ART AND CHEMICAL SCIENCES 981 CHIMIA 2001,55, No. 11

Chimia 55 (2001) 981-989 © Schweizerische Chemische Gesellschatl ISSN 0009-4293

Mass Deacidification of

Agnes BlUhera* and Beat Vogelsangerb

Abstract: Paper, the carrier of our written heritage, decays within a relatively short period of time mainly due to its acid content and the influence of air pollution. More than 50% of the records of our libraries and are already at risk. Mass deacidification allows this problem to be counteracted by significantly slowing down the deterioration of paper and thereby prolonging its life span. An overview of all relevant mass deacidification methods is given. As it was found that the 'papersave' method was the most suitable to meet Swiss requirements, a plant using an optimised version of this method was built and put into service in March 2000. Thanks to this optimisation, not only and loose sheets but also documents in archival boxes could be mass-deacidified for the first time. Due to the necessity to analyse not only test but also original documents, new non-destructive testing methods had to be developed. Findings of basic investigations regarding treatment effects as well as results of routine quality control turned out to be very satisfactory. Keywords: Conservation' Deacidification' Mass deacidification' Non-destructive testing· Paper

1. Introduction

Up until the beginning of the 19th century, Fig. 1. Exampleofse- riously deteriorated paper was a hand-made quality product. documents. At that time rag paper was manufactured from carefully prepared plant fibres and sized using animal glue to provide a sur- face suitable for inscription. As the de- mand for paper grew, the manufacturing process became more industrialised. In 1805 Moritz Illig introduced stock sizing with the rosin-alum sizing process, and in 1844 wood was first used as the new source of raw materials [1]. The poor ageing performance of papers manufac- tured with a pH value of 4.0-5.5 in the presence of aluminium sulphate and sul- phuric acid, and with a high mechanical wood pulp content, has been well-known Acid decay describes the problem that af- ageing of paper is based on the complex for some time (Fig. I). Since around 1990 fects all modern papers from the period interaction of decay and oxidation pro- paper manufacturing has largely 1850 to 1990 which were intended to last cesses at the cellulose and hemicellulose switched to neutral or alkaline production 'forever' in libraries and archives. molecules. The key reactions are [4]: where alkyl ketene dimers are used as Paper is made from cellulose (40- • The acid-catalysed cleavage of ~- size and calcium carbonate as a filler. 100%), fillers and coatings (0-50%), sizing glucosidic bondings, detectable by the This has removed the main cause of acid agents (0-4%) and additives (0-0.5%). decrease in average degree of poly- decay. However, recycled paper presents Groundwood and unbleached papers also merisation (OP). The initial OP of a huge problem, especially for archives. contain hemicelluloses and lignin (0- cellulose is in the range 1000-36000, 50%) [2]. Endogenous factors affecting depending on its source, technological the ageing performance of paper are the processes and other factors. From a 'Correspondence: Dr. A. Bluher" quality of the cellulose fibres, the lignin OP of 400-500 the physical strength ·Schweizerische Landesbibliothek SLB and hemicellulose contents, and the level decreases rapidly [5][6]. Hallwylstrasse 15 CH-3003 of free sulphuric acid and presence of • The oxidation of primary and second- Tel.: +41 31 3222359 acid inscriptions. Exogenous factors are ary hydroxyl functions to carbonyl Fax: +41 31 3228463 the climatic conditions of the storage and carboxyl groups. The resulting E-Mail: [email protected] bNITROCHEMIE WIMMIS AG facilities, light, the acid content of the air acids boost the autocatalytic decay of CH-3752 Wimmis and wear and tear due to usage [3]. The the chains. ART AND CHEMICAL SCIENCES 982

CHIMIA 2001,55, No. 11

Table 1. Mass deacidification methods in use

Proce s Features Agent Operating Company Location Main users Treatedslnces rtol operations

Vlennes Aqueous immersion Effechve a9 nt. calCIum Osterrelchische VIenna OsterrelChlsche 5,300 volum 0 treatment WI h paper hydroxide; since 1999 Natlonelbibliothek, Austrian National- newspapers/nce 1986 [151l16J strengthening "borate buffer" (boric acid + Joselpletz 1, A-1 015 Nalional bibllothek sodium hydrOXIde) Wlen,Ausria Ubrary Ca~aclty: 800 volumes Freeze-drying or .4 I per year Re-sizlng aro;n : methylcellu ose (MC 400). (16)

Paper- Mechanical Effective agen ZFB Zentl1Jm fur ZFB, Various librari ca. 500,000 sheets splillng Itllnlorcement by calcium/magnesium carbonate Bucherhaltung GmbH, LeipZig and archlV sinc 199 [ 8) machln Introduction Of a Mommsenstrasse 7. (17) core paper alter Core paper. four diffel'ent 0-04329 Leipzig splitting 01 the kinds of paper according to Germany ongmal paper damage and format 01 objects

Aqueous Adhesive for he core: deacidification carboxymethylcellulose + methylcellulose (1:1) calclum/magn lum carbonate

Bucke- Aqueous Immersion Fixatllles: Suspension Rewln HanS Neschen AG, NescI1en, State of 2.8 milt sh tsslnc burger EL cationic) and Mesilol Archlvcenler, PO boll BOckeburg Lower Saxony, 7/1998(20) Konservle- FIliation of we ar- NBS (amonlc) 1340,0-31675 Sa onlan Sate ru~s' sol ubi Inks and dyes Buckeburg. Germany Archives ve anren fur Effective agent magnesium Leipzig/Dresden mod mes Paper strengthenmg bicarbonate Archlvgul Buckeburg: Two- Hans Neschen AG. Neschen, Germany Nalfonal Opening In Jun 2001. (19J sIege-treatment R -sIzing agent: Aussens elle: Berlm Archive Annual capacily 13.5 methylhydroxyethylcellulose Dahlwlu mill. sh ets (3 shlhs). B rlln: One-stage- Undenallee 53-57, Planned or 2002' 9 treatment 0- 15655 Dahlwltz- mill. sheets [20] Hoppegarten,Germany

