MEDDELELSER OM KONSERVERING

MEDDELELSER OM KONSERVERING 2-2013 CLEANING PLASTICS IN MUSEUMS A CHANCE FOR DAYDREAMING: CONSERVATION- RESTORATION AND TECHNICAL RESEARCH OF AN OUTDOOR UPR PLASTICS INSTALLATION FROM 1992 NYT PROJEKT OM FOGNING AV STYRENPLAST FUTURE_TALKS 2013 PLAST PÅ MUSEUM NY FORBUNDSORDFØRENDE MEDDELELSER OM KONSERVERING PÅ NETTET? IIC - NORDIC GROUP NORDISK KONSERVATORFORBUND

MEDDELELSER OM KONSERVERING 2 2013 Kære Læsere, I dette tema kommer vi lidt omkring PLAST. Mod sædvane er der blot to artikler, derimod har vi under rubrikken NYT to plast-referater. Plastik omgiver os dagligt, vi forbruger store mængder i vores hverdag, men indsamler også til fremtiden. Det har, for mig, altid stået som en spændende problematik - hvad gør vi når vi vil have ting, som ikke nødvendigvis er ment til at holde længe, til præcis at holde, restaurere og gemme. Med dette temanummer, håber vi, at viderebringe mere ny viden om den foranderlige plastik og lidt status på plast-fronten. Vi er stolte over at kunne bringe nyt indenfor forskningen, fra Yvonne Shashua og Kathrine Segel samt en undersøgende gennemgang af UPR- produkter (Unsaturated Polyester Resin) fra Päivi Kyllönen- Kunnas og Ulla Knuutinen. Det næste temanummer er blevet bestemt af Forbundsrådet og det er Formidling på internettet. En problematik, eller udfordring, om man vil, som ligger os på redaktionen meget på sinde. Vi arbejder på digitalisering af tidligere og fremtidige udgivelser. Der er mange formelle problematikker som vi er begyndt at tænke over, blandt andet, at det offentliggjorte materiale ikke bliver misbrugt eller at der kommer juridiske krav fra genstandsejere eller andre. Vi vil gerne høre om jeres historier, udfordringer og løsninger. Så skriv nogle gode artikler, deadline for temanummeret er 1.august 2014! Der er stadig ledig plads til en spændende artikel til det generelle nummer, så hold jer ikke tilbage. Redaktøren takker som altid forfatterne for deres store arbejde, Liisa Valkeapää for finske resuméer og Caroline Allington for proofreading af bladets tekster. Også stor tak til de anonyme reviewere og redaktionen. Herudover takkes Bladpuljen i Danmark for støtte til forsendelse til danske abonnenter. God læselyst. LOA LUDVIGSEN Publikationsplan 1-2014 Generelt nummer. Deadline 1. marts 2014. Åben for artikeloptag. 2-2014 Temanummer - Formidling på internettet. Deadline 1. august 2014. Stadig plads til artikler og NYT. 1-2015 Generelt nummer. Deadline 1. marts 2015. Dear Readers, In this themed issue we look at PLASTIC. Contrary to custom, there are only two articles. However, we have two reports from plastic meetings in the NYT section. Plastic surrounds us every day, we use large quantities in our everyday lives, but also collect it for the future. To me, It has always been an interesting problem - what do we do when we want to keep things that are not necessarily meant to last, how to restore and preserve plastics? With this themed issue, we hope to convey some new knowledge about how plastics alter, and currentideas on the "plastic front". We are proud to bring new research, from Yvonne Shashua and Kathrine Segel as well as an investigative review of UPR products (Unsaturated Polyester Resin) from Päivi Kyllönen - Kunnas and Ulla Knuutinen. The theme of the next issue has been chosen by Forbundsrådet and it is Communication on the Internet. This can be a problem or challenge; and it is something that the editors have been thinking about. We are working on the digitisation of past and future issues. There are many formal problems which we have only begun to consider, such as that the published material will not be misused, or any sudden legal requirements from the owners of objects or others. We want to hear about your stories, challenges and solutions so please write us some good articles. The deadline for this themed issue is August 1st 2014! If you have an article for the general issue, don t hesitate, there is still space, deadline is March 1. The editor, as always, thanks the authors for their hard work, Liisa Valkeapää for Finnish abstracts and Caroline Allington for proofreading of the English text. Many thanks also to the anonymous reviewers and editors. In addition, we thank the Danish Bladpuljen for subsidising the postage to Danish subscribers. Enjoy your read. LOA LUDVIGSEN Publication Dates 1-2014 General issue. Deadline 1st March 2014. Some room left. 2-2014 Themed issue: Communication on the internet. Deadline 1st August 2014. There is still room for both articles and NYT/News items. 1-2015 General issue. Deadline 1st March 2015.

