|
Chapter 2.4 Laser Cleaning of Paper and Other Organic
Materials Wolfgang Kautek Department of Physical Chemistry, University of
Vienna, Währingerstraße 42, 1090 Vienna, Austria Contents 2.4.1
Introduction Pulsed lasers are becoming new tools in the hands of restorers [1-3]. Most activities, however, were concerned with the laser cleaning of stone artefacts, wall paintings, and facades. There also exist fundamental knowledge and experience on the cleaning of technical surfaces relevant e.g. in the electronic industry [4,5] or the high-precision removal of inorganic and organic films e.g. in biosensoric technology [6]. The contactless laser cleaning of biogenetic surfaces such as parchment and paper, on the other hand, has been approached only in recent years [7-25]. Traditional dry as well as aqueous cleaning methods for paper include mechanical scratching and the use of a brush, eraser or draft clean powder, i.e. a granulated eraser (Fig. 2.4.1). Paper may also be cleaned with water, organic solvents, enzymes etc. [26,27]. Satisfactory results are often obtained by a treatment using cellulose ethers like carboxymethyl cellulose (CMC), methyl cellulose (MC) or hydroxylpropyl 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.
Laser
cleaning of parchment to some extent parallels dermatological laser
applications that are regarded as a multi billion-dollar market (Fig.
2.4.2) [28,29]. There, selective photothermolysis of pigmented
subsurface structures, such as melanin particles, enlarged blood
vessels, and tattoo ink particles, or char-free vaporization of skin
takes place. Then the skin’s natural physiological mechanisms
break down and remove the laser-altered remnants. In parchment
cleaning, however, natural resorption of laser-altered remnants is
absent, and contaminants have to be removed completely without any
irreversible morphological and chemical conversions of the substrate
(Fig. 2.4.3).
The
cleaning objects paper and parchment belong to the chemically most
fragile substrates exposed to high-power laser radiation. The main
constituent of paper is cellulose (Fig. 2.4.4). It is a linear polymer
of The
main constituent of parchment is collagen, which forms long ropes and
tough sheets (Fig. 2.4.5). All contain a long stretch of triple helix
connected to different types of ends. The simplest is merely a long
triple helix, with blunt ends. These “type I”
collagen
molecules associate side-by-side like fibres in a rope to form tough
fibrils. Fig. 2.4.5 depicts a basement membrane, which forms a tough
surface that supports e.g. skin or parchment. A different
collagen–“type IV”–forms an
X-shaped complex
supporting extended networks. Two other molecules
–cross-shaped
laminin and long, snaky proteoglycans- fill in the spaces, forming a
dense sheet.
Laser beam delivery has been realized commonly either via an optical fibre or an articulated optical arm to a hand-held output optics common in facade laser cleaning. In the beginning of laser cleaning of parchment and paper excimer lasers radiating at 308 nm were employed (Fig. 2.4.6) [7-12]. The beam with a rectangulous cross section was delivered by several mirrors and a final lens onto the substrate. This itself was scanned with a computer-driven x-y stage. An evaluation of this near-UV radiation versus visible radiation of a solid state Nd:YAG laser at 532 nm indicated that cellulose and paper degraded after UV-laser treatment as well as after a period of accelerated humid oven ageing due to depolymerisation [13]. No detrimental effects of a Nd:YAG laser treatment at 532 nm was observed. Cellulose is more or less transparent in the visible region, however strongly absorbs UV radiation, e.g. the C-C-bond at 347 nm, the C-O-bond at 333 nm, and the C-H-bond at 289 nm. Photolysis may occur as a result of photon absorption which in turn leads to severe immediate depolymerisation. However, it is more likely that the absorption of a near UV photon results in excitation of electrons in a chemical bond, facilitating photo-oxidation reactions.