Weiro Non a~ueous liquid Effective agent: melhoxy operatinvnstilUtion National Na.lIonal Ubrary 01 1 100,000 books since (211122) phase mpregnation magnesium m thyl carbonate National 'brary of Ubra~ of Caneda. National 1 81 [22] 10 vacIJum Canada. 395 Cana a, Archives of Canad Solvent; telrafluoroethane Wellington Streel. Ottawa Predrying (HFC- 134a), methanol Ottawa, ON K1A ON ethanol Solullon supplier: Wei T'o Assoc ates, Inc., 21750 Main Street, UnIt 27, Matteson, IL 60443-3702, Illinois, U.S.A.

Sable Non aCjueous liqUId Effectiv agent. methoxy Sable-sur- Blbliohequ Since 1987. vaflant fn'ase Impregnation magnesium m thyl carbonat Sarthe. nallonal de (23) n vacuum France France, Paris Solvent. hydrochloro-fluoro Predrying carbons. methanol

Battell -V rfah n [24-27J

'Oas Non-aqeous liquid Effectl\! agent magneSJum- ZFB Zentrum fur Deutsche Vanous libraries Around 250 tons sine Vari hran rrhase Impregn lion tllanlm-alconolate Buchemel ung GmbH, aOcherel and archiVes 1994 [18J der n vacuum Mommsenstrasse 7, Leipzig Deutsch n Solvant: hexamethyl 0-04329 Lelpzlg, Blbliothek' Thorough predrying diSiloxane Germany

papersave Battelle Battelle Various libraries 115 on sfnce09l1996 lng nleurtechnlk Technrkum, and archIVes 1281 GmbH. DOsseldorter Eschbom bel Str 9, 0-65760 Frankfurt Eschbom. Germany

papersave Nltrochemie Wlmmls Nltrochemle. SWISSnational 90 tons sInce 3/2000 SWISS(29J AG. Nlesenstrasse, Wirnmls library. SWISS (145.000 boolls and 4.2 CH·3752 Wimmis, nallonal archives mill. sheets) SWitzerland

Bookkeeper Non-aqeous liquid EffeotlVe agent: magnesIum Preservation Cranberry Library 01 Congree More than 700,000 [30) phase Impregna ion oxide. submlcron powder Technologies L.P•. Tohnshfp Washington, Nat. boo s. More than 25 Thomson Perk Drive. Pennsylvania. Library of Quebec. tonn s of archIves No predrymQ Suspension liQUid' perfluoro cran= Township. US.A. Nat. Archives of 131) heptane With surfactant PA 1606 U.S.A. CanaCla, 60 libraries and archives 10 U.S.A.

Mlddenweg 5768. Heerhugowaard. Dutch Federal Since 1996 375'000 1704BR, The Nelheriands Archivlas. Dutch books or the LC [32) HeerhU~owaard, Royal Library. Th Net rians NI ersachsisches Slaatsarchiv. 30 libraries and archives In Europe

L1bertec Dry process EffectIVe agent 50% Ubertec Ubertec, Library of the Around 100,000 (12)[33J magnesIum oxide - 50% Blbliotnekendlenst NOmberg Bavanan Stale. volum equival nl to Minim I predryfng calcium cilrbonilte. submlcron GmbH. Kilianstr. 86, Library 0 lhe City around 90 tons since powder 0-90425 Numberg, o Munich, Ub:r, of 1996 [33) Germany he Stal of Berl n, Transport: dry stream of air Bundespresseamt ART AND CHEMICAL SCIENCES 983

