Canadian Mineralogist Y ol. 24, pp. 307-322 (1986)

ARCHITECTURALCONSERVATIONAND APPLIEDMINERALOGY RICHARD A. LIVINGSTON Departmentof Geologlt,Universityof Maryland, CollegePark, Maryland20742,U.S.A.

INTRODUCTION

ABSTRACT The field of architecturalconservationconcemsthe restoration and protection of buildings and other structuresfrom environmental agentsof deterioration such as climate, air pollution and acid rain. Although usually thougtrt of only in relation to historic buildings, the principles of architectural conservation can also be applied to contemporary structures. Applied mineralogy is an essential part of architectural conservation. It is usedto study the mechanismsof deterioration,to diagnosethe problemsof individual buildings and to develop methods of dealing with the problem. Typical usesof appliedmineralogyinclude characterization of building materials, identification of altered surface-layers,and the prediction of the durability of both the original material and the alteration products as a function of environmental conditions. Seveqalapplications are reviewed including: inherent vice, air-pollution attack on limestone and marble; failure of iron reinforcements in masoffy, damageby soluble salts to masonry, corrosion of bronze and the performance of weathering steel, Keywords: architecture, building materials, air pollution, corrosion, marble, bronze, reinforced concrete,salt damage,steel, acid rain.

Souuarnr Le domaine de la conservation architecturale porte sru la restauration et la protection d'ddifices et autres structures desagentsdansl'environnqment qui contribuent a leur d6t6rioration, tels que le climat; la pollution atmospherique et les pluies acides. Quoique I'on 6tudie gdndralement cesfacteurs pour les 6difices d'int6r€t historique, les principes de la conservation architecturale s'appliquent aussi bien aux structurescontemporaines.La min6ralogie appliqu6een forme une part essentielle.Elle s'adresseaux m6canismesde d€t6rioration, au diagnostic de probldmes dans descasparticuliers el au d6veloppementde m6thodespour rem6dieraux probldmes.Par'exemple,on s'en sert pour caracteriserles mat6riauxde construction,identifier les couchs alt6r€esd leur surface,et pr6dire la durabilitd desmat6riaux originels et desassemblagesd'alt6ration en fonction du milieu. On discute de plusieurs applications, dont les ddfauts cach6s,l'attaque du calcaire et du marbre par la pollution atmosphdrique,la d6confiture desrenforts de fer dans les pierres tai1l6es,le dommaged0 aux selssolubles, lacorrosion du bronze et le rendementde l'acier traitd pour I'environnement. (Traduit par la R6daction) Mots-cl6: architecture, matdriaux de construction, pollu' tion atmosph6rique,corrosion,marbre, bronze,bdton arm€, dommagepar les sels,acier, pluies acides.

The field of architectural conservation concerns the study of the deterioration of buildings and methods of preventing this damage. Historic buildings have receivedthe most attention, but the principles of architectural conservation can also be applied to modern buildings as well. Sincemany of the problems addressedby architectural conservators could have beenpreventedby more enlightened constructionand maintenancepractices,a knowledge of architecturalconservationcould be usedto advantage at the stageof the specificationof materialsand architecturaldesign. The fundamental problem of architectural conservation concernsthe failure in someway of building materialsto perform asintended. Usually, the situation is the acceleratedweatheringor corrosion of the material, to the point wherethe building's structural integrity may be compromised.However, aesthetic considerationscan also be important. These include the loss of details in sculpture, the staining of a surfaceor the loss of a polished finish. The practice of architectural conservation involvesseveralaspectsof applied mineralogy. The durability of inorganic building-materials is determined largely by the physical properties of the individual mineral phasesof the materials, suchas crystal structure, rnolar volume and hydration states, as well as petrographic considerationsconcerningthe physical relationshipsamongthe phases,including grain morphology and porosity. The materials and their alteration products are thus usually studied by the methodsof optical and electronmicroscopyand Xray analysis.Occasionally,infrared absorption spectroscopy, differential thermal analysis and other scientific tools are also employed(RossiManarese 1982). The actual rates of weathering and deterioration are usually determinedby kinetics rather than thermodynamic equilibria, and involve some difficult problems in applied geochemistry.Theseare beyond the scopeof this paper. Nevertheless,simply from a study of the mineral assemblagesfound in and on building materials, the architectural conservator can make reasonablyaccurateinferencesabout the major causesof deterioration. The causesof deterioration can be grouped into severalmajor categories,beginning with the use of inferior materials in the original constructidn, or inhslsnt yise. The failure ofprotective surface-layers

307

308

THE CANADIAN MINERALOGIST

or the formation of deleteriousalteration-layers are two more categories.The presenceof soluble salts is another. Interactions betweenbuilding materials is a separatecategory,as is biological attack. Finally, measurestaken to protect a deteriorating building can sometimesintroduce additional problems.

ble than others found in Italy. A century ago, Julien (1883)published a listing of the lifetimes of different building stones in New York City, which indicated that gneisswas the most durable and brownstone the least (Table 1). A drawback to this method of specifying durability is that visual appearancecould be deceptive.Thus INITERENT VIcE building stonesof widely differing mineralogy were sometimesmistaken nith each other, creating conThe original choice of building materials is obvi- fusion about true psrformance 3sfuilding materials. ously a major determinant in the durability of the For instance,a building stoneusedwidely in Flanders structure. Materials that fail because.of intrinsic during the Middle Ages was known as "petit granit" tltoftssmings are said to suffer from inherent vice. or little granite from its resemblanseto granite, This is especiallyprevalent in stone, and the impor- although it is in fact a compast grey limssfele (Nijs tanceof choosingdurable stonehasbeenemphasized 1985).Similarly, tle term alabasterwasappliedintersince antiquity. The Roman author Yitruvius, writchangeablywith onyx marble and with gmsum until ing in the first century A.D. (Morgan l9l4), this century Merrill 1908,p. 243). Sincethe soludescribedthe relative durability of building stones biliry of the latter is an order of magnitude greater from quarries in the vicinity of Rome. This is typi- than tle former, an architectural element made of cal of the body of conventionalwisdomthat hasbuilt gypsum where marble was actually intended would up over the centuries concerning the suitability of show unexpectedlyrapid deterioration when exposed stonesfrom a given region for use in construction. to weather. Thus the first step for the architectural Unfortunately; this fype of classificationdepended conservator in diaCnosingthe causesof deterioraprimarily on geographyrather than mineralogyitself, tion is to verify that the mineralogy of the building in the sensethat the durability of the stone was stoneis actually what the property owner tlinks it is. describedas a function of where it was quarried and Nevertheless,certain types ofinherent vice could not of its composition. As long as the stone'waszup- be attributed to easilydistineuishableheterogeneous plied from familiar quarries, inherent vice could be minerals occurring as veins or inclusions (Lewin & avoided. However, in the absenceof a durability Charola 1979).For example, the prominent silicate index basedea minelslsgy, it would be difficult to stylolites in Tennesseemarble will weather out predict the performance of stone from a newly preferentially (Winkler 1975, p. lM). In a somewhat opened quarry. Even in this century, the durability similar case,the marble from Lee, Massassachusfis, of stone has still been specified largely by its place usedon someof the exterior of the U.S. Capitol conof origin (Merrill 1908) rather than in terms of a tains small flakes of tremolite, which weatherto talc quantitative relationship using mineralogical and upon exposure.The associatedchangesin dimension, petrological parameters.This lack of a generalthe- plus the presenseof moisture that encouragesfrost ory of stone decay has been identified as a major damage, create localized centres of deterioration problem in the field of architectural conservation (Winkler 1982).The result is a pockmarked appearClorraca 1982). ance on the stone surface (Fig. l). Another type of This is not to say tlat mineralogy was totally differential weathering has been observedon sculpignored. To the extentthat different rock-typescould ture in India made of "khondalite", a quartzofeldsbe identified by macroscopic properties like color, spathic gneiss. This contains proqinent crystals of texture and hardness,crude estimatesof durability garnet, which tend to weather out preferentially, could be made. Vitruvius observedthat c€rtain t)'p€s leaving noticeablevoids and pockmarks (Lal 1985). ofrocks suchastravertine weregenerallymore duraThe existenceof troublesomeheterogeneouslydistributed minerals is not always so apparent to the naked eye.For example,the limstond that was used 'I. in tlte KansasStateHouse containsvarying amounts TABLE LIFETIMEOF BUILDINGSTONESIN NEII YORKSITT' ofsilicates (Grisafe 1982).The deteriorationofthe Coarse-grained bromstone 5-15 stone appearsto correlate with higher clay and silt Flne-grained broxnstone 20-50 Coflpact bromstone contents (Fig. 2). This was only establishedtlrough 100-200 Nova Scotia sandstone 50-200 a detailedmineralogical examination. Similarly, cer0hlo sandstone Centurles Coarse fossllI ferous llnestone 20-40 tain Egmtian limestone artifacts have been known Fine oolltic llnestone 30-40 to disintegraterapidly when washed.An analysisof Coarse-gralned marble 30-40 Flne-gralned marble 50-100 the stone showedthat the most sensitivecontained Fine-g.al ned dolonitic marble 60-80 relatively high amounts of chlorides as well as siliGranite 75-240 Gnelss Centurles cates @arton & Blackshaw 1976). These findings * Jullen (i883). were used to establish an index of washability, in