With this knowledge, a computerized laser cleaning system suitable for high-precision cleaning of flat large area substrates was developed (Fig. 2.4.7) [23]. It allowed the restoration of artefacts of organic materials, such as paper, parchment, leather, textiles, wood, but also inorganic materials, such as metals, alloys, and ceramics. It was based on a compact high pulse energy diode pumped Q-switched Nd:YAG laser operating at 1064 nm and 532 nm, operating with a pulse duration of 8 ns and a repetition rate of up to 1000 Hz. The laser-processing compartment with an integrated exhaust system provided Laser Class I conditions, so that the operator did not require safety goggles. Objects could be scanned supported by a remote computer control system. The operator followed the process on the computer screen through a camera system. The scanning could be controlled manually or programmed in the computer, defining the pulse energy, number of pulses, laser spot overlap and shape of the area to be treated. As an alternative, an optical fibre with an ergonomic hand piece was used for manual cleaning of objects under Laser Class IV conditions requiring eye protection. The workstation featured on-line visible, ultraviolet and fluorescence imaging for the identification and documentation of visible and chemical changes of the irradiated substrate areas.
The
laser spot is scanned over the objects through a remote computer
control system (Fig. 2.4.8). The operator followed the process on the
computer screen through a camera system and controlled it manually
through keyboard, mouse and/or foot pedal operation or by automatic
laser beam scanning operation. As an alternative, an optical fibre with
an ergonomic hand piece could be used for manual cleaning of 3D
objects.
Lightness changes ΔL quantified by this relative technique could
be correlated with the cleaning status. The overall colour
difference ΔE included also the colour changes for which
the human eye is particularly sensitive.
One of the major challenges of precision cleaning is to avoid areas of ink, printed letters or pigments. A section of an old office document on rag paper was contaminated with pencil scratching almost to unreadability (Fig. 2.4.9). The area to be laser-cleaned was drawn on the computer screen by the mouse movement as a customized lithographic mask. Then, the system automatically scanned the intended area.
One of the major challenges of precision cleaning is to avoid areas of ink, printed letters or pigments. A section of an old office document on rag paper was contaminated with pencil scratching almost to unreadability (Fig. 2.4.9). The area to be laser-cleaned was drawn on the computer screen by the mouse movement as a customized lithographic mask. Then, the system automatically scanned the intended area.
The cleaning system also proved successful on original, such as a sheet of music of a psaltery from 15th century, which was heavily soiled during a bombing attack at very end of Second World War (Fig. 2.4.10). There, hand-written ink letters, notes, and lines of the staff often have to be preserved from the converting action of the laser beam. Again, the cleaning result demonstrates the feasibility of this ultra-precise "contactless rubber". Cotton paper e.g. was treated with increasing fluence on the patches numbered with 1, 2, and 3 (Fig. 2.4.11). The visible image showed no alteration below the ablation threshold fluence above 1.4 J cm-2. At a drastically higher value, at 5.0 J cm-2, the visible image only indicated a faint darkening. Invisible diagnostic tools, however, allow a more sensitive judgement on the degradation status. The reduction of the IR and UV reflectivity indicate an increased absorption in IR and UV due to chemical changes. This UV absorption increase is accompanied by a drastic fluorescence increase.
