Maney Publishing

Long-Term Chemical and Physical Processes in Oil Paint Films Author(s): David Erhardt, Charles S. Tumosa and Marion F. Mecklenburg Source: Studies in Conservation, Vol. 50, No. 2 (2005), pp. 143-150 Published by: Maney Publishing on behalf of the International Institute for Conservation of Historic and Artistic Works Stable URL: http://www.jstor.org/stable/25487732 . Accessed: 05/12/2014 10:58

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp

. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

.

Maney Publishing and International Institute for Conservation of Historic and Artistic Works are collaborating with JSTOR to digitize, preserve and extend access to Studies in Conservation.

http://www.jstor.org

This content downloaded from 160.111.254.17 on Fri, 5 Dec 2014 10:58:14 AM All use subject to JSTOR Terms and Conditions 143

Long-Term Chemical and Physical Processes inOil Paint Films

David Erhardt, Charles S.Tumosa andMarion F.Mecklenburg

Oil paints dryby polymerization. This 'drying'process may be substantiallycomplete and thesurface of thepaint film dry to the touch within but measurable continue slower weeks, changes for years. Other, processes also continue, primarily hydrolysis of esters. as case glyceride This produces carboxylic acid groups eitherfree fatty acids (in the of acid groups that have not reacted or acid bound to the oil matrix case otherwise) groups crosslinked (in the of acid groups that have engaged in polymerization These react with to in the case a reactions). may pigments form carboxylate salts (called soaps of fatty acid). These changes affect the the and the that conservation treatments it. physical properties of paint way affect This paper examines the extent of and in some and hydrolysis soap formation naturally aged drying oil paint films, the extractability of these materials in organic and measured and over time in solvents, predicted changes thephysical properties of naturally aged paint films. Long-term and mechanical due to are minor to or physical changes aging compared those produced by overcleaning excessive exposure to heat.

or INTRODUCTION fatty diacids formed by scission reactions of the unsaturated acids. fatty The acid groups produced by Plant oils are of esters composed primarily triglyc?rides, react hydrolysis may with metal ions from pigments to of glycerol and fatty acids. In drying oils, most of the produce carboxylate salts (referred to as soaps in the case acids from which the are derived are fatty triglyc?rides of a Free are fatty acid). glycerol molecules formed only oils, and oil polyunsaturated. Drying paints compounded after all three are so glyceride bonds hydrolyzed, that from them, dry by a process of autoxidation and subse little free glycerol will be present unless extensive quent polymerization of the unsaturated fatty acid hydrolysis has occurred. The effects of cannot in hydrolysis groups the triglyc?rides. This process is quite complex be understood unless the influence of other processes is the (see, e.g., review by Wexler [1]). An oil paint film also known. In this study the bulk mechanical that is to the touch properties dry within weeks undergoes further of several paint films are followed and the extent of the reactions for decades longer. Chemically this includes hydrolysis and oxidation reactions examined. further crosslinking reactions, oxidation of unsaturated acids and hydrolysisof glyceride bonds. Hydrolysis is a a of oils significant chemical reaction in paint film even in the Hydrolysis first few years [2]. Hydrolysis may yield saturated fatty The reaction that produces the ester bonds in oils can be acids (which lack the functional that react groups during reversed. This process, called hydrolysis, yields the the unsaturated acids that crosslinking process), fatty and component fatty acids glycerin. Partial hydrolysis, of have not yet become part of the crosslinked oil matrix one or two ester bonds, di- and or yields monoglycerides, otherwise reacted, acid groups attached to the matrix respectively. If any of the fatty acids have reacted to bonds formed and short-chain by during polymerization, become of the part polymer matrix, hydrolysis yields an acid attached to group the polymeric matrix by the ReceivedMarch 2003 bond formed during polymerization. That hydrolysis of

STUDIES IN CONSERVATION 50 (2005) PAGES 143-150

This content downloaded from 160.111.254.17 on Fri, 5 Dec 2014 10:58:14 AM All use subject to JSTOR Terms and Conditions 144 D. ERHARDT, C. S. TUMOSA AND M. F. MECKLENBURG

