Table of results Part II of II

Transcription

1 Project reporting in FP6 Project no CONSTGLASS Conservation materials for stained glass windows assessment of treatments, studies on reversibility and performance of innovative restoration strategies and products Specific Targeted Research Project FP 6 Thematic Priority 8.1: Policy-oriented research Part II of II

2 Part I: Bourges, Cathédrale St. Etienne (France) Chartres, Cathédrale Notre-Dame (France) Le Mans, Cathédrale St. Julien (France) Cologne Cathedral (Germany) Klausen, Parish and Pilgrimage church St. Maria Himmelfahrt (Germany) Hannover, Marktkirche St. Georgii et Jacobi (Germany) Part II: Canterbury Cathedral (England) The Burrell Collection, Glasgow (Scotland) Burgdorf (Canton Bern, Switzerland) Academy of Fines Arts, Krakow (Poland) Various fresh materials Anti-reflex-glass

3 Pilot object: Picture Canterbury Cathedral SXXVIII D8 1-Pilot object Identification of the panel: SXXVIII D8 Treatment Product: Polymer coating, possibly Viacryl Application: Probably applied with a brush after the panel was leaded. sample Table_Results_CAN_ S XXVIII D8 Page 1 of 6

4 2-Results sample reference: CAN SXXVIII D8 Questions Techniques Answers Morphology What is the physical appearance of the coating? Is the coating stable? On both areas the remaining coating was on the rougher textured area of glass. Has the coating come off the smooth glass surface? Optical Microscope SEM The coating was delaminating on the test areas. Unstable surface, non uniform covering. Deterioration of the film: Micro cracks and detachment. Loss of transparency. A higher magnification indicates the coating has not taken off the gel layer. n/a Desktop tomography Phase-contrast tomography on Synchrotron n/a n/a CAN D8 Fragment sent for analysis to LHRM CAN D8 Area where fragment was removed for analysis Chemical Composition Organic component composition What is the chemical composition of the coating? Is it the same coating applied to CAN C1? Is it Viacryl? What is the ratio of Viacryl and Desmodur? Has the coating degraded or changed in any way? Microbiology Is there microbial activity? SEM/EDS FTIR (LRMH) RAMAN Molecular biology ATP measurements The composition does not correspond with LRMH reference spectra. Viacryl is the closest. n/a No microbial activity was visible under optical microscope Table_Results_CAN_ S XXVIII D8 Page 2 of 6

5 Reversibility (A) How can we remove the coating without damage to the paint layer? What method and solvent can we use? Is the coating stable? How are the solvents reacting with the glass corrosion, putty and lead? How do we remove all trace of the solvent? Test studies Elimination Area selected for reversibility test on panel SXXVIII D8 D8 Coating D8 Coating under optical microscope before reversibility test Dichloromethane CH2 Cl2 + Klucel G solvent gel in compress / poultice Duration Result 1 minute Coating is lifting off in flakes that can easily be removed with a scalpel or bristle brush. 5 minutes Coating has lifted and turned white and flaky. The solvent has evaporated and a lot of the flakes can be removed with a bristle brush. Remaining coating is mostly on the textured corroded glass surface. Reapplied Most of the remaining coating fresh gel. has flaked and lifted and can be Then brushed off. The last remains observed 5 could be removed with a swab minutes later moistened with dichloromethane Dichloromethane gel compress: Stages 1. Japanese paper application 2. Solvent gel application 3. Absorbant pad application 4. Melinex sheet application 5. Compress removal 6. Dry, flakes of coating remain 7. Flakes brushed off 8. Final swab clean with Dichloromethane Table_Results_CAN_ S XXVIII D8 Page 3 of 6

6 Retreatability Do we need to re-treat the glass Test studies Retreatability No need. No treatment recommended Before reversibility test in reflected light Before reversibility test in transmitted light After reversibility test in reflected light After reversibility test in transmitted light Table_Results_CAN_ S XXVIII D8 Page 4 of 6

7 Reversibility (B) How can we remove the coating without damage to the paint layer? What method and solvent can we use? Is the coating stable? How are the solvents reacting with the glass corrosion, putty and lead? How do we remove all trace of the solvent? Test studies Elimination Area selected for reversibility test from panel SXXVIII D8 D8 Coating D8 Coating under optical microscope before reversibility test Ethanol + Klucel G solvent gel in compress / poultice. Duration Result 1 minute No visible difference to the surface of the coating 5 minutes The coating has softened and it was removable with a scalpel. 10 minutes Softened further, removable with a scalpel Reapplied Some coating could be removed fresh gel. with a swab moistened with Then ethanol. Once the ethanol had observed 10 evaporated the coating was hard minutes later to detect without strong light Reapplied fresh gel. Then observed 10 minutes later Ethanol gel compress: Stages Much of the coating has been removed, what remains is in the textured surface. 1.Japanese paper application 2. Solvent gel application 3. Absorbant pad application 4. Melinex sheet application 5. Compress removal 6. Dry coating flakes remaining 7. Flakes brushed off 8. Final swab clean with ethanol Table_Results_CAN_ S XXVIII D8 Page 5 of 6

8 Retreatability Do we need to re-treat the glass? Test studies Retreatability No need. No treatment recommended Before reversibility test in reflected light Before reversibility test in transmitted light After reversibility test in reflected light After reversibility test in transmitted light Table_Results_CAN_ S XXVIII D8 Page 6 of 6

9

10 (Labelling field) 1-Pilot object Pilot object: Canterbury Cathedral nii 7 Decorative border Picture Identification of the panel: nii 7 internal face in transmitted and reflected light Treatment Product : Microcrystalline Wax no:1129 Application: Microcrystalline Wax diluted with white spirit. Applied with a bush and mechanically polished Table_ResultsCAN_N II 7 Page 1 of 4

11 (Labelling field) 2-Results sample reference: CAN nii 7 Questions Techniques Answers Morphology What is the physical appearance/condition of the wax? Has the wax penetrated into the surface of the glass? Optical Microscope SEM Nano Tomography The wax appears to be intact mechanical polishing marks still visible. A light layer of accumulated dust on wax surface. Appears to remain transparent. Mild surface sheen when illuminated by reflective cold light. Paint appears solid and in good condition under the wax n/a sample Tomography on Synchrotron Chemical Composition What is the chemical composition of the applied mixture of waxes and solvent mix? Organic component composition Has the wax degraded or changed in any way? Microbiology Is there/ has there been any microbial activity on the wax surface? SEM FTIR RAMAN Metabolic activity and taxonomical description of microorganisms Fraunhofer Institute results: Found microcrystalline wax 1129 only in wax shavings sent from panel nii 7 n/a n/a CAN nii 10 (in situ) - low metabolic activity (ATP 164 RLU/25 cm²) - isolated microorganisms: Aspergillus fumigatus, Aspergillus versicolor, Cladosporium sphaerospermum (fungi; medium contamination) CAN nii 6 (in storage) - low metabolic activity (ATP 158 RLU/25 cm²) - no microorganisms isolated Table_ResultsCAN_N II 7 Page 2 of 4

12 (Labelling field) Reversibility How can I remove the wax without damage to the paint layer? What method and solvent can I use? Do we need to reverse the wax? The wax is stable and the panel is protected with protective glazing. Does the dichloromethane solvent react with the glass, glass corrosion and deposits, putty or lead and how? How do I remove all trace of the solvent? Test studies Elimination Sample area Area tested: wax layer on top of stable paint Dichloromethane compress and swabs Compress Duration: 2 mins + Results: removed the wax but left a slight residue, the absorbent pad formed white crystals after their use. Wax residue was removed with a swab moistened in dichloromethane. Cotton wool swabs: 30 seconds 10 seconds using swab moistened with dichloromethane and a further 2 swabs to remove residue. White crystals formed on the surface of the swabs after their use. Is there a reaction to the cotton wool? Or has the solvent evaporated and left chlorine residue? Under magnification there appears to be no residue of the dichloromethane and no softening of the surrounding putty on the glass. Wax coating under optical microscope before reversibility test Reversibility test with dichloromethane swabs After test Before test Retreatability Should I retreat the paint if the wax is removed? Test studies Retreatability Selected area for reversibility test No need. No treatment recommended. Table_ResultsCAN_N II 7 Page 3 of 4

13 (Labelling field) Before reversibility test in reflected light After reversibility test on the left side Table_ResultsCAN_N II 7 Page 4 of 4

14

15 1-Pilot object Pilot object: Canterbury Cathedral NXVII 03 Picture Identification of the panel: NXVII 03 internal face in transmitted and reflected light Treatment Product: Microcrystalline wax no: 1129 and Polythene A Wax (unknown ratio) ; retreatment with Paraloid B72 Application: Wax applied with a brush in dilution with white spirit. Application of wax early to mid-1980s in situ as temporary consolidant of severely flaking painted decoration. No cleaning was carried out prior to application of wax. The wax was not polished (unlike CAN nii 7). Window remained unprotected until Re-treatment with Paraloid B-72 in Fractures bonded with Silicone Rhodia CAF 3 in 1992 Table_Results_CAN_N XVII 03 Page 1 of 4

