Corrosion Protection of Synthetic Bronze Patina

Abstract

Bronze artifacts are generally covered with green or blue coloured corrosion products called patina, which not only enhances the good appearance of the bronze, but also helps to protect it. Because of the increased air pollution and acid rain the large collection of statues and works of art made from bronze exposed in the urban environment could be damaged. The increase of air pollution damages also archaeological bronze objects exposed or stored in a museum. This is why it is necessary to find ways to improve the protection that the patina gives to bronze. In order to preserve metal works from the aggressive atmosphere, organic inhibitors are often employed. The inhibiting effects of two imidazole derivatives (4-methyl-1-phenylimidazole and 4-methyl-1-(p-tolyl)imidazole) on artificial patina were examined. The results of these investigations have shown that both inhibitors studied improve the protective properties of bronze patina in simulated urban acid rains.

Full text

Corrosion Protection of Synthetic Bronze Patina
K. Marušiæ+, H. Otmaèiæ-Æurkoviæ, H. Takenouti,* A. D. Mance, and E. Stupnišek-Lisac
Faculty of Chemical Engineering and Technology, University of Zagreb, Croatia
kmarusic@fkit.hr, hotmac@fkit.hr, admance@fkit.hr, elisac@fkit.hr
*Laboratoire Interfaces et Systèmes Electrochimiques (LISE), CNRS,
Université Pierre et Marie Curie, Paris, France
ht@ccr.jussieu.fr
Bronze artifacts are generally covered with green or blue coloured corrosion prod-
ucts called patina, which not only enhances the good appearance of the bronze, but also
helps to protect it. Because of the increased air pollution and acid rain the large collec-
tion of statues and works of art made from bronze exposed in the urban environment
could be damaged. The increase of air pollution damages also archaeological bronze ob-
jects exposed or stored in a museum. This is why it is necessary to find ways to improve
the protection that the patina gives to bronze. In order to preserve metal works from the
aggressive atmosphere, organic inhibitors are often employed. The inhibiting effects of
two imidazole derivatives (4-methyl-1-phenylimidazole and 4-methyl-1-(p-tolyl)imidazole)
on artificial patina were examined. The results of these investigations have shown that
both inhibitors studied improve the protective properties of bronze patina in simulated
urban acid rains.
Key words:
Archeological artefact, imidazole, artificial patina, electrochemical impedance spectro-
scopy, energy dispersive X-ray spectroscopy
Introduction
When exposed to their environment, bronze
objects, undergo a corrosion process, leading to the
formation of a protective layer of corrosion prod-
ucts on their surface.1This kind of patina is called
natural patina and it forms spontaneously. Depend-
ing on the environment it is exposed to, the patina
has a specific structure.2The second type of patina
is synthetic or artificial patina with defined chemi-
cal composition that can be formed in the labora-
tory allowing appropriate accelerated surface treat-
ment of bronze.3,4
Synthetic patina with a similar composition as
the natural patina is used for the present investiga-
tions. There are two electrochemical methods for
making artificial patina; potentiostatic (under po-
tential control) and galvanostatic (under current
regulation). The patina formation was carried out in
a sulphate/carbonate solution to mimic an urban at-
mospheric conditions.
Because of an increase of air pollution leading
to acid rainfall, the bronze and patina in urban at-
mospheres are being destroyed. Therefore, a similar
solution to the patina formation medium, but acidi-
fied to pH 5, is used for the corrosion test solution
corresponding to acid rain in an urban zone. Previ-
ous investigations5–11 have shown that the environ-
mentally friendly imidazole derivatives are good
copper corrosion inhibitors in different media. It is
reasonable to expect then that these compounds are
also valid for other bronzes, if they showed a
marked protective effect for these two different
copper alloys
Experimental conditions
To obtain a representative patina in a short pe-
riod, the patina was synthesised by two different
electrochemical methods, on two different bronzes:
Cu-8Sn-14Pb (B66 bronze) and Cu-6Sn. The com-
position of these alloys is given in Table 1.