CLEANING PLASTICS IN MUSEUMS YVONNE SHASHOUA AND KATHRINE SEGEL INTRODUCTION Cleaning involves the removal of all materials that either change an object s materiality or meaning and have neither been applied intentionally by its creator nor are important as signs of its historical use (Bollard (2008)). Although all surveys of the condition of plastics conducted in museums in the United Kingdom and Scandinavia since the 1990s conclude that approximately 75% of collections require cleaning, few treatments have been developed (Shashoua and Ward (1995). This is mainly due to the high risk of damaging plastics mechanically or chemically when using an invasive cleaning treatment. Plastics which are in a rubbery phase at ambient temperature, such as polyethylene and plasticised polyvinylchloride (PVC), are visibly abraded by contact with brushes, cloths and sponges. Polyethylene containers including Tupperware, are readily scratched by green scouring pads or nylon brushes during washing. Solvents and detergents applied either alone or as part of mixtures in commercial cleaning products can both extract additives from flexible plastics such as PVC and induce environmental stress cracking in rigid plastics such as polystyrene, polycarbonate and polymethylmethacrylate. Despite the high risks, removal of oily fingerprints, carbonaceous dirt (external dirt) or crystalline, acidic degradation products (internal dirt) from museum objects or artworks comprising plastics is essential to maintain their significance, chemical stability and commercial value. This article summarises the findings of an exhaustive, 2-year evaluation of mechanical, aqueous and non-aqueous cleaning techniques for their effectiveness at removing external dirt and effect on chemical and physical stabilities of selected plastics. The research described was coordinated by the two authors as part of the EU 7th Framework Programme project POPART (Preservation of Plastic ARTefacts in museums) between 2008 and 2012. The project concluded with a conference Preservation Of Plastic ARTefacts in museum collections held in Paris in 2012 and publication of the same name. The highlights of both conference and publication are available freely on the internet http://popart-highlights.mnhn.fr/ POPART was divided into four working groups each focusing on different topics such as plastics' identification, collection surveys, degradation and active conservation. Together with the Cultural Heritage Agency of the Netherlands, Victoria and Albert Museum in London and Paris-based Laboratories: Centre de recherche et de restauration des musées de and the Centre de recherche sur la conservation, the National Museum of Denmark evaluated mechanical, aqueous and non-aqueous cleaning techniques initially for their influence on the long term stability of plastics and afterwards for their effectiveness at removing dirt. The purpose was to assist conservators with making decisions concerning cleaning plastics. This article will present an overview of the state of the art in cleaning plastics prior to the start of POPART, describe the research strategy and processes to evaluate mechanical, aqueous and solvent cleaning of selected plastics and conclude with the application of POPART s findings to two case histories. STATE OF THE ART IN CLEANING PLASTICS Until the launch of POPART in 2008, conservators believed generally that all semi-synthetic and degraded plastics were sensitive to water and solvents. This influenced the typical approach to cleaning them. Mechanical cleaning in the absence of aqueous cleaning agents or solvents was considered the technique offering the lowest risk of inducing permanent damage and, as a result, conservators chose to wipe away dust and loose particles with a dry cloth or brush to minimise damage. Where dirt was attached to the surfaces of an object as exemplified by oily fingerprints or sticky carbonaceous dirt and could not be removed with a dry brush or cloth, conservators often decided to leave it undisturbed to reduce the risk of inducing further damage. Until 2008, all published research into plastic cleaning by conservators or conservation scientists had used visual appearance or optical microscopy alone for evaluating whether cleaning had caused damage to surfaces. - ARTIKEL MED PEER REVIEW 3

POPART CLEANING RESEARCH STRATEGY The research strategy taken in POPART was to first ascertain whether the cleaning materials and techniques evaluated induced changes or degradation to model plastics substrates. Those that did not induce changes were further investigated for the effectivity of cleaning to remove artificial, standard dirt. Finally, the most effective and least damaging cleaning techniques were applied to real museum or study objects to compare the findings from cleaning model plastics with reality. Mechanical cleaning techniques were evaluated first and only the least damaging and most effective techniques were then combined with aqueous and solvent cleaning agents. MODEL PLASTICS New, colourless and transparent or pearlescent model plastics were used as model substrates to investigate whether cleaning would damage the plastics themselves. Initially, polymethylmethacrylate (PMMA), polyvinylchloride (PVC), unsaturated polyester (UP), high density polyethylene (HDPE), polystyrene both cast high impact (HIPS) and extruded (EPS) forms were selected as model plastics with each POPART partner researching one or two plastics for cleaning (see figure 1). Selection of plastics types for inclusion in POPART was based on their high representation in museum collections and requirements for cleaning. Unsaturated polyester was removed early from the project due to its inhomogeneity and replaced with cellulose acetate (CA). Extruded polystyrene proved too difficult to clean due to its uneven texture and was also discounted from POPART. CLEANING MATERIALS, CLEANING AGENTS AND ARTIFICIAL DIRT Published literature and verbal recommendations describing effective and non-damaging materials and techniques to clean plastics for industrial and conservation applications were used to select twenty-two cleaning materials, which were also available in all nine POP- ART partner countries. Materials comprised paper tissues, cotton swabs, natural and synthetic cloths, sponges and brushes and compressed air from can and pressure lines (see figure 2). Published literature was also used to select aqueous- and solvent- based cleaning agents which complied with health and safety requirements in all partner countries. Aqueous cleaning agents included distilled water, Figure 1. A selection of model plastics. From left to right; polymethylmethacrylate, plasticised PVC, unsaturated polyester (replaced with cellulose acetate), high density polyethylene, high impact polystyrene and expanded polystyrene (only included in the first stage). (photograph by Kathrine Segel). Figur 1. Et udvalg af model plast. Fra venstre mod højre polymethylmethacrylat, blødgjort PVC, umættet polyester (skiftet ud med cellulos acetat), højdensitets polyethylen, slagfast polystyren og ekspanderet polystyren (kun med i starten af projektet) (foto af Kathrine Segel). 4 - ARTIKEL MED PEER REVIEW