16 2-Results sample reference: CAN NXVII 03 Questions Techniques Answers Morphology Has the re-treatment coating of Paraloid B72 changed or deteriorated since application in 1992? Paraloid B72 treatment in reflected light Has the Paraloid B72 penetrated into the remaining wax layer? What is the bonding between these two layers? Optical Microscope SEM Observations during conservation in 1992 (window unprotected to that date): Wax coating delaminating and flaking with little or no adhesion to substrate in many places. Good adhesion on other places. Surface deposits of dust on wax. Corrosion processes had continued since wax application, both in the painted decoration and in the glass surface. More corrosion products had developed in existing damage areas underneath the wax coating, adding to its delamination. Observations during examination in 2009 (window protected by internally ventilated protective glazing since 1992): Remains of wax coating stable, no new delamination observed Paraloid -B72 application appears sound and unaltered; no delamination observed. Surface appears clean with no new deposits. The Rhodia CAF3 silicone repairs made in 1992 have not deteriorated and remain stable. Paraloid B72 application on top of remaining wax in reflected light What is the condition of the silicone since application in 1992? Desktop tomography Phase-contrast tomography on Synchrotron Silicone repair and Paraloid B72 treatment Table_Results_CAN_N XVII 03 Page 2 of 4

17 Organic component composition If there is wax present that is significantly different from the microcrystalline wax or polythene A? To determine if the same mixture was used on all windows. FTIR RAMAN Wax shavings were carefully removed under the microscope and sent to the Fraunhofer institute. Results found that Microcrystalline wax no: 1129 and Polythene A wax are present in the wax shavings taken from this panel (unknown ratio) Sample area Sample area wax shaving Chemical Composition Microbiology Is there/ has there been any microbial activity on the surface? (Wax, Paraloid and Silicone on the surface) SEM Molecular biology ATP measurements Very low levels of fungal spores detected, no typical microflora for historic glass, very low metabolic activity. Swab test Table_Results_CAN_N XVII 03 Page 3 of 4

18 Reversibility Re-treatability Test studies Elimination Test studies Retreatability The wax coating was removed mechanically with scalpels under optical microscope in Where adhesion was good the wax coating was thinned down mechanically, but not completely removed. No new reversibility test was carried out in 2009, as the unstable condition of the painted decoration was of concern. Paraloid -B2 in acetone dilution was applied in 1992 to consolidate flaking paint. This was often applied onto remaining wax coatings. Application of Paraloid -B72 with brush. Easy application and good immediate adhesion. Table_Results_CAN_N XVII 03 Page 4 of 4

19

20 Pilot object: Picture Canterbury Cathedral NXVII C1 1-Pilot object Identification of the panel: NXVII C1 internal face in transmitted and reflected light Treatment Product: The panel was used as a test panel for a pilot study on paint consolidation by the Fraunhofer Institut Silicatforschung. Three different consolidation materials have been used on separate glass pieces within the panel: SZA, Ormocer/Paraloid B72 1/1 and Paraloid B72. Numbering system : B1, G1, G4 and W2 tested with SZA B2, G2, and W1 tested with Ormercer and Paraloid B72 B3 and G3 tested with Paraloid B72 Table_Results_CAN_XVII C1 Page 1 of 3

21 2-Results sample reference: CAN NXVII C1 Questions Techniques Answers Morphology: Has there been any deterioration or change to the consolidants since application in 1992 Optical Microscope B1 SZA in reflected light G1 Test SZA in reflected light G4 SZA in reflected light W2 SZA in reflected light Test areas B1, G1, G4 there is no visible deterioration of the coatings. Test area W2 the coating appears to have developed a white opaque surface since application. B2 Orm/Par in reflected light G2 Orm/Par in reflected light W1 Orm/Par in reflected light Table_Results_CAN_XVII C1 Page 2 of 3

22 Test areas for Ormercer and Paraloid 1/1 show no visible deterioration of the coating on all test pieces. B3 Paraloid B72 in reflected light B3 Paraloid B72 in reflected Test areas for Paraloid B72 show no visible deterioration of the coating on all test pieces. Chemical Composition Organic component composition Microbiology Reversibility SEM Desktop tomography Phase-contrast tomography on Synchrotron SEM FTIR RAMAN Molecular biology ATP measurements Test studies Elimination The consolidants applied to the panel are all stable and no change or deterioration to the surfaces has occurred since application in Test area W2 using SZA is the only coating which has altered. The coating now appears to have developed a white opaque surface. Re-treatability Test studies Retreatability No need. No treatment recommended Table_Results_CAN_XVII C1 Page 3 of 3

23

24 Pilot object: Picture Canterbury Cathedral NXVII C2 1-Pilot object Identification of the panel: NXVII C2 internal face in transmitted and reflected light Treatment Product: Microcrystalline wax no: 1129 and Polythene A Wax; re-treatment with Paraloid B72 Application: Wax applied with a brush in dilution with white spirit. Application of wax early to mid-1980s in situ as temporary consolidant of severely flaking painted decoration. No cleaning was carried out prior to application of wax. The wax was not polished (unlike CAN nii 7). Window remained unprotected until Re-treatment with Paraloid B-72 in Fractures bonded with Silicone Rhodia CAF 3 in 1992 Table_Results_NXVII_C2 Page 1 of 4

25 2-Results sample reference: CAN NXVII C2 Questions Techniques Answers Morphology Has the re-treatment coating of Paraloid B72 changed or deteriorated since application in 1992? Paraloid B72 treatment Has the Paraloid B72 penetrated into the remaining wax layer? What is the bonding between these two layers? What is the condition of the silicone since application in 1992? Optical Microscope Observations during conservation in 1992 (window unprotected to that date): Wax coating delaminating and flaking with little or no adhesion to substrate in many places. Good adhesion on other places. Surface deposits of dust on wax. Corrosion processes had continued since wax application, both in the painted decoration and in the glass surface. More corrosion products had developed in existing damage areas underneath the wax coating, adding to its delamination. Observations during examination in 2009 (window protected by internally ventilated protective glazing since 1992): Remains of wax coating stable, no new delamination observed Paraloid -B72 application appears sound and unaltered; no delamination observed. Surface appears clean with no new deposits. No deterioration to the silicone repairs made in 1992, they remain stable. SEM Silicone repair and Paraloid B72 treatment Desktop tomography Table_Results_NXVII_C2 Page 2 of 4

26 Phase-contrast tomography on Synchrotron Dummy samples were made to replicate the glass and consolidation methods used using Microcrystalline wax and polythene A wax with a coating of Paraloid B72. The thickness of the wax coating (made in 3 layers) and Paraloid B72 were increased and coloured with raw umber powder pigment for 2 samples. Sample CAN 1a: Paraloid B72 mixed with pigment Sample CAN 1b: Wax mixed with pigment Results found that the Paraloid B72 had merged with the surface of the wax, therefore there is no risk of delamination of the Paraloid from the wax Table_Results_NXVII_C2 Page 3 of 4

27 Chemical Composition If there is wax present that is significantly different from the microcrystalline wax or polythene A? To determine if the same mixture was used on all windows. Sample area SEM/EDX Wax shavings were carefully removed under the microscope and sent to the Fraunhofer institute. Results found that Microcrystalline wax no: 1129 and Polythene A wax are present in the wax shavings taken from this panel (unknown ratio) Sample area wax shaving Organic component composition Microbiology Reversibility Retreatability FTIR RAMAN Metabolic activity and taxonomical description of microorganisms Test studies Elimination Test studies Retreatability CAN NXVII D2 (in situ): - low metabolic activity (ATP 190 RLU/25 cm²) - isolated microorganisms: Engyodontium album (fungus) and bacteria (medium contamination) CAN NXVII 3 (in storage): - low metabolic activity (ATP 200 RLU/25 cm²) - isolated microorganisms: Aspergillus fumigatus (fungus) and bacteria (medium contamination) The wax coating was removed mechanically with scalpels under optical microscope in Where adhesion was good the wax coating was thinned down mechanically, but not completely removed. No new reversibility test was carried out in 2009, as the unstable condition of the painted decoration was of concern. Paraloid -B72 in acetone dilution was applied in 1992 to consolidate flaking paint. This was often applied onto remaining wax coatings. Application of Paraloid -B72 with brush. Easy application and good immediate adhesion. Table_Results_NXVII_C2 Page 4 of 4

