氦比重瓶法测孔隙度.pdf

POROSITY DETERMINATION WITH HELIUM PYCNOMETRYAS A METHOD TO CHARACTERIZE WATERLOGGED WOODSAND THE EFFICACY OF THE CONSERVATION TREATMENTS*I.D.DONATO and G.LAZZARADipartimento di Chimica Stanislao Cannizzaro,Universit degli Studi di Palermo,Viale delle ScienzeEdificio 17,90128 Palermo,ItalyThe helium pycnometer allows us to measure the cell-wall density of dry woods and the basicdensity of wood samples soaked with water and/or a consolidant solution if a non-volatilesolvent is used.These parameters were correlated to the porosity,which for degraded water-logged wood is related to the maximum water content.Moreover,this has revealed thepossibility of investigating,by means of accurate cell-wall density determination,the efficacyof several consolidants in the treatment of waterlogged woods.KEYWORDS:HELIUM PYCNOMETRY,POROSITY,DENSITY,WATERLOGGED WOOD,CONSOLIDATIONINTRODUCTIONMany properties of archaeological wood,particularly those involving deterioration,sorption,penetrability,swelling and strain-related phenomena,are probably dependent on the porosity ofthe wood,which,in turn,depends on cell-wall and basic densities(Capretti et al.2008).Thus,accurate measurements of these physical parameters are essential for a correct determination ofsuch properties.Several conventional methods have been successfully applied for determinationof the physical and chemical properties of waterlogged archaeological woods(Capretti et al.2008;Pizzo et al.2010).The concept of density is very simple,and the mass and volume of abody would seem to be among the easiest physical parameters to measure.Actually,for a porous,hygroscopic,polymeric material,the measurement of the volume is extremely controversial.Thisis particularly true in the determination of basic and cell-wall densities of wood.For determina-tion of the basic density,the volume is measured by either of two techniques.The volume ofsamples with a proper geometrical shape is obtained from the characteristic dimensions measuredby means of a calliper,whereas for samples with an irregular shape,the volume is obtained bymeans of a Regnault pycnometer,employing mercury as a reference liquid.The measure of thecell-wall volume in order to determine the cell-wall density is more difficult.Gas pycnometry,based on the BoyleMariotte law of the volumepressure relationship,isan effective tool to determine the volume of porous materials because it does not possess thelimitations of other testing methods(Tamari 2004),such as the problem of air entrapment.In gaspycnometry the main approximation in the estimation of the sample volume is the nonideal gasbehaviour and its adsorption on to the solid material.Nevertheless,experimental results(Weberand Bastick 1968)and modern simulations of gas adsorption on solids(Neimark and Ravikovitch1997;Talu and Myers 2001)have demonstrated that helium can reasonably be considered as anideal and non-adsorbing gas at room temperature(300 K)and low pressure(0.5 MPa).A gas*Received 12 July 2011;accepted 7 November 2011bs_bs_bannerArchaeometry 54,5(2012)906915doi:10.1111/j.1475-4754.2011.00657.x University of Oxford,2012pycnometer,using pure helium,was employed in this work to measure the cell-wall density ofdegraded waterlogged wood after dehydration,and of wood finds preserved with enforcingmaterials after exsiccation.Moreover,helium pycnometry could be used to measure the basicdensity when a wood sample soaked with water and/or a consolidant solution of a non-volatilesolvent is used.The primary purpose of this study was to:(a)correlate the Maximum Water Content(MWC%)(Florian 1990,8)with the porosity ofdegraded waterlogged wood;(b)test the modifications of the cell-wall density of softwood and hardwood dehydrated archaeo-logical finds;and(c)study the effects on the porosity of several chemicals used as strengtheners for degradedwoods.EXPERIMENTALMaterialsPoly(ethylene)glycols(PEG 4000,PEG 1500 and PEG 600),hydroxypropylcellulose(Klucel,Mw=80 kDa)and a-d-glucopyranosyl-a-d-glucopyranoside(TREA)are from Sigma.Theacetone was a J.T.Baker product.Colophony(from Phase)is a complex mixture essentiallycomposed of isomers of abietic acid(90%)and the corresponding esters,aldehydes and alcohols(10%).Rosin 100(from Bresciani s.r.l.)is a chemically modified colophony(stabilized ester ofpentaerythritol).Vinavil 8020Sis a commercial copolymer(vinyl acetate and vinyl versate)from Mapei SpA.The densities of the pristine consolidants and their solutions used for theconsolidation treatments are given in Table 1.Wood samples and characterizationWood samples were collected from findings related to the ancient vegetation coeval with theancient ships of Pisa(Italy)from that archaeological site.They are dated from the seventh centuryTable 1Densities of dried consolidants and their solutions at 25CDensity of solutionDensity of dried consolidant(g cm3)(g cm3)20%PEG4000+2%TREA in water1.0421.20,*1.5820%PEG1500+2%TREA in water1.0641.20,*1.5820%PEG600+2%TREA in water1.0441.20,*1.587%Klucel