Nuclear Inst.

and Methods in Physics Research B 461 (2019) 154–158

Contents lists available at ScienceDirect

Nuclear Inst. and Methods in Physics Research B

journal homepage: www.elsevier.com/locate/nimb

Studies of bronze corrosion phenomena by EBS and complementary T

techniques
Domagoj Šatovića, , Vladan Desnicaa, Stjepko Fazinićb, Maja Mičetićb
⁎

a
Academy of Fine Arts, Ilica 85, 10000 Zagreb, Croatia
b
Ruđer Bošković Institute, Bijenička cesta 54, 10000 Zagreb, Croatia

ARTICLE INFO ABSTRACT

Keywords: Two binary bronze alloys were investigated in order to clarify the development of corrosion products on bronze
Elastic Backscattering Spectrometry (EBS) exposed to chloride corrosive media. Bronzes were prepared by magnetron sputtering deposition on glass to
analysis create highly uniform and smooth bronze layers.
PIXE The main goal was to monitor corrosion phenomena in the first and the most pronounced period of corrosion
Bronze
layer formation in duration of 11 days. For this purpose Ion Beam Analysis (IBA), i.e. Elastic Backscattering
Corrosion
Archaeometry
Spectrometry (EBS) with 2 MeV protons and 1.5 MeV Li beams were used to monitor the general stoichiometry
and thicknesses of the corrosion product layers for different aging parameters. Simultaneously, Proton-Induced
X-ray Emission Spectroscopy (PIXE) was used in addition to ATR-FTIR and Raman measurements to find out if,
beside copper oxides, other species were present in the corrosion layer as well.

1. Introduction The aim of this work is to investigate the structural properties of the
corrosion films under development during the first and the most pro-
For efficient preservation of cultural heritage objects, thorough nounced period of corrosion layer formation on two bronzes (Cu-6Sn and
understanding of corrosion of bronze pieces of art, especially clar- Cu-12Sn) widely used as an artistic material [17]. Artificially corroded
ification of corrosion phenomena occurring on outdoor and indoor bronze specimens were investigated in order to gain precise information
bronze exposed to different corrosive media is of high importance on the formation of the corrosion product layers and in-depth elemental
[1–7]. profile of corrosion layers for different aging parameters i.e. different
To get full information of ongoing corrosion processes on the bronze corrosion time. Three types of samples were prepared by magnetron
surface is essential for any conservation-restoration procedure. Bronze sputtering deposition on thick glass substrates: these were about
artifacts are vulnerable to outdoor and indoor corrosion environments – 300–500 nm thick layers of pure copper, copper with 6 wt% of Sn and
from acid rain containing sulfates and carbonates in urban–industrial copper with 12 wt% of Sn. For each of the samples, corrosion layers were
environment, to chlorides in marine environmental exposure, which are prepared artificially by immersion of samples in corrosive medium for a
main cause of “bronze disease” [8–11]. Indoor corrosion environments specified period of time. All the samples were exposed to the chloride
(museums) are less aggressive, however, it has been reported that in- corrosion environment for the duration of 1, 2, 3, 8 and 11 days.
door corrosion seems to be affected by a greater number of air pollu- The samples were first investigated using complementary ion beam
tants than the outdoor corrosion, especially when indoor sources of analysis techniques PIXE (Particle Induced X-Ray Emission) and EBS
corrodents are present [12,13]. (Elastic Backscattering Spectroscopy) using 2 MeV protons and 1.5 MeV
Chloride ions and sulfur compounds are the most common and Li beams at the RBI Tandem Accelerator Facility in Zagreb.
important atmospheric corrosive agents, as has been reported [14,15]. Additionally, ATR-FTIR and Raman measurements were performed at
Bronze artifacts are usually very heterogeneous in composition, and the Academy of Fine Arts in Zagreb Correlation was established be-
thus corrode forming various and inhomogeneous corrosion product tween the results obtained with the EBS and other spectroscopic tech-
surface layers [6,16]. niques (PIXE, FTIR, Raman).

⁎
Corresponding author.
E-mail address: [email protected] (D. Šatović).

https://doi.org/10.1016/j.nimb.2019.09.040
Received 24 June 2019; Received in revised form 17 September 2019; Accepted 25 September 2019
Available online 03 October 2019
0168-583X/ © 2019 Elsevier B.V. All rights reserved.
D. Šatović, et al. Nuclear Inst. and Methods in Physics Research B 461 (2019) 154–158

Fig. 1. EBS spectra obtained with 2 MeV protons (left) and 1.5 MeV Li beam (right) of the measured samples 1a (pure Cu deposited on glass) after different times of
exposure to chloride corrosion environment.