The concentration was normalized so that the
sum of the all elements analyzed by EDS (Energy
dispersive X-ray spectroscopy) is 100 %. This is
also the case for all other EDS results.
K. MARUŠIÆ et al., Corrosion Protection of Synthetic Bronze Patina, Chem. Biochem. Eng. Q. 21 (1) 71–76 (2007) 71
+Corresponding author
Original scientific paper
Received: October 15, 2006
Accepted: December 13, 2006
Table 1 –Composition of the B66 and Cu-6Sn bronze in
atomic %
Sn Pb Ni Zn Fe Sb P Cu
B66 7.91 13.60 0.59 0.52 0.22 0.15 0.02 76.99
Cu-6Sn 6.10 0.01 00.10 0.02 00.11 93.66
The concentration normalized to 100 % for the total of elements an-
alyzed.

Cu-6Sn bronze was the composition represen-
tative of the Transylvanian archaeological patina of
the last Neolithic to Roman period.12 Bronze B66
containing lead as an alloy element is, in contrast,
selected as the representative of bronze coins found
in Morocco for the Post-Roman era.13 Therefore,
the composition of two bronzes correspond to that
of the cultural heritage, rather than bronze statues
exposed in the urban area. The aim of this paper is
to discover whether the anticorrosion effect of
imidazole compounds used are valid for two differ-
ent compositions, one a Cu-Sn binary alloy and the
other Cu-Sn-Pb ternary alloy.
Patina was synthesized, on both bronzes, in
an aerated solution composed of g=0.2gL
–1
Na2SO4+g=0.2gL
–1 NaHCO3at 30 °C. The pH of
this solution, as prepared was ca 8.5. This solution is
expected to produce a patina characteristic for the ur-
ban environment. The SO42– ions are one of the main
pollutants in urban atmosphere due to the industrial
activity and the car exhaust emission.
Patina on the B66 bronze (Pb containing the
bronze above) was synthesised during 24 h under a
constant anodic current density j=30mAcm
–2.
The patina on the Cu-6Sn bronze was synthe-
sised under potential regulation:
– During 60 s at – 0.2 V vs. to the initial open
circuit potential, that varies substantially from one
specimen to another.
– The potential of the open circuit becomes re-
producible and is equal to – 0.05V vs. SCE.
– Patina formation during the next 48 h at 0.09
V vs. SCE.
– Second step patina formation during another
48 h at 0.07 V vs. SCE.
This procedure introduced later in the present
work, though more complicated than the current
regulation method, allowed patina formation of six
electrodes simultaneously. The first step is intro-
duced for two reasons; to increase the reproducibi-
lity and to avoid the formation of pits at the elec-
trode surface. This step allows the reduction of a
native oxide layer, formed during the time separat-
ing between the moment when the electrode was
polished, and the time when it was dipped into the
patina formation solution. The overall current den-
sity is lower, compared to the patina formation at a
constant current. Consequently, so is the thickness
of the patina layer for the same polarization period.
A slow patina formation allowed, in contrast, the
structure closer to the natural patina.14 This process
is therefore well adapted to compare the inhibiting
effect of a few compounds simultaneously.
The morphology and crystallographic structure
of artificially obtained patina were examined with
Scanning Electron Microscopy (SEM) and X-ray
Elemental Energy Dispersion Spectroscopy (EDS)
analyses. SEM studies were performed with a Leica
Stereoscan 440 coupled with EDS elemental
semi-quantitative analyses (Princeton Gamma-Tech,
Inc.) at 20 keV.
The protective characteristics of patinated
bronze with and without the presence of inhibitors
were investigated by electrochemical impedance
spectroscopy (EIS). Measurements were performed
using EG&G potentiostat/galvanostat Model 263A
and Frequency Response Detector 1025 that were
controlled with Power-Sine software. The EIS mea-
surements were carried out in g= 0.2 g L–1 Na2SO4
+g= 0.2 g L–1 NaHCO3acidified to pH 5 by addi-
tion of a dilute sulphuric acid at room temperature
without regulation. The corrosion test solution is at
the thermodynamic equilibrium to ambient air.