28

29 1-Pilot object Pilot object: Canterbury Cathedral svii 1 Decorative border Picture Identification of the panel: svii 1 internal face in transmitted and reflected light Treatment Product : Microcrystalline Wax no:1129 and Polythene A Wax (unknown mixture quantities) Application: Wax diluted with white spirit. Applied with a brush and mechanically polished Sample Table_Results_CAN_S VII 1 Page 1 of 2

30 2-Results sample reference: CAN svii 1 Wax Shavings Questions Techniques Answers Morphology What is the physical appearance/condition of the wax? Optical Microscope SEM The wax appears to be intact mechanical polishing marks still visible. A light layer of accumulated dust on wax surface. Appears to remain transparent. Mild surface sheen when illuminated by reflective cold light. Paint appears solid and in good condition under the wax Selected area for sampling in reflected light (before) Nano tomography Tomography on Synchrotron Selected area for sampling in reflected light (after) Chemical Composition Organic component composition If there is wax present that is significantly different from the microcrystalline wax or polythene A? To determine if the same mixture was used on all windows. Microbiology Reversibility Re-treatability SEM FTIR RAMAN Molecular biology ATP measurements Test studies Elimination Test studies Retreatability Wax shavings sent to the Fraunhofer Institute. Results found that Microcyrstalline wax no:1129, Polythene A wax and Ozokerite wax 7020 are present in the wax shavings taken from this panel (unknown ratio) No need. No treatment recommended. Table_Results_CAN_S VII 1 Page 2 of 2

31

32 1-Pilot object Pilot object: Canterbury Cathedral S VII 4 Decorative border Picture Identification of the panel: S VII 4 internal face in transmitted and reflected light Treatment Product : Microcrystalline Wax no:1129 and Polythene A Wax (unknown mixture quantities) Application: Wax diluted with white spirit. Applied with a brush and mechanically polished Sample Table_Results_CAN_S VII 4 Page 1 of 2

33 2-Results sample reference: CAN S VII 4 Questions Techniques Answers Morphology What is the physical appearance/condition of the wax? Is there/ has there been any microbial activity on the wax surface? Optical Microscope SEM The wax appears to be intact mechanical polishing marks still visible. A light layer of accumulated dust on wax surface. Appears to remain transparent. Mild surface sheen when illuminated by reflective cold light. Paint appears solid and in good condition under the wax. Nano tomography Tomography on Synchrotron Surface of wax under reflected light Chemical Composition SEM Organic component composition Microbiology Is there/ has there been any microbial activity on the wax surface? Reversibility FTIR RAMAN Molecular biology ATP measurements Test studies Elimination Very low levels of fungal spores detected, no typical microflora for historic glass, very low metabolic activity. No significant difference between archive storage and in situ. Re-treatability Test studies Retreatability No need. No treatment recommended. Table_Results_CAN_S VII 4 Page 2 of 2

34

35 1-Pilot object Pilot object: Canterbury Cathedral S VII 4 Decorative border Picture Identification of the panel: S VII 4 internal face in transmitted and reflected light Treatment Product : Microcrystalline Wax no:1129 and Polythene A Wax (unknown mixture quantities) Application: Wax diluted with white spirit. Applied with a brush and mechanically polished Sample Table_Results_CAN_S VII 4 Page 1 of 2

36

37 Pilot object: Picture Canterbury Cathedral SXXVIII 8h 1-Pilot object Identification of the panel: CAN SXXVIII 8h Treatment: Silicone repairs. Bonded with Rhodia CAF 3 silicone during the conservation. Application: Edge bonded. Excess silicone remains on the surface of the glass. Full panel in studio (transmitted light) Full panel in studio (reflected light) Table_Results_CAN_S XXVIII 8h Page 1 of 2

38 2-Results sample reference: CAN SXXVIII 8h Questions Techniques Answers Morphology What is the condition of the silicone since application in 1970s? Optical Microscope The window has been protected by an internally ventilated protective glazing system since The excess silicone appears firm and translucent; probably unchanged since application. The whole surface of the glass, including the excess silicone, is covered with fibrous dust which can easily be removed with a sable brush. SEM Silicone repair (face) transmitted light Desktop tomography Phase-contrast tomography on Synchrotron Silicone repair (face) reflected light Chemical Composition Organic component composition Microbiology Reversibility Retreatability SEM/EDX FTIR RAMAN Molecular biology ATP measurements Test studies Elimination Test studies Retreatability No microbial activity was visible under optical microscope. No need. No treatment recommended. Table_Results_CAN_S XXVIII 8h Page 2 of 2

39

40 Pilot object: Picture CAN SXXVIII A16 1-Pilot object Identification of the panel: SXXVIII CAN A16 Treatment Product: Polymer coating, possibly Viacryl Application: Probably applied with a brush sample Table_Results_CAN_S XXVIII A16 Page 1 of 3

41 2-Results sample reference: CAN SXXVIII A16 Questions Techniques Answers Morphology What is the physical appearance of the coating? Optical Microscope Is the coating stable? Image under optical microscope in reflected light Image under optical microscope in transmitted light An opaque coating inside the pitting only. Slightly lifting out of the pitting. Serious deterioration of the film: flakes of coating remain in corrosion pits only; the coating shows discolouration (slight yellowing) and loss of transparency. A higher magnification indicates the coating has not taken off the gel layer. Selected glass piece in reflected and transmitted light Chemical Composition Results from SEM Not interesting for this investigation Table_Results_CAN_S XXVIII A16 Page 2 of 3

42 Organic component composition What is the chemical composition of the coating? Is it Viacryl? Has the coating degraded or changed in any way? Microbiology Results from LRMH FTIR RAMAN Molecular biology ATP measurements No conclusion on the composition of the polymer. The composition is close to that of Viacryl. It is a mix of Viacryl and another product. The composition of the coating corresponds to an aliphatic amide What remains of the coating is inside the corrosion pits. There is a discoloration (slight yellowing) and loss of transparency of the coating. n/a No microbial infestation visible. Reversibility Re-treatability Test studies Elimination Test studies Retreatability A reversibility test was deemed unnecessary on this medieval sample because the coating had detached itself from the glass surface in most areas without any intervention by conservators. The coating that remains inside the corrosion pits appear under optical microscope to be lifting, so it is predicted that in time these remains will also detach from the glass. Our primary concern is for the object, and so we do not want to carry out any treatments on it that are not necessary. The protective glazing on the window should provide a stable environment so that re-treatment should not be necessary. Table_Results_CAN_S XXVIII A16 Page 3 of 3

43

44 1-Pilot object Pilot object: Canterbury Cathedral CAN SXXVIII C1 Picture Identification of the panel: CAN SXXVIII C1 sample Treatment Product: Polymer coating, possibly Viacryl + Desmodur Application: Probably applied with a brush. Table_Results_CAN_S XXVIII C1 Page 1 of 8

45 2-Results sample reference: CAN SXXVIII C1 External side Questions Techniques Answers Morphology What is the physical appearance of the coating? Is the coating stable? Optical Microscope Smooth, glossy, transparent surface. Fills all the pits and / or texture on the glass. It is very stable and solid. No visible deterioration of the film. It is uniform and smooth. There is no visible discolouration or loss of transparency. The support is a post-medieval glass with an irregular surface but not altered. Microscope observations reveal some little bubbles, but only on the glass interface. The adherence of the coating is good. Sample sent for analysis to LHRM (external surface) CAN C1 Desktop tomography n/a Coating under optical microscope (external) CAN C1 Phase-contrast tomography on Synchrotron n/a Chemical Composition SEM Not interesting for this investigation Organic component composition What is the chemical composition of the coating? Is it Viacryl? Is it Viacryl? What is the ratio of Viacryl and Desmodur? Has the coating degraded or changed in any way? FTIR (LRMH) Molecular biology ATP measurements No conclusion on the composition of the polymer. It is close to Viacryl. It was compared to other possible consolidants (Paraloid, Motena, Sinmast, Primal, Mastic and Silicon spectra ) but none correspond to LRMH spectrum. The composition corresponds to an aliphatic amide. No visible deterioration of the coating. It is uniform and smooth. No microbial infestation visible.. Table_Results_CAN_S XXVIII C1 Page 2 of 8

46 Microbiology Test studies Elimination Area selected for reversibility test with similar coating on CAN C1 C1 full panel reverse C1 Coating Dichloromethane + Klucel G solvent gel in compress / poultice Duration Result 1 minute Softened removable with a scalpel. Gel has already become firmer. 5 minutes The coating has gone white and started lifting from the glass. 10 minutes Lifting off further easily removable with a scalpel. Reapplied Softer and lifting off further. fresh gel. treated Then area observed 10 minutes later 30 minutes Flakier, could be removed with cotton wool swabs moistened with dichloromethane. 60 minutes Gel had dried out, and flaky coating could be brushed away with a bristle brush. Some coating remaining in pits could be removed with a cotton wool swab moistened with dichloromethane. Dichloromethane gel compress: Stages C1 Coating under optical microscope before reversibility test 1 Japanese paper application 2. Solvent gel application 3. Absorbent pad application 4. Melinex sheet application 5. Removal of compress 6. Remaining flakes of coating brushed off Table_Results_CAN_S XXVIII C1 Page 3 of 8