CHI MIA 2001.55. No. 11

• The new formation of hydrogen Table 2. Mass deacidification methods under development bondings and main valency bonds be- tween the cellulose chains leading to Process Features Agents Developer State of compression and embrittlement of the Development paper structure. CSC Non-aqueous Effective agent: CSC, S.L.. Pilot plant since Since the final goal of mass deacidifi- Save liquid phase carbonated Mallorca 269. E- 1999 cation is to prolong the life span, it is nec- [34][9] impregnation magnesium di 08008 Barcelona. essary to investigate the behaviour of under n-propylate Spain deacidified materials under artificial age- pressure ing. Artificial ageing, however, will nev- Solvent: HFC 227 Slight (1.1.1,2,3.3,3- er be able to perfectly simulate natural predrying heptafluoropropane) ageing. Hence the current differences of opinion regarding the best-suited ageing Nanomer- Sol-gel-based Effective agent: sol- Institute for New Experimentalstage method. In recent years the standardised Techno- reinforcing gel silane-system Materials (INM). ISO 5630-3 method based on isothermal logien [35J system modified with Saarbrucken, storage at 80 °C and 65% Lh. was most methacrytoxy Germany frequently used. Recent research, howev- groupings, containing perfluorinated silanes; er, concludes that cycling of temperature addition of MgO and humidity allows better simulation of natural ageing [7]. An excellent review about the problems associatedwith artificial Table 3. Discontinued installations or developments ageing of paper was publishedby Porck [8]. It is agreed that the general require- Process Features Agents Developerl State of ments of a mass deacidification process Operating Realisation are as follows [9]: Company • Long-term stability of the treated ma- DEZ Gas-phase Effective agent: Library of Pilot plant In terial [36][37] process diethylzinc Congress. Houston, Texas • Results are archivally sound and do Washington. from 1990-1994 not produce any detrimental elements Thorough pre- DCfTexas • Neutralise acids and raise pH to a drying under Alkyls (Akzo) specified level exclusion of (U.S.A.) oxygen • Deposit alkaline buffer/reserve • Will process books without damage FMC liquid phase Effective agent: FMC, lithium Piiol plant In to their bindings [38] Impregnation carbonated Corporation Bessemer City. • Will not cause inks and colours to magnesium of America, U.S.A.• since 'bleed' Predrylng dibutoxytriethylene L1THCO. 1990 • Will not leave any residual smell or glycolate (MG-3); U.S.A change the texture or feel of the paper since 1993: mag- nesium butyl • Environmentally safe glycolate (MBG) Safe to current and future users Solvent: Freon 113; since 1993: heptane 2. Review of Mass Deacidification Methods Graft- liqUid phase Effective agents: British Laboratory copoly- impregnation solubilized acrylate Library, chamber meriza- monomers (e.g. ethyl London 2.1. History tion [391 Exposure to acrylate + methyl Paper deacidification dates back to y-rays methacrylate + the 1930s, when 'single sheet' neutralisa- alkaline monomers) tion of paper in an aqueous solution of Paper strength- calcium bicarbonate was invented. Since ening and then, washing and deacidifying of paper deacidification has been a topic of numerous investiga- tions [10]. In the 1960s the first non- aqueous techniques were developed [1 I]. isting deacidification processes and the or volatile amines were unable to cause a These were based on magnesium meth- development of new ones. More detailed lasting deacidification effect and created oxide, solubilised in a mixture of metha- reviews of the history of paper deacidifi- serious health hazards. nol and freon compounds and applied to cation can be found in [12-14]. the paper by spraying. These activities An overview of the most important 2.2. Classification and Comparison formed the basis for the later develop- methods in use, under development or of Methods ment of mass deacidification processes. discontinued is given in Tables 1-3. It From Tables 1-3 it can be seen that In 1981, the world's first mass deacidifi- must be added that numerous other de- magnesium-based deacidificants are gen- cation plant opened in Toronto. During acidification systems were evaluated but erally favoured. The deacidificants can the 1980s and 1990s, a great deal of effort failed for a variety of reasons. In particu- be applied either in dissolved form or as was invested in the improvement of ex- lar, gas-phase treatments with ammonia particles, using different solvent types or ART AND CHEMICAL SCIENCES 984