The investigated corrosion inhibitors were:
4-methyl-1-phenylimidazole (PMI at concentration
c= 5 mmol L–1) and 4-methyl-1-(p-tolyl)imidazole
(TMI at c=1 mmol L–1). The inhibitor concentra-
tion was selected according to our former work on
the copper electrode for Cu-Sn-Pb alloy.10–11
Results and discussion
SEM observation and EDS analyses
The artificial patina formed by electrochemical
synthesis was dried at room temperature and ob-
served by SEM.
Fig. 1 and 2 show the structure of patina on the
surface of B66 and Cu-6Sn bronze. The SEM pic-
tures show different structures of patina layers. The
B66 bronze has three layers (Fig. 1, point 1, 2, 3).
EDS analyses were carried out on these layers and
the results are summarized in Table 2.
72 K. MARUŠIÆ et al., Corrosion Protection of Synthetic Bronze Patina, Chem. Biochem. Eng. Q. 21 (1) 71–76 (2007)
Fig. 1 –SEM image of the B66 bronze covered with a syn-
thetic patina formed by galvanostatic method

The outer layer of the B66 bronze patina
(Point 1. in Fig.1.) is rich in copper (33 %) and
oxygen (64 %) and contains some quantity of
sulphur (3 %) while the content of tin is negligi-
ble. The carbon-content was clearly observed, but
not analysed because the EDS is insufficiently
accurate for this element. Copper is soluble and
by dissolution – precipitation mechanism, it
forms patina with granular structure (point 1 in
Fig. 1) which consists of (Cu4(OH)6SO4) and
Cu2(CO3)(OH)2. This interpretation is essentially
supported by the Raman spectroscopy analyses re-
ported elsewhere.4,17
The intermediate layer (point 2. in Fig.1.) is
also rich in copper and oxygen but has a very dif-
ferent, smooth structure which indicates that the in-
termediate layer contains cuprite (Cu2O), character-
ized by a reddish colour.
The composition of the inner layer (point 3. in
Fig.1.) shows an enrichment in tin (12 %) and de-
crease in copper content (26 %). It is important to
note that the selective dissolution of copper from
the bronze matrix is one of the main features of
bronze corrosion and patina formation. This process
is known as decuprification. This is the same pro-
cess found on newly patinated bronze as well as on
archaeological bronzes2.
The Cu-6Sn bronze also has three layers
(Fig. 2, points 1, 2, and 3). EDS analyses were
also carried out on these layers of the Cu-6Sn
bronze patina and they are summarized in the Ta-
ble 2.
Layer 1 on the Cu-6Sn bronze (Point 1 on Fig.
2) contains a significant amount of oxygen (79 %)
and a small amount of sulphur (0.5 %). This means
that this layer is formed of copper-oxide, with a
small amount of copper-sulphate.
The crystalline structure that can be observed on
the SEM picture as layer 2 (point 2, Fig. 2) contains
a greater amount of copper (36 %) and sulphur (1.
%) than layer 1. Thus, it can be assumed that in this
part of the bronze copper-sulphate crystals are the
main corrosion product (Fig. 2). This layer is rough,
and probably the laser beam collected also the inner
layer formed essentially copper oxide. But, the col-
our of this part is in favour of copper salt.
Layer 3 (point 3, Fig. 2) contains mostly cop-
per (52 %) and oxygen (48 %) which form cop-
per-oxyde (Cu2O) (Fig. 4) like the intermediary
layer on the B66 bronze. It is not completely ex-
cluded that there is an inner most layer rich in tin
oxide, but we did not succeed in reaching this stra-
tum.