ARCHITECTURAI,

CONSERVATION AND APPLIED MINERALOGY

309

Frc. l. Pockmarks on the U.S. Capitol as a result of the selective weathering of tremolite.

terms of these two variables, that could be usd to KEY: determine whether it is safeto wash a given artifact Em RslatlvelYhlgh dog@ ot wdhslhg wlttr @ (Fie. 3). alglflcent tpalllng ot flaklng of don€ A more benip form of inherent vice concernsthe tra"*r,nn *ttl| sllght 8patllng ffi esthetic problems that arise when a stone discolors wutnertng wtth no spaltlng [--l after installation on a building. This is particularly qnwthsrsd tune a problem with stones containing iron hydroxides n"t.totv l--l l2 and oxyhydroxides, which may not be stable when exposed to the atmosphere. Also, many of these stonescontain a certain amount of porewater. After the stone has been quarried and put into place, the evaporation of the porewater can causesignificant oi changesin the mineralogy, leadingto color changes. T The sandstone used in the Smithsonian Castle in Eo AWashington, D.C., for example,was intended to be E a lilac grey, but over time the color of the sandstone IE has deepenedto a dark red (Witherineton 1975). Another example is the white Pentelic marble used o on the Parthenon, whish turns to a golden color when exposed to the atmosphere, owing to trace o- 4 quantities of Fe-bearing pignent deposited on the d a surfaceupon drying out of the stone (Winkler 1975, oo p. 78). A third exampleinvolveswhite Bethelgranite from Maine. This is particularly susceptibleto staining in the presenceof water due to the rapid kaolinization of the feldspars (GaIe et al. 1985). Theseinstancesof inherent vice can be explained ll12l58 1722 4 1 3 I 1 8 1 4 2 13 5 6 1 6 2 0 1 99 r c n 7 in terms of clearly identifiable mineralogical characcon Numld teristics. However, there are other situations where the causesare not so appaxent.There is a great variation in durability among limestones and marbles Frc. 2. Durability of KansasCity Courthouse limestone as a function of mineral composition (Grisafe 1982). that does not appear to be attributable to the !

!

!. g

310

THE CANADIAN MINERALOGIST

less than 35 years before being replaced by larger ones. On the other hand, historic monumentsare expectedto last indefinitely. The longest finite time period that has been specified for the durability of building materialsis 10,0S years.This hasbeenproposedfor markers at disposalsitesfor nuclear waste @erry 1983). Mankind is also inadvertently shortening the lifetime of many buildingsby changingthe environment, so building materials once appropriate are no longer as durable. The effect of increasinglevelsof air pollution on stone deterioration has been well documented(Amoroso & Fassina 1983).Another exampleof the deleteriouseffect of human activites on stone durability concernsthe Temple of Karnak in Egypt. After surviving for thousandsofyears, the temple is being threatened by deterioration caused 12345678910t! by the recrystallization of soluble salts within the A ACID I'ISOLUALE IATIEF stone. This is associatedwith changesin the movement of groundwaterthat occurred after the conFtc.3. Criteriafor hazardsof washingEgyptianlimestone of the Aswan Dam (Hartline 1980). asa functionof insolubleand chloridecontent(Bar- struction ton & Blackshaw1970. Sunrecn LeYeRs presenceof prominent heterogeneous mineralssuch as silicatesor salts. One school of thought holds that the differences are associatedlargely with physical characteristicsrelated to the petrographictexture, such as pore-sizedistribution (Robertson 1982)or internal surface-area(Hudec 1977).Another school leanstoward a mineralogical explanation, including degreeof dolomitization (Caner& Seeley1981,Pellerin 1976).Presumably,both the petrographicand mineralogicaleffects could be related aspectsof a more fundamental phenomenon associatedwith diageneticor metamorphic processes. Finally, it should be uoted that inherentvice is a relative concept, since a stone that is unsuitable for exterior use can still be utilized under shelter. Consequently,whether or not a material is said to suffer from inherentvice dependsto a certainextenton how wisely it is used. Even alabastercan be used for architecturalpurposesif shelteredfrom rain, as in the caseof the windows of Galla Placidia'slomb in Ravenna(Winkler 1975,p.45). Inherent vice is also relative to the desiredlifetime of the material in service.If the material falls significantly short of surviving for the specifiedlifetime, then it can be said to suffer from inherent vice. Some materials like galvanized fencesmay be expectedto last for 20 years, whereasmost structural materials are expectedto last the lifetime of the building. The building's life is itself relative. For economicpurposessuchas amortizing construction costs,a building's life can be setat fifty years(AIA 1977).However, this may not reflect the real lifetime. In some cities like New York, many buildings have lasted for

