Atmospheric Corrosion and Conservation of Copper and Bronze

CHAPTER 4
Atmospheric corrosion and conservation of

copper and bronze
D. Knotkova & K. Kreislova

SVUOM Ltd., Prague, Czech Republic.
1 Introduction
Copper and its alloys are important technical metals both for use as construction
materials and for special use, e.g. in electro-technical industry and crafts.
This chapter deals with the atmospheric corrosion of copper and bronze used
for architectural purposes, constructions and objects of art, especially sculptures
[1]. Such uses rely on a number of properties of copper and copper alloys such as
high resistance to atmospheric corrosion or ability to form characteristic layers of
corrosion products (patina) which not only protect objects, but also usually make
objects look better. High resistance of copper and bronze corrosion product layers
to corrosion allows for greater stability of objects in the atmospheric environment.
Objects and elements made of these materials survive for centuries.
Copper and bronze objects or elements of construction (sculptures, roofs, clad-
ding, artifacts of stone-made sculptural and architectural historical monuments) are
affected by atmospheric environment. Its effect, especially with long-term exposure,
causes gradual changes. The surface of objects is covered with corrosion products,
dirt, biological contamination, etc.
The negative effect is seen mainly in outdoor atmospheric environments where
along with natural climatic factors (changes of temperature, relative humidity, pre-
cipitation, snow, etc.), gaseous and solid pollutants released by industrial activities
are also present. Aggressive components of pollution (gaseous, acid rain) are trans-
ported to long distances, which affect materials in a wider area.
Indoor atmospheres (museums, galleries, depositories, etc.) are less aggres-
sive, although in closed spaces in depositories or storage unfavorable corrosion
WIT Transactions on State of the Art in Science and Engineering, Vol 28, © 2007 WIT Press
www.witpress.com, ISSN 1755-8336 (on-line)
doi:10.2495/978-1-84564-032-3/04
108 Environmental Deterioration of Materials
conditions may be formed which can cause damage to the stored copper and bronze
objects. This chapter does not specifically deal with this area.
Copper and its alloys also show specific corrosion manifestations in the form of
inter-crystalline corrosion and corrosion cracking. Corrosion needs to be eliminated
in bimetallic corrosion cells.
This chapter also focuses on studies of defects and disorders in copper and bronze
objects exposed to atmospheric environment. Selected technological procedures
of cleaning and conserving elements of construction objects namely, historical
monuments are presented too.
2 Atmospheric corrosion of copper and its alloys

Copper
Copper is a suitable material for many constructions and architectural elements
such as wall cladding, roof and copula roofing. Copper has good mechanical and
metalwork properties and is therefore used for construction accessories such as
gutters, window ledges, edging, troughs and chimney facing.
Properties of copper are significantly influenced by production technology. For
centuries copper, like other metals, has been produced on a small scale by artisans.
Metallurgically refined copper contains usually around 99.7% Cu, and silver or
other precious metals are not taken out of it. It can be used directly for rolling.
Electrolytic refinement produces purer copper containing 99.92%–99.97% Cu,
0.004%–0.02% O2 and only 0.01%–0.015% of all metal admixtures, the rest being
solved gases.
Sheets of rolled copper are used mainly for architectural and construction ele-
ments in atmospheric environment. Atmospheric exposure of copper sheets causes
change of mechanical properties (strength, ductility) as corrosion spreads over the
whole surface exposed to the corrosive environment. After 20 years of such expo-
sure, tensile strength decreases by less than 5% and the change in length is 10% of
the original values. Fatigue limit depends on the refinement of crystallization and
can go down by one-third or more if crystallization gets significantly coarse.
Copper alloys
Historical monuments and objects of art are made mainly of bronze; brass is used
only rarely. Bronze of various compositions is used. The differences among portions
of individual alloy elements are big; these also depend on historical trends.
Traditional types of bronze used for making sculptures contain tin, lead and
zinc, the share of each of these being approximately 5%. Modern types of bronze
are represented by a wider variety of alloys containing approximately 90% copper
and other metals such as tin, zinc, lead, aluminum, silver and others. Zinc and tin
increase the breaking strength and hardness of an alloy; it improves castability and
is cheaper than tin. Lead increases resistance to corrosion and improves the ease of
processing and castability. Bronze used for sculptures also contains antimony and
a small amount of iron and nickel (tenths of %). When an alloy is cooled slowly, it
passes through different phases, each being different in crystalline composition and
Atmospheric Corrosion and Conservation of Copper and Bronze 109
content of individual elements. The resistance of the types of bronze used nowadays
to corrosion does not significantly vary from the resistance of copper, evaluated in
terms of corrosion loss. Surface manifestations of corrosion attack are more uneven:
dips and corrosion on the grain edges are present.
Currently, bronze with a high level of resistance to corrosion is being developed
to be used for making copies of valuable sculptures that should be stored in depos-
itories. The increased resistance of newly developed alloys shall be achieved by
additions of tin, silicon, or nickel.
A less-used group of modern bronze alloys is formed by alloys containing tin
(10–14%). These materials are close to the historical types of bronze. Their resis-
tance to corrosion is very good, and they are easy to process containing up to 10%
of tin.
2.1 Corrosivity of atmospheres and guiding corrosion values for copper and
copper alloys
Copper and bronze objects are exposed to atmospheric environment over a long
term. It is important to know the conditions of the environment to estimate and
evaluate the damage being caused and to select proper protective measures, that is,
the means of conservation.
Corrosivity of atmospheres has been rising since the end of the 19th century
together with atmospheric pollution, reaching maximum levels in the 1950s and
1960s. There has been a significant decrease during the last ten years as a con-
sequence of various measures against atmospheric pollution and the change of
industrial structure.
Standard ISO 9224 – Corrosion of metals and alloys – Corrosivity of
atmospheres – Guiding values for the corrosivity categories provide guiding corro-
sion rates for copper for corrosivity categories according to ISO 9223 – Corrosion
of metals and alloys – Corrosivity of atmospheres – Classification (see Table 1).
Corrosion of copper and bronze within Europe was evaluated as a part of the
UN ECE International Co-operative Programme on Effects on Materials including
Historic and Cultural Monuments for the period of eight years (1987–1995) [2].
Table 1: Guiding corrosion rates for copper.
Corrosivity Corrosion rate (µm/a)

category
First 10 years Following years
C1 ≤0.01 ≤0.01
C2 0.01 < 0.1 0.01 < 0.1
C3 0.1 < 15 0.1 < 1.0
C4 1.5 < 3.0 1.0 < 3.0
C5 3.0 < 5.0 3.0 < 5.0
Table 2: Average mass loss (g/m2 ) of copper and bronze on

20 European test sites [2].
Material Exposure period Time of exposure (years)
1 2 4
Copper 1987–95 10.1 16.4 27.8

1997–2001 7.6 12.1 19.7
Bronze 1987–95 7.3 12.8 23.6
1997–2001 4.2 7 12.8
Yearly corrosion loss of copper was between 0.2 µm/year (test site Toledo, Spain)
and 1.6 µm/year (test site Kopisty, Czech Republic). Long-term exposure at the
atmospheric station of SVUOM (1981–96) in the environment of C4 corrosivity
category resulted in a corrosion rate of 2.6–2.8 µm/year. Comparison of corrosion
rates of copper and bronze exposed repeatedly in Europe in the first and second
stages of the UN ECE International Co-operative Programme shows how the corro-
sivity of the observed territory has been changing during this time period (Table 2).
Decay of copper and bronze objects is highly influenced by environmental condi-
tions, mainly by acidifying airborne pollutants. In the past ten years environmental
conditions have changed in many European countries, the effect being a significant
decrease of sulfur dioxide (−75%) and nitrogen dioxide (−40%) concentrations,
whereas the amount of ozone remains nearly the same. Sulfur dioxide is still the
main factor for corrosion processes but the presence and combinations of other
pollutants are becoming more important. Repeated long-term exposure of copper
(99% Cu) and bronze (81–85% Cu, 6–8% Sn, 3–5% Zn, 5–7% Pb) samples for 1, 2
and 4 years were carried out to prove and compare the corrosion effect under actual
environmental conditions [2].
Compared to the exposure carried out between 1987 and 1991, reduced sul-
fur dioxide content in the environment seems to influence bronze surfaces more
than copper surfaces. The mass loss of bronze reaches values lower by 45%, but
copper’s metal loss lower only by 24%. The most significant decrease of values can
be seen in the weight change of bronze, which reaches values lower by 77% than
ten years ago.
The results of the UN ECE International Co-operative Programme were used
for the derivation of damage equations for copper and bronze, which help to cal-
culate corrosion loss from environmental data. Besides the usual components of
atmospheric pollution, the concentration of tropospheric ozone is considered for
the calculation of the corrosion loss of copper and bronze [3].
Long-term corrosion tests of copper and its alloys and the following derivation
of guiding corrosion rates are based mainly on corrosion tests using metal sheet
samples. Casts made of copper alloys have different properties, which result in
differences in the effect corrosion has on them.
Information about corrosion of copper in specific conditions was published in a