2. Materials and methods detector having 8 μm Be window with the installed permanent magnet
system for deflection of backscattering protons to prevent their detec-
Two binary bronze alloys differing in composition were prepared as tion by the SDD detector. Si (Li) detector was positioned at 135° scat-
a thin film on glass by magnetron sputtering deposition. Cu-Sn alloys tering angle in horizontal plane, while SDD detector is positioned at
were prepared to imitate composition used in statuary bronze artifacts 150° scattering angle in vertical plane. An EBS detector installed in
i.e.: Cu-6Sn bronze (94 mass% Cu, 6 mass% Sn) labeled as 3b, and Cu- vertical plane at the scattering angle of 165° was used to record proton
12Sn (88 mass% Cu, 12 mass% Sn) labeled as 4b. For the reasons of and Li backscattered spectra. The accelerated proton beam coming from
comparison, measurements were also done on pure copper film the accelerator into the end-station was collimated by the set of circular
(99.995 mass% Cu labeled as 1a). Thin films consisting of approxi- collimators to the diameter of 3 mm. Data from detectors were collected
mately 350 nm thick Cu + Sn layers were produced by magnetron by using conventional charged particle spectroscopy chain connected to
sputtering deposition (KJLC CMS-18 system) on glass substrate. The the homemade multiparameter data acquisition and analysis system
glass substrates were cleaned with acetone, isopropyl alcohol and SPECTOR [21]. Measured elastic backscattering spectra were analyzed
deionized water before loading into the deposition chamber. The sub- by using reliable and well known ion beam analysis and simulation
strates were placed on a rotational stage with a rotating rate of 10 rpm. software SIMNRA [22]. For the proton EBS analysis of light elements
Pure Cu and Sn(99.995%) were used as targets in DC magnetron dis- we used evaluated elastic backscattering cross sections calculated using
charge mode, with the deposition rate of about 2 nm/min. the on-line SIGMACALC calculator [23] for the scattering angle of 165°.
Corrosion layers were formed artificially by immersion in 0.012 M Measured PIXE spectra were analyzed using PIXE analysis software
sodium–chloride solution (0.7 g/l NaCl aqueous solution) simulating GUPIXWIN, the Windows version of the well known and reliable GUPIX
chloride corrosion environment for a specified period of time (longest software [24].
immersion time was 11 days). In this way we were able to get well
formed and defined corrosion layers. Solution of 0.012 M NaCl was 3. Results and discussion
chosen to simulate the type of pollutants found in museum environ-
ments [18]. Coupons were taken out every 24 h successively for a At first, the results of the measurements performed on pure copper
period of 1–11 days, dried and stored in desiccator. films deposited on glass are presented. Fig. 1 shows the EBS spectra
All corrosion layer formation experiments were carried out in a obtained with 2 MeV protons and 1.5 MeV Li beam related to un-
1 dm3 laboratory vessel, at room temperature (25 ± 1 °C). The volume- exposed pure Cu film deposited on glass and those exposed for 2 and
to surface specimen area was 20 mL cm−2 according to the ASTM G 31 8 days to chloride corrosion environment. Proton EBS spectra show
(1990) standard [19]. clearly the Cu film at the surface of glass and main glass substrate
Attenuated total reflectance Fourier transform infrared spectroscopy components. The figure also clearly shows the development of the Cu
(ATR FTIR) was carried out using Perkin Elmer Spectrum 2 spectro- oxides at the surface, after 2 days as a thin layer and after 8 days spread
meter equipped with a solid state attenuated total reflectance (ATR) over the Cu film thickness. SIMNRA analysis showed that unexposed Cu
sample stage. Spectra were collected from four scans at the resolution of film thickness was about 2.7 × 1018 atoms/cm2 that would correspond
4 cm−1, between 450 and 4000 cm−1. The ATR-FTIR analyses of the to about 350 nm. After 2 days exposure to chloride environment, thin
bronze specimens surface were performed directly on the surface oxide layer is formed at the surface. According to the SIMNRA analysis,
without scraping the surface corrosion products. two layers could be identified, one with the thickness of about
The Raman spectra were obtained with a QE Pro-Raman 1.2 × 1018 atoms/cm2 and with Cu and O atomic concentrations of
Spectrometer from Ocean Optics, USA. about 2:1, and the second layer of more or less pure Cu with the
The measurements were performed using a 785 nm laser excitation thickness of about 2 × 1018 atoms/cm2. After 8 days exposure to
line with spectral resolution of approximately 13 cm−1. Instrument was chloride environment, the first oxide layer covers almost entire Cu film,
equipped with fiber optics probe which allowed very precise posi- with Cu and O atomic concentrations of about 2:1 spread to the depth of
tioning of the sample area. The laser power was set at 150 mW. about 4 × 1018 atoms/cm2. Much thinner layer of more or less pure Cu
PIXE and EBS measurements were performed using 2 MeV protons remained at the interface with the glass substrate with the thickness of
and 1.5 MeV Li beam from the 1.0 MV Tandetron electrostatic accel- only about 0.2 × 1018 atoms/cm2. The tails at the low energy side of
erator at the Ruđer Bošković Institute (RBI) in Zagreb [20]. The mea- the Cu EBS spectra indicate that the roughness of Cu layers increased
surements were performed at the general purpose ion beam analysis (i.e. that they have thickness distribution). This is confirmed by related
chamber, and at the ion microbeam. At the general purpose ion beam 1.5 MeV Li EBS spectra which clearly demonstrate that after 8 days
analysis end-station PIXE spectra were recorded by two solid state si- exposure Cu layers exhibit highly increased roughness, i.e. extended
licon detectors. The detector dedicated for recording higher energy thickness distribution. Fig. 2 shows PIXE spectra of the same samples.
parts of PIXE spectra was 80 mm2 Canberra Si (Li) detector with 25 µm Almost all the elements seen in these spectra are from the glass sub-
thick Be window and 300 µm Mylar absorber. At the same time, low strate, including Mg, Al, Si, S, K and Ca. All the three PIXE spectra are
energy parts of PIXE spectra were detected by 10 mm2 Ketek SDD normalized to the height of the Ca Kα x-ray peak. Ca also originates