47 Reversibility (A) How can we remove the coating without damage to the paint layer? What method and solvent can we use? Is the coating stable? How are the solvents reacting with the glass corrosion, putty or lead? How do we remove all trace of the solvent? Re-treatability Do we need to re-treat the glass? Test studies Retreatability No need. No treatment recommended. The piece of sample glass separated into two parts along the previously bonded break line, when it was under observation in the LHRM laboratory. Before reversibility test in reflected light Before reversibility test in transmitted light After reversibility test in reflected light After reversibility test in transmitted light Table_Results_CAN_S XXVIII C1 Page 4 of 8

48 Reversibility (B) How can we remove the coating without damage to the paint layer? What method and solvent can we use? Is the coating stable? How are the solvents reacting with the glass corrosion, putty and lead? How do we remove all trace of the solvent? Test studies Elimination Area selected for reversibility test with similar coating on CAN C1 C1 Coating C1 Coating under optical microscope before reversibility test Ethanol + Klucel G solvent gel in compress / poultice. Duration Result 1 minute Softened removable with a scalpel 5 minutes Further softened 10 minutes Softer but still in place. Ethanol had been absorbed into pad or evaporated. Reapplied Ethanol had all been absorbed fresh gel. into pad or evaporated, so fresh Then gel was applied using a thinner observed 10 absorbent material. minutes later 30 minutes Soft but still in place 60 minutes Soft but still in place. Gel still moist 4hrs 6hrs Ethanol gel compress: Stages Fresh gel re-applied Coating on surface removed, coating remaining in etched pits 1. Japanese paper application 2. Solvent gel application 3. Absorbent pad application 4. Melinex sheet application 5. Removal of compress 6. Flake of coating remaining 7. Flakes brushed off 8. Final swab clean with Ethanol Table_Results_CAN_S XXVIII C1 Page 5 of 8

49 Retreatability Do we need to re-treat the glass? Test studies Retreatability No need. No treatment recommended. Before reversibility test in reflected light Before reversibility test in transmitted light After reversibility test in reflected light After reversibility test in transmitted light Table_Results_CAN_S XXVIII C1 Page 6 of 8

50 sample reference: CAN C1 (inside) Questions Techniques Answers Morphology What is the physical appearance of the coating? Is the coating stable? Optical Microscope Smooth, glossy, transparent surface. It is very stable and solid. Fragment for Sample sent for analysis to LHRM (internal surface) CAN C1 ResEM No visible deterioration of the film. It is uniform and smooth. There is no visible discolouration or loss of transparency. The support is a glass with black trace line paint and recovered by a brown matt. Glass paint is cracked or missing, with deposits between cracks. coating Coating under optical microscope in reflected light (internal) CAN C1 Desktop tomography Phase-contrast tomography on Synchrotron n/a n/a Coating under optical microscope in transmitted light (internal) CAN C1 Chemical Composition What is the chemical composition of the coating? Is it Viacryl? Is it Viacryl? What is the ratio of Viacryl and Desmodur? Is it the same coating on both the external and internal surface? FTIR (LRMH) The coating could be a mixture of Primal, Viacryl and the component in the polymer on the external surface. Table_Results_CAN_S XXVIII C1 Page 7 of 8

51 Organic component composition Has the coating degraded or changed in any way? Results from LHRM SEM RAMAN No visible deterioration, discolouration or loss of transparency of the coating. n/a Microbiology Molecular biology ATP measurements No microbial activity was visible under optical microscope. Reversibility Test studies Elimination Retreatability Do we need to re-treat the glass? Test studies Retreatability No need. No treatment recommended. Table_Results_CAN_S XXVIII C1 Page 8 of 8

52

53 1-Pilot object Pilot object: Princess Cecily, Burrell Collection, Glasgow Pictures Identification of the panel: Burrell Collection, Glasgow. Registration number: Dimensions: 39.4cm x 29.8cm. Currently (May 2010) in storage. Treatment In the mid-1970 s, this panel was dismantled, stop-gaps of extraneous glass removed, edge-to-edge bonding of broken fragments and gap-filling undertaken, and the panel releaded. Araldite AY103/HY951 epoxy resin was used for bonding and filling. Where relevant, the resin was coloured to match the blue and aubergine glass and the yellow stain. In addition, white insoluble weathering products were mechanically removed from pits using fine tools. Photographs (transmitted & reflected light): May 2010 Table_Results_GLA_Princess Cecily Page 1 of 3

54 2-Results sample reference: Princess Cecily Questions Techniques Answers Morphology Visual Inspection Photograph: May 2010 The panel was inspected in conjunction with other resin-bonded items in the Burrell Collection. The epoxy resin in this panel has experienced a total of 35 years aging; approximately half that period in dark storage and the remaining time ( ) in the strong illumination of the Burrell Collection artificially-lit gallery during opening hours. The polymer appearance has been well-maintained. There has been minimal yellowing of the clear resin used for bonding or filling and the colorants used to match the yellow stain and coloured glass have shown good light-fastness. The degree of yellowing is judged to be just perceptible. The results of this natural aging of Araldite AY103/HY951 are superior to those predicted by accelerated aging (see Norman H Tennent, "Clear and Pigmented Epoxy Resins for Stained Glass Conservation: Light Ageing Studies", Studies in Conservation, 1979, 24, ). Optical Microscope SEM Desktop tomography Chemical Composition Organic component composition Phase-contrast tomography on Synchrotron SEM/EDX FTIR Table_Results_GLA_Princess Cecily Page 2 of 3

55 RAMAN Raman spectroscopy was undertaken using a PerkinElmer IdentiCheck spectrometer with a fibre-optic probe, focusing on an area of clear resin fill. The spectrum is consistent with a standard spectrum of Araldite AY103/HY951. Microbiology Reversibility Retreatability Molecular biology ATP measurements Test studies Elimination Test studies Retreatability Table_Results_GLA_Princess Cecily Page 3 of 3

56

57 1-Pilot object Pilot object: St Francis roundel, Burrell Collection, Glasgow Pictures Identification of the panel: Burrell Collection, Glasgow. Registration number: Dimensions: diameter, 17cm. Currently (May 2010) on display in the artificially-lit gallery. Treatment In the late-1970 s, this panel was dismantled, edge-to-edge bonding of broken fragments and gap-filling undertaken, and the perimeter of the panel re-leaded. Ablebond epoxy resin was used for bonding and filling. Where relevant, the resin was coloured to match the yellow stain with a mixture of Oracet Yellow GN and Chromophtal Blue A3R DOP paste. To complete the restoration missing parts from the pictorial scene, the text and the donors inscription were retouched using opaque pigments in an acrylic medium (Cryla Colour) to replace the missing painting, and pigments in polyurethane medium (Chinaglaze) as a translucent material to replace the missing enamel and stain. The treatment was described in detail in the following article : Janet H Notman and Norman H Tennent, "The Conservation and Restoration of a Seventeenth Century Stained Glass Roundel", Studies in Conservation, 25 (1980), Photographs, before and after conservation: 1979 Table_Results_GLA_ St Francis roundel Page 1 of 3

58 2-Results sample reference: Questions Techniques Answers Morphology Chemical Composition Visual Inspection Photograph: May 2010 Optical Microscope SEM Desktop tomography Phase-contrast tomography on Synchrotron SEM/EDX The panel was inspected in conjunction with other resin-bonded items in the Burrell Collection. The epoxy resin in this panel has experienced a total of 31 years aging; the initial 5 years in dark storage and the remaining time ( ) in the strong illumination of the Burrell Collection artificially-lit gallery during opening hours. The polymer appearance has been well-maintained. There has been minimal yellowing of the clear resin used for bonding or filling and the colorants used to match the yellow stain and coloured glass have shown good light-fastness. The degree of yellowing is judged to be just perceptible. The results of this natural aging of Ablebond and the colorants are consistent with those predicted by accelerated aging (see: Norman H Tennent, "Clear and Pigmented Epoxy Resins for Stained Glass Conservation: Light Ageing Studies", Studies in Conservation, 1979, 24, Norman H. Tennent and Joyce H. Townsend, "The Photofading of Dyestuffs in Epoxy, Polyester and Acrylic Resins", in Preprints of ICOM Committee for Conservation 7 th Triennial Meeting, Copenhagen, 1984, ). Table_Results_GLA_ St Francis roundel Page 2 of 3