The ablation threshold fluences Fth for fresh and pre-aged papers with all laser wavelengths have been determined [20]. The pre-aged papers are more sensitive against laser radiation and therefore, their ablation threshold fluences are lower. Especially for the infrared laser, all papers show in the aged stadium a decrease of resistance against laser radiation. The colorimetric measurements were performed at bleached sulphite softwood cellulose paper with no fillers and no sizing. A comparison of colour measurements on fresh to pre-aged paper for all three laser wavelengths is shown (Fig. 2.4.12). The laser fluences and pulse numbers were varied. Differences of ΔE* > 1 are visible by the naked eye. Treatments with the green and ultraviolet laser resulted in similar colorimetric changes. These were more pronounced after laser irradiation for pre-aged samples. This observation corresponds with the threshold fluence behaviour where normally the threshold of fresh paper is slightly higher than for the pre-aged samples. Size exclusion chromatography (SEC) was used to characterize the molecular mass distribution of cellulose in laser treated areas and untreated background (Fig. 2.4.13). The macromolecules are separated according to their hydrodynamic volume in solution by driving the solution through a porous, three-dimensional network in a chromatography column. From the chromatogram average molecular masses, and the corresponding average DP values, can be determined with good precision and weight-average molecular masses (Mw) can be reported. Cellulose is not soluble in organic solvents typically used as mobile phases in SEC. Cellulose was chemically modified prior to analysis by reaction with phenyl isocyanate (PIC). The resulting soluble cellulose tricarbanilate (CTC) derivative showed enhanced UV detection properties. A comparison between the background and the lasered matrix areas showed that the ultraviolet laser radiation resulted in a significant decrease in the Mw value for higher fluences and higher number of pulses. Whereas the green and the infrared laser light interaction caused no Mw effect. Chemiluminescence
experiments (CL) were carried out in a photon counting instrument. The
CL of the paper samples was recorded during dynamic experiments
performed in nitrogen. Paper discs were conditioned for 10 minutes at
40 ºC, and then heated up to 200 ºC.
A comparison of fresh bleached sulphite softwood cellulose paper and additive-free cotton linters cellulose paper (no fillers, no sizing) to post-aged samples for the laser treatment with the green wavelength showed no significant differences between the CL responses of treated and untreated areas. The bleached sulphite softwood cellulose paper P1 showed the typical CL peroxide decomposition peak centred at 125 ºC. Its position and intensity is not altered by laser treatment and/or artificial ageing. Contrary to the results obtained with the green laser, significant differences between the CL responses of all papers were observed with UV radiation (Fig. 2.4.14). Laser treatment causes the CL intensity of the fresh paper to increase in the temperature interval of 40–125 ºC. Artificial ageing considerably reduced the differences between the CL response of lasered and non-lasered areas. These results suggest the generation of emitting species as a result of UV laser treatment, which react further into non-emitting products in the course of artificial ageing.
More and more, destructionless processing based on off-line and on-line diagnostics based on systematic multi-method approaches become important. On the one hand, Scanning Electron Microscopy (SEM) and Transmission Electron Microscopy (TEM) play a role in observing the dimension of morphological effects on originally intact collagen fibres; on the other hand, Diffuse Reflectance Infrared Fourier Transform Spectroscopy (DRIFT), Pyrolysis Capillary Gaschromatography (Py-CGC), Electron Spin Resonance (ESR), Electron Spectroscopy for Chemical Analysis (ESCA), Energy-dispersive X-Ray Fluorescence (EDX), the determination of the hydrothermal stability by the micro-hot-table technique as well as the analysis of hydrolysis products allow to establish an in-depth chemical diagnosis of laser-induced chemical changes and degradation processes in dependence on laser fluence and wavelength [25]. The sample preparation for the photometric determination of the water-soluble degradation products of collagen in laser-treated parchment required only 10-25 mg of parchment. 1 ml of the extract was mixed with chloramine-T solution. Photometry of the hydroxyproline colour complex was done versus a standard of distilled water. At the visible (532 nm) and the infrared wavelengths (1064 nm), the photochemical phenomena were absent, and the only result was a decrease of the solubility due to thermally induced crosslinking of the collagen fibres. Both photochemical (308 nm) and photothermal (532 nm, 1064 nm) alterations occurred already at fluences below the respective ablation thresholds as indicated in Tab. 2.4.1 by horizontal dashed lines.
Tab. 2.4.1: Photometric determination of the water-soluble degradation products of collagen.