155
D. Šatović, et al. Nuclear Inst. and Methods in Physics Research B 461 (2019) 154–158

presence of the surface oxide layer and of the second layer richer in Cu
than the first layer, with possible interface layer(s) in between with
gradually increased oxides contribution towards the surface fully oxi-
dized layer.
As illustration, Fig. 4 shows the SIMNRA analysis for the EBS spectra
obtained by 2 MeV protons of the Cu-6Sn bronze on glass exposed to
chloride environment for 2 days (left) and 8 days (right). The simula-
tion related to 2 days exposure (left) was obtained taking two layers: (i)
surface layer 1.6 × 1018 atoms/cm2 thick, with Cu, O, Sn and Cl in
atomic concentrations 0.58:0.36:0.02:0.04; (ii) second layer 1.5 × 1018
atoms/cm2 thick, with Cu and Sn in atomic concentrations 0.97:0.03. It
can be assumed that surface layer is composed of copper(i)-oxide and
some tin oxides together with very small amount of copper-chloride
species.
The simulation related to 8 days exposure (right) was also obtained
taking two layers: (i) surface layer 2 × 1018 atoms/cm2 thick, with Cu,
Fig. 2. PIXE spectra of the measured samples 1a (pure Cu deposited on glass) O, Sn and Cl in atomic concentrations 0.622:0.31:0.004:0.064; (ii)
obtained with 2 MeV protons simultaneously with the EBS spectra after dif- second layer 2 × 1018 atoms/cm2 thick, with Cu, O and Sn in atomic
ferent times of exposure to chloride corrosion environment. concentrations 0.62:0.355:0.025.
PIXE spectra of these samples are similar to those shown at the
from glass and its Kα x-ray peak intensity variation due to the ab- Fig. 2. Again, sulfur seems to be present only in glass and Cl con-
sorption within the surface Cu oxide layer can be neglected in the first centrations are very low but they increase with the exposure time (and
approximation. The inset shows the zoomed area related to S and Cl is the largest for 8d yellow sample and similar for 2d and 8d dark
peaks. The unexposed sample (0d) spectrum shows the S and Cl in- samples).
tensities typical for the glass substrate. In case of the samples exposed At the end, the results of the measurements performed on Cu-12Sn
for 2 and 8 days, the content of S does not change, while the Cl content bronze (labeled as 4b) deposited on glass are presented. Fig. 5 shows
increases with the exposure time, indicating a possibility of develop- the EBS spectra obtained with 2 MeV protons and 1.5 MeV Li beam
ment of Cu chlorides at the surface. By the SIMNRA analysis of 8d related to Cu-12Sn bronze deposited on glass and the same kind of
sample, it is estimated that atomic concentration of Cl in the oxidized samples exposed for 2 and 8 days to chloride environment. In this case
Cu layer would be below 0.03%. the samples exposed for 2 days also show visually different areas, cor-
Figs. 3 and 4 present the results of the measurements performed on responding to the level of oxidation (less oxides in ‘dark’ areas). The
Cu-6Sn bronze (labeled as 3b) deposited on glass. Fig. 3 shows the EBS samples exposed for 8 days show similar dark and ‘yellow’ areas as the
spectra obtained with 2 MeV protons and 1.5 MeV Li beam related to previous Cu-6Sn samples. Both proton and Li EBS spectra show similar
Cu-6Sn bronze deposited on glass and the same kind of samples exposed qualitative behavior as in the case of Cu-6Sn samples, i.e. partial oxi-
for 2 and 8 days to chloride corrosion environment. The samples ex- dation of the surface after 2 days exposure and full oxidation after
posed to chloride corrosion environment for 8 days showed two visually 8 days (in this case for both dark and yellow areas) and the loss of Sn
clearly distinctive parts seen as yellow and dark areas. EBS and PIXE from the surface. In this case, the 8 days exposure to chloride en-
spectra were taken from both of these areas. Proton EBS spectra shows vironment results with very pronounced roughness, i.e. extensive
clearly the Cu and Sn contributions of the bronze at the surface of glass thickness distribution, as clearly seen by looking at the low energy part
and main glass substrate components. The figure also clearly demon- of the Cu area in both 2 MeV proton and 1.5Li obtained EBS spectra.
strates the development of the oxides at the surface. After 2 days ex- The ATR-FTIR spectra of 3b and 4b bronzes measured after 2 days of
posure a thin oxide layer is developed at the surface side. After 8 days exposure showed small sharp peak at 646 cm−1 (Fig. 6, left). Literature
exposure the dark areas show intermediate thickness of the oxide layer, reports the Sn(II) oxide band appearing at 632 cm−1 in the infrared
while in case of the yellow areas the oxide layer is spread over the spectra of the yellow area of the corrosion products of bronze [25].
entire Cu-6Sn film. The figure also clearly shows disappearance of Sn At the bronze spectra measured after 8 days of exposure, intensive
from the surface with the increased exposure to chloride environment. band appeared at approximately 605 cm−1 and overlapped the pre-
This is also demonstrated by the EBS spectra obtained by 1.5 MeV Li existing band at 646 cm−1. As already reported this band could be at-
beam. Also, the Li EBS spectrum related to yellow area with 8 days tributed rather to the formation of copper(I) oxide than to vibrational
exposure corresponds to almost fully oxidized bronze with highly ex- energies of tin species [26]. The appearance of the new band at the ATR
tended thickness distribution. The spectra 2d and 8d (dark) show the FTIR spectra of the bronze surface after 8 days of exposure coincided

Fig. 3. EBS spectra obtained with 2 MeV protons (left) and 1.5 MeV Li beam (right) of the measured samples 3b (Cu-6Sn bronze deposited on glass) after different
times of exposure to chloride corrosion environment.

156
D. Šatović, et al. Nuclear Inst. and Methods in Physics Research B 461 (2019) 154–158

Fig. 4. EBS spectra obtained with 2 MeV protons for 3b sample (Cu-6Sn bronze on glass) exposed to chloride corrosion environment for 2 days (left) and 8 days
(right). The spectrum on the right is related to 'yellow' phase. The figures show experimental spectra and SIMNRA simulations (total and for each element).

Fig. 5. EBS spectra obtained with 2 MeV protons (left) and 1.5 MeV Li beam (right) of the measured samples 4b (Cu-12Sn bronze deposited on glass) after different
times of exposure to chloride corrosion environment.