study [4]. Copper is influenced even by a clear atmosphere at room temperature. A
thin layer of invisible oxide is formed; no darkening appears. This layer provides
some protection against low levels of pollution in indoor atmospheres.
If copper is exposed to air at high temperatures, a thin coating of interference col-
ors is formed. If traces of sulfides or hydrogen sulfide are present in the atmosphere,
copper darkens. Mixed layers of oxides and sulfides are very thin and adherent.
The initial corrosion effect in outdoor atmospheres is staining. It is typical
for roofing and cladding and is not considered aesthetically valuable. Sometimes
sharply demarcated stains and stripes are formed, especially at places where water
is held up and dries out. Stains are dark, dark red or turning blue. This effect is
accelerated even by the presence of mild hygroscopic salt deposits.
Stress corrosion cracking (SCC) is typical especially for brass, but is less for
other copper alloys. The effect appears under tension stress above a certain limit
in a polluted (industrial) atmospheric environment, especially when it contains
ammonium compounds. SCC affects copper alloys containing 20% or more zinc, but
only rarely the other alloys. Among the objects presented in this publication, brass
screws connecting construction elements as window frames, metalwork details, etc.
are affected by SCC. In sheltered exposures SCC can be accelerated by cumulating
dust and other dirt.
2.2 Formation mechanism of surface layers
Atmospheric corrosion of copper and bronze results from the presence of air humid-
ity and oxygen, while aggressive gaseous and solid components of atmospheric
pollution are effective too [5–7]. The process, mainly in its initial stage, is of
electro-chemical nature. Since layers of surface electrolytes are thin and dry out
periodically, the reaction products are deposited in the form of solid surface layers
which then take part in corrosion reactions, and the nature of the corrosion process
becomes chemical too. Stability or solubility of surface layers components, espe-
cially the ability to bond aggressive atmospheric components into salts of limited
solubility, have a fundamental influence on keeping corrosion rates of copper and
its alloys low in atmospheric conditions. High resistance of copper and its alloys to
corrosion in atmospheric conditions is then derived from the protective function of
corrosion product layer usually called patina. Surface layers on bronze and copper
are formed by periodically repeated processes during wetting and drop-out of basic
salts from saturated electrolytes at a suitable but not too low pH.
Formation of patina is a long-term process going through a number of stages.
Different mechanisms determine the speed of the process in individual stages. The
corrosion product layer is subject to transformation in the formation period and
eventually it gets into balance with the external environment affecting it. The natu-
ral, stable, green surface layer is formed in 10 to 20 years in atmospheric conditions,
depending on the corrosivity of the atmosphere. At low levels of pollution in unsuit-
able climatic conditions (dry, low temperatures) the green surface layer is formed
in a long time or could not be formed at all. Surfaces remain dark. Different local
Table 3: Stability of basic patina components in pH levels [7].
Compound Formula pH level
Cuprite Cu2 O >4

Brochantite Cu4 (SO4 )(OH)6 3.5–6.5
Antlerite Cu3 SO4 (OH)4 2.8–3.5
Atacamite Cu2 Cl(OH)3 3.8–4.3
Gerhardite Cu2 (NO3 )(OH)3 4.0–4.5
Malachite Cu2 (CO2 )(OH)2 >3.3
degrees of wetness and other microclimatic effects cause uneven coloring of objects,
e.g. roofing and sculptures. Typical properties of surface layers can be estimated
after 3–5 years of exposure.
Primarily, a layer of cuprite is formed on newly exposed surfaces of copper and
bronze after wetting. Surface electrolytes in the next corrosion stage contain sulfate
ions corresponding with their dry and wet depositions from the atmosphere and ions
of copper diffusing through the cuprite layer. Copper sulfates turn into basic sulfates,
which are deposited on the cuprite surface in the form of solid layers.
Formation and composition of patina are significantly influenced by acidity of
rain. Rain of high acidity dissolves the surface layer and washes it off. Rain of usual
acidity influences the composition of patina in its pH, because the thermo-dynamical
stability of minerals differs, e.g. antlerite is more stable in acid environments than
brochantite (see Table 3).
Basic patina components are formed in the presence of surface water layer by the
nucleation and deposition of solid phases in a row of reactions. Brochantite is said
to need at least two weeks to crystallize. During this period insufficiently bonded
copper and sulfate ions can be washed out by rain. Wash-out is effective especially
for thin initial layers.
Relations between time required for the deposition of more stable components of
patina and the occurrence of rain may influence the formation of layers of natural
patina.
2.3 Influence of selected factors on quality of forming layers
Patina on copper or bronze object differs locally in thickness and composition in

visually distinctive areas. The reason for this variety is the shape arrangement of
objects, which creates differences in the access of individual environmental factors
to the material in various parts either as a consequence of un/shelteredness from
the factors or various speeds at which environmental components are transported
to the surface. The surface layer formed manifests as various shades of green, can
be blue and green, gray and green, brown or black. Results of X-ray diffraction
analysis show that color (especially dark) does not provide sufficient information
on composition of a layer. Color borders can be sharp or diffusive.
Occurrence of green stripes formed by streams of water flowing away on dark

background is typical for sculpture surfaces. Dark, coarse and thick crusts on partial
surfaces, mostly on soffits, usually differ in appearance too. These crusts contain
many foreign matters – impurities and crystallized soluble fractions of corrosion
products of copper. The protective function of such layers is limited.
Corrosion attack on bells 100–600 years old that was exposed in various atmo-
spheric environments in the Czech Republic (countryside, cities, highly polluted
industrial areas) was evaluated [1]. All bells were made of bronze containing a
relatively high portion of tin (approx. 20%). Considering the massiveness of the
bells, corrosion damage does not represent any threat to the function of the objects.
Surfaces are covered with a relatively coherent but not compact layer of corro-
sion products. Corrosion attack on sides facing wind is usually more significant.
No corrosion pits occur. Different manifestations of corrosion can be seen on the
inner surfaces of the bells, especially in the bells exposed for long periods. The
corrosion layer contains less impurities, is usually blue-and-green, and crusts are
present locally. Condensation and wet corrosion probably take place inside the bells.
Surface layers contain mainly antlerite, and a high portion of quartz, gypsum and
calcite (dust). Also present is moolooite (copper oxalate) which proves biological
pollution (birds). Brochantite was also found on a bell exposed since 1651.
Properties of layers, which then affect the corrosion rate of copper and its alloys
in the long term, are formed according to conditions of pollution and temperature-
humidity characteristics of an area, but also depend on the exposed surface.
The natural surface layer characteristic of current copper and bronze surface
exposed for tens or hundreds of years in an open atmosphere contains mainly
brochantite, antlerite and cuprite. Formation of antlerite or brochantite is subject
to free access of precipitation. In atmospheres with natural or technological salin-
ity, patina also contains atacamite and paratacamite – Cu2 Cl(OH)3 . Patinas that
have not reached a stabilized state contain posnjakite – CuSO4 · Cu(OH)6 · H2 O.
Gerhardite – Cu2 (NO3 )(OH)3 is found rarely. Patinas nowadays do not contain
malachite – Cu2 (CO3 )(OH)2 which was present in analyses from the past years.
Surface layers often contain admixtures of gypsum and siliceous sand, and pati-
nas also contain portions of formates, acetates and oxalates which are formed in
atmospheres polluted by anthropogenous and biological effects. These components
have been identified only in the last few years by the use of ion chromatography in
this field [8]. Salts of organic acids are reported to form a kind of binding agent in
patina layers. Other substances can be formed in environments with specific effects.
Soluble CuSO4 · 5H2 O (chalcantite) cumulates in soffit areas. An overview of the
occurrence of individual components and their properties is presented below.
Some papers refer to other types of components of patina in the initial stage of
corrosion too.
2.4 Identification of individual corrosion products
An overview of the phases determined repeatedly in surface corrosion layers on

copper and bronze objects in the Czech Republic is presented in Tables 4 and 5.
Table 4: Phases identified in patina in the atmospheric environment of the

Czech Republic.
Name Formula Line
Antlerite CuSO4 · 2Cu(OH)2 7-407

Brochantite CuSO4 · Cu(OH)2 13-398
Posnjakite CuSO4 · Cu(OH)6 · H2 O 20-364
Cuprite Cu2 O 5-667
Tenorite CuO
Atacamite Cu2 (OH)3 Cl 25-269
Paratacamite Cu2 (OH)3 Cl 25-1427
Botallackite Cu2 (OH)3 Cl 8-88
Moolooite C2 CuO4 · nH2 O 29-297
(NH4 )2 Cu(SO4 )2 · 6H2 O 11-660
Chalcantite CuSO4 · 5H2 O 11-646
Gerhardite Cu2 (OH)3 NO3 14-687
Melanothallite Cu2 OCl2 35-679
β-Cu2 S 26-1169
Cu2 S 23–959
Quartz SiO2 33-1161
Gypsum CaSO4 · 2H2 O 33-311
Calcite CaCO3 5-586
Individual components were determined by a phase analysis of samples taken from

objects in Prague and other places within the Czech Republic (approx. 300 samples)
[9–10]. Components shown in Tables 4 and 5 are typical for urban and industrial
atmospheres without any significant salinity effect.
2.5 Composition of surface layers and patina components characteristics
• Brochantite CuSO4 · Cu(OH)2 , antlerite CuSO4 · 2Cu(OH)2 – corrosion prod-

ucts of copper and bronze are mainly brochantite, antlerite or a mixture of both.
Brochantite occurs mainly at places of direct impact of rain, snow and dew, while
antlerite is formed mainly by condensed air humidity. Both products have pro-
tective properties and slow down further corrosion of metal surfaces. From the
aesthetic point of view, they give patina blue-and-green coloring, which, how-
ever, can be devaluated by minimum amounts of darker components as carbon
black, black oxides and sulfides.
• Cuprite Cu2 O, tenorite CuO – these represent an anchoring component on the
metal-patina border. They form thin adhesive layers, mostly black, brown red
or orange.
• Posnjakite CuSO4 · Cu(OH)6 · H2 O – it is a component of young sulfate patina
and turns into antlerite or brochantite later on.
Table 5: Composition of patina layers.
Composition Objects
Statues outside Prague Statues, copper roofs in the Statues in the