Scanning electron microscopy (SEM) detected massive phase changes like melting and boiling that accompanied the ablation (vaporization) of the material above Fth (Fig. 2.4.15). However, no changes were observed at F < Fth, where the water-soluble degradation products of collagen already were affected by the laser treatment and exhibited changes on the molecular level. The hydrothermal stability by the micro-hot-table technique relies on heating collagen above the helix-coil transition temperature, which causes a collapse of the rod-like three-stranded collagen unit into random coils, which then constitute gelatine [25]. When collagen hide fibres are heated in water they will deform and this deformation was seen as shrinkage of the fibres (Fig. 2.4.16). The hydrothermal stability of the fibres decreased in proportion to the state of deterioration. The measurement of shrinkage temperature in the so-called micro-hot-table technique demands small samples of parchment fibres (~0.3 mg), which were heated in water on a microscope slide positioned on a microscope hot table [25]. During heating the fibres started shrinking at the shrinking temperature (Ts), which depends on the stability of the collagen.
The quantity of water-soluble degradation products of collagen is affected by laser radiation already below the ablation threshold fluence, Fth, at which material removal is detected under microscopic inspection (Tab. 2.4.2). The solubility of the modern parchment increased with 308 nm at low fluences (0.3 Jcm-2) possibly due to photochemical degradation (Tab. 2.4.1). The ancient specimen already showed increased solubility at even lower fluence (0.1 Jcm-2), but exhibited a decreased value at 0.3 Jcm-2, where the modern sample still showed molecular degradation. Obviously, the decrease of soluble products indicates cross-linking or other analogous phase changes.
Tab. 2.4.2: Ablation threshold fluences.
This technique proved sensitive to laser irradiation that caused photochemical reactions as e.g. collagen polymer degradation (Tab. 2.4.3; compare to Tab. 2.4.1, 308 nm). The shrinking temperature, Ts, was only decreased under UV laser treatment, also at F < Fth. Wavelengths which caused mainly thermal reactions (e.g. cross-linking, melting etc.) did not affect Ts. even though some laser fluences were chosen above the ablation thresholds.
Tab. 2.4.3: Hydrothermal stability of parchment by the micro-hot-table technique
The photochemical deterioration at 308 nm observed by the increase of soluble products (Tab. 2.4.1) and the decrease of Ts (Tab. 2.4.3) may be related the oxidation of collagen. Both oxidative and acidic deterioration is reflected in the measurements of the increase of soluble products and the shrinkage temperature. These methods are therefore valuable indicators for the total degree of deterioration of a parchment sample. Recently, shrinkage temperatures determined by micro-thermo-mechanical analysis was correlated with the amino acid residue composition of parchment, in particular proline and hydroxyproline [30]. Oxidative
breakdown processes of parchment base on heat and light [31]. Parchment
starts degrading at tripeptides in clusters of charged amino acids
following the pattern: R-C*H2-NH2 → +[ OH] → R-CHO + NH3 → +[O2, H2O] → R-COOH with R as the rest of the Lysine side chain, -(CH2)3-, including the peptide main chain. Polar groups, particularly carboxylic acid functions (besides some conjugated double bonds) are then formed. Biodeterioration of organic cultural heritage materials is a common problem. Particularly the removal of discoloration caused by fungal pigments is yet an unsolved problem in paper conservation. In the following example, cellulose (cotton and linters) and 16th century paper (rag), were incubated with several fungi types, such as Cladosporium, Epicoccum, Alternaria, Chaetomium, Aspergillus, Trichophyton, and Penicillium on agar for three weeks. Then they were immersed in 70% Ethanol for removal of hyphae and mycelia and deactivation of the remaining conidia. These specimens were laser-treated with a Q-switched Nd:YAG laser operating at 532 nm and a pulse duration of 8 ns. Colour differences were determined spectrophotometrically. Best cleaning results were observed with fungi such as Penicillium and Alternaria. Dry laser cleaning generally turned out to be superb over wet bleaching approaches. Biodeterioration of organic cultural heritage materials is a common problem [33, 34]. While the conventional sanitizing techniques for removal of fungal material from paper using chemical