EIS measurements
Electrochemical impedance spectroscopy (EIS)
was applied on both bronzes covered with artificial
patina in order to investigate the protective effect of
two imidazole derivatives; 4-methyl-1-phenylimi-
dazole (PMI) and 4-methyl-1-(p-tolyl)imidazole
(TMI). After the impedance measurements in the
sulphate/carbonate solution, the electrodes were im-
mersed in the sulphate / carbonate solution contain-
ing one of the studied inhibitors. The influence of
PMI on the EIS spectra on the bronze B66 covered
with patina is presented in Fig. 3a and that of TMI
in Fig. 3b. Fig. 4a illustrates the effect of PMI on
the Cu-6Sn bronze with artificial patina and Fig. 4b
that of TMI. These spectra were collected after one
hour of immersion in the corrosion test solution.
First, the EIS was collected in the solution without
inhibitor, then with the inhibitor added to the corro-
sion test solution.
K. MARUŠIÆ et al., Corrosion Protection of Synthetic Bronze Patina, Chem. Biochem. Eng. Q. 21 (1) 71–76 (2007) 73
Table 2 –The elemental composition in atomic % of the
B66 and Cu-6Sn bronze patinas at three different
positions
Position Cu S O Sn
B66 bronze
1 33.1 2.6 64.2 0.1
2 35.2 1.5 63.2 0.1
3 26.4 2.9 57.8 12.9
Cu-6Sn
1 20.1 0.5 79.4 0
2 35.6 1.1 63.3 0
3 51.6 0.1 47.9 0.4
Fig. 2 –SEM image of the Cu-6Sn bronze covered with ar-
tificial patina formed under potential regulation

Though not clearly seen for many diagrams,
there are three capacitive loops involved in the imped-
ance spectra. Therefore, the equivalent electrical cir-
cuit depicted in Fig. 5 was used to carry out the pa-
rameters fitting with a simplex method.
The Rf-Cfcircuit, the contribution of which
will be revealed in the high frequency domain, cor-
responds to the capacitance and resistance of the
surface film, likely the oxide layer which is too thin
to be observed by SEM picture. The Cfis related to
the dielectric property of this layer whereas Rfde-
notes an ionic leakage through this layer. The me-
dium frequency circuit Rt-Cdcorresponds to the
charge transfer resistance and the double layer ca-
pacitance. The low frequency loop represented by
the RF-CFcircuit is allocated to faradic resistance
and faradic capacitance,12 implying the patina layer
and eventually oxidation – reduction reaction in-
duced by the dissolved oxygen.
Table 3 summarizes the results of the regres-
sion calculation.
The results relative to the blank test, that is the
EIS measured before addition of the inhibitor scat-
tered significantly. This phenomenon was explained
by the formation of a native oxide layer between
the moment when the patina formation was made,
and the effect of imidazole compounds was evalu-
ated.12 When recorded successively the impedance
spectra in the corrosion test solution without inhibi-
tor, the low frequency limit of the impedance de-
creased at the initial period corresponding to the
dissolution of the native oxide layer, then increased
with time corresponding to the formation of another
oxide layer at the bronze surface.
The value of Cfis located between 0.1 and
1mFcm
–2, which is in agreement with the forma-
tion of a thin surface film. This capacitance may
correspond likely to the oxide film located between
the substrate bronze and the patina layer. In this hy-
pothesis, the permittivity of many oxides is close to
10, and then if the model of planer capacitance is
applied, the thickness of the oxide film is evaluated
to be d= 10 to 100 nm. This value validates the hy-
pothesis on the origin of the high frequency capaci-
tive loop. In the solution containing the PMI, Rfof
the B66 bronze with patina layer is much higher
than in the uninhibited solution (Sample 1). Also,
the increase of Cfis observed which is in agreement
74 K. MARUŠIÆ et al., Corrosion Protection of Synthetic Bronze Patina, Chem. Biochem. Eng. Q. 21 (1) 71–76 (2007)
Fig. 4–EIS spectra of the patinated Cu-6Sn bronze in sul-
phate/carbonate solution a) with and without the addition of
5 mmol L–1 PMI; b) with and without the addition of 1 mmol
L–1 TMI
Fig. 3–EIS spectra of the B66 bronze covered with an arti-
ficial patina in sulphate/carbonate solution a) with and without
the addition of 5 mmol L–1 PMI; b) with and without the addi-
tion of 1 mmol L–1 TMI
Fig. 5 –Equivalent circuit used for fitting EIS data

with the transformation of the surface film to a new
and a thin protective oxide layer. In contrast, Rfand
Cfof the B66 sample immersed in the TMI contain-
ing the solution are similar to those observed in the
uninhibited solution. The TMI modifies little the
oxide layer. Rtvalue in both studied corrosion in-
hibitors is much higher than that in the uninhibited
solution. That is, these compounds hinder markedly
the corrosion kinetics.