occurs a extent was more in some glycerides has been well known for long time. of hydrolysis than 90% older a The hydrolysisof oils (and fats)with alkali to produce paint samples. However, they also made number of was one statements other of soaps (the fatty acid salts of the alkali) probably unsubstantiated regarding aspects of the first chemical reactions discovered. The pioneer aging, especially regarding the physical and mechanical ing chemistCarl Wilhelm Scheele firstisolated glycerin aspects of aging. It is unlikely that the evaporation of a in the late 1700s (he also discovered oxygen, nitrogen, glycerol, which has boiling point of 290?C (with a at normal chlorine, and other elements and compounds). Scheele decomposition), would be significant process a matrix such showed that glycerin and fatty acids could be produced temperatures, especially from within solid as by the action of lead oxide on fixed (vegetable) oils in the polymerized oil. In fact, their results show that was more a are the presence of water [3] (it not until than significant amounts of glycerol present in quite old was to the hundred years later that the term hydrolysis used paints. Work by the present authors showed that describe the reaction). It was realized very early in the relative amounts of glycerol increased with increasing was due to the fact that all three last century that hydrolysis of oil responsible for hydrolysis, primarily gly some ester must to free many of the changes (some good, bad) in the pro cerol bonds undergo hydrolysis yield time Boon et al. also inferred from their data that perties of paints stored for long periods of (years), glycerol [8]. water are in an ionomeric form without that the presence of promoted hydrolysis and soap the paints mostly the relative amounts of free acid formation [4], and that completely ionomeric paints presenting data showing from of versus cite studies could be prepared directly soaps hydrolyzed carboxylate present. They showing the that old have little extractable material as drying oils [5]. O'Neill and Brett, in examining paintings may at their statement that the free acids are reactions taking place in paint films, looked the supporting present as was amount of material that could be extracted from oil paint soaps; however, the study cited of previously amount ion in the cleaned solvent that had films after aging, and the of metal (i.e., treated) paintings likely little extractable material left.Work the extracted material. They concluded that paint films easily by present more that if not of the free reacted fastest with basic oxide pigments, slowly authors shows many, most, fatty even more if at molecules are as the acids rather than as with carbonate pigments, and slowly, all, acid present that the Boon et al. also mention that con with inert oxide pigments. They also concluded soaps [8]. repeatedly to the ionomeric form is for the formation of metal soaps contributes to the particular version responsible film hard and this properties of white lead and zinc white oil paints [2]. paint becoming brittle, although association has not been demonstrated in their or any the other study. To the contrary, previous work by pre Hydrolysis of dried oil paint films sent authors, summarized in this article, shows that the some free and stiffness of films in a that While hydrolysis and soap formation with strength paint change way oils occurs can be modeled and over fatty acids present in drying very quickly, mathematically, changes long films occurs over a time of can be The rate of extensive hydrolysis of paint periods aging predicted. changes concern Most oil in these slows as the span longer than is of industrially. paint properties considerably paints age, on occur in before and a film hundreds of old should not be research focused changes that paints paint years or the after much stiffer or more brittle than a film application during drying process applica paint only oil was of con decades old. If a is to become brittle, as for tion. Hydrolysis in unpolymerized great paint going instance zinc oxide it does so cern because it affects the mixing, storage and working pigmented paints do, quite on the behaviour of It is that films do properties of the paint. Research early. easily shown, though, paint focused on the oxidative become brittle after to paint after application, though, prolonged exposure high temp since and forma eratures or solvents The authors have in polymerization process, hydrolysis soap [9, 10]. present a number of tion in the dried oil were much slower. Consequently, fact prepared completely hydrolyzed paint Boon and co films from mixtures of and free acids few reports examined the process until pigments fatty These workers reintroduced the concept and showed that (derived from the hydrolysisof dryingoils). films, extensive over the museum which are now several old, are coherent and some hydrolysis could be quite years a method that are as flexible as from the time-scale [6, 7]. Using derivatization comparable paints prepared and non-esterified oils. This is the fact that which distinguished between esterified original despite glycerol, not between free acids and would act as a was removed the carboxylic acid groups (but plasticizer, during of of the oils and is not Studies of these ionic, or soap, moieties), they showed that the degree hydrolysis present. with and concluded that the will be later. hydrolysis increased age, paints reported

STUDIES IN CONSERVATION 50 (2005) PAGES 143-150

This content downloaded from 160.111.254.17 on Fri, 5 Dec 2014 10:58:14 AM All use subject to JSTOR Terms and Conditions LONG-TERM CHEMICAL AND PHYSICAL PROCESSES INOIL PAINT FILMS 145