(investigation in 1994) center of Prague center of Prague
18 samples (investigation in 1994) (investigation in 1998)
57 samples 33 samples
Abundance Frequency Abundance Frequency Abundance Frequency

(%) (%) (%)
Antlerite 9 50 18 31 10 30
Brochantite 6 33 43 75 17 52
Cuprite 3 17 16 28 8 24
Atacamite, paratacamite 0 0 3 0 1 3
Moolooite 0 0 0 0 1 3
(NH4 )2 Cu(SO4 )2 · 6H2 O 0 0 4 7 4 12
Chalcantite 0 0 0 0 3 9
Quartz 12 67 33 58 18 55
Gypsum 4 22 2 4 13 40
Calcite 0 0 3 5 4 12
Entirely amorphous 2 11 1 2 0 0
Atmospheric Corrosion and Conservation of Copper and Bronze
115
• Atacamite, paratacamite Cu2 (OH)3 Cl – dominant component of patina in

maritime environments. Only exceptional occurrence was identified within
Czech Republic in cases of presence of chlorides (e.g. surfaces of fountains
where chlorides are contained in water or on surfaces affected by de-frosting
agents).
• Moolooite C2 CuO4 · nH2 O – its occurrence is attributed to the effect of bird
excrements on metal. It was identified at locations sheltered from a direct impact
of precipitation.
• Ammonium copper sulfate (NH4 )2 Cu(SO4 )2 · 6H2 O – water-soluble corrosion
product of copper which is contained in crusts. It was identified at locations
affected by anthropogenic influences.
• Chalcantite CuSO4 · 5H2 O – was identified as a part of soffit crusts in highly
polluted environments. This component is well soluble, has no protective effect
and its presence indicates increased corrosion of metals.
• Quartz SiO2 – most frequently occurring component of dust, chemically inert
component of surface layers. It increases volume of patina at higher concentra-
tions, decreases adhesion to surfaces and deteriorates the look of patina.
• Calcite CaCO3 , gypsum CaSO4 · 2H2 O – calcite is often a part of dust, having
a negative effect on it. It turns into gypsum by the effect of sulfate ions, causes
expansion of volume and degradation of patina. High amounts of gypsum were
identified in patinas from locations polluted with dust from neighboring lime
works.
Besides copper, bronzes contain other alloy elements (lead, zinc, tin, iron). Cor-
rosion products of these components occur in surface layers in small amounts that
have no significant effect on the protective properties of these layers. Corrosion
products of tin (SnO2 · nH2 O, SnO2 ) used in bronze alloys in the past increase
protective ability of surface layers.
Surface corrosion layers including the protective patina on copper and bronze
do not have a simple composition, not even an unambiguous crystalline structure.
Metallographic cuts document two- or three- layer construction of patina in most
cases, in some cases foreign matters can be seen in the top green layer. Results
[1, 11, 12] are related to the evaluation of layers on copper sheet; surface layers
on bronze can be evaluated less frequently [13].
The closest zone to the metal on the cut is orange, red or dark layer, mostly made
of copper oxides. The next layer is bright green and contains mainly brochantite,
but also antlerite. In this sub-layer, dust and other impurities accumulate. However,
siliceous particles can be in the bottom oxide layer too. Dark top layer, if present,
is formed mainly by carbon black. Varying and atypical composition occurs in case
of voluminous dark or gray and green crusts in soffit locations (high portion of dust
particles, antlerite, various soluble corrosion products of copper).
Individual layers, even those that are relatively chemically homogenous, are
formed by irregular crystalline formations. Besides impurities, cavities and other
defects are also present in layers that allows for the transportation of ions through
layers.
The thickness of green patina (green top sub-layer) formed in an open atmosphere
in long term is between the range of 40–50 and 75 µm. The thickness of the black,
red or dark sub-layer is approximately 10 µm. On the bottom side of the sheets, the
dark sub-layer is thicker (approx. 20 µm), while the green sub-layer is only a few
micrometers thick. The depth of pits (if they occur) on surfaces exposed in long
term ranges from 50 to 70 µm, but also from 100 to 105 µm in some cases.
The thickness of crusts is higher (approx. 200 µm). The boundary of metal surface
under these crusts is usually very uneven, broken, with corrosion penetrations of
various depths. Penetrations have shapes rather different from those of shallow pits.
To sum up, the surface layer of copper is a system that may vary in chemical
composition, composition in the layer cut, protective function and appearance.
3 Typical corrosion manifestations on copper and

bronze objects
Copper and bronze objects or construction parts as roofing, cladding, doors or
sculptures have been exposed to the atmospheric environment for centuries without
any serious corrosion problems in most cases, except for the aesthetic problems
caused by cumulating impurities and the formation of corrosion layers differing in
color locally. In contrast to copper roofing and cladding, where corrosion damage
may affect the service life of an object, damage in bronze sculptures is considered
aesthetically desirable both for relatively higher thickness of the exposed material
and for periodical maintenance and protection of some sculptures [1, 9, 14–16].
Standardization of defects and disorders and evaluation of causes of their occur-
rence are listed individually for sculptures (bronze and copper) and for roofing,
cladding and metal work after general characterization, the reason being not only
partial differences in causes but also differences in technical and aesthetic demands.
3.1 Sculptures and objects of art
Corrosion manifestations (described in more detail below), which may lead to

defects in some cases, can be generally divided into the following groups:
• Manifestations of general corrosion differentiated by space arrangement of an
object,
• uneven or local corrosion attack which may be resulting from space orientation
of a surface but also from construction design (e.g. joints),
• mechanical damage (cracks, changes of shape),
• defects resulting from the way sculptures were made (connection lines, uncov-
ered rivets),
• corrosion of carrying and supporting construction, even bimetallic,
• corrosion damage and formation of deposits resulting from the function of an
object (fountains), and
• negative aesthetic manifestations resulting from previous conservation and var-
ious surface treatments (e.g. pigment coatings).
Main types of corrosion manifestations:

1. Formation of black and green layers of corrosion products. The composition
of most samples of surface layers from various objects corresponds with the
typical composition of patina formed nowadays in urban and industrial areas
(brochantite, antlerite, cuprite, admixtures of quartz and calcite). The surface
layer is bound firmly to the surface. The surface layer is protective, i.e. it
decreases the corrosion rate of a metal.
2. Crusts that are formed mainly in sheltered locations are of usual composition
(admixtures of gypsum, quartz, etc.). Specific components as ammonium
copper sulfate (NH4 )2 Cu(SO4 )2 · 6H2 O and copper formate ((HCOO)2 Cu)
were found only in a small number of cases. Crusts on sculptures located in the
center of Prague often contain chalcantite (CuSO4 · 5H2 O) and moolooite, i.e.
copper oxalate (C2 CuO4 · nH2 O). These fractions do not have a protective
function.
3. Turquoise areas are formed around little pits on surfaces of bronze sculptures
where condensate from inside the sculpture reaches out.
4. Uneven corrosion along grains was observed in a limited number of cases
on horizontal and less inclined surfaces of statues exposed in highly polluted
environments.
5. An interesting finding was the occurrence of places with little pits in over-
lapping areas on sculptures made of hammered copper sheets (e.g. under
coat overhang on a statue of a horse rider). These pits were under a layer of
deposited dirt and corrosion products. It may be assumed that it is a case of
crevice corrosion.
6. Bimetallic corrosion of steel joints and fixation elements and their surround-
ing was found on parts of sculptures that are not fully treated (back side).
Corrosion of steel joints and voluminous corrosion products of iron may
damage and open the joints.
7. Extensive corrosion damage was found on the inner steel supportive con-
struction of sculptures made of hammered copper sheets. Bimetallic corro-
sion occurs at places where steel inner construction is in contact with copper
surface. Corrosion attack on inner supporting construction reaches the degree
of defective stability of an object in some cases, which even increases pene-
tration of environment into the inner space of a sculpture.
8. Various joints and elements used during the treatment of a sculpture become
visible ( joint lines, matching of parts, overlaps of inner joints).
9. Soldered joints on sculptures made of copper sheets are being damaged and
local corrosion manifestations occur around these joints. Dark dots and stains
occur on soldered areas. This is also a case of soldered decorative parts of
cladding and copulas.
10. Cracks in surfaces and deformations or fall-off of parts are rather rare.
11. Corrosion manifestations related to aging of sculpture coatings (conservation,
patination, pigmented and transparent coatings) are aesthetically undesirable.
Coatings change in color, wash away, grow through corrosion products or are
gradually separated from the base.
12. Water deposits in fountains.
13. Corrosion manifestations on bells are significantly influenced by the manner
of exposure (tower spaces) and shape (inner and outer surface). Pollution by
bird excrements is an important factor. Serious damage occurs after mechan-
ical wear-out on the inner surface and corrosion damage on steel parts of bell
hangers.
3.2 Copper roofs, cladding and metalwork elements
Copper sheets and metalwork elements were used in the past only for roofing and
other parts of important buildings (churches, palaces). During the last ten years
there has been a growth in the use of copper for a wider range of purposes (family
houses, business objects).
Both the types of use differ in the material used, construction design and techno-
logical treatment which is reflected by differences in the tendencies for formation
of defects and damages. Material and technological differences of both types of use
are also reflected in the mode of maintenance and repairs.
A great deal of attention has to be paid to the designing and making of copper
roofs or claddings of a modern object. Designing copper (and also titan-zinc) roof-
ing and cladding is a specialized technical work nowadays. Producers of materials
and elements elaborate extensive manuals setting construction designs and techno-
logical treatment of elements, their connections and making them into an object,
especially in relation to its temperature-wetness regime and material composition
in combination with copper parts.
There are four main causes of defects and damages of copper roofs and metal-
work elements – mechanical damage, corrosion impact of the outdoor environment,
corrosion impact derived from the construction design of a building and specific
corrosion impact derived from its indoor regime. The degree of stress caused by
individual effects may differ, and causes are often combined. The method of repair
is chosen according to the type of defect.
Mechanical damage
• parts loosen by wind,
• extensive heat stress, changes of temperature,
• obstacles to compensate movement during heat stress,
• operation stress (walking areas),
• accidental damage (falling objects, perforations).
Corrosion attack (environmental effect and design)