A building material, upon exposure to the atmosphere, will commonly react with various environmentalagentsto producea surfacelayer. This is specially true with metals,but may also occur with certain types of stones.The presenceof this layer can modify the mechanismsand the rates at which degradationoccurs.Furthermore, dependingupon the specificsystemof materialsand environmentfactors involved, the effect of the surfacelayer can be either detrimental or beneficial. The simplest caseis probably that of oxidation of metals.Upon exposureto the afinosphere,a baresurface will begin to react wilh oxygen. As more and more of the surfacebecomescoveredwith a layer of oxide, the rate of reaction slows down. Eventually, the surface becomes coated with a layer of oxides, which introduces a barrier to further reaction. The oxygen must diffuse through this layer in order to react with metal; the rate of reaction will tend over the long term to be proportional 10 the squareroot of time, rather than a linear function, which would be the usual case for the bare metal (Uhlig 1964).The slowingdown of the rate of reaction in effect is the slowing down of the corrosion rate, and, under theseconditions, the metal is said to be passivated@ourbaix 1973).In fact, this is an oversimplification, and the actual mathematical forms of the corrosion rate are still not completely resolved @vans 1968). Nevertheless,the concept remains useful in understanding the importance of the mineralogy of the surface layer in protecting against further deterioration. If the surfacelayer is to be protective,it must form a continuous, tightly adherent layer. Since metal

ARCHITECTURAL

311

CONSERVATION AND APPLIED MINERALOGY

oxides are usually brittle, and thus not able to resist tensile stresses,it is important for the surface layer to be in compressionto avoid cracking. This led Pil'ing and Bedworth (Jhlig 1964)to proposea criterion of protection basedon the ratio of the molar volume of the metal oxide in the surface layer (V) to the molar volume of the original metal (VJ. If the ratio is greater than one, the surface layer would be in compressionrather than tension:

TABLE2.

MTIO OF METALIO METALOXIDE* VOLUME

METAL

OXIDE

-Li th'i un Tltani um Al um'lnum Tin Lead Zi nc Copper Iron

Li 20 Ti0

RATTO 0,58 1.20 1.28 1.32 1.37 1.55 L.72 2.10

Al cOe Snb2" Pb02 Zn0 Cur0 re2u3

*Derived from CRCHandbook(1972).

!b >! trV-

(l) TABLE3.

wheren is the number of metal atomsin the chemical formula of the surface layer. Therefore, the mineralogy of the surface layer, which determines the molar volumenis important in determiningthe metal's resistanceto corrosion. Valuesof the Pilling-Bedworth ratio for a number of metal oxidesare presentedin Table 2. lt can be seenthat the ratio is less than one for lithium, calcium and sodium. Consequently,thesemetalswiII all tend to oxidize completely upon expo3ureto the atmosphere,and thus are not usablefor structural purposes. The metals aluminum, lead, zinc and tin have a ratio in the rangebetweenI and 1.3, and thus are suitable for construction. Copper has a fairly high ratio, and the ratio for iron exceeds2. This amount of expansioncan causesignificant compressive stressesin the surfacelayer, causingit to spall off. In fact, these two metals tend to form thick layers known as scale, which do not adhere as well to the surfaces. Although it was originally developed for oxidation in dry atmospheres,as in the caseof mill scale formed on heatediron slabs,the Pilling-Bedworth ratio is useful as a conceptualtool to understandthe behavior of a surface layer in other circumstances. For metalsused in architecture,outdoor exposure usually meansexposureto moisture.A wet metal surface can develop a variety of alteration products besidesoxides. The most obvious would be the hydroxides, but it is possiblealso to form carbonates through reactions with atmospheric carbon dioxide @eitknecht1959).In coastalareaswheremarine salt is presentin the atmosphere,metal chloridesare also

ONMETALS* TYPICALSURFACE-LAYERS I'4OLAR VOLUME'MTIOS, I'IIlIERAL/I'IETAL MOIAF-TONffTFITIO

Lead Cerussite Hydrocerusslte

PbC0r 2Pbcd3.Pb(0H)2

Zlnc Smlthsonlte 6oslarlte

ZnCo? znso;.7H20

Copper Malachlte Brochantlte Atacmlte

CuC03.Cu(0H)z Cus04.3Cu(0H)z cucl2.3cu(0H)2

Iron Goethite ibl anterlte

Fe00H FeS04.7H20

2.2 2.3 3.0 15.6 3.8 4.0

3.0 10.0

*Derlved frm CRCHandbook(1972)

found. Finally, in areas with significant air pollution, sulfates formed by reaction with sulfur dioxide gas or sulfate particulates may be found on the surface. The most co[lmon of thesesurface-layerminerals and their molar-volumeratio are listed in Table 3. The molar-volume values all easily satisfy the Pilling-Bedworth criterion. Many of the typical mineralsin corrosion layersare not neutral salts, but rather the basic compoundsthat incorporate additional hydroxide molecules.Thus, on copper sruis found rather faces,brochantiteCuSO4.3Cu(OH)2 than chalcanthiteCuSO4.7H2O. This prevalence of basic compounds can be explained by two factors. One is the low concentrations of SOr a4d chloridesrelativeto oxygenin the

TABLE4.5oLuBItITIEs0FsIt'{PLEANDBA5lcsALTs0Fl.t!@ .

SIMPLE SALTS

I'lETAL OXIDE AI Pb Zn Cu Fe Sn

64.0t 1.61 -

HYDROTIDE CARBOMTE

1.05 2.6 lo-tt 2.s 5 0.079 2.0$

1.8i 10.0t 30.05 67.Ot

SULFATE.

CHLORIDE

iu+fs zo.r ro+fs ru.r.r 26.010.?r 70.0lurTt 78.010r:t 36.0 10'qi re.oro+j5 1 . s 1 0 ' 1 5 42.0 10+:S 3e.0iotls 16.0 10'{E 73.0lo*c$

r Edrth-surface atnosphere and 25oC; values ln mg/L Llnke(1965).5 " Stephen& Stephen (1963). Sources: t.

BASICSALTS

SULFATE CHLORIDE CARBONATE

1.31

22.or

79.01

59.01

143,0i

169.0t

3t2

'

"T*WAir:ADIMDIERSTGIST

FIc.4. Streaks of color on bronze statue due to differences in contact with rainwater. Also, stains on marble pedestal due to runoff of soluble copper salts. Coicoran Gallery of Art, Washington, D.C.

atmosphere. This results in a low ratio of these anions on the surfacecomparedto the hydroxides. The other reason is that the solubilities of most of the neutral salts are much higher than those of their basic counterparts, as can be seenin Table 4. Consequently,in the presenceof flowing water, the simpler minerals would be dissolved rapidly, whereas the basis compounds would tend to remain, forming a protective layer. The inability ofzinc to form sparingly soluble basic sulfates may explain why its corrosion rate remains linear with time in polluted areas, whereas other metals show a more typical slowing down with t:me as the surface layer builds up (Haagenrud et sl. 1983). The effect of wetting can be seenmost dramatically on copper or bronze objects that are pafily sheltered from rain. The dry areas will retain their blackish brown character associatedwith the oxides cuprite and tenorite. However, tle areasfrequently washedby water will displry a variety of bluish green colors characteristic of chloride or sulfate compounds(Fig. 4). In most urban situations,the copper patina will be predominantly sulfates,exceptnear marine areas where chlorides may dominate. This

is a reflection of relative concentrationsof chlorides versassulfur speciesat the site. However, in the case of the Statue of Liberty, surrounded by the sea,the patina contains mostly brochantite with very little atzarnite (Wdker 1980).This indicatesthe relatively high prevailing levels of sulfur dioxide in the vicinity, as well as the fact that for thermodynamic reasons, the sulfate will displace the chloride anion (Evans 1968). The prediction of which minerals will appear on a given metal surfase under a particular set of envhonmental conditions is extremely complex. Generally, electrochemicalprocesseson the surface must be taken into account. Since the exposure usually involves intermittent wetting by rain rather than continuous immersion in water, steady-$ate equilibrium conditions are not maintained, and chemical thermodynarnics must be modified by kinetic aspects@ourbaix 1973).Fiually, the composition of the metal must be taken into account in the caseof alloys. Observantreaderswill have noticed that the ratios in Table 3 are usually much laxger than unity. According to the discussionof scaling given abovs,