as SAMPLES AND EQUIPMENT relative amounts of fatty acids present free acid, soap, ester. were and Small amounts of glycerides observed in Paint consisted of films of and linseed samples pigment extracts the chromatograms of the of the unhydrolyzed oil unless otherwise with no (cold-pressed specified) were samples; however, these small enough not to affect other additives. were in the labora They naturally aged the calculations and were not taken into account. at tory 40-50% RH and 22?C under ambient light. were cast on were Paint films polyester sheets, and as were RESULTS AND DISCUSSION removed needed. Oil specimens the same cold oil from which the were pressed paint samples prepared. Hydrolysis were Paint specimens also removed from fragments of at was panel paintings least 200 years old. These fragments Hydrolysis within two years measured to be as are from French from the as panel paintings dating extensive 20% in specimens of pure cold-pressed or same eighteenth century that had previously been donated for linseed oil in the pigmented oil stored under were on a research purposes. Specimens weighed conditions of 40-50% RH and 22?C, within the envi model to museums. Mettler balance, AT201, within 0.01 mg. ronmental range typically recommended for Extractions were in mL performed 50 of reagent-grade Analysis of oil paint films from unprovenanced solvents for 24 hours. measurements were Stress-strain European panel paintings at least 200 years old indicates made on screw-driven tensile the extent some testers, samples being that the of hydrolysis approaches 80% in 0.2 x 6 x 125 mm. The cases. a can approximately procedure has Hydrolysis is process that be detected early been described previously [11]. Fatty acid composition in the drying of paint films. Increases in the amounts of was determined and can quantified by gas-chromatography free saturated fatty acids (palmitic and stearic) be as previously described [8]. detected within a year. These acids are not involved in The of ionized and esterified percentages free, (soap), any oxidative process (e.g., crosslinking) and although acids were determined as Free acids were account fatty follows. they generally for less than 10% of the total fatty removed from a small extraction a are paint sample by with acid composition of linseed oil, they preferentially excess of large dichloromethane, accompanied by located in triglyc?ride positions 1 and 3 and are more of the with a rod to ensure com grinding sample glass easily hydrolyzed than acids located at the 2-position extraction. The solid was an plete residue separated and [12]. Therefore, their hydrolysis rate represents upper rinsed with more which was also and limit on rate solvent, separated the total hydrolysis of the polymerized combined with the first extract. This combined extract glyceride units. Azelaic acid is formed by oxidative contains the free fatty acids. Several milliliters of diethyl scission at the C9 double bond in the 18-carbon unsatu ether were then added to the A of solid residue. drop rated acids (oleic, linoleic, and linolenic). Figure 1 plots dilute acid was to convert amount a (1 N) hydrochloric added any the total of azelaic acid in cold-pressed linseed to are soaps free acids, which extracted (with agitation) into the ether, which is removed. The acidified sample two more is extracted with aliquots of ether, and the are ether extracts combined. This step is conducted to quickly minimize any hydrolysis of the esterified water acids. The is evaporated from the solid residue, and the remaining esterified acid groups are removed from the residue by standard procedures, i.e., hydrolysis with KOH in methanol, acidification, and extraction with ether. The three extracts (free, soap, and ester are acids) evaporated. A known amount of tetratriacon tane (the 34-carbon straight-chain alkane) in hexane is added to each of the three extracts as an internal stand are ard. The samples trimethylsilylated with N,0 6/s(trimethylsilyl)trifluoroacetamide (BSTFA) and Years as analyzed by gas-chromatography previously described areas to areas Figure 1 Proportion of azelaic acid ina cold-pressed linseed oil film [8]. The ratios of the of fatty acid peaks the over time,expressed as the ratio of the amount of azelaic acid to the of the tetratriacontane are used to calculate the peaks (constant) amount of palmitic acid.

STUDIES IN CONSERVATION 50(2005) PAGES 143-150

This content downloaded from 160.111.254.17 on Fri, 5 Dec 2014 10:58:14 AM All use subject to JSTOR Terms and Conditions 146 D. ERHARDT, C. S. TUMOSA AND M. F. MECKLENBURG