59 Organic component composition FTIR RAMAN Microbiology Reversibility Retreatability Molecular biology ATP measurements Test studies Elimination Test studies Retreatability Table_Results_GLA_ St Francis roundel Page 3 of 3

60

61 1-Pilot object Pilot object: Picture Ref. Parish Church south Panel of fragments Identification of the panel: Panel of fragments Internal face, transmitted light External face, reflected light Treatment: , by Konrad Vetter - Araldite binder AY103 by 100 parts, Hardener HY951 by 9 parts (Astorit AG 8840 Einsiedeln). - For thin fragments (1-1.5mm) with multiple fractures, the simple edge bonding was not considered to be sufficient and the method of back-plating was used. TableResults_Fragments Page 1 of 19

62 sample reference: Doubling red flashed glass with Araldite. Questions Techniques Answers 2-Results Morphology Why is the Araldite deterioration so different (yellowing in different stages / crizzling and stable surfaces) on the same piece of glass? When peeling off, does the Araldite hurt the glass surface? How far did the Araldite penetrate into the painting surface? Can you detect and differentiate several preparations of Araldite on the samples from these objects? On these samples you can see several steps of this process, as well as our corresponding, provisional classification and cartography of this phenomena. We propose that the various stages of changing of the material and of its properties which occur on this large sample are investigated and described according to the possibilities of the analyzing methods available in the project: visual microstructure, chemical and physical properties, interfaces properties. Optical Microscope The results of the microscopic analyses can be found at the end of the document ANNEXE 1. TableResults_Fragments Page 2 of 19

63 Stage 3: Clear and transparent resin filling, beginning of or advanced yellowing. The plated glass compound has still a dark shining aspect. The whitish micro-bubbling is getting denser. Due to the yellowing of the material, the bubbles can have a brown-ochre colour, under reflected light the areas can also look milky. The adhesion starts to weaken in these parts, but in general it is still very strong. Stage 4: Changing aspect from bright, white, shiny aggregations of points to iridescent surface areas. The Araldite surface does not look deep dark any more, but rather white on dark (mostly to be seen on edgings or cracks, where mechanical impacts and movements may have occurred). TableResults_Fragments Page 3 of 19

64 Stage 5: Transparent detachment of the Araldite, which looks grey-bluish in reflecting light. The filling resin starts to peel of from the smoother carrier glass but is still attached to the original glass surface. The compound can look less yellow in transparent light. Stage 6: The Araldite takes an even brighter whitish aspect, smoky or foggy. The layer may have been shrinking. It seems to be detached from the old glass, but still adhering to the doubling glass. TableResults_Fragments Page 4 of 19

65 Stage 7: The Araldite looks golden, amber until ochre-yellow. The layer begins to break up, by lined-up cracks or flakes. It seems that at this step it has been detached from both glass surfaces. Stage 8: Yellowed Araldite with a shining whitish surface aspect looks therefore as detached from both surfaces. Broken up by cracks or flakes, also shrunken flake-insulate which can even overlap. (Air bubble from process.) SEM n/a TableResults_Fragments Page 5 of 19

66 Desktop tomography Phasecontrast tomography on Synchrotron n/a Sample CSRIV_01 The sample has been analysed with phase-contrast microtomography, with photon energy of 27 kev and a sample-to-detector distance of 66 cm, in order to enhance the contribution of the low-absorption consolidant. TableResults_Fragments Page 6 of 19

67 The two images are related to two different positions on the bottom tip from the sample photo on page two of the data sheet. The colour code is the following: - the original glass is yellow - the Araldite is pale blue - the metal parts are red - the degraded glass is green With the exception of the edges of the fragments, the Araldite sticks rather well to the glass, even if there are some cracks from the surface to the glass. The green parts could be, in principle, either Araldite or glass, but the second hypothesis is more correct. The small metal parts can come both from the grisaille or, in general, from some pigments, but sometimes come from the lead frames. The situation is pretty much the same both for the opaque and the transparent Araldite forms, and the loss of adhesion is usually confined to the edges of the fragment, even on the other tip of the glass segment. Particular observations: The plating glass has been detached, its surface seems to have been smoother than the surface of the original. The crack in the epoxy layer occurs at the border between parts of different thickness. This confirms an observation on larger samples: The effect could be due to shrinking, but also to the different mechanical stresses due to thermal expansion (glasses, but especially the resin itself). In the crack area (top of the image, old crack reopened for sampling), the well adhering epoxy infill has stripped off a part of the adjacent glass evidence for the risks of de-restoration. TableResults_Fragments Page 7 of 19

68 Optical computer tomography OCT (Piotr Torgorski, Turn Poland) For the results of OCT, see the report of Pavel Karaskiewicz below ANNEXE 2 Chemical Composition SEM/EDX n/a Organic component composition FTIR n/a Surface of the sample is too rough for FTIR-spectroscopy. RAMAN n/a Microbiology Molecular biology ATP measuremen ts No results TableResults_Fragments Page 8 of 19

69 Reversibility We found out by taking out the test samples, that even when the Araldite seems to be in a bad condition, it s still sticking to the carrier glass. Test studies Elimination On these cartography you can see several steps of detachment. - In the green zone, the adhesion is lost. - Zone yellow shows parts where the adhesion starts to weaken, but in general it is still very strong. - The red zone shows Araldite in very good condition. It would be hard or even dangerous to taking out this glass pieces. That s why we decide to keep them like it is and not to reverse the back plating. Re-treatability In this case, we don t re-treat the panel. Test studies Retreatability TableResults_Fragments Page 9 of 19

70 ANNEXE 1: Optical Microscope Preface to the damage characterization of Araldite plating The historic glass segment of a church window exists of three fragments of red flashed glass. The plating was made at the non-flashed side. The thickness of flat glass was about 1.5 mm; the Araldite plating was around 200 µm thick. The following damage characterization was made with light microscopy with a 50 fold magnification. Both reflected light (RL) and transmitted light (TL) was used; also combined with dark field (DF). The corrosion description of the plating contains the phenomena between the cover glass and the Araldite as well as the one between the Araldite and the original glass. Also the original glass shows damages, caused by corrosion. These must not be confused with the one from the plating, but could be a cause of defective plating. To clarify this comparison with the original glass before the plating would be necessary. Corrosion phenomena, which can be found under the yellow aged Araldite plating, looking yellowed discoloured, depending on the light. Without plating these phenomena looks grey-whitish. The original glass shows often this damage. Microscopic pictures of original glass with plating with two different lights. Transmitted light (TL) Dark field with transmitted light (TL/DF) Microscopic pictures of original glass without plating with two different lights. Reflected light (RL) Dark field with reflected light (DL/RF) TableResults_Fragments Page 10 of 19

71 Damage characterization of Araldite plating Samples: Switzerland, Parish Church west Panel of fragments 3a, 3b, 3c Stage 3 - Cover glass-araldite : Good connection between cover glass and Araldite. Araldite shows yellowing and conditional of manufacturing bubbling (RL/DF and TL/DF), locally small iridescent areas (RL). - Araldite -original glass: no abnormalities visible Area: Cover glass/araldite Area: Araldite /original glass No abnormalities RL Same area, RL/DF Same area, TL/DF TableResults_Fragments Page 11 of 19

72 Stage 4 - Cover glass-araldite : Connection between cover glass and Araldite is partially broken. The new very thin gap between both media looks iridescent (RL) or milky (TL/DF), because of light optic phenomena. - Araldite -original glass: Connection between Araldite und original glasses is in smaller areas broken. The Araldite looks whitish crystalline; the phenomena are only visible in dark field. Area: Cover glass/araldite Area: Araldite /Original glass RL RL Same area, DF Same area, TL/DF Stage 5: TableResults_Fragments Page 12 of 19

73 - Cover glass-araldite : Proceeding of gap formation between cover glass and Araldite. Iridescent decrease because of the increasing of gap width. Grey-blue emerging damage phenomena (RL). In dark field only the border line are visible, on other areas it is transparent (DF). - Araldite /Original glass: Araldite is dissolving in bigger areas, but not laminar, from the original glass. Increasing of the whitish crystalline areas mainly visible in dark field. In reflected light as dark and partly iridescent areas weakly visible. Area: Cover glass/araldite Area: Araldite /Original glass RL RL Same area, DF Same area, DF Stage 6: TableResults_Fragments Page 13 of 19

74 - Cover glass/araldite : The adhesion properties of the Araldite have changed significantly: uneven appearance, formation of rough segments which adhere either at the cover or on the original glass (RL+DF). Maybe the appearance has changed because of the contraction respectively the expansion of the plating material. On the inner side of the cover glass partly crystalline structures are visible (DF). - Araldite /Original glass: Araldite adheres mainly on the original glass. Only isolated smaller golden yellow areas visible. These could be a hint for the beginning of delamination (image presentation is difficult). Area: Cover glass/araldite Area: Araldite /Original glass RL DF Same area, DF DF TableResults_Fragments Page 14 of 19