ARCHITECTURAL

313'

CONSERVATION AND APPLIED MINERA'LOGY

this suggeststhat the corrosion layerswould continually spall off. This does occur in some cases;for example,if the basiccaxbonatelayer on zinc becomes thick enough it will break off (Waite 1980). However, the buildup of compressivestressescan be avoided in a number of ways. Often the corrosion product incorporates enough water to make it gel-like and, thus, more flexible. In other cases, up, so severaldifferent corrosionJayersare-stacked that eventhough the total changein iolume is large, the differential changegoing from one layer to the next is not drastic. For instance, the stable greenish blue patina found on outdoor copper or bronze consists of malachite. Table 3 indicates that the malachite,/copper ratio is 3.8, which suggeststhat the layer would be under considerablecompressive stress.However, the patina actually consistsof two layers, with an inner layer ofcuprite Cu2Oseparating the malachite from the native copper (Walker 1980).The volume ratio for the pairs copper,/cuprite and cuprite/malachite equals 1.7 and 2.3, respectively. Thus the effective compressivestressesare not as severe. Another example of the importance of the stacking of corrosionlayersis the performanceof weatheriqg steel. This, unlike ordinary carbon steels,forms rust layers that are stable, so that the rate of rusting slows down with time. This effect appearsto be due to the presenceof specific trace elements, notably copper, in concentrationsaround 0.390.Theseelementsaffect the crystalstructureof the rust products, which, in the caseof the ordinary non-weathering steel, are composed mainly of alpha ferric oxyhydroxide, goethite and hematite. However, on weatheringsteel,a distinct inner layer develops,consisting of cubic phases, including m?gnetite and lepidocrociteFeOOH aswell ascryptocrystallineferrous hydroxide Fe(OH)r (Keiser et al. 1982). Thts inner layer seemsto be more compactand lesspermeable to diffusion of oxygen and moisture. The electrical resistanceof the cubic phase is also higher, thereby reducingthe rate of electrochemicalreaction (Suzuki et al. 1980). Regardingthe role of copper in this phenomenon, Inouye (1968)found that small amounts of it stabilized the ferrous hydroxide crystallites. A definite peak in the size of the crystallites (3.5 pm) was reached at a copper content of 20/0.This is consistent with field experiencethat weathering steel initially rusts rapidly, but in the process, the copper remains behind, becoming enriched in the surface layer. When a coppercontentof about 2Vois reached in the surfacelayer, the rust layersbecomestabilized. The supply of ferrous iron may account for the formation of magnetite, which givesthe resulting layer of rust its characteristic dark brown color, in contrast to the normal reddish color of rust due to hematite.

IVIATERIALS SunTaCTLAYERSONPOROUS One important difference between metallic and nonmetallic surfacesis that the latter generallyhave a significant pore-structure; thus the material is no longer two-dimensional, and quite different results occur. The most typical problem is the formation of crustsofgypsum on carbonatestone.This is due to the deposition of sulfur compounds from the atmospherein various forms via acid rain or sulfur dioxide gas (Amoroso & Fassina1983).As Table 5 indicates, there is a significant increase in molar volume in going from calcite to gypsum. Over time, this meansthat the gypsum layer becomesso thick that it spalls off, exposing fresh calcite to the surfacefor further reaction. Moreover, gypsumwill also crystallize within the pore spaces. As the pores becomecompletely filled, the gypsum will begin to exert stresseson the pore walls, eventually leading to brittle failure of the calcite. Also, the g;ypsum tendsto entrap dust particles, which give it a characteristic black color (Camuffo et ol. 1983).Thesefeatures are shown in Figure 5. As Table 5 also shows, the solubility of gypsum exceedsthat of calcite by a factor of forty. This meansthat on those areasof the stone surface that are exposedto frequent rr6hing by rain' the gypsum and dust will generallybe entirely removed,leaving behind an apparently cleanwhite calcite surface. In the process,however,the erosionrate ofthe stone hasbeenincreaseddramatically over that for natural rainwater, and fresh calcite has also been exposed for further reaction. SoLLTBLE SALTs The previous section indicated that the crystallization of gypsumwithin the porous material can give rise to severeexpansivestresses.This problem is not limited solely to gypsumcreatedby reaction with air pollution. There are a variefy of soluble salts that can be transported into the material by water movement, there to crystallize, expandand causedestruction. Vitruvius pointed this out as a major factor in stone deterioration (Morgan 1914). One of the salts most commonly found in such casesis sodium chloride (halite), which is ubiquitous in marine coastalareas.In Venice, for example,this

AT 25"C OF PmTLA|{DIIE,CALCIII At{DCYpSlJl'l PROPERTIES TABLE5, PHYSICAL

r{Ar,rE

FoRnuLA T*ffi

Portlandlte

C6(0H)2

31.6

Calclte Cyps@

CaC03 CaS04.2H20

31.1

1'650 50

b

60.0

2,080

c

chrlstoffe.ss

E christoffe6en (1976).

a CRc(1972), b

Butlef (1982), c

ff;iit

* I

314

THE CANADIAN MINERALOGIST

rr#

i$

l:a::

S.:'

Ftc.5. Deteriorationof marble figuresdue to gypsumcrusts.Also, spallingof stonedue to useof consolidants.Church of S. Giovanni e S. Paolo, Venice, Italy (courtesyG. Helms).

creates serious deterioration of buildings. The damage does not occur simply at the waterline. Owing to capillary forces in the porous buildingmaterial, the water can rise as high as severalmetres. Although rising brackish groundwater is a major sourceof halite depositedin porous material$, there are other ways in which the salt can appear. One is the useof salt-contaminatedbuilding materials. For instance,the lime usedfor mortar to constructthe Taj Mahal apparentlycontainedsignificant amounts of salt. This now appears to be a graver threat to the building than prevailing air pollution (Lal 1984). In somecases,the building itself may havebeenused to store salt, with the result that salt depositscan be found whereverthe merchandisecamein contact with the wall. The Magazinv di Sala in Yenice is a clear exampleof this kind of problem. Away from marine areas, other types of soluble salts becomeimportant. Arnold (1984)has identified, among others, Epsom salts, sylvite and sodium

nitrate and potassium nitrate as causing damagein Switzerland.Theseappearto originate from groundwater containingMdt+ , C**, Na+ and K+ in solution with CO32-,Cf and NOr-. The nitrate and, to some extent, the carbonate ions arise from biological activity. As the groundwater risesthrough walls it can come into contact with gyp$um, either in the form of wall plaster or as the result of the reaction of air pollution with carbonate materials. Consequently, the more exotic sulfate speciescan form. Beyond all these,one of the most prevalent sources of damaging salts in North America is antarcticite CaCl2.6H2Oused for de-icing roads. The greatest damagewith all these salts occurs when they periodically dissolve and then recrystallize. The recurring cycle createsa wedging action that can be very destructive. In addition to dissolution-recrystallization cycles, certain salts can also undergo significant changesin volume by changing their state of hydration.