Fig. 6. ATR FTIR (left) and Raman (right) spectra of the 4b sample (Cu–12Sn bronze) surface after different times of exposure to chloride corrosion environment.

with the growth of the third layer observed at Li beam EBS spectra. layers where the corroded fraction gradually decreases with depth.
The Raman spectra reveal that the corrosion compounds are very FTIR spectral bands (peak at 605 cm−1) confirmed that Cu2O is the
poorly crystallised or non-crystallised. However, a visible band max- main compound present in the corrosion layers of samples exposed for 2
imum at about 615 cm−1 and broad band in the 200–220 cm−1 region and 8 days to corrosive environment.
correspond to the presence of cuprous(I) oxide (Cu2O) but could also be A comparison of the two EBS with 1.5 MeV Li beam and 2 MeV
related to Sn–O vibration bands (Fig. 6, right). As reported [27] Sn(IV) proton beam together with ATR-FTIR and Raman methods has shown
and Cu(I) oxides have characteristic peaks at the same positions, which that these methods complement one another in giving fully describable
makes difficult the separation of peaks in the case of weak intensities corrosion product layers. It is reasonable to assume that the simulta-
and in the presence of poorly crystallised particles. neous application of both EBS methods could prove particularly valu-
able in determination of corrosion layer formation on bronzes.
4. Conclusion
Declaration of Competing Interest
The EBS results clearly shows that for both bronzes (Cu-6Sn and Cu-
12Sn) after 2 days exposure to the chloride corrosive environment three The authors declare that they have no known competing financial
distinct layers are formed: corroded surface layer, back layer interfacing interests or personal relationships that could have appeared to influ-
the glass that is mainly Cu, and the interfacing region between the two ence the work reported in this paper.