75 Stage 7: - Cover glass/araldite : The damage phenomena of stage 6 are increasing and the segmentation gets finer. The connection between Araldite and cover glass is only locally existent. This is good visible on the light-grey and large-scaled areas (RL). - Araldite /Original glass: The Araldite appears golden yellow, large-scaled and crumbling. This damage is a hint for the increasing brittleness of the Araldite. The connection of the Araldite to the original glass was also decreased. Area: Cover glass/araldite Area: Araldite /Original glass RL DF Same area, RL Same area, TL/DF TableResults_Fragments Page 15 of 19

76 Stage 8: - Cover glass/araldite : Connection between cover glass and Araldite is totally lost. Brittleness of Araldite leads to a fine segmentation and to a bigger crack structures and holes. Small whitish structures can partly be found on the Araldite (DF). - Araldite /Original glass: Connection between Araldite and original glass is totally lost. The demonstration of this corrosion phenomenon is only hardly possible with microscopic techniques. Area: Cover glass/araldite Area: Araldite /Original glass TL DF TableResults_Fragments Page 16 of 19

77 ANNEXE 2: OCT: Optical Coherence Tomography Optical Coherence Tomography is based on the recording of changes of optical properties of materials transparent for chosen wavelength (in the case of this research IR radiation 810nm was used). The scans are recorded as series of.jpg pictures (about 150 for one scan) which are shown as an.avi movie from which an interesting.jpg picture can be extracted for interpretation as well as a set of protocols for separated pictures. The sample analyzed was a piece of flashed red glass from Romont with a painted layer on it sticked with an adhesive (probably epoxide) to the glass support approx. 2 mm thick. Both sides of sample were scanned: 1. The original glass surface: named face 2. The supporting glass named: back The 10 scans of the Romont sample were executed: number scan number side additional interpretation back back face face face back back.avi back.avi back.avi Three scans (7, 8, 10) for clarity the scans have been reinterpreted and those scans are most suitable for interpretation. From the.jpg pictures two were chosen for interpretation. They have to be treated as an example and help for possible further evaluation of the OCT results. 1 2 The typical scanning result is depicted on fig.1 where: 1. The final scan 2. The scanned area (yellow line depicts the scanning line) 3. The picture of the sample showing the spot of spot of scanned area 4. The written information is the description of scanning conditions. The most important one is: Wymiar (X,Y,Z) [mm] with information on of dimensions of scanned area. In example below: X=11.6 mm; Y=12 mm and Z=1.85 mm. Z is the deep of light penetration. Bear in mind that the X value is on vertical axis and Y on horizontal one. 3 fig. 1. Typical OCT scan report of the sample of Romont glass TableResults_Fragments Page 17 of 19

78 The raw results depicted above do not show the real scan view and need to be adjusted. The final picture is shown below (fig.2). surfaces 1 adhesive 2 old glass surface fig. 2 Scan of the Romont sample from the back Fig 2 shows the OCT scan from the back of the sample, i.e. from the backing glass. The depth of penetration of IR radiation is about 2 mm (cf. the scale) and reaches through the glass and adhesive to the back surface of the old glass. The interface backing glass - adhesive gives only information that there is no delamination between them, but on the old glass surface several phenomena are seen: 1. corrosion 2. delamination of the glass surface which might have been caused by adhesive's contraction. The OCT scanning from the face of the sample is easier to analyze as IR do not penetrated deep into the glass, probably because of red copper containing layer (copper absorbs IR radiation). Typical example of such a scan shows fig.3. air 1 glass surface 2 fig. 3. OCT scan from the face of the sample TableResults_Fragments Page 18 of 19

79 The scan depicted on fig.3 shows a (1) glass breaks with slightly uneven joint of glass edges and (2) the copper flashed surface. Above depicted examples show only the possibilities of the method. The OCT for multilayered samples is still on its development stage and interpretation is difficult and time consuming. These two examples however show the way of interpretation and the potential of the method. dr Pawel Karaszkiewicz Krakow TableResults_Fragments Page 19 of 19

80

81 1-Pilot object Pilot object: Picture Ref. Parish Church south Heraldic stained glass panel Identification of the panel: Panel of fragments Internal face, transmitted light External face, reflected light Treatment: , by Konrad Vetter - Araldite binder AY103 by 100 parts, Hardener HY951 by 9 parts (Astorit AG 8840 Einsiedeln). - For thin fragments (1-1.5mm) with multiple fractures, the simple edge bonding was not considered to be sufficient and the method of back-plating was used. 2-Results sample reference: Has not been taken. Questions Techniques Answers TableResults_Heraldic Page 1 of 7

82 Morphology Why is the Araldite deterioration so different (yellowing in different stages / crizzling and stable surfaces) on the same piece of glass? When peeling off, does the Araldite hurt the glass surface? How far did the Araldite penetrate into the painting surface? Can you detect and differentiate several preparations of Araldite on the samples from these objects? On these samples you can see several steps of this process, as well as our corresponding, provisional classification and cartography of these phenomena. We propose that the various stages of changing of the material and of its properties which occur on this large sample are investigated and described according to the possibilities of the analyzing methods available in the project: visual microstructure, chemical and physical properties, interfaces properties. Optical Microscope We made visual and microscopic analyses of its appearance to establish the characteristics of the decomposition process when Araldite has been used. Stage 1 Clear and transparent resin, no or very little yellowing. The plated glass compound has a dark shining aspect, due to the intact adhesion of the filling material with both glass surfaces. There are no air bubbles or other structural changes. TableResults_Heraldic Page 2 of 7

83 Stage 2 Clear and transparent resin filling no or very little yellowing. The plated glass compound has a dark shining aspect. With the microscope, small air bubbles can be detected between the resin and the old glass (white glimmer, not to be compared with air bubbles from the plating process). The adhesion to both glass surfaces is still good, except from the areas with micro-bubbles. TableResults_Heraldic Page 3 of 7

84 Stage 3: Clear and transparent resin filling, beginning of or advanced yellowing. The plated glass compound has still a dark shining aspect. The whitish micro-bubbling is getting denser. Due to the yellowing of the material, the bubbles can have a brown-ochre colour, under reflected light the areas can also look milky. The adhesion starts to weaken in these parts, but in general it is still very strong. Stage 4: Changing aspect from bright, white, shiny aggregations of points to iridescent surface areas. The Araldite surface does not look deep dark any more, but rather white on dark (mostly to be seen on edgings or cracks, where mechanical impacts and movements may have occurred). TableResults_Heraldic Page 4 of 7

85 Stage 5: Transparent detachment of the Araldite, which looks grey-bluish in reflecting light. The filling resin starts to peel of from the smoother carrier glass but is still attached to the original glass surface. The compound can look less yellow in transparent light. Stage 6: The Araldite takes an even brighter whitish aspect, smoky or foggy. The layer may have been shrinking. It seems to be detached from the old glass, but still adhering to the doubling glass. TableResults_Heraldic Page 5 of 7

86 Stage 7: The Araldite looks golden, amber until ochre-yellow. The layer begins to break up, by lined-up cracks or flakes. It seems that at this step it has been detached from both glass surfaces. Stage 8: Yellowed Araldite with a shining whitish surface aspect looks therefore as detached from both surfaces. Broken up by cracks or flakes, also shrunken flake-insulate which can even overlap. (Air bubble from process.) SEM n/a TableResults_Heraldic Page 6 of 7

87 Desktop tomography Phase-contrast tomography on Synchrotron n/a n/a Optical computer n/a tomography OCT Chemical Composition SEM/EDX n/a Organic component composition FTIR n/a RAMAN n/a Microbiology Molecular biology ATP measurements Reversibility We found out by taking out the test samples of the fragment panel, that even when the Araldite seems to be in a bad condition, it s still sticking to the carrier glass. Re-treatability In this case, we don t re-treat the panel. Test studies Elimination Test studies Retreatability TableResults_Heraldic Page 7 of 7

88

89 1-Pilot object Pilot object: Picture AFA Smolensk, Cracow Identification of the panel: Bay: 1 Panel: 1 Internal face Treatment Cleaned with water with unknown detergent Adhesive: epoxy resin Epidian 5 with amino hardener (triethylenetetramine) Application: from inside only, directly on glass breaks without panel dismantling Note: according to FTiR analysis the material is a epoxide resin but probably not Epidian Table_Results_KRA_AFA Page 1 of 4

