научная статья по теме CONTROLLED REMOVAL OF OVERPAINTING AND PAINTING LAYERS UNDER THE ACTION OF UV LASER RADIATION Физика

Текст научной статьи на тему «CONTROLLED REMOVAL OF OVERPAINTING AND PAINTING LAYERS UNDER THE ACTION OF UV LASER RADIATION»

ОПТИКА И СПЕКТРОСКОПИЯ, 2011, том 111, № 2, с. 319-324

= ЛАЗЕРЫ ДЛЯ СОХРАНЕНИЯ ПРОИЗВЕДЕНИЙ ИСКУССТВА

УДК 535.212

CONTROLLED REMOVAL OF OVERPAINTING AND PAINTING LAYERS UNDER THE ACTION OF UV LASER RADIATION

© 2011 I. Apostol*, V. Damian*, F. Garoi*, I. Iordache*, M. Bojan*, D. Apostol*, A. Armaselu**, P. J. Morais***, D. Postolache****, and I. Darida

* National Institute for Laser, Plasma and Radiation Physics, 077125 Magurele, Jud. Ilfov, Romania,

e-mail: ileana.apostol@inflpr.ro ** Transilvania University of Brasov, Department of Physics, Brasov, Romania *** Instituto de Soldadura e Qualidade, Taguspark, Oeiras, 2740-0120Porto Salvo, Portugal **** DANARTSrl, Bucuresti, 012062, Romania Received January 21, 2011

Abstract—Laser material removal applied to selective overpaintings and subsequent painting layers detachment was studied in order to select the best cleaning practice of painted artworks. The ablation depth as a function of incident laser fluence/intensity and irradiation pulse number was considered as a reference parameter. We have measured the ablation depth with both a contact microprofilometer and a white light interferometer as a function of laser irradiation parameters. The measurements have evidenced that the ablation depth in our experiments varied between 2 and 100 ^m making possible selective removal of painting layers.

INTRODUCTION

Laser material removal from the surface of different objects with direct application in surface laser cleaning is already demonstrated [1, 2]. Applications in specific cases have been reported as well [3]. Generally, the main parameters discussed to establish the working range for laser cleaning of different materials are: the ablation threshold (the lowest value of the incident fluence/intensity, to remove a significant quantity of material from a surface) and the substrate damage threshold (the highest value for which the material is completely removed without the damage of the substrate). These are the limiting values for a working range which is recommended to be known or determined before the laser cleaning process begins. Another parameter which influences the laser cleaning process is the number of subsequent laser pulses which are incident at the same place. In case of laser irradiation pulses with incident fluence below the ablation threshold the pigments discoloration in the irradiation region can be an important effect [4, 5].

Therefore the scale of the phenomena appearing in the region of pulsed laser radiation interaction with painting materials depends on irradiation parameters and goes from pigments discoloration, surface vaporisation and material removal to substrate damage.

To use the best cleaning practice, irradiation parameters selection is the key of success in the process of controlled and selective material removal from the surface. In this study we have tried to present in an evident manner the importance of the irradiation parameters selection in case of laser removal of painting layers.

We have selected to measure the ablation depth as an important parameter characterizing the laser re-

moval of the material from a surface. In case of laser cleaning of a real art object the ablation depth is difficult to be measured in situ, so it is important for the restorer to have a very clear representation of the effects induced under the action of laser radiation.

Generally ablation depth and the related ablation rate was studied for varnish removal as a method to demonstrate the selectivity of laser removal of thin layers from surfaces and also the ablation efficiency [6, 7].

In this work we have analyzed the removal of subsequent different painting layers as a function of the irradiation parameters in order to evidence the importance to select the proper irradiation conditions.

EXPERIMENTAL SET-UP

Laser cleaning experiments were realized on artificial samples (e.g. mock-ups) developed in order to simulate as better as possible different painting techniques and conditions. Mock-ups preparation respected the specificity of easel paintings as substrates, materials, and method of preparation, following traditional recipes. The support for the easel paintings was wood or canvas. In order to simulate old paintings the prepared samples were dried and artificially aged for a period of eight months in controlled conditions of temperature, humidity and especially of illumination.

Experiments (Fig. 1) were realized in conditions that respect the main requirements for a controlled and reproducible removal of painting layers: selection and control of incident laser fluence and intensity, subsequent irradiation pulses number or irradiation time, etc. In order to have a uniform transversal distribution of the laser fluence in the irradiation area an

Fig. 1. General view of the laboratory setup developed for laser cleaning experiments.