oil film as a function of time. The amount of azelaic acid 0.25 as amount to is plotted the ratio of the of azelaic that of palmitic acid (which remains constant, since palmitic acid does not react). If azelaic acid is formed from unsaturated fatty acids bound by glyceride ester bonds, freepalmitic acid to the resulting azelate group is also attached the azelate or must ?rolyzed glyceride polymerized oil matrix, and hydrolysis occur before the azelate is present as the free acid. Two jl patmitatesoap """' separate reaction processes (oxidation at the C9-position azelatesoap of unsaturated acid and of the ester * '00>m groups hydrolysis j** *" freeazelaic acid bond) must occur before free azelaic acid is formed, so the rate of formation of free azelaic acid will be lower rate than the of the hydrolysis step. Hydrolysis of azelaic 10 15 acid precursors can occur before the oxidation step, but the free azelaic acid produced by this reaction sequence Years seen will only be after subsequent oxidation. Thus the Figure 2 Proportions of azelaic and palmitic acids present as free rate of formation of free azelaic acid (relative to the final acid, soap, and total hydrolyzed (freeplus soap) in a lead white in total amount of azelaic acid is a lower limit of produced) cold-pressed linseed oil paint filmas a functionof drying time. The rate one the of the reaction, hydrolysis of unsaturated amounts are expressed as fractions of the total acid present in the acid bonded fatty acids that oxidize to azelaic acid. Unsaturated paint (freeplus soap plus by glyceride esters). acids (or their polymerized forms) fill the less easily to hydrolyzed 2-position in glycerol preferentially the Oxidation rate saturated acids, and the of hydrolysis of their ester can glyceride linkages should be slower than that of The autoxidation process of paint films be monit the saturated fatty acids, but faster than that of the ored by measuring the initial weight and weight changes over curves as as appearance of hydrolyzed azelate. The measurement of time (so-called Weger [13-15]) well hydrolyzed acids (palmitic, stearic, azelaic) is compli by determining the amount of azelaic acid formed. a common react cated by process. Free fatty acids will Azelaic acid may be formed by oxidative scission of the to at with many pigments form soaps. Soap formation ties double bond the C9-position of the unsaturated fatty to cross up any free acid and prevents it from being easily acids and is initially bound glyceride (or the or extracted by solvents lost by thermal evaporative pro linkedmatrix) by the original glyceride ester linkage. cesses. Free azelaic acid, with two acid groups, should On hydrolysis it becomes a free acid and may bind to as as a be approximately twice likely to be present salt. metal ions in the pigment. In pure Unseed oil, azelate treat was to an By carefully selecting solvents and by mild acid formation found have incubation period of 5 as ment the relative proportions of fatty acids present 10 days. The amount increased rapidly for about two can more glycerides, free acids, and soaps be determined. months, and then slowly for the next two years. amounts Figure 2 shows how the of azelaic and palmitic This parallels nicely the measured weight gains of the acids present as free acid and as soap increase over time bulk paint films. Figure 3 shows a typicalweight gain a curve in lead white paint film. The total fraction of hydro for lead white in cold-pressed linseed oil, which can seen can lyzed acid (free plus soap) is also shown. It be be compared with the rate of formation of azelaic amount as an that for palmitic acid, the present the free acid acid shown in Figure 1. After incubation period of as two is about twice that present the soap. As expected, the several days, the paint gains weight rapidly for about more fraction of hydrolyzed azelaic acid is less than that of weeks, and slowly for several months. Longer a a palmitic acid, and greater percentage of the hydrolyzed aging results in slow weight loss. It is clear from the as a difunctional azelaic acid (about half) is present soap. initial rapid weight gains that uptake of and reaction one or are (An azelate 'soap' may have both carboxylic acid with oxygen primarily responsible for the initial to curve. once groups converted the carboxylate salt; the experi shape of the Weger However, most of the not are mental method does distinguish between the two sites that readily oxidized have reacted, the contribu were as possibilities.) Data for stearic acid also determined tion of other, concurrent processes such the loss of are to can seen as a and nearly identical those for palmitic acid, but volatile compounds be gradual loss of are omitted for clarity. weight. These low molecular weight compounds may

STUDIES IN CONSERVATION 50(2005) PAGES 143-150

This content downloaded from 160.111.254.17 on Fri, 5 Dec 2014 10:58:14 AM All use subject to JSTOR Terms and Conditions LONG-TERM CHEMICAL AND PHYSICAL PROCESSES INOIL PAINT FILMS 147

occur two changes within years, with slower changes over next continuing the 10 years. Since the weight of or the paints is either increasing stable during this time, the reduction in the amount of extractable material to (mobile phase) must be due the incorporation of some of this material into the polymer matrix (through or continued crosslinking), the conversion of free fatty acids to insoluble metal salts, rather than a loss of volatile material.