CHIMIA 2001,55, No. 11 the gas phase as carrier. This leads to the tivity of the employed deacidificant with telle process ('papersave') was conclud- following classification of deacidifica- water, a certain amount of pre-drying is ed to be the most suitable because it is ef- tion methods: necessary. This entails physical stress for ficient, environmentally friendly and en- A) Aqueous processes, usually with dis- paper, leather and parchment and in- ables treatment not only of books but also solved alkaline earth carbonates as creases the costs of investment and treat- of loose documents in boxes. In Summer a neutralisation agent (Bi.ickeburg, ment. Type Band E processes have the ]998, the Swiss parliament decided to in- Viennese process, ) broadest range of application, and there- vest CHF 13 million in the construction B) Non-aqueous processes using organo- fore come closest to a real 'mass' process. and operation of the Swiss deacidifica- metallic agents, usually magnesium In the methods with ultra-fine parti- tion plant. This plant was built in Wim- alcoho]ates, with organic solvents cles (C and D), bound documents - books mis and became operational in March such as alcohols, freons, perfluoro or clipped-together sheets - are placed in 2000. alkanes or si]oxanes (Wei To, FMC, a machine separately and fanned out. The Swiss mass deacidification plant Sable, Book Saver, Battelle). This ensures that the particles can reach uses the third generation of the Battelle C) Treatment with ultra-fine particles of the interior of the books. As no predrying process - this improved procedure, called agent, usually magnesium oxide, ap- is needed, moisture sensible materials 'papersave swiss', is distinguished from plied as a suspension in perfluoro like leather and parchment can be treated its predecessors through greater variabili- alkanes (Bookkeeper). at lower risk. However, the ultra-fine par- ty and better control of process parame- D) Treatment with ultra-fine particles of ticles are inherently unable to penetrate ters as well as through an active recondi- agent, usually magnesium oxide, ap- the paper as effectively as dissolved de- tioning facility. The plant belongs to the plied directly from a stream of air acidificants. These can also form powdery Swiss Confederation and is privately run (Libertec). deposits. Furthermore, the stream of air by NITROCHEMIE WIMMIS AG. With E) Gas-phase processes with organo-me- or solvent respectively used to introduce two treatment chambers allowing a ca- tallic agents, namely diethyl zinc the particles causes physical stress on the pacity of 120 t per year, it is the largest (DEZ). pages and the book as a whole. and most modern of its kind. 40 tons of All described methods, if correctly From the 'side-effect point of view', capacity each are utilised by the SNL and applied, allow not only complete neutra]- the solvent-free methods are clearly fa- SFA, while a further 40 tons are available isation of acidic papers but also the incor- voured. Unfortunately, they cause other for other customers. poration of an alka]ine reserve. problems. In the case of the only true gas- As a rule, the aqueous methods (A) phase method (E), the handling of the 3.2. The Process and the Chemistry are suitable for the treatment of single deacidificant diethyl zinc, a highly explo- Crates containing the books, docu- sheets and contain a paper-strengthening sive gas, is extremely dangerous - acci- ments and archival boxes to be treated aspect. The paper splitting method is de- dents have happened, and the pilot plant (500-1200 kg) are placed in the treat- signed for the mechanical reinforcement was put out of operation in 1994. ment chamber (Fig. 2 and Fig. 3). Here of very fragile single sheets. The Bi.icke- the materia] is pre-dried, i.e. the normal burger method is designed for archival moisture content of 5-8% is reduced to material which has to be taken apart in 3. The Swiss Mass Deacidification <1% by mild heating in a vacuum. Drying single sheets as a first step. Since paper Project is followed by the actual deacidification. clips and plastic covers are removed in In this step, the chamber is completely this process, the materia] is given com- 3.1. Background and History flooded with deacidification solution prehensive conservation treatment at the Due to world-wide surveys one can which impregnates the material. After same time. The Vienna process is de- assume that on average 80-90 % of the the solution is drained, the material is signed to deacidify and re-strengthen holdings in libraries and archives are again vacuum-dried in order to remove newspaper after removing the covers. threatened by acid decay. Of this, 10% is the solvent. The evaporated solvent is As the bleeding of inks and dyes con- badly damaged or unusable, 30% affect- condensed and recycled. These three stitutes the major problem associated ed and an additional 40% is at risk [41]. steps of the process take three to four with all aqueous methods, inks and dyes These figures apply equally to the Swiss days. During the subsequent recondition- have to be either absent or chemically (SNL) and the Swiss ing period, air of a defined temperature, fixed before the treatment. Federal Archives (SFA) which are rela- humidity and quantity is blown through In the non-aqueous processes (B) im- tively young institutions. They each con- the material for about three weeks. The pregnation with the treatment solution tain 3000 t of holdings at risk from acid paper regains the lost humidity and reac- takes place in a vacuum chamber. Books decay, which would take 20 to 30 years to tions take place between treatment agent as well as loose sheets and archival mate- deacidify at the rate of 40 t per annum. and material, releasing alcohol. The dea- ria] in boxes can be treated. The more The Swiss National Library has adopted cidified material is then returned to the modern methods use non-polar solvents a strategy for mass deacidification to pre- customers. Different treatment programs in order to minimise the 'bleeding' of al- vent further decay of original stock were developed for library material coho]-sensitive inks and dies. For ecolog- which is at risk but still usable. Micro- (mainly books) and archival material ical reasons, however, solvents such as filming is used to preserve more badly (documents with inscriptions, mainly in freons or halogenated hydrocarbons are damaged stock. boxes). In particular, the concentration of inappropriate in the long term. Siloxane In 1990, both SNL and SFA started a deacidificant (METE) had to be reduced solvents are an environmental friendly al- joint venture aimed at establishing a by one-third in the program for archival ternative, but since they are highly flam- Swiss deacidification plant. Between materials. mable they require extensive safety 1991 and 1994, the world's leading sys- The main problem in developing a measures. One drawback of the solvent- tems (Wei To, Battelle, FMC, DEZ and new mass deacidification method was to based methods (B) is that due to the reac- Bookkeeper) were evaluated. The Bat- find an environmentally friendly, non- ART AND CHEMICAL SCIENCES 985 CHI MIA 2001,55, No, 11 polar solvent able to dissolve magnesium Fig. 3. Treatment cham- aicoholates, The papers ave method is ber; charging with frame based on Battelle's invention, according containing crates with books is in progress. to which a complex of magnesium ethox- ide and titanium ethoxide (METE) can be easily dissolved in hexamethyl disiloxane (HMDO), a highly inert, low-viscous and (with a boiling point around 100°C) rela- tively volatile solvent [25]. During treatment the magnesium ethoxide immediately neutralises free ac- ids present in the paper as follows:

During reconditioning, the excess of magnesium ethoxide reacts with humidi- ty and carbon dioxide, thus forming the desired alkaline buffer consisting of mag- nesium carbonate which will protect the paper from future acid attacks:

Treatment Chamber Note: It has not yet been determined Pre- Impregnation After· Recon- whether the alkaline buffer consists of drying 1·2 Hours drying dItioning 48 Hours 24 hours 3-4 Weeks MgC03, 4Mg(C03)·Mg(OH)2 or even Mg(OHh· The titanium ethoxide also present in Water Water the paper reacts with humidity to titani- 1. Pre-drylng um hydroxide, which will decay into in- ert titanium dioxide and water. There- 2. Treatment fore, titanium ethoxide does not contrib- ute to the deacidification effect but has to 3. After-drylng be incorporated in the solution since 4. Reconditioning magnesium ethoxide alone is not soluble inHMDO.