or physical means have proved sufficient in many cases [34], the removal of discoloration caused by fungal pigments is yet a problem in paper conservation. There are very rare reports on the removal of fungi from paper by lasers [35, 36]. A far-UV krypton fluoride (KrF) excimer laser was able to remove Aspergillus niger mould from filter paper while viable spores and mould fragments were released into the atmosphere [37]. Actually UVB radiation of a 308 nm XeCl excimer laser can be used for the treatment of skin mycosis fungoides [38]. That means that UV laser radiation can deteriorate fungi. However, irradiation of cellulose with even a near-UV excimer laser at 308 nm resulted in photo-oxidative degradation of the paper substrate, accompanied by an increase in oxidized groups content (carbonyl or carboxyl) and a severe decrease in degree of polymerisation [13,14]. That means that successes of anti-fungal laser treatments with UV lasers are accompanied by photochemical paper deterioration. There are several types of fungal damage on
paper, such as Conventional anti-fungal treatment consists of Laser irradiation after the ethanol leaching resulted in drastic improvements of the specimens' appearance to the naked eye in respect to discoloration with all fungi types except Cladosporium and Aspergillus (Fig. 2.4.17). Alternaria caused substantial problems with mechanical and chemical treatments but yielded good results by laser treatment. This grows in surface-near regions and therefore can be removed without paper ablation. In contrast, Chaetomium grows deeply in the paper matrix with its hair-like aspendices. Therefore, conventional approaches were completely unsuccessful, whereas the laser could yield an acceptable cleaning affect. Epicoccum exhibits leaching remnants deeply distributed in the paper matrix with various colours. Bleaching resulted in relatively good in-depth discoloration. The laser could remove coloured material near the surface, and deep contaminant regions were left over causing a brownish colour. Penicillium material was converted into almost black material by the alcohol treatment. The laser was very successful at least in the surface-near regions (Fig. 2.4.17). Cladosporium grows deeply in the fibre material forming dark leaching products that could not be affected by neither mechanical nor chemical treatment. The laser yielded a limited success near the surface.
Multispectral imaging allowed documenting laser-cleaning results in comparison to cut paper sections, which have undergone bleaching in permanganate instead. Laser cleaning examples of Penicillium on cotton paper are represented in Fig. 2.4.18. This fungus shows that remnants after the removal of hyphae and mycelia and the deactivation of the remaining conidia by ethanol could be removed by the laser at least in surface-near regions. The ablation threshold fluences determined by microscopic inspection were at least Fth ~1.4 Jcm-2 for the pure cotton paper, and at least Fth ~1.0 Jcm-2 for the rag paper. The applied laser fluences ranged below the ablation thresholds and above. In the latter case, the laser was used as a "contactless scalpel". An appreciable deep cleaning action is documented by the IR reflectivity images (Fig. 2.4.18). The laser turned out to be more efficient than the aggressive bleaching process. The fluorescence images may be used to identify possible irreversible material changes. The spectral imaging system was further used for semiquantitative colour change measurements according to the CIE L*a*b* formalism. This allowed an even more detailed analysis. This approach was used to identify irreversible changes of the paper types caused by the chemical treatment, alcohol vs. KMnO4. Both treatments cause conceivable colour changes expressed by ΔE ~ 5 - 9. The cotton darkened it both cases (ΔL > 0) whereas the old rag underwent a slight cleaning. The change of the entire fluorescence spectrum could be expressed by a ΔL evaluation. Interestingly, the pure cotton type showed a drastic increase of fluorescence in both chemically very different liquid treatments suggesting irreversible chemical alterations. A high negative ΔL indicates a strong relative cleaning effect. Clearly, the laser showed the higher efficiency versus the chemical bleaching. The laser exhibits the most negative ΔL values when used on Cladosporium, Alternaria and Penicillium, which species show the dark material remnants. On Epicoccum, Chaetomium and Trychophyton the laser is less efficient but still comparable to the chemical bleaching process.