in water1.0091.2760%colophony in acetone0.9701.0860%Rosin 100 in acetone0.9571.077%Vinavil 8020S in acetone0.8170.93*PEG.TREA.The temperature is 30C.Porosity determination with helium pycnometry907 University of Oxford,2012,Archaeometry 54,5(2012)906915bc to the second century ad(Giachi et al.2011).Samples of waterlogged wood were character-ized according to the procedure of the Italian standard UNI 11205:2007(see UNI 2007).Theidentification of wood taxa was done by means of optical microscopy on thin sections along thethree characteristic directions of wood;from this analysis and the comparison with the taxonomictables UNI 11118:2004(see UNI 2004),the wood samples were classified as softwood(Larixdecidua Mill.,Pinus nigra Arnold and Pinus pinaster Aiton)and hardwood(Arbutus unedo L.,Ulmus cf.minor,Fagus sylvatica L.and Quercus sp.caducifolia).Wood treatmentsThe wood desalination was performed by washing samples with deionized water at 25C;theprocedure was halted when the conductivity was?10mS cm1:(a)Waterlogged archaeological wood was consolidated after desalination by impregnationcarried out by immersing the samples in a solution of water-soluble polymers(PEGs+TREAorKlucel).Refractive index(nD)measurements were utilized to follow the evolution of the processof impregnation.The impregnation was considered complete when the refractive index of thesolutions in which the sample was immersed approached the concentration value of the impreg-nating mixture before its use,within the limits of the measurement error.Drying after treatmentwas carried out by freeze-drying(45C,104atm).The latter operation was performed by usinga Heto FD 2.5 freeze-dryer.(b)Different samples of waterlogged archaeological wood were consolidated by immersion inacetone solutions of:(60 wt%)colophony,(60 wt%)Rosin 100 and(7 wt%)Vinavil 8020S.Inthis case,the impregnation water was replaced by acetone,and the refractive index(nD)mea-surements were used to follow the kinetics of the process.Finally,acetone-filled samples wereimmersed in the consolidating solutions.The diffusion of these solutions into wood was followedby measuring the variation of the consolidant concentration with time through viscosity mea-surements.Flow-time was measured at constant temperature by means of a Ubbelohde-typecapillary micro-viscosimeter,equipped with an optical sensor and an AVS 440 automatic unitfrom Schott-Gerte.A second impregnation step was carried out after the first one,and in somecases the temperature was increased in order to speed up the process.After treatment,each sample was placed into a box at room temperature(25C and 1 atm)toslowly remove the acetone;the box was opened periodically to let the acetone vapours out(Giachi et al.2010).MethodsThe Accupyc 1330 Micromeritics gas pycnometer,which used 99.995%pure helium,wasemployed to determine the volume of the samples by measuring the pressure change of heliumin a calibrated volume.The apparatus had a 10 cm3cell,well suited for the small availablequantities of many of the studied materials.Standards for the volume calibration(balls purchasedfrom Micromeritics,Vcal=6.371684 cm3)were used at 25C.The experiments were performed by using the cell with a 75%filling ratio.The results showedthat density values and standard deviations(corresponding to the 10 repetitions for each analysis)decreased with the number of purges.The repeatability of the technique was investigated byconducting three independent analyses on the same sample.The measurements were performedon the same day with a new sample for each analysis,because of the possible evolution of theproduct during purges and measurements.The pressure was considered constant if the rate was908I.D.Donato and G.Lazzara University of Oxford,2012,Archaeometry 54,5(2012)9069150.34 mbar min1.The variation in the pycnometer density,due to the operating parameters,canaffect the accuracy of the result up to 0.01 g cm3.Before each cell-wall density measurement,thewood samples(dried at 105C and freeze-dried,or air dried at 25C)were kept in a desiccator for2 h under vacuum.RESULTS AND DISCUSSIONWaterlogged degraded archaeological woodWithin the field of cultural heritage,wood plays a relevant role.These finds can survive better inwet environments where microbial and fungal activities are limited,as in the case of underwatershipwrecks.In such environments,anaerobic bacteria are primarily responsible for the depletionof wood(Bjrdal and Nilsson 2000,2008).Waterlogged woods are characterized and classifiedbased on physical,chemical and biological parameters,by analysis of the wood structure and byoptical and electron microscopy.The water content and density are the physical parameters mostfrequently determined,in order to obtain a rapid classification of decay for waterlogged wood(Florian 1990,8).The maximum water content(MWC%)is determined as follows:MWCmass of water in wet wood sampleoven-dry mass o%=()100f f wood sample()(1)The wet wood sample is dried at(105 1 