Fig. 2. Flow diagram of the 'papersave swiss' mass deacidification process. 4. Quality Control

4.1. Quality Requirements was implemented for the quantitative de- pressed as %MgCOruptake - can be The basic contract between the pri- termination of colour changes. The anal- determined by the increase in both Mg or vate operator and its main customer, the ysis of treatment intensity, however, Ti. This is very important since, due to Swiss Confederation, contains numerous turned out to be more difficult. Surface- the physics of XRF, Mg is detected only provisions governing the quality of the pH measurements only indicate whether on the paper surface (Mg X-rays pene- treatment as well as the safety, infrastruc- the document is acidic or alkaline but trate only about 5% of paper sheet thick- ture and logistics of the process [29][42]. give no quantitative information about ness), whereas for the Ti-measurement a According to these quality standards, the the amount of incorporated MgC03. few sheets are penetrated. The new tech- general suitability of the 'papersave Therefore, a quantitative technique based nique was validated according to SN EN swiss' process must be established. Fur- on X-ray fluorescence (XRF) was devel- 45001 and allows now the intensity and thermore, compliance with the require- oped. A commercial XRF spectrometer homogeneity of mass deacidified original ments must be verified for each single was fitted to a huge, specially constructed documents to be analysed for the first batch. vacuum chamber. The system was time on a routine basis. equipped with an xy-table which allows 4.2. Non-destructive Analysis each position of a book or document to be 4.3. Results of Basic Investigation As destructive testing of original ma- located above the XRF detector (Fig. 4). Several investigations were perform- terial is not possible, suitable non-de- At each position, an area of about 1.5 cm2 ed in order to establish basic findings structive analytical techniques had to be of the paper page is measured. Since it on the effectiveness of the 'papersave found. Most side effects can be deter- was established that Mg and Ti are de- swiss' process. First it was demonstrated mined qualitatively by visual and physi- posited on the paper in stoichiometric ra- that the treatment intensity (uptake of cal assessment. A spectrometric method tios, the treatment intensity - usually ex- magnesium) is highly constant both with- ART AND CHEMICAL SCIENCES 986

CHIMIA 2001,55, No. 11

in and between the batches and does not depend on the location of the document within the treatment chamber. More detailed XRF analysis proved that all paper sheets were fully deacidi- fied over the entire area. In addition, good to excellent homogeneities of Mg/ Ti-deposition were established (Fig. 5). It was then demonstrated that the 'paper- save swiss' deacidification acts in a three-dimensional way: the paper sheets are neutralised not only on the surface but Fig. 4. Sample chamber of XRF spectrometer for non-destructive analysis of Mg and Ti- throughout the paper matrix. This was uptake; test book is located on xy-table. demonstrated by (i) direct REM-EDXRF measurements which found Mg and Ti throughout the diameter of a treated pa- per sheet, (ii) XRF measurement of Mg in split paper which determined even slightly higher Mg-contents on the interi- or of the sheet than at the surface, and (iii) the fact that sheets treated in fully sealed paper envelopes receive the same MgC0 uptake as open-treated sheets. • 0.90-1.00 3 1.0 Mechanical testing on artificially .0.80-0.90 aged papers established that the 'paper- B' 0.9 0.70-0.80 0 0.60-0.70 save swiss' treatment slows down paper l:J) 0.8 00.50-0.60 ageing at least by a factor of 2-7 (depend- == 0.7 ';ft. .0.40-0.50 ing on the test) and therefore significant- 0.6 0.30-0.40 ly prolongs the document's life span. For -.!!! ni .0.20-0.30 this purpose two different paper types ~ 0.5 0.10-0.20 were deacidified with the two different CC 0.4 '5 0.00-0.10 deacidificant concentrations and aged for Q) 0.3 24 and 48 days at 80°C at 65% r. h. (ISO ~ ClI 0.2 5630-3). All non-aged and aged samples Q. ;:) 0.1 were tested with respect to tensile strength (EN-ISO 1924/2), tensile 0.0 strength after Bansa-Hofer folding, tear- ing resistance (Elmendorf; EN 21974), and folding index (Schopper; ISO 5626). Data analysis showed that the treatment Fig. 5. Homogeneity oftreatment of an archive document. The alkali uptake ranges between 0.7% itself does not significantly alter the me- and 0.85% MgC03 over the whole paper sheet, proving complete and homogeneous deacidi- chanical properties of the two papers fication. (neither strengthening nor weakening). During ageing, however, the deacidified samples show a much slower deteriora- tion rate of mechanical properties than the untreated ones. non-aged 24 days 48 days All papers undergo a small but meas- BO°C/ 65% r.h. 800C /65% r.h. urable colour change during treatment (slight yellowing and darkening). This not treated treatment-induced change, however, is DD I negligible compared to the discoloration during ageing. In case of the investigated ProgramI.'" papers, deacidification also showed a ~LfBRARY" positive effect regarding ageing-induced DOD discoloration - the rate of discoloration was significantly reduced (Fig. 6). Prog~" ~ARCHIVE· DOD 4.4. Results of Routine Analysis Since the start of the operation more than 12000 analytical results have been Fig. 6. Influence of treatment and/or ageing on paper made from sulphate cellulose, The collected. In the first four months of op- 'papersave swiss' treatment causes only small colour changes. During ageing, strong discolor- eration ten test and ten original docu- ation occurs however at a much lower rate for the two deacidified papers than forthe untreated one. ments per batch were analysed. In the ART AND CHEMICAL SCIENCES 987 CHIMIA 2001,55, No. 11

2.5 • Program"LIBRARY"

'6 2.0 • Program"ARCHIVE" () C> :iE • • ~ 1.5 L • ~ ••~ M a; .!a .••.,... 1.- ~ ... «o 1.0"...,•• ~".,.6. '"\ ,. .., ., ••., Q). .- "'"a 0.5 ------~ Fig. 7. Process chart showing MgC03 uptake of 34 batches library books and of 27 batches 0.0 archival material, from XRF analysis of three to ten test books per batch. The good stability of process as well as the difference between the two programmes are clearly visible.