ARCHI.TEq"IIJRAL CONSERVATION AND APPLIED'MINER,ALOGY TABLE6.

315

HYDRATION RELATIVEHUIIIDIIY OF TYPICIL sOL$LE SALTS, 25OCT RH

IRAT{SITION - N!2C03.7HZ0 [ta2C03.H20 ttla2C03.7H20- Na2C03.10H20 NaS04 - NaS04.l0H20

485 791 7A

rPuehrlnger et al.(1985)

HYGROSCOPIC POIN1SOF SOLUBLE 5ALTS(25'C)T SALT KrS0a Kfloe' Nacdoa. loHz0 NaS04:10H20 KCI llaS04 NaCl l{aN0r nqtU3 Mg(N0)g) z.6Hzo Ca(NQ)2.4H20 K2C03.2H20 ItgCl2.6H20 CaClz.6Hz0

R H ,% 97.0 92.5 92.0 (18.5'c) 87.0 84.3 81.0 7a o

61.8 52.9 5!.0 42..4 33.0 29.0

z0m

B b

B B 6

a

c

S a l l r a l e l y c e l l q u e s c e n l 'E s a l r r l u c f u d l e s r r m 5 0 l l o E l dellquescent, C salt usually deliqu€scent a Arnold(1982).

A

Table 6lists severalthat may be imponant under surfaceconditions.The sodiumcarbonatesare notable in this respect as weU as the mirabilite-thenardite pair. Arnold & Kueng(1985)reportedseeingnatron and mirabilite dehydrate and rehydrate on a daily cycle on a building in Switzerland. A third aspectof the solublesaltsis their hygroscopicity. As Table 7 shows, many of the important soluble salts have hygroscopic points at relative humidities considerablybelow 10090.Thus ttrey can become deliquescent when their critical relative humidity is exceeded;when the relative humidity drops below the critical point, they will recrystalliee. Therefore, this makesit possibleto'haye many more dissolution-recrystallization cycles than would be predicted simply from the number of rain eventsper year. Arnold (1982)haspointed out an interestingcorollary. The prevailingrelative humidity is a predictable function of the lbcal climate. In many temperate climates,the relative humidity rarely drops below 55Voand also rarely exceeds8590.Therefore, salts such as antarcticite practically never dry out.On the other hand, saltswith a high critical relative humidity rarely becomedeliquescent.However, thosewith intermediate valuescan cyclebetweendeliquescence and crystallinity in responseto day-to-dayvariations in the relative humidity. The results can be seen in the form of distinct zoneson walls. The least soluble salts will precipitate first from rising groundwater and thus will not rise very high. Consequently, the lower parts of the wall will show crystals of such minerals as magnesiteand gypsum.The most solubleand hygroscopic

asa functionof Frc.6. Arnold'szonesof saltdeposition relaA. Saltsareseldomdeliquescent' hygroscopicity. varies,regionof tivelylittle damage.B. Deliquescence dargxeatest C. Saltsareusuallydeliquescent, dtrmage. ker colored,moderatedamage.D. No solublesalts present;undamaged. SeeTable7 for dataon individual salts(Arnold 1982). salts,suchas antarcticiteor nitrocalcite, will be transported to the higher Zone of the wall, which may appeiu darker owing to the continuous presenceof moistue. In betweenwill be the zone of greatest damage,where salts of intermediate hygroscopicity are found. Thesezones are illustrated in Figure 6. Thus a critical part of diagnosing the sondition of a building is taking surface samplesand analyzing the saline minerals pre$ent.Arnold (1984)has shownthat this can done readily using thetraditional methods of optical mineralogy and microchemistry. MaTSRIALINTERAcTIoNS An evaluation of the damage to buildings must also take into account the possible interaction betweenthe materials. Galvanic corrosion between two dissimilar metalsin contact in the presenceof an aqueoussolution is the classicexample. The corrosion due to the galvanic cell created betweenthe copper skin and the wrought iron armature of the

316

THE CANADIAN MINERALOGIST

Statue of Liberty is the major reason for the restoration work now in progress(Neilsen 1984). Galvanic corrosion does not occur betweennonmetallic materials, but they can still interact in other ways. As noted above, many of the alteration products are soluble and can be washed from one part of the building to another, causing further problems.For example,gypsumis producedby the reaction between calcite and sulfur dioxide in the atmosphere. The gypsum can then be transported from the limestone or marble to another material, where it can recrystallize. This has beenobservedin the caseof a limestonewall on top of a sandstone course, where the latter is deteriorating from airpollution attack on the former (Melville & Gordon 1979). Another problem often encounteredis the use of iron bars as clamps or reinforcements in ma$onry and concrete.As shownin Table 3, creationof rust leads to a major increase in volume. When this occurs within the wall, the result can be expansive forcesthat can break off the masonry.A well-known example of this effect involves the buildings on the Acropolis in Athens. As originally built in the 4th century8.C., the marble blocks werefastenedwith iron clamps,but the Greekarchitectswereawareof the problem of rusting. Consequently,the bars were all wrapped with lead, which servedto exclude air. However, during the restorationsat the start of this century, the stoneswereput back together with bare iron rods. Thesehave rusted and causedcracking of the stone.As a result, it has beennecessaryto dis-

mantle the buildings and remove the iron (Skoulikides197Q. The buildings are being reconstructed using titanium fittings (Angelides 1970. A related problem occurs in reinforced concrete where the iron reinforcements eventually begin to rust and expand (Gonzilez et al. 1983).In this case the situation is more complicatedbecausethe mineralogy of the cement paste itself changesover time. After it has initially hardened, the cementstill contains some residual calcium hydroxide. However, over time this hydroxide reacts with CO2 to form calcium carbonate, resulting in a volume decrease (Neville 1981).Thus the alterationproduct doesnot move outward to form a crust. Instead, a carbonation front is producedwithin the material that moves progresssivelyinward from the surface. When the carbonation front reachesthe reinforcemenl, rusting beginsand the concreteis broken open, as shown in Figure 7. The rusting is often attributed to the carbonationprocessitself, which reducesthe pH in the viciniry of the rod from about 13 to 8. However, Gonzalez et al, (1983) have shown that evenwhen the carbonationfront is at the reinforcement, significant corrosion doesnot take placeunless the relative humidity is above 5090 or chlorides are present. This suggeststhat the main role played by carbonation is to open the pores in the concrete, allowing the usual agents of rusting, oxygen and moistureto reachthe iron. Consequently,by measuring the progres$of the carbonation front after severalyears, it is possibleto estimatethe remaining lifetime of the concrete(Sentler 1984).

Fl$lsSrS$ E.Olid sisi€ ESrS

!r)EEa,!;Eid

i-+tsi$isiq sSeisG

elsdd$Gs s6i6iis 4 $t.$s

etsr$ @id$

!toirE 9.1q9 $

$o4'iq

ora{li;j.{

Si+.r$E q4sqs.l

s$tilji+sl +t4t!q!ql! $,itssi{ s€E6i4

qqet$ +i66d

9lss6s 4Slldebl

.Fid6s 6dss e.t6 EE

+E+n+$ isligsi$ M6E

Frc.7. Bursting of concretebalustraderail due to ru$ting of iron reinforcements' Washington, D.C.