anionic and nonionic detergents, synthetic saliva and triammonium citrate. THE AQUEOUS CLEANING AGENTS EVALUATED WERE: Dehypon LS45, 1% (w/w) in distilled water, non-ionic detergent, cloud point 22 C, fatty alcohol C12-C14, experimental critical micelle composition (CMC) 0.0598 g/l, supplied by Conservation Resources Ltd. (UK), www. conservationresources.com distilled water Judith Hofenk de Graaff detergent, 1% (w/w) of concentrate in distilled water, non-ionic detergent, concentrate comprises 50 g sodium dodecylbenzenesulphonate, 50 g tri-sodium citrate and 5 g sodium carboxymethylcellulose in 1000 ml distilled water Orvus WA Paste, 1% (w/w) in distilled water, anionic detergent, sodium lauryl sulphate. Orvus WA paste has an experimental CMC of 0.29 g/l, supplied by University Products. The Archival Company, www.universityproducts. com Synthetic saliva, used as supplied, mucin (35mg per ml), unspecified quantities of xylitol, methylparaben, dinatrium, benzalkoniumchloride and EDTA, supplied by Pharmachemie BV (NL), www. tevapharmachemie.com Triammonium citrate (TAC), 2.5% (w/w) in distilled water, supplied by VWR & Bie & Berntsen, www.vwr.com Solvents comprised ethanol, isopropanol, white spirit (mixture of paraffins, cycloparaffins and aromatic hydrocarbons with boiling ranges which typically lie between 150 C and 220 C) and xylene. Two types of artificial dirt developed by cleaning researchers to mimic fingerprints or sebum soil and oily, carbonaceous soil which represented what is commonly known as museum dirt were prepared from published Figure 2. The 22 cleaning materials evaluated in POPART (photograph by Kathrine Segel). Figur 2. Rensningsmaterialer (22 i alt) evalueret i POPART (foto af Kathrine Segel). recipes and applied by brush or bar coater to model plastics. Sebum soil comprised palmitic acid and 1-propanol (Kuisma et al. (2005)). Organic oil soil was prepared by dispersing carbon black in paraffin oil (Koponen et al. (2007)). After application of artificial dirt, samples were allowed to dry for a week at room temperature in a covered box before cleaning was performed. CLEANING PRACTICE Initially, the effects of each mechanical cleaning procedure on clean model plastics were - ARTIKEL MED PEER REVIEW 5

Visual examination of soiled PMMA after cleaning SEBUM SOIL Dehypon LS45 Distilled water Judith Hofenk de Graaff detergent Orvus WA Paste Synthetic saliva Tri-ammonium citrate Ethanol Isopropanol White spirit Xylene PVA sponge Cotton bud Cotton cloth Leather chamois Spectacles cloth Microfiber cloth ORGANIC OIL SOIL Dehypon LS45 Distilled water Judith Hofenk de Graaff detergent Orvus WA Paste Synthetic saliva Tri-ammonium citrate Ethanol Isopropanol White spirit Xylene PVA sponge Cotton bud Cotton cloth Leather chamois Spectacles cloth Microfiber cloth Not tested Very good cleaning Good cleaning Poor cleaning Bad cleaning Figure 3. Visual evaluation of the effectiveness of cleaning soiled model PMMA. A traffic light colouring system has been used to represent very good cleaning as green and bad cleaning as red. Effectiveness is a measure both of cleaning techniques abilities to remove dirt while not damaging the plastic substrate. Figur 3. Visuel bedømmelse af effektiviteten ved rensning af snavset model PMMA. Farverne fra et lyssignal er blevet brugt til at repræsentere meget god rensning som grøn og dårlig rensning som rød. Effektiviteten er både et udtryk for rensningsteknikkernes evne til at fjerne snavs uden at beskadige plastmaterialet. investigated. From literature and discussions with experienced conservators it was clear that mechanical action can be applied in either linear or circular directions and 5 repetitions of each were investigated separately. Twelve cleaning products which were found to abrade or otherwise change surfaces were removed from further participation in POPART. The remaining 10 cleaning materials were used to apply the aqueous and solvent-based cleaning agents. Finally, the six, least damaging aqueous, four solvent- based cleaning materials and six agents were tested on model plastics with artificial dirt. Selected samples were exposed to accelerated light and thermal ageing to investigate the chances of degradation induced by cleaning with time. EVALUATION TECHNIQUES Visual examination of surfaces before and after cleaning has long been the primary form of evaluation used by the conservation profession to determine whether changes have been induced by treatment producing such symptoms as scratches, deposited residues or reduced reflectivity. While visual examination is one of the most important techniques to evaluate a conservation treatment, it is highly subjective and, therefore, qualitative and poorly reproducible. In POPART, it was decided to supplement visual evaluation with quantitative changes in gloss, contact angle and percentage area scratched after each cleaning. Gloss is the ability of a surface to reflect specular light. Materials with smooth surfaces are highly reflective (glossy), while very rough surfaces reflect no specular light and therefore appear matt. Gloss of test substrates before and after cleaning was determined using a reflectometer. Photomicrographs were examined for scratches by converting coloured images to black and white with a computer photo or paint program such as Adobe Photoshop and then by calculating the area occupied by scratches per unit area using the free program Image J (http://imagej.en.softonic.com/). Changes in surface energies of model plastics induced by cleaning were likely to be caused by contamination from residues of cleaning 6 - ARTIKEL MED PEER REVIEW

Visual examination of soiled PVC after cleaning SEBUM SOIL Dehypon LS45 Distilled water PVA sponge Cotton bud Cotton cloth Leather chamois Spectacles cloth Judith Hofenk de Graaff detergent Orvus WA Paste Saliva TAC Ethanol Isopropanol White spirit Xylene Microfiber cloth ORGANIC OIL SOIL Dehypon LS45 Distilled water Judith Hofenk de Graaff detergent Orvus WA Paste Synthetic saliva Tri-ammonium citrate Ethanol Isopropanol White spirit Xylene PVA sponge Cotton bud Cotton cloth Leather chamois Spectacles cloth Microfiber cloth Not tested Very good cleaning Good cleaning Poor cleaning Bad cleaning Figure 4. Visual evaluation of the effectiveness of cleaning soiled model PVC. A traffic light colouring system has been used to represent very good cleaning as green and bad cleaning as red. Effectiveness is a measure both of cleaning techniques abilities to remove dirt while not damaging the plastic substrate. Figur 4. Visuel bedømmelse af effektiviteten ved rensning af snavset model PVC. Farverne fra et lyssignal er blevet brugt til at repræsentere meget god rensning som grøn og dårlig rensning som rød. Effektiviteten er både et udtryk for rensningsteknikkernes evne til at fjerne snavs uden at beskadige plastmaterialet. agents or surface inhomogeneity caused by scratches. They were quantified by determining changes in contact angle formed between a 20 microliter drop of distilled water and surfaces of model plastics. Because changes in gloss, contact angle and percentage of scratches before and after cleaning were each independently related to surface properties of the plastic being cleaned, they could be mathematically combined to calculate a Cleaning Vector, M where M = ((% area scratched) 2 + (%change in contact angle) 2 + (%change in gloss) 2 for each cleaning material/agent combination. Cleaning was a measure of both how effective and how damaging cleaning processes were. There was an attempt to incorporate visual assessments into Cleaning Vectors but the result was unsatisfactory. However, the trends exhibited by Cleaning Vectors were mirrored by visual assessment. POPART RESULTS FOR CLEANING MODEL PMMA AND PLASTICIZED PVC Although mechanical cleaning has been generally perceived as the least damaging technique to remove dirt from plastics, POPART results suggest that the risk of inducing scratches or residue deposits using a dry brush or cloth is measurable. It is less damaging to remove loose dust particles by blowing clean filtered air from a can or a pressure line, although the latter can be contaminated with lubricant oil. For PMMA and PVC the five following materials were found to be least damaging to surfaces; cotton bud, cotton -, microfiber - and spectacles cloths, natural (ostrich) and synthetic feather dusters. Feather dusters cannot be used in combination with aqueous cleaning products or solvent. It was concluded that linear rubs removed dirt from surfaces, while circular rubs merely redistributed it. - ARTIKEL MED PEER REVIEW 7