• impact of surrounding local atmospheric environment,
• local effect of smoke,
Table 6: Types of defects on bronze and copper objects in Prague.
Age of Type of object Defect

objects
(years) Roof, Sculpture Aesthetica Corrosionb Mechanicc
cladding (bronze,
(copper) copper)
450–240 1 5 3 3 2
150–110 – 4 2 1 1
100–80 3 28 25 21 4
75–60 3 17 13 13 6
50–15 1 39 24 24 6
<15 2 4 3 5 1
Total 10 97 70 66 20
a Nonuniform patina, differences in color.
b Pitting,crusts, etc.
c Holes, damaged joints and supporting constructions.
• corrosion effects caused by contact with unsuitable materials (bimetallic corro-

sion, soldering),
• corrosion effect of rinsing of surrounding materials (acidic fractions and others),
• corrosive and other negative effects of nesting birds, mainly pigeons,
• pitting on roofing overlaps and board welts (depth to 50 µm),
• formation of non-protective surface layers (ammonium copper sulfate con-
tained) around ventilation inlets where exhalations from residential and opera-
tion spaces of buildings leave,
• rusty areas on roofs surrounding steel elements (e.g. lightning-conductor
connections).
Damages caused by indoor effects

• condensation,
• indoor exhalations,
• damaged base or inner construction as a result of decay or insect attack.
Table 6 gives an overview of the different types of defects on copper and bronze
objects in Prague [1, 14].
4 Selected methods of cleaning and conserving copper and

bronze objects
This section lists methodical knowledge, conditions of use and properties of treated
surfaces for these technologies:
• physical (mechanical) and chemical cleaning methods,

• patination and properties of artificial patina including their transformation after
exposure in atmospheric environment,
• conservation.
The main benefit is quantification of changes caused by selected technologies of
cleaning on surfaces of objects [1].
4.1 Cleaning of copper and bronze objects, namely sculptures
Techniques of cleaning and removal of corrosion products [1] can be divided into
three groups:
• cleaning by water under pressure,
• mechanical or abrasive cleaning (blasting),
• chemical cleaning – ‘draw-off’ and pickling.
4.1.1 Cleaning by water under pressure

Removal of corrosion products and dirt by water under pressure is often used in
combination with abrasive techniques or chemical cleaning. Pressures between 50
and 1000 psi are used for water cleaning. Particles that fall off or are not fixed to the
surface and soluble parts of corrosion crusts and layers are removed or redistributed
on the surface by pressure washing. The effectiveness of washing depends mainly
on the conditions and configuration of a surface and the abilities and experience of
staff.
4.1.2 Mechanic (abrasive) cleaning

Various methods of mechanical cleaning such as different types of abrasive and pol-
ishing agents, pastes, metal wool and brushes made of various materials including
mechanical brushes, scalpel, special abrasive steel wool, etc. are used by conser-
vators. Cleaning using hand-operated mechanical equipment is difficult and time-
consuming. These methods can remove crusts, deposits and growth from a surface
and keep a thin layer of patina. Whether it is possible and desirable to maintain
a thin layer of patina depends on the degree of pollution of a sculpture, previous
surface treatment, shape of an object, etc. Usually not all parts of a sculpture can
be cleaned in this way, especially those with no access.
A big problem during the mechanical cleaning of sculptures is the removal of
a layer of previous improper ‘conservation’ by various coatings and waxes or the
removal of the remnants of partially deteriorated previous conservation.
Mechanical cleaning using abrasives – blasting – is used on historical objects
rather rarely, but there are cases when this method is the one to be used (e.g.
inner surface of fountains, bell surfaces). Blasting is a process during which solid
particles of metals, minerals, synthetic resins or plants, sometimes combined with
water, are thrown at high speed against objects in order to clean, reduce or harden
Table 7: Material loss of copper with a thin brown layer of corrosion products after
blasting by various abrasive agents (blasting normal to the surface).
Abrasive agent Pressure Nozzle Material loss

(MPa) diameter
(mm) Copper Brass
(g/m2 ) (µm) (g/m2 ) (µm)
Corundum 120 0.4 1 1.5 0.18 1.8 0.22

2 4.5 0.55 3.5 0.43
0.2 1 1.1 0.13 0.7 0.09
2 2.1 0.26 2.9 0.36
Black corundum 180 0.2 1 1.0 0.12 2.2 0.26
Balotine 60 0.2 2 0.5 0.06 0.1 0.02
their surfaces. Compressed air (pressure between 140 and 550 kPa) and abrasives of
different sizes, usually between 175 and 1000 µm, are used for blasting. Different
combinations of speed, distance from surface and angle of incidence are used.
The selection of these parameters is determined by the character of the surface
to be cleaned, especially by its coarseness, hardness and thickness of crusts and
corrosion layers. These methods have a great impact on the topography of a surface,
its deformation and the deformation of microstructure, etc., although very soft
abrasives are used, e.g. glass balls, sodium carbonate, aluminum oxide, olivine
sand, plastic with different hardness and size of particles, some natural products
such as ground walnut shells or crushed corn.
Blasting gives rise to unstable surfaces sensitive to humidity, which makes corro-
sion processes continue. These methods remove neither the active centers of chlo-
rides or the sulfates from deep surface pits. The blasting process can leave parts of
abrasives on a surface, which can negatively affect it after exposure to environment.
Aggressive abrasive techniques can uncover more active pits and centers on sur-
faces than less aggressive procedures. The impact of blasting on surfaces of copper
and its alloys was evaluated on samples of copper sheet at a different degree of
corrosion attack (Tables 7 and 8).
Surfaces are changed after blasting in a way resulting from the technology
used:
• Blasting agents significantly increase the coarseness of cleaned surfaces depend-
ing on the hardness of the agent and the shape of its particles.
• Bigger changes of macrostructure are caused by corundum, but the surface is
cleaned more efficiently; blasting by balotine leaves remnants of surface layers.
• Blasting contaminates surfaces with elements or impurities contained in the
blasting agent. Higher portion of trace amounts of impurities occurs when
balotine is used.
Table 8: Weight and thickness loss of copper with natural green patina exposed for
60 years after blasting by various abrasive agents (blasting normal to the
surface).
Abrasive agent Pressure Nozzle Material loss

(MPa) diameter
(mm) (g/m2 ) (µm)
Corundum 120 0.4 2 1.2 0.15

0.2 1 11.2 1.38
2 1.0 0.13
Black corundum 180 0.2 1 11.2 1.37
Balotine 60 0.4 2 1.2 0.15
0.2 2 1.0 0.13
Balotine 134 0.2 1 10.7 1.31
4.1.3 Chemical cleaning