ARCHITECTURAL

CONSERVATION AND APPLIED MINERALOGY

A third type of material interaction can be observedwhere a bronze statue is placed on a marble pedestal.The result is frequently a greenishblue staining of the marble due to soluble copper compounds being dissolved from the statue (Fig. ). As the runoff washesover the marble, the copper is depositedon the stone.Althoueb this effect hasbeen known and studiedfor many years(Kessler& Anderson 1953), very little has been published on the mineral species involved. In some cases, marble deterioration has beenobserved,but in other cases, it has not. Consequently,it is not possibleto say

3t7

whether the deposition of copper causesthe deterioration or is merely a symptom of an underlying effect associatedwith the flow of the water over the stone. Another type of staining problem oscurs with weathering steel.As noted above, its durability derives from the stability of the rust layer, which develops €ls copper is concentrated in the surface layer. During the initial period, while the layer is being built up, there is a significant amount of iron that washesoff and can stain concrete foundations (Fig. 8). Thesewash stainsare typical of weathering

Ftc.8. Staining of concreteby runoff from weatheringsteel.CollegePark, Maryland.

318

THE CANADIAN

$teelconstruction and, in somecases,may be objectionable on estheticgrounds (Evans 1968). BIOLOGICAL

A]TACK

Building materials are subjectedto attack from a variety of organismsranging from algaeand bacteria to birds and rodents. Generally, the causeof the attack can be identified by visual inspection or appropriate biological analytical methods. Mineralogy can be appliedin a few instances,however.For example,unknown surfacedepositscan be classified as to their origin, either as inorganic or organic, e.9., pigeondroppings,by the presenceof phosphatesor nitrates. If present in significant quantities, these anionsareunlikely to havebeenderivedfrom air pollution. Instead,an organic sourcewould be probable. In somecases,a mineral form of uric acid, urinite, can also be found @el Monte & Sabbioni 1980. Many forms of plant life attack stone surfacesby secretingoxalic acids in order to €xtract nutrients. In a particular caseon the island of Torcello near Venice, calcium oxalatecrystalswere found on limestone columns in the form of whewellite and weddellite. Thesemineralsare found in a variety of plant and animal tissues,but rarely in an inorganic setting. In the caseof Torcello, it was confirmed that the minerals are due to the activity of specific bacteria (Del Monte & Sabbioni 1983). TRSATMENTS

Given all the problems discussedin this article, it is not surprisingthat man has attemptedto find ways to mitigate damage to buildings. This most commonly takes the form of a protectivs ssaling that is intended to isolate the material from the aggressive environment. This approach has worked reasonablyfor metal surfacesand for wood, but has not been very successfulfor stone (Price 1982): This has not been for lack of trying. Ever since stone deteriorationwas observedon the Housesof Parliamentin the 1850s,treatmentshavebeenproposed, sometimeswith a very questionablescientific basis. The proposals have ranged from esthetically impossible approachessuch as applylnC tar, to the mysterious method put forward by Szerelnay, who claimed the ancient Egyptians used it on the Pyramids (Lewin 1966).In most cases,the proposers werewell meaning,but ignorant about the long-term effects, which sometimescould be more detrimental than the problem they were meant to cure. Generally, the treatments involved applymg organic compounds to the stone, either to serye as a water repellent or as a binder to consolidatecrumbling stone.However, there havebeensomemethods that imitated petrological processes.The oldest is lime washing,which hasbeenusedon limestonesince

MINERALOGIST

the Middle Ages. This consistssimply of washingthe stone periodically with water that is saturated with calcium carbonate(Ashurst 1984).As the solution dries, some calcite is left behind in the pores and cracks. However, the depositedcalcite can be readily washed out again by rainwater. Consequently, rather than being a permanent treatment, the lime wash servesin effect as a sacrificial coating @rice & Ross 1984). As long as more calcite is being depositedon the stoneby this method than is being rwashedoff by rainwater, the original stone should / be to a large extent protected. This is really only I feasibleon unpolishedsurfaceswherechangesin the finish due to the limewash layer are not noticeable. Church, in 1862,was one of the first to try the barium hydroxideapproach(Lewin 1966).This was basedon the conceptthat if the weatheringof calcite is due to its dissolution, perhaps it could be slowedby replacingthe surfacewith a more insoluble carbonatesuchaswitherite. This would be done by applying a barium hydroxide solution to the stone, which would react with the calcium carbonate to form insoluble barium carbonate.This approachhas been demonstrated in the laboratory; witherite has been found to define a solid solution with calcite, binding the grains together (Lewin & Baer 1974). However, it has not worked very well in the field. The chiefdrawbackliesin the rapid rate ofthe reaction, which occurs very near the surface. Consequently, the barium hydroxide solution does not penetratevery far into the stone. The result is a hard surfaceJayeroverlying a soft'and friable interior, possessing'Sifferent coefficientsof thermal expansion. Over d certain number of thermal cycles,the harder surfacelayer will tend to spall off (Hosek & Panek 1985). have attemptedto A largenumber of researchers get around this difficulty by modifying the solution to slow down the reactionand thus increasethe depth of penetration. The best known is the barium hydroxide - urea method (Lewin & Baer 1974).Even with this refinementthere are problems if sulfates are present,which is ustally the casewherestoneis deterioratingowing to air pollution. The resulting reaction-productwill tend to be barium sulfate rather than barium carbonate,which doesnot act to bind the grains together @adfield 1984). An applied mineralogy approach has also been attemptedfor silicaterocks, particularly sandstones. This usesvarious compounds built around the SiOo tetrahedron,linked by any of a number of organic groups to form polymers. After injection into the stone,the organicgroupsvolatilizs, leavingbehind the SiOatetrahedra,which then bind to the available silicate surfacesto consolidatethe stone. This approachalsodatesback to the 1860s(Lewin 1966). One of the simplestformulations usesethyl groups to link the tetrahedra, giving the formula:

ARCHITECTURAL

CONSBRVATION AND APPLIED MINERALOGY

Si(OC2H5)4.Upon hydrolysis, this yields hydrous silica SiOrr2IltO as the cement and ethyl alcohol, C2H'OH, which evaporates(Amoroso & Fassina 1983). However, as with the barium hydroxide approach for carbonate stone, there is a trade-off between penetration and volatility. A compound that volatilizeseasily,and tJrusdoesnot leavebehind an organic residue,is also one that tends to react rapidly, limiting its penetration.A large number of organic formulations have been tried. The most promising appearto be the alkylosilanes.However,the cost of the material as well as the special precautions required because of the potential hazards to the health of workers limit their use to the more valuable sandstonecarvingsand sculptures.The possibility of applying them to large wall-areas thus seems remote (Lewin & Wheeler 1985). In view of the ingeniousapplication of mineralogy in this method, it is discouraging to learn that some manufacturers of thesecompoundsare now recommendingtheir use on limestones,which cannot be justified from mineralogical principles. Finally, mineralogical analysishas also beenused to identify past treatmentson particular monuments. For instance,Zehnder& Arnold (1984)discovered that efflorescencesof soluble salt that causedamage to the sandstoneof the Erlacherhof in Berne,Switzerland, are composedof calcium and magnesium formates. These compounds were presumably createdin the processof cleaning the building with a formic acid solution. In another case,peculiar oxalateand silicatelayers were discoveredon the surfaceof the marble of Trajan's Column in Rome (Guidobaldi et al. 1982). Thesewereultimately recognizedasthe residuefrom restorationwork around theturn ofthe century. The oxalates are apparently due to oxalic acid used in cleaningand polishing the marble. The silicatesmay be tracesof water glassremaining from a protective coating applied after the stone was cleaned. Thesefindings are a reminderof the fact tlat many of the most prominent monumentsmay have undergone sometype of treatment in the past. The percentageof monumentsttrat havebeentreatedin some way remains unknown becauseno centralized databank has been established. In any event, thrs type of activity was often regardedlargely asmaintenance work, and detailed records were not kept. What makesthe situation worse is that in many cases,the treatments were experimental. That is, the inventor of the treatment had not tested it extensivelybefore applying it to a major monument. The resultingtrialand-error process of treatment testing has unfortunately revealedmany failures, to the detriment of the monumentsusedas test beds.Thus someof the deteribration now being attributed to other causes may in fact be due to someundocumentedtreatment