A further criterion for the laser cleaning result is the comparison of L of the infected areas after laser irradiation with pure and uninfected samples ("Laser vs. pure"). The best result would be the resemblance of the cleaned surface with the original pure surface ("Success"), i.e. ΔL ~ 0. In this respect, the substrates exhibit different behaviours. On cotton, the best success was observed again with the dark types Alternaria and Penicillium, but not Cladosporium. Even Chaetomium and Trichophyton could be removed so that ΔL almost resembled that of the pure cotton paper. On rag, the darkest species Penicillium shows the best cleaning success followed by Cladosporium and Alternaria. Soiled biogenetic fibrous substrates such as paper and parchment consist of a phase mixture of particulates, condensed foreign phases and cellulose or protein polymers, respectively. Particle decontamination on flat solids such as silicon wafers has already been studied systematically [5,40-43]. The phase separation of these condensed films and particles from the matrix is the purpose of laser cleaning. Model experiments with microspheres on silk fibres allowed a first insight into the fundamental phase separation processes on fibrous substrates (Figs. 2.4.19 and 2.4.20) [20].
Pulsed lasers can be promising new tools in the hands of restorers. Know-how and knowledge about the contactless laser cleaning of biogenetic surfaces such as parchment and paper has been reviewed. Current aqueous cleaning methods are not always sufficient, in particular if the paper is coated with sensitive printing as well as writing media. Colour changes after laser irradiation were most pronounced with pre-aged samples than for fresh ones. The weight-average molecular mass of cellulose significantly decreases towards stronger laser interaction with ultraviolet wavelengths. The effect is in the same order of magnitude as that of the artificial pre-ageing. Treatment with visible (532 nm) and infrared laser radiation (1064 nm) is negligible and gave the most promising results, with no discolouration and no other visible alteration, nor detectable chemical changes. UV laser light causes photochemical degradation i.e. molecular degradation also on parchment. Photometric determination of the water-soluble degradation products of collagen is a sensitive method to detect laser-induced alterations of parchment. These occur already below the ablation threshold fluence and are indicative for changes on the molecular level. Thermally induced crosslinking of the collagen fibres is increasingly observed with increasing fluence if IR radiation. The hydrothermal stability and shrinking temperature measurements by the micro-hot-table technique rendered itself sensitive to laser irradiation that caused photochemical reactions (i.e. UV). Scanning electron microscopy (SEM) is only sensitive to massive phase changes like melting and boiling accompanying ablation (vaporization) above the threshold fluence. The photochemical deterioration in the UV observed by the increase of soluble products and the decrease of the shrinking temperature may be related to oxidative breakdown processes, which occur only in the presence of light. Laser
cleaning of fungi-overgrown cotton and rag paper sanitized by ethanol
showed success in contrast to chemical bleaching by KMnO4.
Remnants after the removal of hyphae and mycelia and the deactivation
of the remaining conidia by ethanol could be removed by the laser at
least in surface-near regions.
The
EUREKA project "Laser Cleaning of Paper and Parchment (LACLEPA)" !
1681 served as an umbrella for part of this research. The author
acknowledges partial financial support by the EU TMR project "Modelling
and Diagnostic of Pulsed Laser-Solid Interaction: Applications to Laser
Cleaning", No. FMRX-CT98-0188, and the EU CRAFT project "Paper
Restoration using Laser Technology", EVK4-CT-2000-30002.
[1]
W. Kautek, E. König (Eds.):
Lasers in the
Conservation of Artworks I, Restauratorenblätter (Special
Issue),
Verlag Mayer & Comp., Wien, 1997.
There are no dedicated providers for organics
yet. However, a first approach to manufacturers of cleaning lasers may
be helpful:
Department of Physical Chemistry |
-(1→4)-D-glucopyranose
units in 4C1
conformation with two coexisting phases, cellulose I
(triclinic) and cellulose I