current routine operation three test books in the ] 8th century from textile fibres, than 98% (SF A) rsp. 96% (SNL) sur- and five original books per batch of 1850 even exhibited much lower colour chang- vived the treatment without a change, and library documents are analysed and the es than modem, wood-free high-quality the required quality standards were ex- results returned to the customers together papers. ceeded. A slight change was visible on an with the treated material. Analysis of this average of 1 to 2% of the documents, al- huge amount of data provides an excel- though both legibility and usability re- lent overview of the main features of the 4.5. Results of Visual and Physical mained unimpaired. The main things to 'papersave swiss' process: Assessment be affected were several unstable red inks From XRF analysis of testbooks it can While the deacidification treatment in covers, printing inks and in coloured be concluded that the process is highly described above is applied only to paper, pens. Iridescent reflections which are re- stable and robust: the quality require- the process involves the library and ar- ported to appear sometimes in liquid ments (95% of test papers between 0.5% chival materials as a whole. In an ideal processes [14][23], were not observed. and 2.0% MgC03) can easily be fulfilled. situation a mass treatment process must Impairments were detected on fewer than For the library programme the MgC03 not affect 0.] % (SFA) rsp. 2% (SNL) of the docu- uptake is around 1.35%, while for the ar- • cover and binding including synthetic ments, but never complete loss. Ongoing chive program it amounts to around and natural glues investigations are conducted to identify 0.90% (Fig. 7). The ditference in MgC03 • covers such as leather, parchment, reasons for side effects and ways of uptake between the two programs per- textiles, plastics, metals avoiding them or at least further reducing fectly represents the difference in deacid- • printing inks, modern printing materi- their intensity. ificant concentration in the treatment so- als and stamping inks, ferro-gallic Despite the excellent results achieved lution (factor 0.67). ink, historic inks and pigments, col- by the 'papersave swiss' process it is still More than 99.5% of analysed original our and gilt edge, marbled paper etc. not possible to dispense with the prepara- and test papers resulted in pH values • photographs, prints, graphics, plastic tion and selection of the holdings in the >7.0 following treatment. Hence the sleeves, overhead slides etc. SNL. For example, the following library quality requirements (98% of documents Since it is impossible to check each stock is not treated: parchment bindings, fully deacidified) are fulfilled. pH values document (1850 library documents or leather bindings, certain synthetic folders of untreated original documents ranged 64000 sheets of archives per batch), spot and some red-covered books. Out of the between 2.3 and 9.0. After treatment, pH- checks are carried out (between I % and ]45000 documents treated, 964 (0.65%) values between 7.0 and 10.0 were ob- 25% depending on the material being were excluded from the deacidification tained. processed). From 20 library documents treatment for conservation reasons. In ad- Routine measurements using the col- per batch a full protocol is drawn up. dition, some tightly-packed documents our spectrometer showed that colour The condition of the papers and docu- are opened and sometimes foil covers change is highly reproducible for a given ments before and after deacidification is and plastic covers removed. The SFA paper type. Again, the impact of the type documented according to quality standards stock is deacidified in the archive boxes of paper was found to vary: wood-pulp- governing paper structure, function and without any kind of selection [43]. free, high-quality papers undergo much appearance of the cover, print and inscrip- smaller colour changes than wood-con- tions, smell and dimensional stability. taining papers (primarily due to lignin, Between the start of the project in 5. Conclusions and Outlook which inherently changes its colour be- March 2000 and June 200] 145,000 doc- tween the acidic and alkaline regions). A uments (43 t) belonging to the SNL and After decades of world-wide research surprising finding was that several pa- 4.2 million documents (48 t) belonging to and development on the subject of mass pers, all of them carefully manufactured the SFA have been deacidified. More deacidification, four different principles ART AND CHEMICAL SCIENCES 988 CHI MIA 2001, 55, No. 11 of process have become well established es, however, the deposition of such high [4] E.g. A. Kaura, T. Krause, Das Papier on the market, serving a wide range of li- amounts of alkaline buffer is either diffi- 1978, 31, 9; R.L. Feller, Joumal of the braries and archives. Although the origi- cult to obtain, increases the severity of American Institute for Conservation (JAlC) 1994, 33, 91; S. Margutti, G. Co- nal goals have almost been met it is still side effects, or causes excessive costs. As nio, P. Calvini, E. Pedemonte, Restaura- clear that there is not, and never will be, a described above, we (and others) also ob- tor 2001,22,67. single universal process. The range of tained good results with much lower [5] R.D. Smith, 'Massdeacidil'ication Tech- materials is too wide, and the require- buffer concentrations. Recent research nologies', in 'Paper Preservation, Current ments of different libraries and archives even indicates that excess alkaline buffer Issues and Recent Developments', Ed. P. too varied. As a rule, mass deacidifica- might increase the degradation of treated Luner, Tappi Press, Atlanta, 1988, p.103. tion generally constitutes part of the papers [44]. Further research is required [6] K.B. Hendriks, 'Haltbares Papier', in 'In- overall conservation concept. In some in- in order to assess the optimum amount ternationale Arbeitsgemeinschaft der Ar- chi v- Bibliotheks- und Graphikrestaura- stitutes it is integrated logistically into a of buffer content, and a