319

appliedsometime in'the last 150years.As a result, the U.S. National Academyof SciencesCommittee on the Conservationof Historic StoneBuildings and Monuments has recommendedagainst the application of experimentaltreatmentsto registeredhistoric landmarks. It has also recommendedthe creation of a censusof treated monuments(Baer 1982). In situations where it is not possible to conserve the original material, particularly whereinherentvice is involved, it may be neces$aryto substitute some other material. An example of the drastic measures that may be taken is Cologne Cathedral, where an entire section of the original sandstonewas replaced by basalt. Another saample is the Renwick Gallery in WashinglonD.C., whereartificial stoneis being used to replacedeterioratedsandstone(Stevens& Lewin 1983). CoNctustoN Mineralogical methods are applied extensivelyin the field of architectural conservation to diagnose causesof damageto buildings and other structures. Optical and electron microscopy are used as well as X-ray diffraction. Someof the ways in which mineralogy is applied include characterizationofthe original material for its suitability for a particular building. Alteration layers can also be studied to determinewhether they protect the brrilding or cause damage.Agents of deterioration such as air pollution or soluble salts can be identified. The efficacy of protective treatments can also be investigated. There are certain aspectsof the field that distinguish it from the more conventional sectorsof geology. One is that obtaining samplescan be extremely difficult, sincemost curators are extremelyreluctant to permit eventhe most minute disruption of the surface of suchmonuments. Another is that when samples are obtained, the phasesof most interest are often thin surface-layersthat may be amorphous or cryptocrystalline, making mineralogical analysis difficult. Also, the practitioner in this field, in addition to knowing mineralogy, must also have some familiarity with the building trades, architecruralhistory and art conservation. Finally, as can be seen from the list of references,researchin this field is generallynot published in mainstream geological or mineralogicaljournals. Most of the work has been publishedin journals for the architect or the art conservator, or in the proceedingsof specializedconferences on stone. Recognition of this field as a branch of applied mineralogy would havethe benefit of bringing it to the attention of a wider audience of geologistsand mineralogists. AclnlowLrncnN4ENTS The author is indebted to a great many workers

320.

TITE CANADIAN

in this field for their adviceand instruction. As much as possible, their contributions have been indicated in the list of references. However, special acknowledgementmust go to Dr. Norbert Baer, Institute of Fine Afis, New York University, for his advice and help over the years. Also, spesialthanks to Dr. Luke Chang, Chairman of the Department of Geology of the University of Maryland, for his encouragementof research in this area of applied mineralogy. Finally, ThomasTaylor, Chief Architectural Conservator, Colonial Williamsburg, has been extremely helpful in presenting the insights of the practicing architectural conservator.

MINERALOGIST

Bunrn, J. (1982): Carbon Dioide Equilibria and Reading,MasTheirApplicatiora. Addison-Wesley, sachusetts. Carvnrrro,D., DruMolqrB,M. & Sanmom,C. (1983): of the sulfatedcrusts Origin andgowth mechanisms on urbanlimestone.Water,Air & SoilPollutiont9, 351-359. CANER, G. & Srerv, N. (1981):Decayzonesof limestone& dolomite. In The Conservationof Stone. ed.). Centroper la ConserII (R. Rossi-Manaresi, vrzione delleScultureall'Aperto, Bologna,Italy. M. (1976):The CnnrsrorT enssN,J. & CspusroFFERsEN, kineticsof dissolutionof calciumdihydratein water. J. Cryst. Growth 35, 79-88.

REFERENCES

AIA [American Institute of Architecls] (1977): Life Cycle Cost Analysis: A Guide for Architects. Washington,D.C. Auonoso,G. & FassrNe, V. (1983):StoneDecoyand Consemation. Elsevier,Amsterdam. Arcrtrors, S. (1970: Replacanentof steelconnections by titanium alloy. In Proc. SecondInt. Symp.on Deteriorationof Building Stones(N. Beloyannis, ed.). Ministry of Culture and Science,Athens, Greece.

CRC lChemicalRubberCompany](1972):Handbook of ChemistryandPhysics(S6thedition).Cleveland, Ohio. Dsr.MoNrs,M. & SansroNr, C. (1983):Weddelliteon Iimestonein the Veniceenvironment.Environmental Sci. Tech, 17,518-522. (1986): Chemicaland bioloeical & weatheringof an historical building: ReggioEmilia cathedral.TheScienceof the Total Environment(tn press).

Ar.r.ror-o, A. (1982):Rising damp and salinematerials. InProc. FourthInt. Congress on Deteriorationand Preservationof StoneObjects(L. Gauri& J. Gwinn, eds.).Univ. Louisville,Louisville,Kentucky.

Evars, U. (1968\: The Corrosion and Ortdafion of Metols: First Supplementary Volume.' Edward Arnold Ltd., London.

(1984): Determinationof saline materials. Studiesin Conservation29, 129-138.

W. (1959):Studiesof the influenceof FerrKNEcHr, chemistryon the corrosionof metals. Chemistry& Industry,Sept.5, l102-1108.

& Kuruc, A. (1985):Crystallizationand habits of salt efflorescences on walls. In Proc. Vth Int. S. (1985):An invesGarn,F., Wstss,N. & HoneNsrEIN, Congresson Deteriorationand Conservationof tigationof the deteriorationof BethelWhite Granite. Stone (V. Furlan, ed.). PressesPolytechniques InProc. Third North AmericanMasonryConf. (J. Romandes,Lausanne,Suisse. Matheny& J. Borchelt,eds.).Univ. Texas,Arlington, Texas. Asnursr, J. (1984): The cleaning aud treatment of 'lime limestones by the method'.I. Monumentum, GoNzarez,J., Ar,arr.lo, C. & Arpnenr, C. (1983):Corx4452. carrosionrate ofreinforcementsduringaccelerated bonation of mortaxsmade with different types of Baen,N. (1982):Reportof the Committeeon Conserin Concrete cemenl./z Corrosionof Reinforcements vation of Historic StoneBuildingsand Monuments. Construction(A. Crane,ed.).Ellis HarwoodLtd., /z Conservationon Historic StoneBuildings and Chichester,U.K. Monuments.NationalAcademyPress,Washington, D.C. GnrsarB,D.A. (1982): Weatheringof the Kansas Capitol building: a studyin limestonedeterioration. Benron, N. & BlacrsHAw, S. (1976):A statistical Technologyand Conservation7(l), 2G31. evaluationof the analysescarried out on Eglptian limestone.Lithoclastia 2, ll-16. Gurooualpt,F.. Msucct,C. & TasAssoLeunntzt,M. (1982):Marblemonumentsof the RomanImperial Bnnnv,W. (1983):Durability of markermaterialsfor Age: pastsurfacetreatments.In Ptoc, Fourth Int. nuclearwasteisolationsites.OfJiceof NuclearWaste Congressotr the Deteriorationand Conservationof Isolation OIYWI-474,National TechnicalInformaStone(L. Gauri& J. Gwinn,eds.).Univ. Louisville, tion Service,Springfield,Va. Louisville, Kentucky.