0# Mechanical 0.5 1.5 2 2.5 3.5 properties

Years Fully hydrolyzed paint films can be produced relatively quickly by heating at both highRH and high tempera Figure 3 Change inweight of a lead white in cold-pressed linseed oil ture. Paint heated under conditions paint filmas a functionof drying time. samples ranging from 75 to 100% RH and 80 to 100?C hydrolyze and become brittle and fractured. The as completely easily be present in the original oil, formed byproducts brittleness of these however, is due to the loss or as samples, during the initial oxidative polymerization, later of low molecular weight plasticizers rather than changes the film oxidizes over time. paint slowly to a brought about by conversion theoretical ionomeric composition. A fully hydrolyzed paint film can be and means a Polymerization crosslinking created through synthetic starting with hydro lyzed linseed oil, grinding in pigment, and casting the In the of and cross early stages drying, polymerization as a resulting mixture film. Such paint films have been can be inferred from the decrease in the amount linking some are more prepared, and flexible than the equivalent of extractable low molecular as the film weight polymers even paint prepared from unhydrolyzed oil. This is true This is determined solvent extrac ages. easily through no are though they contain glycerol ester bonds and tion studies. Figure 4 shows the changes in the amount potentially totally ionomeric. More complete results will of extractable material over the first several for a years a be reported elsewhere. Thus, hydrolysis of paint film from lead white and cold paint prepared pigment over time will not cause a film to a necessarily paint pressed linseed oil, and commercial paint containing become brittle. A paint film will become stiffer and titanium white in alkali-refined linseed oil. The major more brittle, however, if it is heated (at any relative or humidity) treated with solvent. Both procedures remove low molecular weight material, by evaporation was and extraction, respectively. It shown previously that evaporation of these plasticizing components does not occur to extent at any significant normal tempera tures a [10]. Thus, properly prepared paint film should remain flexible unless it is mistreated by overexposure to or heat solvents and loses low molecular weight statement plasticizing compounds. This is supported by curves Figure 5, which shows stress-strain measured for films of lead white in linseed oil one 6 12 paint cold-pressed as as curves and 10 years old, well later extrapolated from Years these and other measured curves. The mathematics of are Lead white Titanium dioxide the extrapolation described by Mecklenburg and was stress at Tumosa [16]. Briefly, it found that plots of a Figure 4 Total of extractable material in paint prepared from lead a versus were specific strain In time linear for each paint, white pigment and cold-pressed linseed oil, and a commercial paint and could be to stress at that strain containing titaniumwhite inalkali-refined linseed oil, as a functionof extrapolated predict for strain be drying time. The values are the extractable fractionof the totalpaint, films of any age. Breaking could predicted and are not corrected for versus pigment weight. similarly by plotting breaking strain time. By

STUDIES IN CONSERVATION 50(2005) PAGES 143-150

This content downloaded from 160.111.254.17 on Fri, 5 Dec 2014 10:58:14 AM All use subject to JSTOR Terms and Conditions 148 D. ERHARDT, C. S. TUMOSAAND M. F. MECKLENBURG

at be reached for hundreds of years, if all. Hydrolysis a of the glyceride ester bonds is slow process relative to even the initial polymerization reactions, and samples hundreds of years old may have substantial portions of ester the glyceride bonds intact. Plotting data for the changes in mechanical behaviour also shows that the occur or major changes within years decades, but also occur over predicts that measurable changes for well 50 a years and approach limiting value after about 150 years. can The major processes in oil paints be grouped into polymerization, hydrolysis, oxidation, and soap form O 0.01 0.02 0.03 0.04 0.05 0.06 0.07 a ation (if pigment capable of reacting to form soaps is Dimensional change (strain), mm/mm present), along with accompanying changes in the mechanical properties. Within the first few years the Figure 5 Stress-strain curves measured forpaint filmsof lead white amounts of extractable materials decline as the smaller in linseed oil one and 10 and later curves cold-pressed years old, molecules continue to crosslink. In addition, the amount extrapolated from these and other measured curves. From of soluble (free) fatty acids increases due to hydrolysis of Mecklenburg and Tumosa [16]. the triglyc?rides. A portion of this may react with metal ions present in many pigments to form insoluble metal stresses at as calculating various strains (as well breaking soaps. Hydrolysis reduces the number of crosslinks for a film of a a stress-strain curve strain) specific age, between the glycerides. However, other crosslinking can be constructed. These are shown as the theoretical reactions must occur since the total amount of extract curves in 5. It can be seen over Figure that the properties able material decreases time. Since the free fatty about 50 and even a to change very slowly after years, that acids tend keep the paint film flexible, any stiffening not or 250-year-old paint film will be substantially stiffer of the paint film must come from further non-ester more a as brittle than film only decades old. It is significant crosslinking and/or possibly from soap formation that the lead white film is is occur 10-year-old already almost well. One possibility that crosslinks through 25% hydrolyzed, and almost half of the freed fattyacids multivalent metal ions. Azelaic acid groups still bound converted to salts In a to can have been (see Figure 2). study of the glyceride by the original glyceride ester bond acid ethylene-acrylic copolymers, Bonotto and Bonner form salts through the free acid group formed by on as found that effects physical properties such stiffness oxidation. A multivalent metal that formed a salt with were more and strength readily apparent by 10% conversion than one such acid group could function as a to at ionomer (salt form), and peaked about 30% crosslink, and possibly increase the stiffness of the paint. is not a conversion [17]. While the system they used However, in study of synthetic polymers containing to a seems was identical dried oil, it that if hydrolysis and acrylic acid groups, it found that multivalent metal formation were to on no more on soap have significant effects ions had effect physical properties than the mechanical properties, they would be apparent in the equivalent amount of monovalent ions, and that the are seen was a 10-year-old sample. That only minor changes change in physical properties solely function of supports the extrapolation showing that much older percent conversion to ionomer, and not of the type or a extent films, with higher of hydrolysis and degree of valency of the metal ion [17]. Whatever the case, these a on soap formation, will still be relatively flexible. processes do not have major effect the mechanical tests properties, since of untreated paint films kept under CONCLUSIONS moderate environmental conditions show that changes in the mechanical properties have already slowed The existence of long-term processes in oil paints is significantly within the first few decades. some readily demonstrated by simple experiments. Plots The implications for conservation follow, in part, of over time show that the of on weight process weight from previously reported work calculating ranges of two were museum gain is quite slow after years, but if the data relative humidity appropriate for the environ the theoretical ment a extrapolated maximum weight gain [18, 19]. If paints retain reasonable elastic region amount can (derived from the weight gain of pure linseed oil and (the they be deformed without permanent not correcting for the weight of pigment present) would change) throughout their lifetime, then the allowable