Fluence 90 mJ/cm2

120 pulses 250 pulses

200 mJ/cm2

250 pulses 120 pulses

280 mJ/cm2

120 pulses 250 pulses

350 mJ/cm2

120 pulses 250 pulses

400 mJ/cm2

250 pulses 120 pulses

450 mJ/cm2

120 pulses 250 pulses

Fig. 2. Laser irradiation effects induced during cleaning tests on a mock-up with canvas support and the following painting layers: I. Ground-mixture of powder of chalk; II. Plaster + ochre oxide; III. Pigment-Ultramarine; IV. Ultramarine + titan white; V. Varnish.

optical system was realized to deliver the laser beam on the mock-ups. Each mock-up sample was able to be translated in front of the laser spot to make irradiations in adjacent positions or in a continuous movement with a controlled translation speed. The irradiation laser system was an Nd: YAG laser working on 1064, 532, 355 or 266 nm, with pulse length of 6 ns for 1064 nm and a frequency of 10 Hz.

Custom software was developed to integrate all the command and control systems and was implemented in the experimental setup to ensure a centralized command. In this way laser energy (and consequently the incident fluence/intensity), repetition rate, number of subsequent irradiation pulses or irradiation time and also the movement of the mock-up in front of the laser beam as step size or displacement speed (e.g. in the case of the continuous movement of the sample in front of the irradiation spot) were selected and commanded in a controlled manner. The irradiated area was monitored with a CCD camera and the image displayed on a monitor in real time.

The irradiated area was analyzed (ex situ) from the point of view of the ablated surface morphology and the thickness of the removed layer with optical microscopy, white light interferometry (WLI) and contact profilometry.

ABLATION DEPTH ANALYSES

Experiments were done on a great number of mock-ups. We will discuss here only some representative samples from the point of view of painting structure and laser radiation effects on them.

In Fig. 2 a study of laser radiation effects on a mock-up with several layers of pigments in a binding medium is presented. Laser beam wavelength was 355 nm and pulse length of 6 ns. At low incident laser fluence even for a rather great number of subsequent irradiation pulses (250 pulses) the induced effect in the irradiated area is a slight discoloration. For higher incident fluence the surface layers (layer V—varnish and layer IV-ultramarine and titan white pigments) are removed up to a complete removal and the dark blue layer appears free of covering layers. But we can see that if the incident fluence is higher than 300 mJ/cm2 the boiling of the painting layers appears with the consequence of their mixing. In this case, for a high incident fluence, it is no more possible to remove in a selective manner the subsequent painting layers.

Generally the first layer to be removed from a painting is the old varnish layer or dirt covering layers, like soot (due to candle smoke or fireplaces), bees-wax (used in restoration to brighten or to protect paintings), overpaintings oil, tempera or casein based, but also additional varnish.

In Fig. 3. we present a mock-up covered with soot fixed with a layer of bees-wax diluted in oil essence and varnish, artificially aged (treated to light and moderate

CONTROLLED REMOVAL OF OVERPAINTING

321

heat for six month in a special chamber). For a laser radiation of 355 nm wavelength and 6 ns pulse length and an incident fluence of 90 mJ/cm2 the covering layer of 5 ^m was removed under 1 irradiation pulse. If the incident laser fluence was the same but the number of the subsequent irradiation pulses is higher the ablation depth is higher, up to the complete removal of the dirt layer, but after its complete removal also the underlay-ers can be removed.

For a fixed incident laser fluence/intensity as a function of the subsequent irradiation pulses number painting layers removal evolution is presented in Fig. 4. The thickness of the removed layer was from 5 ^m for 1 irradiation pulse to 20 ^m for 20 pulses. The mock-up structure, in the section r2 of the mock-up, were the laser ablation effects are presented, was constituted of:

— Support: wood,

— Ground: powder of chalk with finings in fish glue + (white layer)

— Painting layers: ultramarine blue pigment (i) covered with a layer of intermediate tone of ultramarine with ochre and white and on half sample tempera all in yellow egg emulsion (1/4) in two layers: one consistent matter and another transparent (ii),

— Covering layer: Fine yellow pigment — binder: oil + dammar varnish 11% (appears green), approximately 5 ^m thick.

Removal of the covering layer considered as a dirt layer was realized under the action of an incident laser fluence (wavelength 35 nm, pulse length 6 ns) of 90 mJ/cm2 in one pulse, with the light blue layer remaining clean (layer ii). For the same incident fluence but more (subsequent) laser pulses the light blue paint layer is gradually removed and the ultramarine blue layer appears (layer i). Removal of the covering layer begins at 60 mJ/cm2, value conside

Для дальнейшего прочтения статьи необходимо приобрести полный текст. Статьи высылаются в формате PDF на указанную при оплате почту. Время доставки составляет менее 10 минут. Стоимость одной статьи — 150 рублей.

Показать целиком