-

ARCHITECTURAL

CONSERVATION AND APPLIED MINERALOGY

321

T. (1983): Lrwrs, W. (1965\: Solubilities: Inorganic.and-Metal' Heeorxnup,S., Kucnne,V. & HaecsNRUD' ChemicalSociety' Organic'Compounds.Ameri.car, Atmosphericcorrosionof unalloyedsteeland zinc: Washington,D.C. 4 yearsexposureat testsitesin Scandinavia.InPtoc. Ninth ScandinavianCorrosion Congress(CopenMsLvrLLn,I. & Gonpor' I. (1979):The Repair,and hagen). Maintenanceof Houses.The EstatesGazette,Ltd'' London. HanrLNe, B. (1980): Irrigation threatens Egmtian temples.Science204,796, G. (1908):Stonesfor Building and DecoraMnRRTLL, WileY& Sons,New York. (1985): Iohn of impregnation tion. Depth PeNer, J. J. & Hospr, asthe criterionfor durability of consolidatedstones. InProc, Fifth Int. Congresson Deteriorationand MonceN,M., translator (1914): Vitruvius: The Ten PolyPreservationof Stone(V. Furlaq'€d.). Presses Books on Architecture. Dover Press,New York. techniquesRomandes,Lausanne,Suisse. (1984):Keepingthe torch lit. Materials Huosc,P. (1977):Rock weatheringon the molecular Nsu.seN,N. April, 78-83. Periormance, level. Geol. Soc,Amer., Engng. Geol. CqseHist. 17.47-51. Nsvrr-Ls,A. (1981):Propertiesof Concrete'Pitman PublishingCo., New York. INouve,K. (1968):The role of copper(Il)in the oxidation of ferroushydroxidecolloid with specialreferenceto the corrosion of iron in an SO2containing -Nrrs,R. (1985):Peuoeraphicalcharacterizationof calenvironment.J. Colloid Interiace Sci.27' 17l'179. .at.oitt u'iitding itones in northern Belgium' ft Proc. Int. Congresson Deteriorationand ConserPolytechJwreN,A. (1883):The decayof building stonesof New vation of StonJ(V. Furlan, ed.).Presses York City. Trans.N.Y. Acad. Sci,,67'68. niquesRomandes,Lausanne,Suisse. C. &IlBrrrnssecn, R. (1982):The Kensn,J., Bnowrrl, Peprrcr-p,T. (1984):Barium hydroxide treatment of electrochemicalreductionof rust filns on weatherthe facadeof the old PatentOffice bwldtng. Smithing steelsurfaces.J, Electrochem.Soc.,288G2889. sonian Inst. ConservationAnalytical Lab.' Open' File Rep. 146. R. (1953):Influenceof the Krssr.en,D. & ANpensoN, wash from bronze on the weatheringof marble. F.M. (1970: Influencede la dolomiedans Notionol Bureau of Standards,Buildine Materials Psu-nRIN, d'altdrationde tunnel ferroviaire.InProc. un cas 137. Report and Structures SecondInt. Symposiumon the Deterioration of Building Stones(N. Beloyannii,ed.). Ministry of problemsin lndia. Lar, B. (1984):Stoneconservation Culture and Science,Athens, G5eece. Proc. Indo-U.S. Workshopon StoneConsemation (Lucknow,India; in press). PounsAlx, M. (1973): Lectures on Electrochemical Corrosion. Plenum Press,New York. (1985):weatheriag and disintegrationof stone monuments.InProc. Fifth Int. Congresson Deterioration and Conservationof Stone(V. Furlan, ed.). Pnrce,C. (1982):The evaluationof stonepreservatives. .Iz Conservationof Historic StoneBuildings Presses PolytechniquesRomandes, Lausanne, and Monuments. National Academy Press, Suisse. Washington,D.C. I-nwrll,S. (1960: The preservationof natural stone, & Ross,K. (1984): The cleaning and treatment 1839-1965- an annotatedbibliography'Art and 'lime method'. II. A technical of limestone by the ArchaeologyTech.Abstr. 6, 185-n7. & Bann,N. (1974):Rationaleof the barium

hydroxide - urea treatment of decayedstone. Sladies in Consemation 19' 2[35.

- & Cnanor.a,E. (1979):Stonedecaydueto foreign inclusions..In The Conservationof Stone. II @. Rossi-Manaresi,ed.). Centroper la Conservazione delle Sculture all'Aperto, Bologna, Italy.

evaluation

of

stone conservation techniques

employed at Wells Cathedral. Monumentum' 301-307. L. & Hrpstnc, B. (1985): J., BsRNrssoN' PueunntcBn, Hydratesaltsanddegradationof materials..InProc. Fihh tnt, Congresson Deterioration and ConserPolytechvatior of Stone(V' Furlan, ed.).Presses niquesRomandes,Lausanne,Suisse.

G. (1985):Alkorysilane chemistry RosenffoN,W.D. (1982):Evaluationof the durability - & WnBeLEn, of limestoneof historic buildings. -IlzProc. Fourth and stoneconservation.InProc, Int. Congresson Int. Congressoir the Durability and Preservatio-nof DeteriorationandConservationof stone(V. Furlan, StoneObjects(L. Gauri & J. Gwinn, eds.).Univ. ed.). PressesPolytechniquesRomandes,Lausanne, Louisville, Louisville, KentuckY. Suisse.

322

THE CANADIAN MINERALOGIST

Rossr-MaNannse,R. (1982): Scientific investigation in relation to the conservation of stone. Preprints Int, Inst. Conservation Congress, 3945. Sevrlsn, L. (1984): Stochastic characterization of carbonation of concrete. In Proc. Third Int. Conference on the Durability of Building Materials and Concrete (Espoo, Finland) 3, 569-580. SrouLxroBs, T. (1976)t Attaque atmospherique (goujons en acier) et mesuresd prendre..Iz Proc. Second Int. Symposium on the Deterioration of Building Stones (N. Beloyannis, ed.). Ministry of Culture and Science, Athens, Greece. Srrpnex, R. & SrnpneN, J. (1963): Solubility of Organic and Inorganic Compounds. Permagon Press, London SrsvENs,H. & LBwrN,S. (1983): Selectionof materials for the Renwick Gallery. ASTM Spec. Tech. Publ. on Rehabilitation and Protection of Buildings (in press).

Uqrrc, H. (1964):Corrosion and Corrosion Control. John Wiley & Sons,New York. Werrs, J. (1980): Deterioration and methods of preservingmetals.ft Metals in America's Historic Buildings.National Park Service,WashingtonD.C. Walreq R. (1980):Corrosion and preservationof bronzeartifacts.J. Chem.Educ. 57, n7-280. Wrr.n