Methods of chemical cleaning include mainly drawing-off and pickling using dif-
ferent chemical solutions. The difference between drawing-off and pickling is that
pickling removes all layers of corrosion products down to pure metal.
Drawing-off is a method that can be carried out in standardly equipped work-
shops or even in the field. It is not quite possible to achieve the defined and even
effect of agents on a sculpture surface. Corrosion products are removed only partly
and locally which however might be suitable for some purposes.
Chelation or complex-forming solutions or pastes are used for drawing off,
e.g. [17]:
• alkaline Rochelle salt (sodium potassium tartrate) removes corrosion products
without affecting the base metal;
• applications based on EDTA (ethylenediaminetetraacetic acid) selectively remove
unadhesive or active corrosion products from bronze surfaces, leaving them
enriched with copper oxides allowing for future repatination;
• application of ammonium hydroxide solution and simultaneous brushing with
wool removes unadhesive layers of corrosion products,
• application of a solution of water and ammonium hydroxide (1:1) and simulta-
neous brushing with a metal brush or pumice removes thick incrustations;
• application of a paste containing 1 volume part of powder soda, 5 volume parts
of calcium hydroxide and 2 volume parts of sawdust removes corrosion products
for strongly deteriorated and corroded surfaces.
Pickling is a process used in technical practice especially as a pre-treatment
before the application of various types of protective coatings. Cleaning is thor-
ough but surfaces are often more affected. Loss of base metal during chemical
cleaning processes equals to 2.5–9.0 years of exposure in an urban atmosphere.
For technical purposes, pickling is defined in ISO 8407 Corrosion of metals and
Table 9: Mass loss of copper with thin layer of oxides after pickling in various
solutions (temperature 20◦ C).
Pickling solution Concentration Time Total mass loss

(M) (min) of copper
(g/m2 ) (µm)
Citric acid 0.10 5×3 0.45 0.05

Chelatone 3 0.20 4 × 20 0.40 0.04
Sodium hexametaphosphate 0.50 4×3 0.28 0.03
Sulfuric acid 0.10 5×2 0.43 0.05
Sulfamidic acid 0.05 5×2 0.85 0.10
Phosphoric acid 5.00 4×2 0.73 0.08
Sodium hydroxidea 0.25 4 × 10 0.48 0.05
a Temperature of solution 70◦ C.
alloys – Removal of corrosion products from corrosion test specimens. The standard
includes recommended solutions for individual metals and alloys. According to this
standard, pickling periods differ for different pickling solutions, metals or degrees
of corrosion. When pickling is used on historical objects it is recommended to apply
pickling solution repeatedly, thoroughly wash the objects after each application and
remove loosened corrosion products and dirt, mechanically.
Comparison of pickling agents and procedures for copper and brass was carried
out both from the point of view of corrosion loss and morphology of surfaces
after pickling in solutions (citric acid, Chelatone 3, sodium hexametaphosphate,
sulfuric acid, sulfamidic acid, sodium hydroxide, phosphoric acid) [18]. Pickling
procedures were verified on copper and brass samples with thin layer of corrosion
products formed in indoor environment (Table 9), and copper samples with thick
layer of green corrosion products formed after exposure in condensation chamber
with SO2 and salt spray (layer of approx. 0.7 g/m2 – Table 10) [1]. It is important
that during pickling undesirable layers of corrosion products are removed and the
base metal is affected to the minimum.
Intermediate loss of copper after individual intervals of pickling by Chelatone 3,
sulfuric acid and sodium hydroxide are shown in Fig. 1. Intervals of pickling are
defined in Table 9. It is obvious that after pickling in Chelatone 3 solution and
sulfuric acid was finished, all corrosion products were removed from the surface
(curve gradually turns into straight line). In the case of pickling in sodium hydroxide
solution, not all corrosion products were removed after four intervals. All corro-
sion products were removed by pickling in sulfuric acid already after three intervals
(6 min); the base metal has been slightly dissolved in the following intervals. Chela-
tone 3 solution dissolved corrosion products more slowly, but after four intervals
(80 min) all corrosion products were removed.
Table 10: Mass loss of copper with relatively thick layer of corrosion products after
pickling in various solutions (temperature 20◦ C).
Pickling solution Concentration Time Mass loss of copper

(M) (min)
(g/m2 ) (µm)
Citric acid 0.1 5×3 22.9 2.6

Chelatone 3 0.2 4 × 20 36.5 4.1
Sodium hexametaphosphate 0.5 4×3 56.3 6.3
Sulfuric acid 0.1 5×2 4.0 0.4
0.6
Mass loss (g/m2)
0.4
0.2
Chelatone 3
Sulphuric acid
Natrium hydroxide
0.0
1 2 3 4 5
Interval of pickling
Figure 1: Corrosion loss of copper during pickling.
In the case of the samples covered with a thick layer of corrosion products, the
course of pickling is similar: only the absolute values of material loss are higher.
Chelatone 3 solution acts more strongly in the first interval, but after the second
interval the intensity of dissolving corrosion products is similar to that of the sulfuric
acid solution. The total material loss after pickling in Chelatone 3 solution is lower
than in the case of pickling in the sulfuric acid solution, while the degree of cleaning
the surface is the same (Table 11).
Surfaces are changed after pickling in a way resulting from the technology used:
• etching of surfaces after pickling does not increase the coarseness of surfaces,
but the structure of metals is more marked, and
• trace amounts of elements contained in pickling solutions (Na, P) can be found
on pickled surfaces even after proper wash-off in some cases.
4.1.4 Other cleaning technologies

Other options are use of high-pressure steam or blasting with little dry ice balls.
Electrolytic methods are suitable for cleaning copper and bronze objects, especially
Table 11: Condition of copper surface after pickling in various solutions.
Pickling solution Condition of copper surface after pickling
Chelatone 3 Original corrosion products removed, products of

reaction with the pickling solution present, their
occurrence is related to crystallographic orientation
of grains of the base metal
Sodium Areal removal, grain borders slightly marked,
hexametaphosphate occurrence of undefinable stains which can be a
result of selective pickling or products formed
during pickling or material impurities
Sulfuric acid Original products removed, course of pickling and
also formation of new products depends on crys-
tallographic orientation of grains of the base metal
Citric acid Original products removed, course of pickling and
also formation of new products depends on
crystallographic orientation of grains of the base
metal, less marked than in case of sulfuric acid
pickling
archeological findings. The aim of cleaning is to remove aggressive ion fractions

(chlorides) from surface layers. These methods allow to keep the layer of patina
and other components on a treated object.
4.2 Artificial patination
Artificial patination is an important step in the process of creating and restoring

bronze and copper historical objects and objects of art [19]. Artificial patination is
not suitable for big outdoor surfaces as copper roofs or copper parts of cladding
of buildings. Because of the long time it takes for a natural green patina to form
there is a need to create patina artificially. Artificial patination is used mainly when
a desired look of a surface needs to be reached quickly. Protective properties or the
resistance to corrosion of created layers vary; in most cases they are not known or
considered.
4.2.1 Creation, composition and properties of artificial patinas

Natural and artificial patinations are processes differing in principle and so they
need to be understood and evaluated. Possibility of use of artificial patina needs to
be assessed case by case.
There are a number of formulas for patination containing various components.
Most of them are complex procedures, and their success is conditioned by expe-
rience and practice. The effect of chemical components is completed by other
procedures that mechanically remove unadhesive portions of patina, mechanically
increase the homogeneity of remaining layers or cause secondary reactions at raised

temperature. The process of creation of artificial patina is fairly difficult. Develop-
ment of methods of artificial patination is supported by the need for fast formation
of relatively evenly green or differently colored surfaces. Requirements for artificial
patinas, especially aesthetic ones, are highly specific; treated surfaces are usually
not very big and various craft procedures may be used to finish the surfaces. Data
on composition of artificial patina are limited, especially data mentioning the pati-
nation process together with the final composition of a layer of artificially created
products.
Patination is a process of controlled corrosion followed by further treatment of the
corroded surfaces. Patination causes loss of the base material; all types of patination
cause corrosion loss of bronze several times higher than the average corrosion loss
in urban atmosphere (3 to 10 times higher). The highest corrosion loss is caused by
two-step procedure (e.g. black patina created by solution of sulfurated potash, green
patina created by solution of copper nitrate). However, corrosion loss of 3–10 µm
is not important for loss in the thickness of either sculptures or sheets of objects of
art on an object.
Mixtures of different chemical substances are used for patination, and creation
of aesthetically desired even surface layers of artificial patina is very complicated.
Brown and black patina is created in water solutions of sulfurated potash and ammo-
nium sulfide (amorphous mixture of polysulfides). A thin layer of copper sulfide
is formed on surfaces of bronze objects. Green patina is usually created in water
solutions of copper nitrate or chloride together with other oxidants and complex-
forming agents. There are also industrial processes of patination of monuments or
semiproducts (sheets).
Longer service life of layers created this way is reached by consequent applica-
tion of waxes, microcrystalline waxes or transparent coatings. Regular maintenance
must be carried out to preserve the final aesthetical effect. However, repeated con-
servation causes changes in color.
Artificial patinas are more porous and less compact than natural patinas. Layers
of artificial patina cannot be sufficiently bound to the copper or bronze base of an
object. That is reached gradually by the repeated impact of climatic conditions,
especially precipitation and dry-out. Uneven impact of outdoor environment also
creates color tones typical for natural patina.
Artificially created green patinas on restored historical objects have a number of
disadvantages:
• insufficient adhesion,
• partial wash-out by precipitation (artificial patinas contain more water-soluble
components), and
• undesired changes of look during the process of transformation of artificial
patina into patina the composition of which is closer to natural patina.
Artificial patinas vary in composition, and the contained components depend on
the composition of patination solutions or pastes used and the manner of consequent
surface treatment (mechanical treatment – brushing, washing, raised temperature).
4.2.2 Artificial patinas produced in workshops