2)C to constant mass(accurate to 0.0001 g);the massof water is determined from the difference in the mass of the sample before and after the dryingtreatment.The basic density(db)was calculated from oven-dry mass and water-saturated volume;thismethod is appropriate for degraded waterlogged woods.The basic density of archaeologicalwood is constrained between the value for identical sound taxon and zero as the wood approachestotal destruction.In other words,the density deviation from the value of sound wood is a measureof the extent of deterioration.The cell-wall density(dcell-wall)was based on oven-dry weight and dry volume,namely thevolume occupied by the cell wall.The porosity of wood(Z%)is related to the cavity volume(Vc)and to the volume of the water-saturated sample(Vtotal).Moreover,(Vcell-wall)is the volume of thecell wall.The following equations correlate these parameters:VVVctotalcell-wall=()(2)ZVVVdd%=()=()1001001totalcell-walltotalbcell-wall(3)Mercury intrusion porosimetry(MIP)is used,in general,to evaluate the pores volume and theirsize distribution.However,this instrument cannot be used for porosity measurement of water-logged woods,because of the necessary high pressure,which leads to compression of wood andconsequently to the collapse of a number of pores or voids.Instead,the accurate determinationof the cell-wall density and,therefore,of the porosity is made possible by using the gaspycnometer.At present,the instrument considered to give the closest approximation to the truedensity of the cell wall is the helium pycnometer.The accuracy is due to the fact that heliumpenetrates into the smallest pores and crevices,approaching the real volume.The use of thisinstrument for cell-wall density determination offers the advantage of being easy to use andPorosity determination with helium pycnometry909 University of Oxford,2012,Archaeometry 54,5(2012)906915rapid,especially with a fully automated apparatus.The accuracy and the reproducibility of thetechnique are sufficient to reveal minute variations.For each wood sample,MWC%,dband dcell-wallwere determined at 25C,and the porosity,Z%,was calculated from the density values shown in equation(3).These quantities arereported in Table 2.The investigated waterlogged woods are very degraded,as the calculatedvalues of MWC%show,in agreement with results from the holocellulose-to-lignin ratioreported elsewhere(Pizzo et al.2010).The dbvalues,experimentally determined by usingthe helium pycnometer,are related to those of MWC:high water contents are linked to alow basic density of decayed waterlogged woods.The dcell-wallvalues for each sample vary.They differ from the literature values for sound woods(dcell-wall=1.50 or 1.53 g cm3)and forwaterlogged woods(dcell-wall=1.48 g cm3),as reported by Jensen and Gregory(2006).From amore accurate analysis of the results,it emerges that the taxon and the degree of deteriora-tion can affect the structure of the cell wall;indeed,the dcell-wallvalues are in the range from1.41 to 1.38 and from 1.37 to 1.50 g cm3,for archaeological softwoods and hardwoods,respectively.The softwoods contain more lignin than hardwoods,and lignin has the lowest specific gravity(1.30 g cm3)compared to cellulose(1.54 g cm3)and hemicellulose(1.53 g cm3)(Pfriem et al.2009).The dcell-wallvalues of waterlogged woods are markedly dependent on the nature ofdegradation of the cell-wall components.Biological and chemical degradation cause depoly-merization of the polysaccharide matrix against a limited degradation of the lignin fraction;this explains why the cell-wall values approach the lignin value upon increasing degradation.A multi-analytical study of degradation of lignin in waterlogged wood reported that ligninfrom archaeological wood can also be altered,leading to a higher amount of free phenol unitscompared to lignin from sound wood of the same species taken as a reference(Colombini et al.2009).A loss of cellulose and hemicellulose and the modified lignin caused the formation ofwater-filled cavities and resulted in a porous and fragile structure.Table 2The taxa and physical characteristics of the waterlogged wood samplesTaxaMWC%dbdcell wallZ%(g cm3)(g cm3)s 1 0.1s 1 0.01s 1 0.01SoftwoodEuropean larch,Larix decidua Mill.5460.161.4088.7European larch,Larix decidua Mill.5050.181.4187.2Stone pine,Pinus pinaster Aiton6890.131.3890.6Stone pine,Pinus pinaster Aiton5590.161.3888.4Austrian black pine,Pinus nigra Arnold3750.221.4183.4Austrian black pine,Pinus nigra Arnold3900.221.3984.2HardwoodStrawberry tree,Arbutus unedo L.7040.131.3490.3Elm,Ulmus cf.minor5490.171.3687.5Elm,Ulmus cf.minor4790.191.4186.5Oak,Quercus sp.caducifolia4710.191.4486.8European beech,Fagus sylvatica L.6780.141.3789.8910I.D.Donato and G.Lazzara University of Oxford,2012,Archaeometry 54,5(2012)906915The MWC%values are strictly correlated with the cavities present in the samples for boththe hardwoods and the softwoods.The MWC%determination based on both gravimetry andthermogravimetry(Donato et al.2