STUDIES IN CONSERVATION 50 (2005) PAGES 143-150

This content downloaded from 160.111.254.17 on Fri, 5 Dec 2014 10:58:14 AM All use subject to JSTOR Terms and Conditions LONG-TERM CHEMICAL AND PHYSICAL PROCESSES INOIL PAINT FILMS 149

7 van den van den and RH fluctuations calculated using data from the available Berg, J.D.J., Berg, K.L., Boon, J.J., to to 'Chemical in curing and ageing oil in Preprints paint samples up 20 years old will apply paint changes paints' of the 12th Triennial Meeting of the ICOM Committee for throughout its normal aging process. The amount of ed. & occur Conservation, Lyon, J. Bridgland, James James (Science dimensional change that will in unrestrained glue Ltd, London 248-253. or canvas most Publishers) (1999) gesso (the materials in paintings respons 8 Erhardt, D., Cunningham, R., and R?s?nen, S., 'Extraction of to or ive RH changes) tangential wood (the material in material from oil paints by solvents' inMaterials Issues inArt most to RH in a panel paintings responsive changes) and Archaeology VI, Volume 712, ed. P.B. Vandiver, M. of ?15% RH about a median of 50% RH is about range Goodway and J.L. Mass, Materials Research Society, or can seen even 0.004, 0.4%. It be from Figure 5 that Pittsburgh (2002) 43-52. to 250-year-old paint is predicted require about four 9 Tumosa, C.S., Millard, J., Erhardt, D., and Mecklenburg, on times this much distortion before it breaks, while the M.F., 'Effects of solvents the physical properties of paint films' in the 12th Triennial the ICOM elastic limit is similar to newer paints (about 0.4%), and Preprints of Meeting of even Committee for Conservation, Lyon, ed. J. Bridgland, James & would not be exceeded by such RH fluctuations London 347-352. case James (Science Publishers) Ltd, (1999) assuming worst conditions (attachment of the 10 Erhardt, D., Tumosa, C.S., and M.F., 'Natural to a Mecklenburg, relatively RH-unresponsive paint highly responsive and accelerated thermal aging of paint films' in Tradition and to extremes substrate that is exposed the RH long Innovation: Advances inConservation, ed. A. Roy and P. Smith, to It was shown enough respond fully). previously that International Institute forConservation of Historic and Artistic oil films retain their elastic even after paint properties Works, London (2000) 65-69. extreme treatment solvent (24-hour immersion). While 11 Mecklenburg, M.F., and Tumosa, C.S., 'An introduction into or the loss of part all of the plastic region upon aging, the mechanical behavior of paintings under rapid loading or treatment conditions' inArt inTransit: Studies in the heating solvent does have implications for Transport ofPaintings, ed. M.F. National of Art, handling and transport (a paint film would be more Mecklenburg, Gallery Washington to (1991) 137-171. likely break rather than permanently distort if 12 'Effect of the in ?t Drozdowski, B., unsaturated acyl position mistreated), it does affect the choice of appropriate on the rate', theAmerican environmental concluons. triglyc?rides hydrog?nation Journal of Oil Chemists' Society 54(12) (1977) 600-603. 13 Weger, M., Die Sauerstoffaufnahmeder Oele undHarze, Eduard REFERENCES Baldamus, Leipzig (1899). 14 Sabin, A.H., 'Linseed oil', The Journal of Industrial and 1Wexler, H., 'Polymerization of drying oils', Chemical Reviews Engineering Chemistry 3 (1911) 84?86. 