kronborg Castle, Elsinore, Denmark

Summary. In this study, tests were undertaken to ascertain whether the laser could achieve a better level of cleaning on polyurethane foam than vacuum cleaning. Optimum laser parameters were found using statistics on data from color measurements. The laser proved to be very effective regarding the removal of dust, but also caused damage on some PU-variants. The laser cleaning has been carried out at National Workshops for Arts and Crafts, Copenhagen, Denmark.

34.1 Introduction With the introduction of polyurethane foam (PU-foam) as an artistic medium, a new challenge is posed in the field of conservation of modern art. The cleaning of PU-foam using traditional methods often causes damage or achieves an insufficient level of cleaning. Consequently, tests were undertaken to ascertain whether PU-foam in three variants can (1) be cleaned with an Nd:YAG laser (Phoenix 1200 from Lynton Lasers), and, if so, (2) whether the laser can achieve a better level of cleaning than vacuum cleaning. PU-foam is commercially produced in numerous variants – roughly speaking, in variants of stiff and flexible PU-foam. Generally, the physical structure of stiff PU-foam is closed cells, while the cells of flexible PU-foam are usually open. PU is generally formed by a reaction of the main components isocyanate and polyole. The polyoles are usually based on polyether or polyester (90% are based on polyether) [1]. The isocyanate, TDI (toluen-2, 4-diisocyanat), is usually used in flexible PU-foam. MDI (diphenylmethan-4, 4-diisocyanat) is the most commonly used isocyanate in stiff PU-foam [2]. In this study, three commercial PU-products were used: PU-F: Flexible PU-foam with open cells – based on TDI and polyether polyol

296

U.S. Dinesen and M. Westergaard

PU-S: Stiff PU-foam with closed cells – based on MDI and polyether polyol PU-SB: Stiff PU-foam (same as PU-S, but with the smooth surface cut off) Most PU-products are degraded at temperatures above 200◦ C. Yellowing and brown/black discoloration can be observed at higher temperatures (>240◦ C) [3]. Yellowing can be a result of the formation of quinon-imides or “photo-Fries rearrangement.” In addition, tests have shown that radiation ( 240◦ C). A reason that PU-F does not discolor to the same extent as PU-S and PU-SB could be that there are differences in the material absorptivity (the thermal conductivity of the PU-products are low – and roughly the same). This observation is supported by the fact that the stability of PU-products (concerning radiation) is related to the isocyanate; the stability of PU-products based on TDI (PU-F) is considered better than the stability of PU-products based on MDI (PU-S and PU-SB) [5]. With the ATR-FTIR analysis it was possible to detect that laser cleaning caused removal of dust (changes at 1,073 and 796 cm−1 – corresponding to Si–O compounds). The analysis, however, showed only few indications of

298

U.S. Dinesen and M. Westergaard Wavelength*Material - ∆E*ab, clean ref 20 15 10 5 0 PU-F

PU-S

PU-SB

532

1064

Fig. 34.1. The two-factor interaction, Wavelength∗ Material, with ∆E ∗ ab as response and 0 representing the clean reference, illustrate that the cleaning potential on PU-F is better using 532 nm. In contrast, 1,064 nm is better on PU-S and PU-SB. Comparing PU-S and PU-SB, it becomes apparent that it is more important to use 1,064 nm than 532 nm on PU-S np*Material - ∆E*ab, clean ref 20 15 10 5 0 PU-F

PU-S

PU-SB

np 3

np 30

Fig. 34.2. The two-factor interaction, np∗ Material, with ∆E ∗ ab as response and 0 representing the clean reference, illustrate that it is important to clean PU-SB and especially PU-S with few pulses per area. On PU-F, the number of pulses is less critical np*Energy density - ∆E*ab, clean ref 20 15 10 5 0 0,25

0,5

1 np 3 ∗

2 np 30

Fig. 34.3. The two-factor interaction, np Energy density with ∆E ∗ ab as response and 0 representing the clean reference, illustrate that the best cleaning potential was achieved at 1 Jcm−2 and three pulses per area

34 Laser Cleaning of Polyurethane Foam

299

Table 34.3. Optimum laser parameters for each PU-variant

PU-F PU-S PU-SB

wavelength (nm)

energy density (J cm−2 )

pulses per area (np )

532 1064 1064

2 1 1

12 3 5

Fig. 34.4. On all PU-variants, the laser cleaned area appears cleaner than the vacuum cleaned area. However, the laser cleaned area does not appear as “clean” as the clean reference

changes in the chemical structure of the PU-matrix, i.e. changes that could relate to the formation of the above-mentioned chromophore groups. With the SEM analysis, however, it became evident, that laser cleaning at given parameters, caused damage to the cell walls in two of the three PU-variants (Fig. 34.7). The damage could only be observed on the laser cleaned areas of the stiff PU-variants (closed cells); on PU-F, the open cell structure appears unaffected.