Figure 5. Plasticised PVC model plastics deformed one week after applying organic oil soil (upper). The deformation persisted despite cleaning (lower) (photograph by Kathrine Segel). Figur 5. Blødgjort PVC model plast blev deformeret en uge efter organisk olieholdig snavs var påført (øverst). Deformationen vedblev efter rensning (nederst) (foto af Kathrine Segel). The concentration of scratches was severely reduced when combining cotton bud or any of the cloths with liquids in the form of water, aqueous detergent solutions and solvent. On clean model plastics the aqueous cleaning agents 1% anionic detergent Orvus WA Past, 1% non-ionic detergent Dehypon LS45, saliva and 2.5% triammonium citrate minimised scratching and the opportunity for residues to form. However, when cleaning dirty model plastics, aqueous cleaning agents failed to dissolve artificial dirt, particularly that representing finger prints (see figures 3 and 4). Instead of cleaning dirty plastics, aqueous cleaning caused further damage, because any undissolved particles were moved across surfaces causing abrasion. This means that identifying dirt types is of great importance when selecting cleaning agents. Visual examination suggested that both artificial dirts had started to migrate into the plasticised PVC within a week after application, causing deformation of the material (see figure 5). A month after cleaning had been performed the sebum soil was no longer visible whereas the organic oil soil appeared more particulate indicating that the paraffin oil had migrated into the PVC. Deformation of the material was also evident. It is likely that the plasticiser in PVC had dissolved the organic components of the artificial soils and transported them between the polymer chains and away from surfaces. The POPART results emphasize the importance of cleaning plastics especially PVC, immediately after soiling takes place. CLEANING PMMA A CASE STUDY In order to examine how cleaning research on model, new PMMA related to real museum objects, Light Sculpture without Light, made in 1976 by the Danish artist Gunnar Aagaard Andersen was selected as a study object. The sculpture, which belongs to the National Gallery of Denmark, comprises several white acrylic sheets, each ca. 5mm thick. These have been Figure 6. Cleaning tests were performed on the edges of Gunnar Aagaard Andersen s polymethylmethacrylate sculpture Light Sculpture without Light (1976) using triammonium citrate (TAC), Judith Hofenk de Graaff detergent, Orvus WA Paste and Dehypon LS45 (photograph by Kathrine Segel). Figur 6. Der blev udført renseprøver med triammoniumcitrat (TAC), Judith Hofenk de Graaff detergent, Orvus WA Pasta og Dehypon LS45 på kanten af Gunnar Aagaard Andersens skulptur Lysskulptur uden lys (1976) (foto af Kathrine Segel). 8 - ARTIKEL MED PEER REVIEW

adhered to form a series of concentric, open boxes. When placed in front of a natural light source, as originally intended by the artist, the transparency of the acrylic sheets admits the light and thus reveals the internal spaces of the object. The sculpture was chosen as a study object for two reasons. It consisted of colourless acrylic sheets and was evenly covered with a layer of dark oily soil and several fingerprints could be observed in raking light. In addition the lower part of the sculpture was severely scratched. Most scratches were filled with soil which darkened the object. Light Sculpture without Light represented both the model plastic and the artificial soils tested in POPART. Unfortunately the sculpture cannot be viewed as originally intended, because some of the joints have failed. Cleaning research into model PMMA suggested that solvents had performed slightly better than aqueous cleaning agents at removing both sebum and organic oil soils. However because the artwork was more than 30 years old and the plastic was visibly, physically damaged and therefore more vulnerable to stress cracking when in contact with solvents, caution was required. A selection of aqueous cleaning agents was tested along the lower edge of the sculpture (see figure 6). Although it was difficult to see a difference in effectiveness with the naked eye, under magnification it was clear that Orvus WA Paste and triammonium citrate were both the most effective at removing dirt and least damaging for new, model PMMA. It was decided to clean the sculpture with Orvus WA Paste 1% applied by microfiber cloth. This technique was successful in removing both the even soil layer and fingerprints but not soil trapped inside scratches. In these recessed areas a small cotton bud was found to be the most effective tool but cleaning had to be performed in the same direction as the scratch. Most of the soil was removed this way. Although the researchers were aware of the Figure 7. Migrating plasticiser from Crash Test Dummy Sierra Sammy s head is sticky and traps dirt from the surrounding environment (photograph by Yvonne Shashoua). Figur 7. Snavs fra omgivelserne sidder fast i klæbrig migrerende blødgører fra Crash Test Dummy Sierra Sammys hoved (foto af Yvonne Shashoua). risk of environmental stress cracking when using solvents in contact with plastics, trapped soil was successfully removed with ethanol using a microscope to help focus on scratched areas and to detect any microscopic damage including spreading of cracks. The sculpture was rinsed with distilled water applied by microfiber cloth after cleaning and no damage was detected. CLEANING PVC A CASE STUDY A crash test dummy, Sierra Sammy, belonging to the Science Museum in London was selected as our second case study. The dummy was made in the mid 1970s and acquired by the museum in 1979. Crash test dummies were used to model the motion of a human driver or passenger when applying the brakes of a vehicle and to predict any injuries arising from not wearing a seat belt. The object comprised a mechanical, steel structure to replicate the weight (60kg in this case) and movement of a human being and was covered with plasticized PVC skin to add textural and cosmetic properties. In contrast to Gunnar Aagaaard Andersen s sculpture where the soil layer was clearly only superficial, the PVC study illustrated different challenges. After 10 years in a well-sealed - ARTIKEL MED PEER REVIEW 9