15 64(6) (1964) 591-611. Eibner, A., Sprung- und Rissbildung antrocknenderOelfarbenauf 2 O'Neill, L.A., and Brett, R.A., 'Chemical reactions in paint stricheund auf Oelhildern, Verlag der Technischen Mitteilungen films', Journal of theOil and Colour Chemists' Association 52(11) f?rMalerei, Munich (1920). (1969) 1054-1074. 16 Mecklenburg, M.F., and Tumosa, C.S., 'Traditional oil paints: 3 Watts, H., A Dictionary of Chemistry and theAllied Branches of the effects of long-term chemical and mechanical properties on Other Sciences, Longmans, Green and Co., London (1875) 616. restoration efforts',MRS Bulletin 26(1) (2001) 51-54. 4 Gardner, H.A., 'Notes on phenomena in paints and varnishes 17 Bonotto, S., and Bonner, E.F., 'Effect of ion valency on the bulk induced by colloidal reactions' in Proceedings of the Scientific physical properties of salts of ethylene-acrylic acid Section, Educational Bureau, American Paint and Varnish polymers', Macromolecules 1(6) (1968) 510-515. Manufacturers' Association, Circulars Numbers 266-295, ed. H.A. 18 Erhardt, D., and Mecklenburg, M.F., 'Relative humidity re in Gardner, Institute of Paint and Varnish Research (1927) 352 examined' Preventive Conservation: Practice, Theory and 369. Research, ed. A. Roy and P. Smith, International Institute for 5 Gardner, H.A., 'Toxic compositions to prevent the fouling of Conservation of Historic and Artistic Works, London (1994) steel to ships and preserve wood bottoms' in Proceedings of the 32-38. 19 Scientific Section, Educational Bureau, Paint Manufacturers' Erhardt, D., Mecklenburg, M.F., Tumosa, C.S., and Association of theUS National Varnish Manufacturers' Association McCormick-Goodhart, M., in The InterfaceBetween Science and (Co-operating), Circulars Numbers 241?265, ed. H.A. Gardner, Conservation, BritishMuseum Occasional Paper Number 116, ed. Institute of Paint and Varnish Research (1926) 231-271. S. Bradley, The British Museum, London (1997) 153-163. 6 Boon, J.J., Peulv?, S.L., van den Brink, O.F., Duursma, M.C., and Rainford, D., 'Molecular aspects of mobile and stationary phases in ageing tempera and oil paint films' in Early Italian AUTHORS Paintings Techniques and Analysis, ed. T. Bakkenist, R. and H. David Erhardt is a Hoppenbrouwers Dubois, Limburg Conservation currently senior research chemist at Institute (1996) 35-56. the Smithsonian Center for Materials Research and

STUDIES IN CONSERVATION 50 (2005) PAGES 143-150

This content downloaded from 160.111.254.17 on Fri, 5 Dec 2014 10:58:14 AM All use subject to JSTOR Terms and Conditions 150 D. ERHARDT, C. S. TUMOSAAND M. F. MECKLENBURG

Education (formerly the Conservation Analytical tion. He received his PhD fromVirginia Polytechnic as Laboratory of the Smithsonian Institution)which he Institute and State University in 1972. Address: for a joined in 1979. He has BS in mathematics and chem Erhardt. Email: [email protected] istryfrom theUniversity ofMiami (1972) and a PhD in a organic chemistry from the University of Maryland Marion F. Mecklenburg is senior research scientist at (1977). Address: SCMRE/MSC, 4210 SilverHill Road, the Smithsonian Center for Materials Research and Suitland, MD 20746, USA. Email: [email protected] Education. He received his PhD in engineering from as the University of Maryland in 1984. Address: for a Charles S. Tumosa is senior research chemist at the Erhardt. Email: [email protected] Smithsonian Center for Materials Research and Educa