300

U.S. Dinesen and M. Westergaard Main interaction, Material, ∆E*ab, clean ref 20 15 10 5 0 Laser cleaning PU-F

Vacuum cleaning PU-S

PU-SB

Fig. 34.5. The main interaction, Material, with ∆E ∗ ab as response and 0 representing the clean reference, illustrate that laser cleaning achieves a better level of cleaning than vacuum cleaning on all PU-variants – especially on PU-F Main interaction, Material, ∆E*ab, soiled ref 20 15 10 5 0 Laser cleaning PU-F

Vacuum cleaning PU-S

PU-SB

Fig. 34.6. The main interaction, Material, with ∆E ∗ ab as response and 0 representing the soiled reference, illustrate that laser cleaning achieves a better level of cleaning than vacuum cleaning on all PU-variants – especially on PU-F

Fig. 34.7. With SEM analysis it became clear that laser cleaning of PU-variants with closed PU-cells (PU-S (left) and PU-SB (right) resulted in damage to the cells (hole in cell walls)

34 Laser Cleaning of Polyurethane Foam

301

34.4 Conclusions Consequently, laser cleaning has proven to be a very effective method for the removal of dust from PU-foam. However, laser cleaning at given parameters cannot be recommended on two of the three PU-variants as a result of the above-mentioned damage to the foam structure. Contrarily, laser cleaning at given parameters has proven to be very promising on one PU-variant (PU-F); on this PU-variant, laser cleaning did not cause damage to the foam structure and the yellowing was less significant than on the other two PU-variants. In addition, the experiments have shown that laser cleaning is much more effective than vacuum cleaning with regards to dust removal. Acknowledgments National Workshops for Arts and Crafts, Copenhagen and Director Ulf Horak for use of laser. Conservator Bent Eshøj, Lic.techn. Jan Jørn Hansen, Geologist Jørn Bredal-Jørgensen, Ph.D. Judith Jacobsen, Designer Lene Harbo Pedersen, Scott Green, Conservator Børge Igor Brandt.

References 1. J. Bützer and K. Kessler, in Kunststoff als Werkstoff: Celluloid und PolyurethanWeichschaum. 105, 2001. 2. G. Oertel, in Polyurethane Handbook, 2. ed. 17, 1993. 3. Y.Shieh, H. Chen, K. Liu, and Y. Twu, in Thermal Degradation of MDI-Based Segmented Polyurethanes. In: Journal of Polymer Science: Part A: Polymer Chemistry, Vol. 37, 4126, 1999. 4. J. F. Rabek, in Photodegradation of Polymers – Physical Characteristics and Applications, 82, 1996. 5. B. Rånby and J. F. Rabek in Photodegradation. Photo-oxidation and Photostabilization of Polymers, 241, 1975.

35 Excimer Laser Ablation of Egg Tempera Paints and Varnishes ∗

P.J. Morais , R. Bordalo, L. dos Santos, S.F. Marques, E. Salgueiredo, and H. Gouveia Instituto de Soldadura e Qualidade, Taguspark, Apartado 012, 2741-901 Porto Salvo, Portugal ∗ [email protected] Summary. In this work a series of egg tempera paint and varnish systems have been prepared, artificially aged and irradiated with KrF excimer laser at a wavelength of 248 nm. The samples were prepared with pure pigments and selected mixtures. It was found that, for some pigments, the colour changed upon laser irradiation even at low energy densities, below the ablation threshold while for other inorganic pigmented egg temperas the degree of discoloration is very small at moderate fluence of ∼ 0.30 J cm−2 . The varnish systems did not present signs of discoloration. The thickness, superficial roughness and magnitude of the colour changes of the samples were measured. X-ray diffraction, Raman spectroscopy and UV/visible spectroscopy were used in order to investigate the changes induced by the KrF excimer laser radiation.