showcase, the phthalate plasticiser included in PVC s formulation, had migrated to surfaces, leaving it shiny, discoloured and sticky. The plasticiser could be considered an internal degradation product. Fifteen years later, phthalate plasticiser started dripping from the PVC covering the object s head, hips and feet (see figure 7). The high concentration of phthalate vapour within the sealed showcase caused the vapour to condense and drip. Dust adhered to the sticky surfaces, thereby altering the object s appearance. Fourier Transform Infrared spectroscopy revealed that the dummy s soil layer comprised phthalate, carbon particles and red pigment which had been used to represent blood in the dummy s construction. POPART research had shown that PVC was cleaned more effectively by solvents than by aqueous cleaning products. However, because removal of internal forms of dirt, such as plasticisers, had not been investigated in POPART, additional information was required to develop a suitable cleaning strategy. Phthalate plasticizers are esters and their carbonyl (C=O) groups induce polarity but they cannot hydrogen bond with other materials readily. As a result they are partially soluble in water but more soluble when detergents or solvents are present. Because the crash test dummy was too large to be placed easily in a fume cupboard, solvent cleaning was not an optimal solution from a health and safety perspective and therefore aqueous cleaning was preferred. Initial cleaning trials using microfiber cloths and aqueous cleaning products were conducted by the authors. Afterwards, Science Museum conservation staff, none of whom had participated in POPART were instructed in how to perform the actual cleaning in order to examine whether conservators could apply POPART results to their own collections. Oily dirt and plasticizer could be removed using both non-ionic detergent Dehypon LS45 and anionic Orvus WA Paste applied by polyester microfiber cloth followed by a rinse in distilled water. Detergent was applied by cotton bud in recessed areas which were harder to access using a light microscope to focus the treatment. After cleaning, the dummy appeared paler in colour due to the removal of black dirt on its surfaces. The PVC skin had neither measurably shrunk nor embrittled despite the removal of some plasticiser. The dummy was then returned to display. Examination of the object 18 months after cleaning, suggested that phthalate plasticiser has begun to migrate to the object s surfaces once again which indicates that the treatment will require repeating and perhaps reviewing. CONCLUSION Research into conservation cleaning of plastics is an area which had not been extensively researched prior to POPART s launch in 2008. Some of the major findings have been counterintuitive and are likely to change conservation cleaning practice for plastic design objects and artworks. For example, mechanical cleaning using a dry brush, cloth or sponge, has long been perceived as the least damaging technique to remove dirt. However it is ineffective for oily or biological dirt such as fingerprints and induces scratches on both thermoplastic and thermosetting materials. Adding sufficient lubricant in the form of water, detergent solution or solvent to the mechanical cleaning material was more effective at removing dirt and less damaging to the plastic substrate. Of the 22 cleaning products evaluated, only canned air, natural and synthetic feather dusters left all plastics unchanged. Duzzit sponges, Scotch Brite sponges and all paperbased products induced more scratching than brushes and cloths. Akapad yellow and white sponges, compressed air, latex and synthetic sponges and goat hair brushes deposited residues. Based on Cleaning Vectors, distilled water, Orvus WA Paste, Judith Hofenk de Graaff and Dehypon LS45 were the least damaging cleaning agents for polymethyl methacrylate, plasticised PVC, high density polyethylene, high impact polystyrene and cellulose acetate. Organic solvents were effective at removing both types of artificial dirt from model plastics but the increased risk of removing plasticisers or inducing stress crazing counted against their use where aqueous alternatives were possible. Transferring the results obtained by cleaning model plastics to real or study objects required some adaptations. Dirt on real objects was rarely homogeneous or from one source as evaluated