90 2-Results sample reference: AFA-1 Questions Techniques Answers Morphology -What is the morphology of the weathered coating? -What is the process of the weathering? - How is the bonding between coating and glass? Optical Microscope Discolouration (yellowing, browning), loss of transparency, adsorption of dirt 1. Loss of adhesion (bubbles) on surface 2. Strong adhesion inside breaks 3. No detachment of glass flakes SEM small fissures on the surface Desktop tomography Good visualization of partial penetration into break Phase-contrast tomography on Synchrotron Good visualization of poor penetration of adhesive into breaks Table_Results_KRA_AFA Page 2 of 4

91 OCT tomography Good view of adhesive - glass bond, bubbles of air, and break in glass Chemical Composition SEM/EDX Organic component composition - What is its chemical evolution? FTIR RAMAN epoxide resin - probably no Epidian Microbiology -Is there a biological contamination? -Is there an active infestation? Molecular biology ATP measurements Low biological contamination on glass and adhesive High contamination in putty - possible source of further contamination Reversibility -How can I remove the film without damage? -Which kind of method and of solvent, can I use? Retreatability -Should I retread the panel? -With which product? -What is its durability? (light, temperature, microorganisms) - What is the compatibility between old and new consolidant? Test studies Elimination Test studies Retreatability On the glass surface: spontaneous delamination Inside breaks: resistant to simple solvents Weakening of a bond after moderate heating Softening of the adhesive in commercial paint stripper (ethylene chloride based in gel form) possible removal with wooden spatulas The retreatment is necessary - removal of old adhesives, cleaning, partial regeneration of leading The treatment will be performed without dismantling the window. The new adhesive of lower viscosity will be used probably Paraloid B % No surface consolidant is necessary Preparation for full treatment of all window Table_Results_KRA_AFA Page 3 of 4

92 Climate measurements Is climate impact important? T RH Jan Feb March April May June AFA Smolensk climate RH -T July Aug Sept Oct AFA Smolensk climate T vs RH Nov Dec C %RH trend RH trend T Because there is no protective glazing, the influence of climate, mostly temperature and humidity is important, particularly due to night-day changes. Because, however, the climatic conditions are not very aggressive, and the window itself has no delicate parts such as painting layer, corrosion etc. it is suggested to perform conservation as soon as possible, mostly for technical reasons (weak leading). This is also connected with sophisticated ironwork and window opening construction, which is the important object itself. The installation of outer glazing without destroying the ironwork is not possible. The owner of the building does want to save the original structure RH T Table_Results_KRA_AFA Page 4 of 4

93

94 Data sheet for samples Various fresh materials Sample no. : LBW Microscope slides with various fresh materials Date of sampling: Febr./March 2008 Sample information Building Data window Investigated panel Manufacturing technique Exposition / storage Conservation treatment Samples for LBW As we discussed at the Constglass meeting in Canterbury, new test glasses should be covered with different materials, so that they may be investigated in the Institute LBW Bioconsult with respect to their microbial contamination. The follow test glasses have been prepared: 1. 3 blank microscope slides of each of the following: Ormocer OR-G SH1 SZA Araldite Paraloid B72 5% in Toluene 2. 3 microscope slides sandblasted with SZA Hxtal -Nyl-1 Fynebond SampleData_various fresh materials Page 1 of 8

95 Data sheet for samples Various fresh materials Designation of test samples: 3 blank microscope slides with fresh coatings: Ormocer LBW/Orm/1/ -2/ -3 SH 1 LBW/SH1/1/ -2/ -3 SZA LBW/SZA/1/-2/-3 Araldite LBW/Aral/1/ -2/ -3 Paraloid B72 (5%) LBW/Para/1/ -2/ -3 3 sandblasted microscope slides with fresh coatings SZA LBW/SZA/ gestr/1/ -2/ -3 Hxtal -Nyl-1 LBW/Hxtal/ gestr/1/ -2/ -3 Fynebond LBW/Fyn/gestr/1/ -2/ -3 Conservation treatment Samples for UGhent New test glasses should be covered with different materials, so that they may be investigated in the University UGhent to find the right parameters for the CT. The follow test glasses have been prepared: 2. 1 blank microscope slide of each of the following: Ormocer OR-G SH1 SZA Araldite Paraloid B72 5% in Toluene 3. 1 microscope slide sandblasted with SZA 4. 1 microscope slide (with outlines) with: Araldite Paraloid B72 (5%) (application of Paraloid : 1x, 2x, 3x, see sketch) 5. 2 microscope slides (with outlines) without application: foreseen for an application with doped Paraloid (Jodine or Brome). This might be better for detection in CT. The Paraloid should be doped in Ghent. Designation of test samples: 1 blank microscope slide with fresh coatings: SampleData_various fresh materials Page 2 of 8

96 Data sheet for samples Various fresh materials Ormocer SH 1 SZA Araldite Paraloid B72 (5%) UGent/Orm/1 UGent/SH1/ 1 Ugent/SZA/1 UGent/Aral/1 UGent/Para/1 1 sandblasted microscope slide with fresh coatings SZA UGent/SZA/ gestr/1 1 blank microscope slide with outlines and fresh coatings: Araldite Paraloid B72 (5%) Ugent/Aral/Kontur/1 Ugent/Para/Kontur/1 (application: 1x, 2x, 3x) 2 blank microscope slides with outlines application with doped Paraloid : Paraloid +????? Paraloid +????? Ugent/Kontur/ Ugent/Kontur/ foreseen for an application with doped Paraloid (Jodine or Brome). This might be better for detection in CT. The Paraloid should be doped in Ghent. Treatment of the microscope slides: coated uncoated Microscope slides with outlines, for example Ugent/Para/Kontur/1 1x 2x 3x SampleData_various fresh materials Page 3 of 8

97 Data sheet for samples Various fresh materials SZA (test glasses no.: LBW/SZA/-1/-2/-3 LBW/SZA/gestr/ 1/ -2/ -3) SZA is a purely inorganic binding material based on Silicium-Zirconium-Alkoxide. SZA builds an inorganic network, which reacts with the network of the glass. Developed by Fraunhofer Institute FhG ISC. SH 1 (testglasses no.: LBW/SH1/ -1/ -2/ -3) thermoplastic colourless 2-component epoxy resin, darkened with black pigment; out of market. Developed by Glasbau Hahn, Frankfurt. Araldite 2020 (formerly XW396/XW 397) (testglasses no.: LBW/Aral/1/ -2/ -3) Two component epoxy resin (bisphenol-a ) with hardener (cyclic aliphatic amine), mix ratio 3 : 1. Manufacturer: Ciba Geigy. Paraloid B72 (5%) (testglasses no.: LBW/Para/1/ -2/ -3) methacrylate/ethyl methacrylate (70/30), 5% in Toluene. Supplier: Kremer Pigmente, Aichstetten. Hxtal -Nyl-1 (testglasses no.: LBW/Hxtal/gestr/1/ -2/ - 3) Two component epoxy resin, fully syntetic polyaddition resin, resin based of Epichlorohydrin; Hardener: Alkyletheramine, Imidirole Supplier: Gosman & Kraan, Conservation supplies dennenlaan 28, NL-1160 AA Zwanenburg. Ormocer (testglasses no.: LBW/Orm/1/ -2/ -3) Inorganic-organic hybrid polymer, a heteropolysiloxane mixed with Paraloid B 72. Developed by Fraunhofer Institute FhG ISC Fynebond (testglasses no.: LBW/Fyn/gestr/1/ -2/ -3) Two component epoxy resin, Component A viscous at RT, liquid at 80 C; Component B at RT liquid, mix ratio 10g : 3,2g; (mix-ratio must be exactly!); Manufacturer: Fyne Conservation Services, Airds Cottage, St. Cathrine`s, Loch Fyne, Argyll PA25 8BA, Scotland, UK. SampleData_various fresh materials Page 4 of 8

98 Data sheet for samples Various fresh materials Results of previous investigations / analyses Description of sample (size, outline, colour, condition) Sample photo 1 (LBW) Sample photo 2 (LBW) SampleData_various fresh materials Page 5 of 8

99 Data sheet for samples Various fresh materials Sample photo 3 (UGhent) Description of sampling The microscope slides were not fully coated (see sketch). In order to create a visible layer on the microscope slides, the following materials had to be applied in multiple coatings: Ormocer, Paraloid B72 and SZA three times each; SH1 und Araldite only once. The glasses were allowed to dry for twenty-four hours between the coatings. As SZA is not adhesive, but rather builds an inorganic network that reacts with glass, the microscope slides were sandblasted and thereby roughened, so that SZA would react more favourably (LBW/SZA/ gestr/1/ 2/ 3 and UGent/SZA/ gestr/1). To achieve the necessary humidity, at which SZA begins to bind, a climate box was built out of plexiglass and wood. Thanks to a small tray of water, humidity of over 70% was reached. The application with Hxtal -Nyl-1 was carried out with older materials, whose hardener had become slightly yellowed; both components were viscous. During application and drying, the material pulled away from the edges of the microscope slides, and collected along a narrow strip in the middle. Even an intensive cleaning with acetone before the application failed to help. For this reason, three slides were sandblasted before being coated with Hxtal -Nyl-1. Sandblasting the slides allowed the material to be applied over two-thirds of the surface of the slide. SampleData_various fresh materials Page 6 of 8