Several techniques are used for application of artificial patinas [20]:
• Direct patination – direct application of chemical solutions on surfaces of objects
using moistened cloths, soft brushes, etc. so that a thin coating is formed on a
surface,
• Patination in vapors – objects are exposed to impact of chemical vapors in a
closed environment. Areas differing in color can be formed during patination
on surfaces of different orientation where gaseous vapors are condensed.
• Dipping patination – the whole object is dipped into chemical solution, often at
higher temperatures. Solutions contain oxidizing substances, copper compounds
and also acids or alkaline substances in some cases. On big objects, solution
can be applied by pouring; however this method does not produce sufficiently
homogeneous and even-in-color patina layers.
• Patination at higher temperatures – a wide variety of mainly dark colors of
copper and bronze can be produced by the use of heat only. Patina layer is
relatively thin. Such dark layer is often used as preparation treatment before
further patination to green color.
Surfaces of objects can also be sprayed with suspension of basic copper salts.
Suspended particles are then fixed to a surface by wax-binding agents applied after
the suspension dries out.
The biggest problems occur in the case of patinating only parts of objects after
repairs, exchanges of damaged parts, etc. Even if the color of artificial patina is
the same as the color of natural patina, the new layer becomes visible due to the
transformation of artificial patina.
4.2.3 Industrially produced artificial patina (examples)

Producers of copper sheets offer sheets with a layer of artificial patination – dark
and green – too. Patinas are made continually on a production line by mechanical-
chemical heat process or by application of gel containing copper nitrate and basic
sulfates or copper chlorides. Extensive changes of patina composition can be
expected to occur during its aging.
TECU® -Patina made by KM company (Germany) was used for cladding and
roofing on both modern and historical buildings in Europe [21]. Sheets with Nordic
Green patina made by Outocumpu company (Finland) were used on many historical
and modern buildings around Scandinavia [19]. In the Czech Republic, both types of
artificial patinas has been used on private or commercial buildings; artificial patina
Nordic Green was used in roofing of Belveder (Queen Anna’s summer palace) in
Prague.
Sheets of artificial patina Nordic Green have been exposed for a long term at
the atmospheric test sites of SVUOM and the changes of patina layer are being
observed. A gradual wash-out occurred during the exposure but after approximately
four years the green surface began to spread again. There is a gradual formation of
natural patina, which grows through the layer of artificial patina. Computer picture
99% of green area 69% of green area 81% of green area
Figure 2: Changes of the layer of artificial green patina Nordic Green.
100
80
green area (%)
60
40
20
0
1 11 21 31 41 51 61 71
exposure (months)
Figure 3: Changes of the layer artificial green patina Nordic Green during long-term
exposure at a field test site Prague.
Table 12: X-ray diffraction analysis of patina layers.
Exposure Time (years)
0 1 2 3 5
Composition Brochantite Brochantite Brochantite Brochantite Brochantite

of layer Posnjakite Posnjakite Posnjakite Posnjakite Cuprite
Cuprite Cuprite Cuprite
analysis was used to evaluate these changes which allow for exact objective eval-
uation of the decreasing portion of the surface covered with green patina and the
expression of gradual changes by numerical values (Figs 2 and 3). Transformation
of the patina layer was observed at the same time (Table 12).
Table 13: Transformation of artificial patinas in various environments [13].
Patination agent Composition of Composition of patina after 2 years

patina before of exposure
exposure
Stockholm Götteborg Kopisty
Mixture of Very thin layer– Cuprite Cuprite Cuprite

polysulfides unidentified Gerhardite Atacamite Posnjakite
Brochantite
Mixture of Gerhardite Gerhardite Cuprite Cuprite
polysulfides Atacamite Posnjakite
+ Cu(NO3 )2 Posnjakite Brochantite
Brochantite Antlerite
Cu(NO3 )2 Gerhardite Cuprite Cuprite Antlerite
Gerhardite Atacamite
Stockholm – low-pollution urban atmosphere influenced by traffic; Götteborg –

urban atmosphere partially influenced by sea; Kopisty – industrial atmosphere,
prevailing influence of SO2 .
4.2.4 Transformation of artificial patinas

Results of long-term tests prove that artificial patinas (produced in workshops or
industrially) are transformed into natural patinas, which are in balance with the
surrounding environment. The rate and degree of transformation depends on the
type of atmosphere, its corrosivity and access of outdoor environmental impacts
(Table 13) [22].
4.3 Conservation
Conservation of copper and bronze is used on historical objects and objects of

art; it is used only rarely on construction elements (objects of arts on buildings).
Conservation decreases the negative effect of outdoor atmosphere on copper and
bronze surfaces and reduces the speed of deterioration of historical monuments.
Conservation in the case of bronze and copper historical monuments means mainly
treatment of surfaces by various types of waxes. Conservation should produce
water-proof, hydrophobic, chemically stable, solid, elastic and even protective layer.
Regular re-conservation at 1-2 years intervals can significantly reduce formation of
patina on bronze and preserve original look of the material. Conservation agents are
applied on objects covered with artificial green patina to protect it from wash-out
and transformation.
Protective effect of waxes is the result of a barrier effect of the layer, mainly its
low permeability for humidity and gaseous pollutants. Waxes provide only short-
term protection (depending on type of wax, thickness of coating and corrosivity of
atmosphere – 2 years at the maximum) and gradually grow old and loose elasticity
and flexibility. They then need to be removed, and the surfaces must be re-conserved
or treated with another protective system. Although very diluted waxes are used
and the coatings formed are very thin, solvents and mechanical treatment have to
be used to remove them.
Application of waxes and similar conservation agents changes the color of
surfaces – mainly those covered with green patina, but partially also those
covered with dark patina. Smallest changes of color are caused by beeswax.
Beeswax is the most often used traditional protective agent for copper and bronze
sculptures. For application on historical objects beeswax is diluted with technical
petrol. Beeswax applied by this way forms thin transparent coatings of 50–100 µm.
Protective effect of beeswax results mainly from the barrier effect of its layer.
A very important property of beeswax for its use in conserving historical mon-
uments is its water-repellent ability (hydrophobicity). Beeswax has a number of
other advantages for protection of bronze monuments:
• very little changes of look, color and character of surfaces and patinas are
observed;
• can be easily removed;
• does not change its look during gradual degradation and has no influence on the
look of a monument;
• its protective effect is well known and verified by long-term use in practice.
The disadvantage of beeswax is a relatively short protective effect in outdoor
atmospheres. The effectiveness of beeswax, especially its hydrophobic properties,
is lowered by the impact of light (esteric bonds are being destroyed).
Microcrystalline waxes are mixtures of carbohydrates with linear, branched or
cyclic chains. Waxes are esters of aliphatic acids C20 –C30 with high-order mono-
function alcohols. Waxes are produced by cooling melted waxes in suitable sol-
vents to form a structure of very small crystals, close to amorphous structure.
These microcrystalline waxes can be applied in thin layers and produce smooth
and bright coatings after polishing. Microcrystalline waxes are not as hydropho-
bic as beeswax. The protective effect of microcrystalline wax itself does not last
longer than one year. Re-conservation is more difficult than in the case of beeswax
layers. Gradual degradation creates an ugly-looking layer that should be removed
before re-conservation. A number of commercially produced conservation waxes is
based on microcrystalline waxes. Higher effectiveness of commercially produced
conservation waxes is a result of the corrosion inhibitors contained by them.
Application of waxes coated at high temperatures is preferred and tested in
Germany [23] and Sweden. Wax applied at high temperatures was also used in
the Czech Republic during last three years. The protective effect of the conserva-
tion layer after approximately two years is satisfactory. Waxes applied in this way
penetrate surface layers more easily and form thicker coatings in most cases.
Application of organic coatings with prevailing barrier effect seems to be effec-
tive only in few cases, one of the reasons being less suitable aesthetic effect after
application on bronze monuments. General repair of old coatings, especially remov-
ing them, is hard and almost impossible in case of applying coatings over patina.
The problem of coatings is tried to solve them by new hybrid organic-inorganic

coatings, which are applied in very thin layers. International task EUR 16637 EN
to develop Ormocer coating was finished in 1995 [24]. The coating has not been
sufficiently tested in practice so far. The aesthetic effect of the new coating may not
be the most suitable one in all cases (high brightness, change of color of the base
when applying the coating over green patina).
Fixation and anti-graffiti coatings on copper sheet semiproducts covered with
industrially produced artificial patina can also be put into this category of protective
measures of temporary effect.
5 Economic impacts
The process of atmospheric corrosion causes permanent damage in materials and
objects. Considerable sums have to be spent on the maintenance or replacement of
elements or damaged parts of objects. Economic considerations are highly influ-
enced by the relation between the price and service life of materials/elements. Costs
related to replacements need to be considered too.
It is very difficult to give statements about the economy of maintenance and
conservation or restoration of sculptures, historical objects and objects of art. Prices
differ a lot; they are often set individually; their service life is considered in time
periods of centuries; and there are also specific relations between service life
expressed as stability and service life considered in different aspects. The process
of atmospheric deterioration of an object is often stopped by making a copy of an
object and placing the original in a depository.
There are, therefore, only present examples of direct costs for repairs of copper
elements of buildings. Data need to be understood rather in their relations than in
terms of direct values, which differ both in time and place (Tables 14 and 15). Prices
shown below include costs for categories other than direct material costs:
• price of material/element,
• design and technological preparation,
• amount of work, including its organization, and
• storage and transport of waste.
Comparing data in both tables leads to the conclusion that using copper for
constructional and architectural purposes is economical. Although the material is
relatively expensive, the service life of elements is long and the requirements for
maintenance are small. Long service life is supported also by the fact that copper
elements and bigger areas (roofing) are used mainly for offices, banks or impor-
tant city buildings where professional treatment and proper maintenance can be
expected.
If maintenance and replacement of copper elements is carried out on historical
objects, costs are higher, usually by multiples of ten times.
Transfer of copper into the environment as a result of atmospheric corrosion,
which is an ecological burden especially for water and soil, is considered a negative
effect having economic consequences. Green corrosion products of copper and its
Table 14: Prices for repair and maintenance of roofing materials in the CR (in 1996)
[25, 26].
Materials Repair type Repair cost (€/m2 )
Galvanized steel Replacement 14.1