? ? ? et R?sum? S?chage de peintures ? l'huile par polym?risation. Ce processus de s?chage peut ?tre pratiquement complet, la au continuent de se surface du film peint para?tre s?che toucher durant des semaines, mais des changements notables produire Par un continue de se en lieu des esters pendant des ann?es. ailleurs, processus plus lent d?velopper, premier Vhydrolyse glyc?rides. cas acides n'ont Ceci produit des groupements d'acides carboxyliques tels que des acides gras (dans le d?groupements qui pas r?agi ou avec cas par ailleurs), encore des groupements acides pont?s la matrice huile (dans le de groupements acides impliqu?s dans des avec des sels ? savons ? dans r?actions de polym?risation). Ceux-ci peuvent r?agir les pigments pour former carboxyliques (appel?s cas et la dont les traitements de le des acides gras). Ces changements affectent les propri?t?s physiques de la peinture fa?on et savon dans siccatives conservation la modifient. Cet article examine l'extension de l'hydrolyse laformation de quelques huiles ces et les vieillies naturellement et desfilms de peinture, V extractabilit? de mat?riaux dans des solvants organiques, changements et des de de vieillies naturellement. Les pr?vus mesur?s dans le temps propri?t?s physiques films peintures changements physiques terme au sont minimes en de ceux un ou et m?caniques ? long dues vieillissement comparaison produits par nettoyage g?n?ralis? une exposition ? la chaleur.

? Malschichten trockenen durch Dieser binnen Zusammenfassung ?lhaltige Polymerisation. "Trocknungsproze?" mag doch treten noch Wochen imWesentlichen angeschlossen sein und die Oberfl?che trocken erscheinen, messbare Ver?nderungen Prozesse schreiten vor allem die der Dabei entstehen Jahre danach auf. Auch andere, langsamere fort, Hydrolyse Glyceridester. die nicht oder als Carboxylgruppen, entweder als freie Fetts?uren (in dem Fall, da? S?uregruppe anderweitig reagiert hat) an vernetzte sind. Diese k?nnen mit zu S?uregruppen, die die ?lmatrix gebunden Pigmenten Carboxylaten reagieren (Salze, Bei diesen Prozessen werden die die im Fall derfreien Fetts?uren Seifen genannt werden). physikalischen Eigenschaften wird der ver?ndert, aber auch die Art und Weise wie sie auf Restaurierungsma?namen reagieren. In dieser Arbeit das Ausma? von an nat?rlich und Malschichten untersucht sowie die Hydrolyse und der Bildung Seifen einigen gealterten ?lfilmen werden erwarteten und die Extrahierbarkeit der entstehenden Materialien durch organische L?sungsmittel. Au?erdem die miteinander Dabei sind die der mechanischen gemessenen ?nderungen der physikalischen Eigenschaften verglichen. ?nderungen zu vonHitze undphysikalischen Eigenschaften gering imVergleich denen,die durch?berm??iges Reinigen oderden Einflu? verursacht werden.

? secan Este de 'secado' a en Resumen Las pinturas al ?leo por polimerizaci?n. proceso puede llegar completarse gran medida, seca tacto en los cambios sustanciales contin?an durante a?os. Otros m?s consider?ndose la superficie al semanas, aunque procesos de los esteres Esto tanto ?cido carbox?licos lentos tambi?n contin?an, principalmente la hidrolizaci?n glic?ridos. produce grupos caso no ?cido unidos a la matriz como ?cidos grasos libres (en el de grupos ?cidos que han reaccionado anteriormente) y grupos se en relaciones de Estos reaccionar con oleosa entrecruzada (en el caso de grupos ?cido que han incluido polimerizaci?n). pueden en el caso de ?cidos Estos cambios las pigmentos para formar sales carboxiladas (llamadas jabones grasos). afectan propiedades la manera en la los tratamientos de conservaci?n Este art?culo examina el de f?sicas de las pinturas y que pueden afectarlas. grado en de aceite secativo de al ?leo tambi?n se hidr?lisis y deformaci?n de jabones pel?culas y pinturas envejecidas naturalmente; de estos materiales en disolventes los cambios en el ser considera la capacidad de extracci?n org?nicos y tiempo que pueden previstos en causadas el de las de las propiedades f?sicas y mec?nicas, principalmente por envejecimiento pel?culas pintura envejecidas a al son menores con naturalmente. Los cambios f?sicos y mec?nicos largo plazo debidos envejecimiento comparados aquellos excesivas o al calor. producidos por limpiezas por sobreexposici?n

STUDIES IN CONSERVATION 50 (2005) PAGES 143-150

This content downloaded from 160.111.254.17 on Fri, 5 Dec 2014 10:58:14 AM All use subject to JSTOR Terms and Conditions