The RF-CFloop corresponds, as mentioned
above, to a redox process involving corrosion prod-
ucts or dissolved oxygen. From the data in Table 3,
it can be seen that this process is slowed down in
the presence of both inhibitors. The formation of a
complex with the patina and the inhibitor may sta-
bilize the patina layer. This is an interesting feature
since the imidazole not only hinders the corrosion
rate of the substrate bronze, but also slows down
the degradation of the patina layer with time.
For Cu-6Sn bronze, the addition of both PMI
and TMI modify scarcely the value Rfand Cf.
Therefore, these molecules do not modify substan-
tially the surface oxide film. In contrast, Rtin-
creases significantly in the presence of both sub-
stances, therefore they slow down dramatically the
corrosion process. This is also true for RF. In other
words, PMI and TMI allow first the inhibition of
bronze corrosion and second the stabilization of the
patina layer leading to the protection of the sub-
strate bronze.
Figure 6 illustrates, for example, the time
change of the impedance spectrum of B66 bronze
covered with artificial patina in the corrosion test
solution containing 1mmol L–1 TMI. A marked de-
crease of the low frequency limit of the impedance,
polarization resistance, can be noticed. Beyond this
initial change, the polarization resistance increased
continuously up to a couple of weeks. At the third
week of immersion, a slight decrease of this resis-
tance was observed. That is, the inhibiting effect of
this compound improves slowly with time. A final
decrease of the polarisation resistance may be ex-
plained by the roughening of the electrode surface.
It is important to note that a slow increase of inhib-
iting effect was observed in all four cases examined
in this study.
Both imidazoles examined in this study
showed an inhibiting effect for Cu-8Sn-14Pb ter-
nary bronze and Cu-6Sn binary bronze covered
with artificial patina. The corrosion test in the urban
K. MARUŠIÆ et al., Corrosion Protection of Synthetic Bronze Patina, Chem. Biochem. Eng. Q. 21 (1) 71–76 (2007) 75
Table 3 –EIS data of the bronze electrodes covered with artificial patina in sulphate / carbonate solution with and without addi-
tion of the inhibitors examined
Bronze B66 Cu-6Sn
Sample Blank PMI Blank TMI Blank PMI Blank TMI
Rf,kWcm20.613 9.29 2.11 2.11 5.66 3.92 7.27 4.60
Cf,mFcm
–2 0.11 0.81 0.12 0.13 0.05 0.07 0.07 0.04
nf0.63 0.70 0.73 0.74 0.80 0.75 0.69 0.71
Rt,kWcm23.0 5.2 28.0 30.9 8.2 35.1 5.3 23.5
Cd, mF cm–2 221 24 60 19 0.7 5.6 1.0 3.2
nd0.59 0.50 0.59 0.67 0.52 0.53 0.61 0.52
RF,kWcm21.9 10.6 2.9 25.0 22.1 33.5 16.5 18.6
CF,mFcm
–2 41 0.49 4.5 0.51 0.36 0.36 0.38 0.62
nF0.97 0.51 0.92 0.80 0.50 0.51 0.53 0.58
Blank: Data from the EIS determined before inhibitor addition in the solution.
Fig. 6 –Time evolution of the impedance spectrum in the
corrosion test solution containing 1 mmol L–1 TMI

atmosphere with artists’ bronze was not carried out,
but the experimental method used in this work will
allow a preliminary selection of inhibitors expected
to be efficient towards the corrosion of bronze in-
duced by acid rain.