100 Data sheet for samples Various fresh materials To prepare Fynebond, Component A was heated to above 40 C and then mixed with Component B. In this case, too, the coating pulled away from the edges of the prepared surface and collected, sausage-like, in the middle of the slide. As before, the slides were sandblasted and then applied with Fynebond. The microscope slide with outlines named Ugent/Para/Kontur/1 got a one time, two times and three times application (from above to below). Perhaps the more times application will be better to detect. 2 microscope slides with outlines were not coated because they were foreseen for an application with doped Paraloid (Jodine or Brome). This might be better for detection in CT. The Paraloid should be doped in Ghent (Ugent/Kontur no. ). Conservators questions Conservators notes (ISC Bronnbach) Co-ordinator s instruction to the scientists Sample must not be destroyed can be destroyed, if necessary Remarks The test glasses were freshly prepared. After application, the glasses were allowed to bind for several days before being packed up for shipping on 11 February The test glasses with Fynebond and Hxtal -Nyl-1 were prepared in late February and early March and sent off on 10 March to LBW. Additional, unapplied samples of the respective materials were also sent (ca. 3 5 g). Responsible conservators for sampling, packing, sending (name, phone, ) Person 1 Person 2 Person 3 Hildegard Stocksiefen hildegard.stocksiefen@dombau-koeln.de Carola Mueller-Weinitschke carola.mueller-weinitschke@dombau-koeln.de SampleData_various fresh materials Page 7 of 8

101 Data sheet for samples Various fresh materials Curriculum Sample Institute Date of arrival Recieved from Returned to Date of return Applied analytical method Responsible scientist (name, phone, ) Responsible person for packing / sending (name, phone, ) Remarks SampleData_various fresh materials Page 8 of 8

102

103 Various fresh materials 1-Pilot object (Additional test) Pilot object: Various fresh materials Identification: Object holders: internal face Treatment: Microscope object holders were coated with various materials. Several surfaces have been sandblasted before being coated (= gestr ). Some materials were also used as paint-layer consolidation materials, so we painted some of the object holders. Materials: ORMOCER SZA Araldite Paraloid Hxtal-Nyl-1 Fynebond coated uncoated outlines are coated uncoated sample reference: Samples for U-Gent UGent/Orm/1 UGent/SH1/ 1 Ugent/SZA/1 UGent/Aral/1 UGent/Para/1 UGent/Aral/Kontur/1 UGent/Para/Kontur/1 2-Results Samples for LBW LBW/Orm/1/ -2/ -3 LBW/SH1/1/ -2/ -3 LBW/SZA/1/-2/-3 LBW/Aral/1/ -2/ -3 LBW/Para/1/ -2/ -3 LBW/SZA/ gestr/1/ -2/ -3 LBW/Hxtal/ gestr/1/ -2/ -3 LBW/Fyn/gestr/1/ -2/ -3 TableResults_various fresh materials Page 1 of 6

104 Various fresh materials Questions Techniques Answers Morphology How is the morphology of the coating? How is the bonding? Chemical Composition Organic component composition Optical Microscope (DBV) SEM (Institute) Desktop tomography (U-Gent) Phase-contrast tomography on Synchrotron (Institute) SEM/EDX (Institute) FTIR (Institute) RAMAN (Institute) The bonding of ORMOCER, SH1, Paraloid and Araldite seemed to be very good. The other materials needed rough surfaces for a good application; so we used sand blasted object holders. The samples were market with "gestr". n/a The samples were used only for CT parameters. n/a n/a n/a n/a TableResults_various fresh materials Page 2 of 6

105 Various fresh materials Microbiology Microscopical analysis, metabolic activity and taxonomical description of microorganisms (LBW) ORMOCER (after one year of moist incubation): - no visible microbial infestation - low metabolic activity (ATP 261 RLU/25 cm²) - isolated microorganisms: Aspergillus versicolor, A.. fumigatus and Penicillium expansum (fungi; medium contamination!), no bacteria After nine month of moist incubation the microscopical analysis gave no evidence of an active fungal growth on ORMOCER SH 1 (after one year of moist incubation): - slightly visible, punctual fungal infestation - medium metabolic activity (ATP RLU/25 cm²) - isolated microorganisms: Penicillium expansum (fungus; medium infestation!) and one bacterium (not identified) On SH1 after nine month of moist incubation some punctual fungal infestations could be stated, but never reached a considerable distribution over the consolidant coating. SZA (after one year of moist incubation): - no visible microbial infestation - low metabolic activity (ATP 221 RLU/25 cm²) - isolated microorganisms: Botritis cinerea and Penicillium rugulosum (fungi; low contamination!), no bacteria After nine month of moist incubation the microscopical analysis gave no evidence of an active fungal growth on SZA TableResults_various fresh materials Page 3 of 6

106 Various fresh materials Araldite (after one year of moist incubation): - encrusted, laminar fungal infestation - high metabolic activity (ATP RLU/25 cm²) - isolated microorganisms: Penicillium expansum (fungus; considerable high infestation!), no bacteria Araldite Paraloid B72 (after one year of moist incubation): - slightly encrusted, moderate fungal infestation - medium metabolic activity (ATP RLU/25 cm²) - isolated microorganisms: Penicillium expansum, P. chrysogenum and Botritis cinerea (fungi; medium infestation!), no bacteria Paraloid B72 SZA on sandblasted surface (after one year of moist incubation): - encrusted, moderate fungal infestation - high metabolic activity (ATP RLU/25 cm²) - isolated microorganisms: Penicillium expansum (fungi; high infestation!), no bacteria SZA on sandblasted surface TableResults_various fresh materials Page 4 of 6

107 Various fresh materials Hxtal-Nyl-1 on sand blasted surface (after one year of moist incubation): - punctual, moderate fungal infestation - high metabolic activity (ATP RLU/25 cm²) - isolated microorganisms: Penicillium expansum and Botritis cinerea (fungi; medium infestation!), no bacteria Hxtal-Nyl-1 on sand blasted surface Fynebond on sand blasted surface. (after one year of moist incubation): - slightly encrusted, moderate fungal infestation - high metabolic activity (ATP RLU/25 cm²) - isolated microorganisms: Penicillium expansum, P. chrysogenum and Botritis cinerea (fungi; medium infestation!), no bacteria Fynebond on sand blasted surface Reversibility Test studies elimination (Institute) n/a Re-treatability Test studies re-treatability (Institute) n/a TableResults_various fresh materials Page 5 of 6

108

109

110 Anti-reflex-glass 1-Pilot object (test samples for additional investigation) Pilot object: Cologne Cathedral, Anti-reflex-glass (used as outer protective glazing) Picture Samples of modern solar glass, coated with a nano-porous layer of silicon dioxide (SIO 2 ), developed at the Fraunhofer ISC Würzburg. The tests were carried out for LBW only. L Remarks TableResults_Antireflex Page 1 of 3

111 Anti-reflex-glass 2-Results sample reference: LBW/Antireflex/1 5 completely coated LBW/Antireflex/I - V incompletely coated LBW/Antireflex/A E completely coated, one edge framed with lead (silicone fixed) Questions Techniques Answers Morphology Optical Microscope SEM Desktop tomography Phase-contrast tomography on Synchrotron completely coated incompletely coated completely coated, one edge framed with lead (silicone fixed) TableResults_Antireflex Page 2 of 3

112 Anti-reflex-glass Chemical Composition Organic component composition SEM/EDX FTIR Microbiology The microscopical analysis indicates a considerable development of fungal hyphae on Antireflex E after 6 month moist incubation RAMAN Microscopical analysis, metabolic activity and taxonomical description of microorganisms Antireflex 1 (after one year of moist incubation): - slightly encrusted, moderate fungal infestation - low metabolic activity (ATP 247 RLU/25 cm²) - isolated microorganisms: Penicillium expansum (fungus; considerable high contamination!) Antireflex I (after one year of moist incubation): - surficial, slight fungal infestation - low metabolic activity (ATP 325 RLU/25 cm²) - isolated microorganisms: Penicillium expansum (fungus; considerable high contamination!) Antireflex E (after one year of moist incubation): - surficial, slight fungal infestation - low metabolic activity (ATP 290 RLU/25 cm²) - isolated microorganisms: Penicillium expansum (fungus; considerable high contamination!) After nine month of moist incubation a considerable fungal growth and spore formation can be stated Antireflex 1. Reversibility Re-treatability Test studies Elimination Test studies re-treatability TableResults_Antireflex Page 3 of 3