Removal 0.5
Aluminum Replacement 12.9
Removal 0.7
Copper Replacement 74.6
Removal 0.5
Concrete tiles Replacement 18.9
Removal 2.3
Clay tiles Replacement 12.6
Removal 1.1
Special clay tiles Replacement 24.6
Removal 1.8
Slate Replacement 46.0
Removal 2.0
alloys are washed away damage also neighboring areas of construction materials.
It is very difficult and expensive to remove traces of this effect and so it is carried
out only exceptionally.
6 Impact of atmospheric corrosion of copper and its alloys on

the environment
As a result of weathering of roofs, facades and other parts of constructions, which
is accelerated due to acidifying pollutants, a significant part of metals is emitted to
the biosphere. The main accumulation of metals occurs in urban areas where the
influx of metals is the greatest. Elevated levels of heavy metals in the sludge used
as fertilizers, in bottom sediments or in potable water may in the long run have
adverse effects on biological systems and human health. Gravimetric measurement
of exposed samples was the methodology used for assessment of release of heavy
metals due to corrosion of materials within the UN ECE International Co-operative
Programme on Effect on Materials including Historic and Cultural Monuments
[27].
A layer of corrosion products is formed on surfaces of metals and thickness of
metal or metallic layer goes on decreasing. Part of the corrosion product remains
on the surface and other parts are washed away by precipitation or flown into
the environment by wind as soluble salts or particles. Results of copper exposure
show that a significant part of the corroded metal remains on the surface as a part
of corrosion products: for example, as green-colored patina on copper roofs or
Table 15: Material corrosion costs by type of building [25] in thousands NOK06 (1995).
Type of Material Total Percentage

building
Galvanized steel Aluminum Copper Wood Plaster Concrete Roofing Brick
felt
Untreated Wire, profile Treated

Small house 2 616 0 3 663 136 63 27 030 3 896 2 869 2 060 317 42 650 21.6
Apartment block 5 169 0 9 434 547 97 4 636 16 994 7 307 2 319 597 47 101 23.8
Manufacture 1 055 0 1 260 109 31 326 321 387 1 212 71 4 771 2.4
Office 10 620 0 10 042 463 103 11 240 4 169 2 148 7 175 415 46 375 23.5
Hotels 91 0 98 5 1 110 41 42 70 8 465 0.2
Services 1 457 0 1 430 66 14 1 600 594 474 1 021 91 6 748 3.4

Agriculture 2 267 0 2 219 123 0 5 850 116 163 23 40 10 803 5.5
Other 1 711 0 1 433 65 23 2 760 1 212 827 701 84 8 815 4.5
Infrastructure 0 3 452 26 552 0 0 0 0 0 0 0 30 005 15.2
Total 24 983 3 452 56 133 1 515 332 53 552 27 342 14 216 14 581 1 623 197 732 100.0
Percentage 12.6 1.8 28.3 0.8 0.2 27.1 13.8 7.2 7.4 0.8 100.0
Table 16: Ranges of copper corrosion rates and metal release rates based on results
on corrosion attack from ICP Materials during the period 1987–95.
Exposure time Corrosion ratea Metal release ratea Metal release

(g/m2 a) (g/m2 a) (%)
1 6.7–13.4 0.3–1.2 3–21

2 5.2–10.3 0.5–1.6 7–33
4 5.0–9.3 0.7–1.8 10–33
8 3.8–6.8 0.8–2.0 20–42
aAverage yearly value during the measuring period.
350 8
total corrosion
300 total metal release 7
corrosion rate 6
250 113 g/m2
metal release rate
Rate (g/m2a)
5
Total (g/m2)
200
4
150
3
100
2
50 1
0 0
0 20 40 60 80 100
Time (years)
Figure 4: Calculated copper corrosion and metal release amounts and rates.
bronze sculptures. Results of average corrosion rates and metal release rates are
shown in Table 16 after 1, 2, 4 and 8 years of exposure. Metal release rate is always
lower than corrosion rate and the proportion increases with time due to decreasing
corrosion rates and increasing metal release rates.
In constant environmental conditions, corrosion rate and metal release rate even-
tually reach the same value, corresponding to the so-called steady state where
the composition and amount of corrosion products do not change with time. In
Fig. 4 calculated copper corrosion and metal release amounts and rates as a func-
tion of exposure time based on dose-response functions and constant environ-
mental conditions ([SO2 ] = 10 µg/m3 , [O3 ] = 40 µg/m3 , RH = 75%, T = 10◦ C,
Rain = 700 mm/a, and pH = 4.5) are illustrated. The difference between the amounts
of corrosion and metal release (113 g/m2 at steady state) corresponds to the amount
of metal retained on the surface in corrosion products.
7 Case studies
During the period 1992–98 in compliance with EU 316 COPAL, inspections on
about 200 copper and bronze objects have been performed by SVUOM specialists.
Results are summarized in a database. Following are three examples:
• Copper decoration element from the roof of Museum of Applied Arts, Prague,
• Bronze statue of St. John of Nepomuk, Charles bridge, Prague, and
• Copper roof of Queen Anna’s Summer Palace, Royal Garden, Prague.
All examples document the stage of non-treated copper and bronze cultural objects
exposed over a very long period in very polluted urban areas.
7.1 Case study 1 – Decoration element on the roof of Museum of Applied

Arts in Prague
Material: copper sheet, partly gilded, sheet remaining thickness 0.43–0.60 mm

Exposure: 100 years in urban atmosphere with corrosivity C4–C5
Defects:
• isolated cracks in area of forming and joints
• release of joints, non-functional joining by steel nails
• bimetallic corrosion
• occurrence of black spots in the area of soldered joints
• occurrence of crusts
• degradation of gilded surface
8
6 9
7
4
3
X-ray analysis of surface layer
Area of removal of samples Appearance of layer Composition
1 Upper hem Bright green, turquoiseBrochantite

2 Upper edge of eye Bright turquoise Brochantite, antlerite
3 Mouth Grey and green Antlerite, quartz, gypsum
4 Profiled part of hem Dark gray Cuprite, antlerite
5 Lower dark hem Red and brown Cuprite, antlerite
6 Lower part of inside Bright turquoise with Antlerite, chalcantite,
surface with dust gray parts to dark gypsum,
and crusts gray and green (NH4 )2 Cu(SO4 )2 · 6H2 O
7 Lower part of Turquoise, gray parts Chalcantite, antlerite,
inside surface brochantite, gypsum
8 Upper part of Grey-and-green to dark Antlerite, quartz
inside surface green, dark gray
9 Inside surface of hem Brown, gray and green Cuprite, quartz, antlerite
Microprobe analysis: chlorides in surface layers (samples 4 and 5)

Metallography cross: uniform microstructure of copper with patina layer
7.2 Case study 2 – Sculpture of St. John of Nepomuk on the Charles bridge
in Prague
Material: bronze alloy, gilded steel decorative elements

Exposure: 300 years in urban atmosphere with corrosivity C4
Defects:
• small holes on the most open areas (forehead, cap)
• dark surface with dust deposits in shelter position
• corrosion on the bonders of grain of alloy
• application of lead as filling material for holes and openings
Measurement of thickness of bronze

4.8 mm
2.4 mm
3.2 mm
2.4 mm
3.8 mm
31.8 mm
Upper surface of cap Green-and gray Brochantite, cuprite

Lower part of coat, shelter Black (NH4 )2 Cu(SO4 )2 · 6H2 O,
graffite
Open surface of body Green Brochantite, quartz, cuprite,
anglesite (PbSO4 )
7.3 Case study 3 – Roof of Queen Anne’s Summer Palace in

Royal Garden in Prague
Material: copper sheet, sheet remaining thickness 0.80 mm on the most of roof
sheet remaining thickness 0.35–0.40 mm on the east saddle part
Exposure: 100–400 years in urban atmosphere with corrosivity C4–C5

Defects:
• cracks and perforations on the sheets with minimum thickness
• occurrence of black spots on sheets
East side of roof
1 Cover of window Middle green Brochantite

2 Sheets on east side Yellow and green Brochantite, quartz, gypsum
3 Area with black spots Black Cuprite
Microprobe analysis of spots: high amount of Fe