157
D. Šatović, et al. Nuclear Inst. and Methods in Physics Research B 461 (2019) 154–158

References [14] E. Johansson, C. Leygraf, B. Rendahl, Comparison of different methods for assess-
ment of corrosivity in indoor environments, 13th International Corrosion Congress,
Melbourne, Australia, (1996).
[1] M.C. Squarcialupi, G.P. Bernardini, V. Faso, A. Atrei, G. Rovida, Characterisation by [15] A.R. Mendoza, F. Corvo, Outdoor and indoor atmospheric corrosion of non-ferrous
XPS of the corrosion patina formed on bronze surfaces, J. Cult. Herit. 3 (2002) metals, Corros. Sci. 42 (2000) 1123–1147.
199–204. [16] C. Chiavari, K. Rahmouni, H. Takenouti, S. Joiret, P. Vermaut, L. Robbiola,
[2] E. Paparazzo, L. Moretto, X-ray photoelectron spectroscopy and scanning Auger Composition and electrochemical properties of natural patinas of outdoor bronze
microscopy studies of bronzes from the collections of the Vatican Museums, monuments, Electrochim. Acta 52 (2007) 7760–7769.
Vacuum 55 (1999) 59–70. [17] C. Chiavari, A. Colledan, A. Frignani, G. Brunoro, Corrosion evaluation of tradi-
[3] G.M. Ingo, T. de Caro, C. Riccucci, E. Angelini, S. Grassini, S. Balbi, P. Bernardini, tional and new bronzes for artistic castings, Mater. Chem. Phys. 95 (2006) 252–259.
D. Salvi, L. Bousselmi, A. Cilingiroglu, M. Gener, V.K. Gouda, O. Al Jarrah, [18] D. Hallam, D. Thurrowgood, V. Otieno-Alego, D. Creagh, A. Viduka, Studies of
S. Khosroff, Z. Mahdjoub, Z. Al Saad, W. El-Saddik, P. Vassiliou, Large scale in- commercial protective petrochemical coatings on ferrous surfaces of historical and
vestigation of chemical composition, structure and corrosion mechanism of bronze museum objects, metal, in: I. MacLeod, J.M. Theile, C. Degrigny (Eds.), Proceedings
archaeological artefacts from Mediterranean basin, Appl. Phys. A 83 (2006) of the International Conference on Metals Conservation, Santiago (Chile), Western
513–520. Australian Museum, Welshpool, 2001, pp. 297–303.
[4] A. Climent-Font, G. Demortier, C. Palacio, I. Montero, J.L. Ruvalcaba-Sil, D. Diaz, [19] ASTM G 31-72, Standard Practice for Laboratory Immersion Corrosion Testing of
Characterisation of archaeological bronzes using PIXE, PIGE, RBS and AES spec- Metals, 1990, pp. 102–109.
trometries, Nucl. Instr. Meth. B 134 (1998) 229–236. [20] M. Jakšić, I. Bogdanović Radović, M. Bogovac, V. Desnica, S. Fazinić, M. Karlušić,
[5] A. Kratschmer, I. Odnevall Wallinder, C. Leygraf, The evolution of outdoor copper Z. Medunić, H. Muto, Ž. Pastuović, Z. Siketić, N. Skukan, T. Tadić, New capabilities
patina, Corros. Sci. 44 (2002) 425–450. of the Zagreb ion microbeam system, Nucl. Instr. Meth. B 260 (2007) 114–119.