project aimed at toren, 8th International Congress, 19th- stock maintenance or process. this is currently under way in our labor- 23th September 1995, Ttibingen, Pre- The choice of the most suitable process is atories. prints', Eds. M.S. Koch, K.J. Palm, The dictated by the stock, the financial re- Whereas mass deacidification allows Royal Danish Academy of Fine Arts, Co- sources, and of course the local availabil- further deterioration to be stopped or penhagen, 1995, p. 75. ity of a provider of deacidification ser- slowed down, it does not restore the orig- [7] M. Klisberger, G. Dessauer, H. Stark, Das vices. inal strength of the paper. Books and doc- Papier 1998, 52, 529; M. Kasberger, Preparation and selection of the stock uments which are already been badly 'Vorg1inge im Papier bei dynamisch be- schleunigter Alterung', Dissertation, TH is generally unavoidable. The effort in- damaged before treatment will remain Graz, 1998. volved depends largely on the deacidifi- unusable. Until now, re-strengthening of [8] H.J. Porck, 'Rate of Paper Degradation: cation process selected. In this respect the paper was limited to 'single sheet' treat- The Predictive Value of Artificial principle of the Battelle process has the ment. For several years efforts have been Tests', European Commision on Preserva- widest field of application, which is why under way to develop 'mass re-strength- tion and Access (ECPA), Amsterdam, and it was regarded as the most suitable pro- ening' techniques, preferably in combi- Commission on Preservation and Access cess for the Swiss National Library and nation with mass deacidification [45]. (CPA), Washington, 2000. the . It was further The problem, however, is not straightfor- [9] 1. Rhys-Lewis, 'The Enemy Within! Acid Deterioration of Our Written Heritage', developed into the 'papersave swiss' ward - all approaches have so far been Mass Deacidification Feasibility Project process in order to make it more suitable unsuccessful [39][40] or are still at the Report to the , 2001, 22. for processing archive materials. basic stage of research (e.g. the use of a [10] A. Lienardy, P. Van Damme, Restaurator Under the present circumstances, the sol-gel based reinforcing system [35]). 1990, 11, I; K. Bredereck, A. Haberditzl, Swiss mass deacidification project must The reinforcing of paper is a field of re- A. Bltiher, Restaurator 1990,11, 165; H. be regarded as a complete success. The search in which considerable progress is Bansa, Zeitschrift fiir Bibliothekswesen quality of the treatment exceeded our ex- still anticipated. U1ulBibliographie ZfBB 1980, Sonderheft pectations. The plant, including logistics Additional research projects carried 31, 34; W.J. Barrow, R.C. Sproull, Sci- ence 1959, 129, 1075. and analytics, was successfully put into out elsewhere concern the best methods [II] A.D. Baynes-Cope, ZfBB 1980, Sonder- operation. Within one year of operation, of selecting stock prior to processing heft 31,55. experience and findings on mass deacidi- [46], including the development of new, [12] H.J. Porck, 'Mass Deacidification - an fication in general and the 'papersave non-destructive analysis methods for mi- Update of Possibi lities and Limitation', swiss' process in particular was broadly cro-analytic evaluation of the condition European Commision on Preservation and expanded. Nevertheless, several ques- of the paper [47]. Access (ECPA), Amsterdam, and Com- tions are still unresolved. Despite all these promising research mission on Preservation and Access (CPA), Washington,1996; H. Bansa, Zeit- One question concerns the effect of projects, it must be noted that the deterio- schrift fiir Bibliothekswesen und Biblio- deacidification on leather covers. At a pH ration of paper is advancing rapidly, and graphie ZfBB 1999, 46, 127. value of over 6, leather is no longer stable waiting for the perfect system presents a [13] A. Lienardy, Restaurator 1991, 12, 75; and the drying process after wet process- much greater risk to our written heritage A.Ch. Brandt, 'Mass Deacidification of ing causes the leather to embrittle beyond than using a less perfect method. Paper - a Comparative Study of Existing the point where it can be fully restored. Processes', Bibliotheque Nationale, Paris, Although up to now no damage to the Received: September 19,2001 1992. leather could be ascertained [26], re- [14] Further informations are available on the homepage of the European Commission search including conservation aspects is on Preservation and Access (ECPA), Am- still necessary in this field. A project was [1] Ullmanns Encykloplidie der technischen sterdam, htlp:\Vww.ecpa.knaw.nl, Interna- therefore launched to investigate this is- Chern ie, Eds. E. Bartholome, E. Biekert, tional Conference 'Mass Deacidification sue in depth. H. Hellmann, H. Ley, W.M. Weigert, E. in Practice', Blickeburg, 2000. Another question concerns the opti- Weise,Veriag Chemie, Weinheim, New [15] G. Banik, W.K. Sobotka, in 'PaperPreser- mum amount of alkaline buffer to be in- York, 1979, Volume 17, p. 577; 'Paper vation, Current Issues and Recent Devel- corporated in the paper. Since DIN 6738 Chemistry', Ed. J.C. Roberts, Blackie Ac- opments', Ed. P. Luner, Tappi Press, At- recommends a minimum content of 2% ademic & Professional, Glasgow, 1996. lanta, 1988, p.l37. [2] Groundwood paper contains mechanical [16] W. Ruhm, personal communication. CaC0 (equivalent to 1.68% MgC0 ) in 3 3 wood pulp which is ground wood with all [17] W. Wachter, J. Liers, E. Becker, ICP Con- newly produced papers in the highest components apart from the bark. Chemi- ference Papers London 1997, Ed. J. Ea- life-span class, many conservators be- cal wood pulp is - more or less completely gan, The Insitute of Paper Conservation, lieve that this concentration should also freed from lignin and hemicelluloses, and Leigh (UK), 1998, 224; I. Briickle, J. Dam- be the target value in deacidification. yields 'wood-free' paper. bragio, Papierrestaurienmg 2000, 1, Sup- With most mass deacidification process- [3] G. Dessauer, Das Papier 1978, 32,32. pl.,75. ART AND CHEMICAL SCIENCES 989