in POPART experiments, but could include internal dirt such as migrated plasticiser from PVC and its associated degradation products. Plastics themselves were often degraded and discoloured. With time, dirt had migrated further into the body of the plastic. Despite the additional challenges, the general findings for poor and effective cleaning materials and agents on the model plastics and artificial dirt evaluated in POPART were also true for real objects. It is clear that the flow charts designed to guide conservators when cleaning plastics offer useful starting points and may be found freely on http://popart-highlights.mnhn.fr/ solvent-based cleaning of both new, model plastics and real objects is presented. A PMMA sculpture from 1976 by Danish artist Gunnar Aagaard Andersen and a Crash Test Dummy in plasticised PVC from the mid 1970s are selected as study objects and successfully cleaned using the guidelines developed for model plastics and artificial dirt. KEY WORDS POPART, plastics, polymethylmethacrylate, polyvinylchloride, mechanical cleaning, aqueous cleaning, solvent cleaning ACKNOWLEDGEMENTS The authors wish to thank conservators Jannicke Langfeldt and Natsumi Henzan from the Science Museum in London and Louise Cone from the Danish National Gallery for their helpful collaboration on this project. The research leading to these results has received funding from the European Community s 7th Framework Programme FP7/2008 2013 under grant agreement no. 212218 (http://popart. mnhn.fr/). SUMMARY Although all surveys of the condition of plastics conducted in museums in the United Kingdom and Scandinavia since the 1990s conclude that approximately 75% of collections require cleaning, few treatments have been developed mainly due to the high risk of damaging plastics mechanically or chemically using an invasive cleaning treatment. Despite the high risks, removal of oily fingerprints, carbonaceous dirt (external dirt) or crystalline, acidic degradation products (internal dirt) from museum objects or artworks comprising plastics is essential to maintain their significance, chemical stability and commercial value. This article presents the results of structured research into cleaning polymethylmethacrylate (PMMA) and plasticised polyvinylchloride (PVC) which comprised part of the collaborative EU 7th Framework project POPART (Preservation Of Plastics ARTefacts in museums) between 2008 and 2012. An exhaustive study of the risks associated with mechanical, aqueous and YHTEENVETO Iso-Britannian ja Skandinavian museoissa tehdyt muovisten museoesineiden ja taideteosten kuntokartoitukset osoittavat, että arviolta 75 % kokoelmista on puhdistusta vailla. Puhdistusmenetelmiä on kehitetty vain vähän, koska voimakas mekaaninen tai kemiallinen puhdistus voi aiheuttaa vaurioita. Muoviesineiden merkityksen, kemiallisen vakauden ja kaupallisen arvon säilyttämiseksi on riskeistä huolimatta kuitenkin välttämätöntä poistaa rasvaiset sormenjäljet, hiilipitoinen lika (ulkoinen lika) ja happamat hajoamistuotteet (sisäinen lika). Tämä artikkeli kertoo järjestelmällisestä tutkimuksesta, jossa puhdistettiin polymetyylimetakrylaattia (PMMA) ja plastisoitua polyvinyylikloridia (PVC). Tutkimus oli osa EU:n seitsemättä Framework-yhteistyöprojektia, POPARTia, joka toteutettiin vuosina 2008-2012. Artikkelissa esitellään tutkimustulokset kaikista mekaanisen puhdistuksen ja vesi- ja liuotinpuhdistuksen riskeistä. Tutkimuksessa puhdistettiin sekä uusia muovisia mallikappaleita että varsinaisia esineitä. Mallikappalemuovin ja keinotekoisen lian avulla kehitettiin puhdistussuositukset, joiden mukaisesti puhdistettiin Gunnar Aagaard Anderssenin PMMAveistos vuodelta 1976 ja plastisoidusta PVC:stä valmistettu kolarinukke 1970-luvun puolivälistä. AVAINSANAT POPART, muovit, polymetyylimetakrylaatti, polyvinyylikloridi, mekaaninen puhdistus, vesipuhdistus, liuotinpuhdistus - ARTIKEL MED PEER REVIEW 11

RESUMÉ Selv om alle undersøgelser af plastmaterialers bevaringstilstand udført på museer i Storbritannien og Skandinavien siden 1990'erne konkluderer, at cirka 75% af samlingerne kræver rensning, er der kun blevet udviklet ganske få behandlingsmetoder, hvilket bl.a. skyldes risikoen for at beskadige plastmaterialet mekanisk eller kemisk som følge af en invasiv rensningprocedure. På trods af disse risici er det altafgørende at kunstværker og museumsgenstande renses for fedtede fingeraftryk, kulstofholdig snavs (ydre snavs) eller sure og krystallinske nedbrydningsprodukter (indre snavs), hvis genstandenes betydning, kemiske stabilitet og kommercielle værdi skal bibeholdes. Denne artikel præsenterer resultaterne af en systematiseret forskning indenfor rensning af polymethylmethacrylat (PMMA) og blødgjort polyvinylchlorid (PVC), som mellem 2008 og 2012 udgjorde en del af det europæiske 7th Framework samarbejdsprojekt POPART (Preservation Of Plastics ARTefacts in museums). En udtømmende undersøgelse af de risici, der er forbundet med mekanisk, vand- og solventbaseret rensning af henholdsvis ny model plast og rigtige genstande præsenteres. En PMMA skulptur fra 1976 af den danske kunstner Gunnar Aagaard Andersen og en Crash Test Dummy i blødgjort PVC fra midten af 1970 erne blev udvalgt som test objekter, og begge genstande blev renset med tilfredsstillende resultater ved at følge de udviklede retningslinjer for model plast og kunstig snavs. Saarikoski, I., Sjöberg, A.-M. and Hautala, M., Utilization of profilometry, SEM, AFM and contact angle measurements in describing surfaces of plastic floor coverings and explaining their cleanability. Surface Science, vol. 584, pp. 119-125. Shashoua and Ward (1995): Shashoua, Y. and Ward, C., Plastics: modern resin with aging problems. American Journal of Physics, 76 (11), 1995, pp. 1074-1077. Shashoua et al (2012): Shashoua, Y., Balcar, N., Barabant, G., Bollard, C., Kuperholc, S., Keneghan, B., Laganá, A., Oosten, T. van and Segel, K., Studies in active conservation of plastic artefacts in museums in Lavédrine, Fournier and Martin, G. Preservation of Plastic Artefacts in Museum Collections. Éditions du Comité des travaux historiques et scientifiques, 2012, pp. 219-269. CORRESPONDING AUTHOR YVONNE SHASHOUA PHD, SENIOR RESEARCHER NATIONAL MUSEUM OF DENMARK yvonne.shashoua@natmus.dk & KATHRINE SEGEL CAND.SCIENT.CONS NATIONAL GALLERY OF DENMARK kathrine.segel@smk.dk NØGLEORD POPART, plast, polymethylmethacrylat, polyvinylchlorid, mekanisk rensning, vandbaseret rensning, solventbaseret rensning REFERENCES Bollard (2008): Bollard, C., Quelles substances endogens trouve-ton à la surface des matières plastiques? Faut-il les éliminer? Conservation Restauration des Biens Culturel 26, p. 33-38 http://popart-highlights.mnhn.fr/ Highlights of European Community s 7th Framework Programme FP7/2008 2013 Preservation of Plastic ARTefacts in museums project Koponen et al (2007): Koponen H.-K., Suvanto, M. and Pakkanen, T. A., Soiling of plasticized poly (vinyl chloride). Journal of Applied Polymer Science, vol. 105. pp. 3047-3053. Kuisma et al. (2005): Kuisma, R., Personen- Leinonen, E., Redsven, I., Kymäläinen, H.-R., 12 - ARTIKEL MED PEER REVIEW