Metallography cross: layered patina – lower layer cuprite; upper layer brochantite,
antlerite pits
200×
Reconstruction:
• realized in 2000–03
• replacing of damaged sheets by new copper sheets
• application of artificial green patina on replaced sheets
References
[1] Knotkova, D., Kreislova, K., Vlckova, J. & Had, J., Effect of Atmospheric
Environments on Copper and Bronze Cultural Monuments, COPAL I–VI,
SVUOM technical reports to project EU 316 EUROCARE COPAL,
1994–1999.
[2] Stöckle, B., Krätschmer, A., Mach, M. & Snethlage, R., Corrosion Attack on
Copper and Cast Bronze. Evaluation after 8 Years of Exposure, Report No
23 of UN ECE ICP on Effects on Materials Including Historic and Cultural
Monuments, May 1998.
[3] Kucera, V., Tidblad, J., Faller, M. & Singer, B., Release of Heavy Metals due
to Corrosion of Materials, Report of UN ECE ICP on Effects on Materials
Including Historic and Cultural Monuments, 2001.
[4] Mattsson, E. & Holm, R., Atmospheric corrosion of copper and its alloys,
Atmospheric Corrosion, ed. W.H. Ailor, John Wiley and Sons: New York,
NY, pp. 365–381, 1982.
[5] Graedel, T., Nassau, K. & Franey, J.P., Copper patinas formed in the
atmosphere – I. Introduction. Corrosion Science, 27(7), pp. 639–649, 1987.
[6] Graedel, T., Copper patinas formed in the atmosphere – II. A qualitative
assessment of mechanisms. Corrosion Science, 27(7), pp. 721–740, 1987.
[7] Graedel, T., Copper patinas formed in the atmosphere – III. A semi-
quantitative assessment of rates and constraints in the greater New York
metropolitan area. Corrosion Science, 27(7), pp. 741–769, 1987.
[8] Proc. of ICOMOS Conf. ‘Metallrestaurierung’, Mnichov, 23–25 October
1997, ed. Martin Mach, Arbeitshefte des bayerixchen Landesaimtes fur
Denkmalpflege, Band 94, Mnichov, 1998.
[9] Knotkova, D., Vlckova, J., Kreislova, K. & Had, J., Atmospheric corrosion
of bronze and copper statues, roofs and cladding on Prague’s monuments,
Proc. of Art ’96, Budapest, October 1996.
[10] Had, J. & Knotkova, D., Patiny a korozní produkty medi a medenych slitin.
Zpravy pamatkove pece, 60(1), pp. 8–10, 2000.
[11] Proc. of Workshops of Project Eurocare – EU 316 Copal, Praha, 1995, and
Wien, 1996, Institut fur Silikatchemie und Archeometrie: Wien.
[12] Franey, J.P. & Davis, M.E., Metallographic studies of the copper patina
formed in the atmosphere. Dialogue/89 – The Conservation of Bronze
Sculpture in the Outdoor Environment: A Dialogue Among Conservators,
Curators, Environmental Scientists, and Corrosion Engineers, ed.
T. Drayman-Weisser, NACE: Houston, TX, 1992.
[13] Riederer, J., The present level of techniques for conservation of outdoor
bronze sculptures. Proc. of Copal EU 316 Workshop, Budapest, May 1994,
and Vienna, 1995.
[14] Kreislova, K., Knotkova, D. & Had, J., Database of corrosion damage of
copper and bronze monuments in Prague. Proc. of the 5th International
Congress on Restoration of Architectural Heritage, Florence, Italy, 17–24
October 2000.
[15] Ashurst, J. & Ashurst, N.., Practical Building Conservation English Her-
itage Technical Handbook, Volume 4: Metals, Gower Technical Press Ltd.:
Aldershot, UK, 1988.
[16] Gullman, J. & Tornblom, M., Bronze Sculpture. The Making and Unmak-
ing, ed. The Central Board of National Antiquities and National Historical
Museum, Stockholm, 1994.
[17] Pago, L., Konzervovanie nezelezných kovov v muzejnej praxi, Archeologický

ústav, CSAV Brno, 1984.
[18] Kuttainen, K., Metal Losses During Cleaning and Patination of Cast Bronze,
Report SCI 1991-1, Stockholm.
[19] Outokumpo copper Nordic Green – method for patination of new-laid copper.
Instruction for use, technical specification by producer, 1996.
[20] Hughes, R. & Rowe, M.( eds.), The Colouring, Bronzing and Patination of
Metals, Thames and Hudson Ltd.: London, 1991.
[21] Med v architekture, VUT Brno, technical specification by producer, 2000.
[22] Johansson, E., Korrosion samt restaurering och konservering av bronsstatyer,
Report SCI, Stockholm, 1995.
[23] Meissner, B., Doktor, A. & Mach, M., Bronze- und Galvanoplastik.
Geschichte – Materialanalyse – Restaurierung, Landesamt fur
Denkmalpflege Sachsen, Mnichov 2000.
[24] Römich H., (ed.), New Conservation Methods for Outdoor Bronze Sculp-
tures, Final report of EUN 16637 EN Protection and conservation of
European cultural heritage, 1995.
[25] Hamilton, R. et al., Rationalized Economic Appraisal of Cultural Heritage
(REACH), Second Annual Report and Deliverables, Project EU, May 2000.
[26] Knotkova, D., Sochor, V. & Kreislova, K., Evaluation of Damages to Mate-
rials and Buildings Caused by Pollution, report SVUOM, Prague, 1997.
[27] Singer, B., Woznik, E., Krätschmer, A. & Doktor, A., Results from the Mul-
tipollutant Programme: Corrosion Attack on Copper and Bronze after 1 and
2 Years of Exposure, Report No. 37 of UN ECE ICP on Effects on Materials
Including Historic and Cultural Monuments, June 2001.

Atmospheric Corrosion and Conservation of Copper and Bronze

Uploaded by

Copyright:

Available Formats

Atmospheric Corrosion and Conservation of Copper and Bronze

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Atmospheric Corrosion and Conservation of Copper and Bronze

Uploaded by

Copyright:

Available Formats

CHAPTER 4

Atmospheric corrosion and conservation of

D. Knotkova & K. Kreislova

2 Atmospheric corrosion of copper and its alloys

Table 1: Guiding corrosion rates for copper.

Corrosivity Corrosion rate (µm/a)

Table 2: Average mass loss (g/m2 ) of copper and bronze on

Material Exposure period Time of exposure (years)

Copper 1987–95 10.1 16.4 27.8

Information about corrosion of copper in specific conditions was published in a

2.2 Formation mechanism of surface layers

Table 3: Stability of basic patina components in pH levels [7].

Compound Formula pH level

Cuprite Cu2 O >4

2.3 Influence of selected factors on quality of forming layers

Patina on copper or bronze object differs locally in thickness and composition in

Occurrence of green stripes formed by streams of water flowing away on dark

2.4 Identification of individual corrosion products

An overview of the phases determined repeatedly in surface corrosion layers on

Table 4: Phases identified in patina in the atmospheric environment of the

Name Formula Line

Antlerite CuSO4 · 2Cu(OH)2 7-407

Individual components were determined by a phase analysis of samples taken from

2.5 Composition of surface layers and patina components characteristics

• Brochantite CuSO4 · Cu(OH)2 , antlerite CuSO4 · 2Cu(OH)2 – corrosion prod-

Statues outside Prague Statues, copper roofs in the Statues in the

Abundance Frequency Abundance Frequency Abundance Frequency

www.witpress.com, ISSN 1755-8336 (on-line)

• Atacamite, paratacamite Cu2 (OH)3 Cl – dominant component of patina in

3 Typical corrosion manifestations on copper and

3.1 Sculptures and objects of art

Corrosion manifestations (described in more detail below), which may lead to

Main types of corrosion manifestations:

3.2 Copper roofs, cladding and metalwork elements

Corrosion attack (environmental effect and design)

Table 6: Types of defects on bronze and copper objects in Prague.

Age of Type of object Defect

• corrosion effects caused by contact with unsuitable materials (bimetallic corro-

Damages caused by indoor effects

4 Selected methods of cleaning and conserving copper and

• physical (mechanical) and chemical cleaning methods,

4.1 Cleaning of copper and bronze objects, namely sculptures

4.1.1 Cleaning by water under pressure

4.1.2 Mechanic (abrasive) cleaning

Abrasive agent Pressure Nozzle Material loss

(g/m2 ) (µm) (g/m2 ) (µm)

Corundum 120 0.4 1 1.5 0.18 1.8 0.22

Abrasive agent Pressure Nozzle Material loss

Corundum 120 0.4 2 1.2 0.15

4.1.3 Chemical cleaning

Pickling solution Concentration Time Total mass loss

Citric acid 0.10 5×3 0.45 0.05

Pickling solution Concentration Time Mass loss of copper

Citric acid 0.1 5×3 22.9 2.6

Figure 1: Corrosion loss of copper during pickling.

4.1.4 Other cleaning technologies

Table 11: Condition of copper surface after pickling in various solutions.

Pickling solution Condition of copper surface after pickling

Chelatone 3 Original corrosion products removed, products of