CHI MIA 2001.55. No. 11

[18] J. Liers, persona] communication Stroud, The Paper Conservator 1994, ]8, [19] W. Feindt, H.V. Rudolph, S.Schiewe, B. 57. Werthmann, Restauro 1998, 104, 120; K. (37) K.E. Harris, C.J. Shahani, 'Mass Deacidi- Bredereck, A. BIUher, Restauro 1992, 98, fication: An Initiative to Refine the Di- 49. ethy I zinc Process', Library of Congress, [20] K.D. Vogt, personal communication. Washington D.C., 1994. [2\] R.D. Smith, 'Mass Deacidification at the [38] R.S. Wedinger, Restaurator 1993, ]4, Public Archives of Canada', in 'Conserva- 102; R.S. Wedinger, Restaurator 1991, tion of Library and Archive Materia]s and 12, 1. the Graphic Arts', Butterworths, London, (39) D.W.G. Clements, in 'Paper Preservation, 1987, p. 125; 1. Holmes, 'Mass Deacidifi- Current Issues and Recent Develop- cation of Books at the National Archives ments', Ed. P. Luner, Tappi Press, Atlan- and the National Library of Canada', in ta, 1988, p.149. 'Paper Preservation, Current Issues and [40] B.I. Humphrey, in 'Paper Preservation, Recent Developments', Ed. P. Luner, Current Issues and Recent Develop- Tappi Press, Atlanta, 1988, p.144. ments', Ed. P. Luner, Tappi Press, Atlan- [22] P. McCormick, R.D. Smith, personal ta, 1988, p.129. communication. [41] Bund-Uinder-Arbeitsgruppe Papierzerfall: [23] P. Vallas, Restaurator 1993, ]4, I. Bericht Uber Ursachen, Ausmass, Wirkun- [24] Personal communication gen und Folgen des Papierzerfalls im Bib- [25] J. Wittekind, Restaurator 1994, ]5, ]89. liotheks-, Archiv- und Verwa]tungsbe- [26] J. Liers, 1. Wittekind, C. Theune, Interna- reich sowie Gegenmassnahmen und Emp- tional Council of . Committee fehlungen, Deutsches Bibliotheksinstitut, for Conservation. II th Triennal Meeting Berlin, 1992. Edinburgh, Scotland, 1-6 September [42] Full text: www:http.snl.ch. 1996, Preprints Vol. II, Ed. 1. Bridgland, [43) M. Piller, personal communication. James & James, London, 1996,533. [44] 1. Dufour, 1.B.G.A. Havermans, Restau- [27] 1. Liers, P. Schwerdt, Restaurator 1995, rator 2001,22,20. /6, 1; 1. Liers, B. Vogelsanger, Das (45) M. Anders, K. Bredereck, A. Haberditzl, Papier 1997, 5/, \l8. International Council of Museums. Com- [28] C. Theune, persona] communication. mittee for Conservation. I Ith Triennal [29] R. Nebiker Toebak, M. Reist, S. Herion, Meeting Edinburgh, Scotland, 1-6 Sep- M. Piller, A. BHiher, Arbido 2000, 2000/3, tember 1996, Preprints Vol. II, Ed. J. 20. Bridgland, James & James, London, 1996, 130] J. J. Kozak, R.E. Spartz, 'Deacidification 481. of Paper by the Bookkeeper Process', in [46] 'Kriterien zur Entscheidung tiber die An- 'Paper Preservation, Current Issues and wendbarkeit von Massenkonservierungs- Recent Developments', Ed. P. Luner, verfahren', project leader G. Banik, by the Tappi Press, Atlanta, 1988, p.129; 'An DFG (Deutsche Forschungsgemeinschaft) granted project III N2-557 22/98. Evaluation of the Bookkeeper Mass De- [47] T. Doering, P. Fischer, U. Binder, 1. Liers, acidification Process', Technical Evalua- G. Banik, Advances in printing science & tion Team Report for the Preservation Di- technology, 2000, 27, 27; J.L. Pedersoli, rectorate, Library of Congress, Washing- 'The Development of Micro-ana]ytical ton, D.C., Ed. S.M. Melnick, Pittsburgh, methodologies for the characterisation of PA, 1994; S. ZumbUhl, S. WUlfert, publi- the condition of paper', in Internationale cation accepted for Studies in Conserva- Arbeitsgemeinschaft der Archiv- Bibli- tion. otheks- und Graphikrestauratoren, 9th In- [31) J.E. Burd, J. van Dorsten, personal com- ternational Congress, I5th-21 th August munication. 1999, Copenhagen, Preprints, Ed. M.S. [32) K. Harris, personal communication. Koch, The Royal Academy of Fine Arts, (33) O. Bell, personal communication. Copenhagen, 1999,p. ]07. [34] R.A. Guerra, J.A. Latlin, J.R. Mendo Escoda, "Desacidificai6n en masa: nuevos desarrollos. Aplicaci6n simultanea de agentes desacidificantes e inhibidores del crecimiento microbioI6gico." Terceras Jornadas de A vances en la Restauraci6n de Papel, 22-25 Mayo del 200], Preprints, Universitat Politecnica de Catalunya, 2001. (35) H. Schmidt, C. Becker-Willinger, A. Sau- er, 'New Techniques to Combine Deacidi- fication and Restrengthening of Paper in One Process', ECPA Conference 'Mass Deacidification in Practice', BUckeburg, Germany, 18.-19. Oct. 2000, http:www. ecpa.knaw.nl. [36] G.B. Kelly, 'Non-aqeous deacidification of books and papers', in 'Conservation of Library and Archive Materials and the Graphic Arts', Ed. G. Petherbridge, But- ~rworths, London, 198~ p. 137; J.