A CHANCE FOR DAYDREAMING: CONSERVATION-RESTORATION AND TECHNICAL RESEARCH OF AN OUTDOOR UPR PLASTICS INSTALLATION FROM 1992 PÄIVI KYLLÖNEN-KUNNAS AND ULLA KNUUTINEN INTRODUCTION This article is based on the material and technical research, and conservation project, which together combined to form the basic strategy after the arm of the sculpture titled The Mermaid was broken off and thrown into the Oulu river estuary, where a passer-by found the arm a week later in 2003. The initial impetus for the article was the writers interest in technical workmanship and material examination of artworks made from UPS composites. In the article we will examine the sculptor and performance artist Jan-Erik Andersson s (1954 - ) artwork, an outdoor installation 'The Last Milk Platform' (1992) as a case study project. The artwork is particularly accessible, because by-passers are welcome to sit on the platform and are invited to write fragments of dreams on the inside of the blue walls of the installation. The project involved co-operation between a polymer chemist and a paintings conservator who specializes in contemporary artwork and created a new theoretical and scientific approach for pioneering the technical research and the conservation/restoration of contemporary artworks made from non-traditional materials. The analytical research project Fig 1.Fig 1. Jan-Erik Andersson s the Last Milk Platform (1992) was reinstalled in the art square beside Oulu City Art Museum in 1995. During a celebration on 15.6.1995 Jan-Erik Andersson gave a speech and the student group Exema and the cow Rousseau gave an performance. Mrs. Jaana Laamanen gave alecture on the significance of milk platforms. Photographer: Päivi Kyllönen-Kunnas - ARTIKEL MED PEER REVIEW 13

has brought to light more detailed information on commercial unsaturated polyester products and UP composites. Through careful documentation, which included interviews, examination and analysis, knowledge was acquired to enable informed conservation/restoration of the art installation in question. Through the identifying characteristics of modern materials we are able to classify each single construction of the artist s oeuvre, for appropriate exhibition, storage and conservation of the artwork. MATERIAL TECHNICAL RESEARCH AND ANALYSIS TECHNIQUES UPR MATERIALS IN ART Unsaturated polyester resins, UPR, have been used as contemporary art materials for the last 20-40 years because they represent inexpensive, easy to use materials and form rigid but light-weight structures. In artworks it has been popular to use unsaturated polyester resin to create strong three-dimensional shapes of fibreglassreinforced polyester. Unsaturated polyester resins are formed from the reaction of polyhydric alcohols with dibasic acids or anhydrides, and then subsequently dissolved and cross-linked in an unsaturated monomer, usually styrene, to produce a rigid three-dimensional polymer network, that has important applications in glass-fibre reinforced plastics (Bergens L., 1995; Whelan A., 1995). In practice, things are much more complicated in commercial products, where several different alcohols and unsaturated acids may be used in the preparation of the polyesters, and many different monomers are used as diltents. Further, the addition of a monomer unit, styrene, can be as high as 50% depending on the product, and is always added in excess. Further two-pack systems of unsaturated polyesters are cured by a free radical mechanism started by adding initiators and accelerators based on peroxides. The physical properties and chemical stability are adjusted using plasticizers and inhibitors (Hare C.H., 1994; Penczek P., 2005). When unsaturated polyesters are used as part of a composite material with gypsum and fiberglass, they are considered stable materials with good chemical resistance and weathering qualities. However, prolonged exposure to ultraviolet light, as well as changes in temperature and humidity, can cause chemical and physical degradation to unsaturated polyesters (Norman N.S. and Edge M., 1992). This is especially true of artefacts made of unsaturated polyester resin, reinforced with fiberglass, and exhibited in the open-air. The ageing of synthetic composite resin materials occurs faster than more traditional craft materials, such as stone and ceramics. Museum objects made of polymers like UPR can sometimes degrade very rapidly, and may have a useful lifetime of only a few decades. The chemical composition and manufacturing process of the polymers play a major role in shaping the degradation behavior. For this reason analytical techniques were needed that could identify UPR polymer materials, and study the effects of degradation on their chemical and physical properties. UPR MATERIAL RESEARCH In order to study the causes of degradation and to plan the conservation project, the authors organized a pilot project for UPR composite materials in which three different commercial UPR products and samples of the Mermaid from the installation of the Last Milk Platform by Jan Erik Andersson were studied. The choice of commercial products was made on the basis of the information provided by the artist in a series of interviews. Similar UPR product formulations that were known to have been used by the artist for the construction of the art installation were studied as references. Preliminary results of the pilot project, which were published in the eps (e-preservation Science) in 2006, showed that there are differences in the stability and ageing properties of different commercial UPR products. Damage, yellowing, degradation and photo-oxidation were verified through VIS (Visible light) stereo-microscope investigations, CIE L*a*b* colour measurements and FTIR (Fourier Transform Infrared Spectroscopy) analyses. VIS microscope studies showed that UPR and their fibreglass composites are very porous materials which allow increased physical and chemical degradation. IR spectroscopy indicated some variations in the chemical composition of both unsaturated polyester resin products and samples from the artwork(knuutinen U. and Kyllönen P., 2006). 14 - ARTIKEL MED PEER REVIEW