35.1 Introduction Laser radiation has been used to remove unwanted layers from surfaces, not only in technological applications but also in items of cultural value. The first examples of lasers into art conservation were explored during the 1970s [1–3], although it was from the early 1990s that the interest and progress in laserbased techniques for the restoration of artworks increased [4–7]. There is a broad range of art materials that have been already investigated, being the stone-based materials extensively studied with the most successful results [8– 10]. The interaction of laser radiation with painting materials has been also a field of research due to its potential advantages over traditional methods [6,11– 14]. It has been proven that UV excimer lasers working at λ = 248 nm with proper selection of process parameters are suitable to remove polymerised varnishes or even over-paintings [4, 6, 11, 13–15]. Nevertheless, paint materials can undergo physical and chemical changes during laser irradiation such as discoloration phenomena in pigments [6, 16]. Thus, due to its nature each paint material reacts differently to the UV excimer laser radiation, which makes it necessary to evaluate the changes induced by the laser light in the

304

P.J. Morais et al.

broad variety of artist’s materials (varnishes, pigments and binding media), in order to provide a safe and efficient process that can improve and complement traditional conservation and restoration practices. The objective of this work was to investigate the effect of KrF excimer laser irradiation (λ = 248 nm) on a series of pigmented egg tempera paints (isolated and mixture), not all previously studied, to the best of the author’s knowledge. In addition, some varnishes were also under investigation. The prepared samples were partially characterised before and after laser ablation, in order to analyse the physical and chemical changes induced by the laser radiation.

35.2 Experimental Methods The presented experiments were carried out on two types of samples: inorganic pigments on egg tempera (as binding medium) and natural varnishes. Isolated and mixed pigmented egg tempera systems were prepared according to the procedure described by Van den Brink [17, 18], which is based on Cennini recipe [19]. In general terms, a homogenous mixture of powder pigment and egg is achieved and mastic resin was added to enhance the adherence to the support. Both types of systems were deposited on sheets of Melinex using a Neurtek film applicator capable of depositing 200 µm of wet film thickness. The varnishes were also deposited on microscope glass slides. The prepared samples were cured in a dark chamber during one month and artificially aged with fluorescent daylight lamps for another month, at a constant temperature and humidity. Inorganic pigments and varnishes of historical interest were selected: barium sulphate (BaSO4 ), yellow ochre (Fe2 O3 · H2 O), lapiz-lazuli (complex natrium-aluminium-silicate, containing sulfur), malachite (Cu(CO2 ) · Cu(OH)2 ), white lead (2Pb(CO3 )2 · Pb(OH)2 ), red lead (Pb3 O4 ), Prussian blue (Fe4 [Fe(CN)6 ]3 · 14H2 O) and cobalt green (CoO · ZnO). Representative varnishes were selected: shellac, dammar and mastic. The laser treatment consisted in the irradiation of the samples with a KrF excimer laser (Lambda Physic LPX 240i) emitting 248 nm wavelength with pulses of 14 ns of duration and 160 mJ of energy. The samples were mounted onto a computer-controlled motorized X–Y translation stage. The final energy density–fluence (in J cm−2 ) was determined by dividing the incident energy by the irradiated area. The distance from the lens to the sample determined the irradiated area. In order to evaluate the morphological changes, the sample’s roughness and thickness were measured by a mechanical profilometer (Tencor Alpha-Step 200). Scanning electron microscopy (Hitachi S4100) was used to obtain photographs of the egg tempera sample’s surface. Their colour value was measured using the CIELAB system procedure [20]. The crystal phases of the pigmented samples were evaluated by means of X-ray diffraction (XRD) measurements (Rigaku Geigerflex D/max) with CuKα radiation. Some

35 Excimer Laser Ablation of Egg Tempera Paints and Varnishes

305

of the samples were also analysed by Fourier transform Raman spectroscopy (Bruker RFS 100/S, Nd:YAG 1,064 nm excitation). UV/visible spectroscopy (Shimadzu 3101) was used in the varnished samples in order to evaluate their transparency in the visible region. The laser treatments consisted in the irradiation of the samples with several different energy densities. Beginning with a low-energy density, below the ablation threshold, it was stepwise increased to a highest value, well above the ablation threshold. In this way, even when no information was available about the ablation threshold of a specific material, it was assured that irradiated areas below and above the threshold were present. The tests were stopped each time a colour change was detectable or when a colour change was detected and did not suffer any further alteration after successive pulses or even when no physical alteration was observed in the samples.

35.3 Results and Discussion Table 35.1 summarises the laser treatment conditions used on the test systems and the calculated fluences based on the distance between the lens and the samples. After laser treatment it was observed that lead white, yellow ochre, malachite and red lead changed their colour even at low energy densities. The chromatic analysis by colorimetry indicates ∆E (magnitude of the colour change) values of 8.0 and 7.8 for yellow ochre and malachite at low energy densities (