[6] K.P. FitzGerald, J. Nairn, G. Skennerton, A. Atrens, Atmospheric corrosion of copper [21] M. Bogovac, I. Bogdanović, S. Fazinić, M. Jakšić, L. Kukec, W. Wilhelm, Data ac-
and the colour, structure and composition of natural patinas on copper, Corros. Sci. quisition and scan control system for nuclear microprobe and other multiparameter
48 (2006) 2480–2509. experiments, Nucl. Instr. Meth. B 89 (1994) 219–222.
[7] L. Robbiola, J.M. Blengino, C. Fiaud, Morphology and mechanisms of formation of [22] E. Rauhala, N.P. Barradas, S. Fazinić, M. Mayer, E. Szilágyi, M. Thompson, Status of
natural patinas on archaeological Cu–Sn alloys, Corros. Sci. 40 (1998) 2083–2111. ion beam data analysis and simulation software, Nucl. Instr. Meth. B 244 (2006)
[8] T. Drayman-Weisser, Dialogue/89. The Conservation of Bronze Sculpture in the 436–456.
Outdoor Environment: A Dialogue Among Conservators, Curators, Environmental [23] SIGMACALC calculator (http://wwwnds.iaea.org/sigmacalc/).
Scientists, and Corrosion Engineers, Nace Houston, 1992. [24] M. Blaauw, J.L. Campbel, S. Fazinić, M. Jakšić, I. Orlić, P. Van Espen, The inter-
[9] T.E. Graedel, K. Nassau, J.P. Franey, Copper patinas formed in the atmosphere – I. comparison of PIXE spectrum analysis software, Nucl. Instr. Meth. B 189 (2002)
Introduction, Corros. Sci. 27 (1987) 639–659. (2000) 113–122.
[10] D. Mudronja, M. Jakšić, S. Fazinić, I. Božičević, V. Desnica, J. Woodhead, Z.A. Stos- [25] L. Robbiola, T.T.M. Tran, P. Dubot, O. Majerus, K. Rahmouni, Characterisation of
Gale, Croatian Appoxiomenos alloy composition and lead provenance study, J. anodic layers on Cu–10Sn bronze (RDE) in aerated NaCl solution, Corros. Sci. 50
Archaeol. Sci. 37 (2010) 1396. (2008) 2205–2215.
[11] A. David Scott, New insights on the corrosion of ancient bronzes using X-ray powder [26] D. Šatović, L. Valek Žulj, V. Desnica, S. Fazinić, S. Martinez, Corrosion evaluation
diffraction: the importance of paratacamite, sampleite, and connellite, Stud. and surface characterization of the corrosion product layer formed on Cu–6Sn
Conserv. 62 (7) (2017) 410–418. bronze in aqueous Na2SO4 solution, Corros. Sci. 51 (2009) 1596.
[12] D. Šatović, V. Desnica, S. Fazinić, Use of portable X-ray fluorescence instrument for [27] L. Robbiola, K. Rahmouni, C. Chiavari, C. Martini, D. Prandstraller, A. Texier,
bulk alloy analysis on low corroded indoor bronzes, Spectrochim. Acta, Part B 89 H. Takenouti, P. Vermaut, New insight into the nature and properties of pale green
(2013) 7–13. surfaces of outdoor bronze monuments, Appl. Phys. A 92 (2008) 161–169.
[13] B. Schmutzler, G. Eggert, C.F. Kuhn-Wawrzinek, Copper(II) hydroxide on artefacts:
Corrosion, conservation, colourants, Stud. Conserv. 62 (2) (2017) 61–67.

158