NEW RESEARCH METHODS FOR UPR STUDIES Research was continued after the pilot project, in order to gain more detailed information about the chemical composition and its effect on curing as well as cross linking density, and the deterioration mechanism of commercial unsaturated polyester products and UPR composites. As a reference material one more commercial UPR product was added. Because the degradation of UPR materials is both a chemical and physical process, all four products: A, B, C and D 1 were first studied with physical breaking strain tests. After that, chemical analyses were performed with NMR (Nucleus Magnetic Resonance) Spectroscopy, and the original UPR samples from the Mermaid were analysed using NMR as well (Stamatakis G. et all., 2010). Although for 50 years NMR has been a very powerful tool in many fields, its application in the field of cultural heritage is rather recent. Usually in this field either liquid or gas chromatography followed by mass spectrometry have been applied to separate molecular mixtures and their components. In fact, complex mixtures of organic compounds may be successfully analysed by NMR using acetone extraction in a very simple preparation. NMR spectroscopy is able to provide insight into both the chemical composition of UPR formulations and of the extent of the degradation processes that affect the physical properties during aging (Capitani D.et all., 2011; Stamatakis G. et all., 2010). For the mechanical tests 2 samples of four different commercial reference UPR products were subjected to different weathering conditions including: humidity, melt/freeze cycles and UVB exposure. It was interesting to note that before aging all four UPR products had similar mechanical properties. This observation stressed the importance of aging studies to determine the suitability of plastics and polymers for use as cultural heritage materials. The mechanical tests showed that UVB radiation has a significant effect on the strength of UPR materials. In general the mechanical data suggested that one of the tested UPR resins (symbolized by C) was the one most susceptible to the deteriorating effects of aging. This conclusion was corroborated by the NMR analysis, which showed that UPR sample C extracts contained increased amounts of phthalates as a plasticiser, and 1.2 propylene glycol (PG) units after UVB aging, compared to the other resin samples, whose acetone extracts were slightly lower in phthalate and PG. UPR formulation C also contained the lowest amount of styrene in the original commercial formulation of the UPR products. UPR resins of low styrene content have been shown to possess inferior mechanical properties compared to traditional UPRs. The reduction in styrene content requires a reduction in the molecular weight of the polyester pre-polymer to maintain the low viscosity of the formulation prior to cross -linking. This increases the number of free polar end groups (-COOH, -OH) available to interact with water, with the higher water absorption leading to more facile osmotic cracking of the UPR formulation (Stamatakis G. et. all., 2002; Spyros A., 2002). Interesting NMR results were especially prevalent in commercial UPR product C, because the art installation sample had almost identical NMR features with the aged UPR C samples. This indicated that the UPR resin C or a resin of very similar composition had been used by the artist. in the construction of this work. Furthermore, the NMR analyses showed that UPR resins contain several low molecular weight organic compounds, including plasticizers and low molecular weight polymers, known as oligomers, rich in polar OH groups that play a significant role in the degradation behaviour of the composite UPR materials. With NMR it was possible to reveal more detailed secrets behind the chemical degradation process in The Mermaid. Polystyrene (PS) fragments which were produced during UPR degradation were found. The presence of PS fragments was expected of UPR materials that have suffered heavy weathering, and that have been exposed to high humidity and extreme temperatures as a result of being exhibited outdoors for several years. Extended degradation can lead to large fractions of polymeric chains being chopped off the three-dimensional macromolecular network that includes polystyrene fragments created during cross-linking. The surface of The Mermaid was exposed to weathering and was therefore continuously subject to chemical alteration by environmental factors. At the same time the surface was the interface through which volatile chemical components, which formed during the ageing, escaped from the polymer matrix (Stamatikis G. et all. 2010). - ARTIKEL MED PEER REVIEW 15

THE ARTIST S MATERIALS FOR THE INSTALLATION The installation The Last Milk Platform (1992) consists of a white, blue and black wooden milk platform, with an aluminium milk churn and a red reinforced-polyester mermaid on its roof. A metal stick is fastened to the milk churn pointing outwards with a black and white bat and a yellow ball. The artist and a group of assistants built the installation for an exhibition in Sotkamo. It was later purchased for the Oulu City Art Collection in 1993. Andersson repaired some damage to the artwork before its re-installation besides the main building of Oulu City Art Museum in 1995, with pots of new UPR paint and other materials. Later, in 2001, he was interviewed and asked to provide step-by-step, detailed structural and material information about the artefact. It was revealed that the form of The Mermaid was made from plaster-of-paris bandages, used in sculpting the human figure. Polyurethane foam was extruded into the inside of the cast figure and some polystyrene, wood and metal particles were added. Then unsaturated polyester resin, as a composite with fiberglass, was used over a hand laid mould and covered with pigmented unsaturated polyester paint. (Kyllönen P., 2005). UPR with fibre-glass (chopped-strand matting) has been one of the preferred plastic materials used by Finnish artists for three dimensional work in art pieces. On The Mermaid a reinforced UP resin matrix over a positive mould was made with a brush while working outside in the summer 3. The layering process using cross-linked polyester resins is often cold-cured, and for this purpose curing at room temperature is desirable (Nicholson J.W. 1997). Laterin the process a two component paint was added to the polyester using a chemical initiator, which forms the rigid chemical bonds of the polymerised polyester. This results in a hard and shiny/smooth finish. In the case of The Mermaid, it is the Mermaid herself which is used as the open (positive) mould for structures made using a manual technique. The glass fibre is laid first, and then brushed over with resin until it is saturated, forming the top,coat of paint.(saarela et al. 2003; Kessler K., 2003; Birley et al. 1994) 4. Fig 2. Artist s sketch for the Last Milk Platform. The artist writes in the drawing over the artwork that it is meant to highlight the feelings of daydreaming. He says that I myself have spent time on several occasions taking part in philosophical discussions at the milk platform. THEORETICAL APPROACH ON CONSER VATION DECISIONS ETHICS IN CONSERVATION AND RESTORATION OF OUTDOOR SCULPTURES What makes an artwork original? For heritage preservation, the ethical and moral (critical) perception of decisions and recommendations made by conservators instil the decisions with a certain ethical and moral guarantee because of their professional qualifications (for example ICOM International Council of Museums membership). Conservation theory is unambiguous, where the importance is to maintain the physical, aesthetic and historical integrity of the ob- 16 - ARTIKEL MED PEER REVIEW