Conclusions
– Artificial patina was formed electrochemi-
cally on bronzes Cu-8Sn-14Pb (B66) under con-
stant current regulation, and Cu-6Sn under three
step potential regulation in aerated sulphate/carbon-
ate solution which simulates the formation of patina
in urban atmosphere.
– The results obtained by SEM and EDS analy-
ses have shown that patina consists of three layers
of different textures on both bronzes. On the B66
bronze the outer layer (1) is turquoise blue of
granular structure that consists of brochantite
(Cu4(OH)6SO4) and of malachite Cu2(CO3)(OH)2.
The intermediate layer (2) is rich in copper as well
as oxygen. This layer may be likely constituted by
the cuprite (Cu2O) layer, characterized by reddish
colour. At the inner layer (3) the copper content is
low indicating the decuprification process. The tin
was transformed into SnO2which protects copper
from further oxidation. The Cu-6Sn bronze layer 1
is formed of copper-oxide and copper-sulphate.
Layer 2 contains copper-sulphate crystals together
with cuprite. Layer 3 consists of copper-oxide.
– The protective efficiency of two non-toxic
imidazole derivatives (4-methyl-1-phenylimidazole
and 4-methyl-1-(p-tolyl)imidazole) on bronze pati-
nas was investigated by electrochemical impedance
spectroscopy. The results of the investigation have
shown that both organic compounds have protec-
tive properties on patinas on both bronzes in a me-
dium simulating urban acid rain due to atmospheric
pollution. Its protective effectiveness increases with
the immersion time in the inhibitor containing the
corrosion test solution.
ACKNOWLEDGEMENTS
The financial support from the Ministry of Sci-
ence and Technology of the Republic of Croatia un-
der Project 0125012 is gratefully acknowledged. The
authors acknowledge EGIDE for financial support
of this work through EcoNet and Cogito projects.
Symbols
c–concentration, mmol L–1
w–mass fraction, %
j–current density, A m–2
Re–electrolyte resistance, Wm2
Rf–film resistance, Wm2
Cf–film capacitance, F m–2
g–mass concentration, mmol L–1
d–thickness, nm
nf–constant phase element coefficient associated
with the film resistance
Rt–charge transfer resistance, Wm2
Cd–double layer capacitance, F m–2
nd–constant phase element coefficient associated
with the double layer capacitance
RF–faradaic resistance, Wm2
CF–faradaic capacitance, F m–2
nF–constant phase element coefficient associated
with the faradaic capacitance
Z–impedance, kWcm2
References
1. Fitzgerald, D. P., Nairn, J., Atrens, A., Corros. Sci. 40
(1998) 2029.
2. Robiola, L., Blengino, J.-M., Fiaud, C., Corros. Sci. 40
(1998) 2083.
3. Rosales, B. Vera, R., Moriena, G., Corros. Sci. 41 (1999)
625.
4. Rahmouni, K., Joiret, S., Robiola, L., Srhiri, A., Takenouti,
H., Vivier, V., Bulg. Chem. Communications, 37 (2005) 26.
5. Stupnišek-Lisac, E., Cinotti, V. Reichenbach, D., J. Appl.
Electrochem. 29 (1999) 117.
6. Gašparac, R., Martin, C. R., Stupnišek-Lisac, E., J. Electro-
chem. Soc. 147 (2000) 548.
7. Gašparac, R., Martin, C. R., Stupnišek-Lisac, E., Mandiæ,
Z., J. Electrochem. Soc. 147 (2000) 991.
8. Stupnišek-Lisac, E., Galiæ, N., Gašparac, R., Corrosion,
56 (2000) 1105.
9. Stupnišek-Lisac, E., Gazivoda, A., Madarac, M., Electro-
chim. Acta, 47 (2002) 4189.
10. Otmaèiæ, H., Stupnišek-Lisac, E., Electrochim. Acta, 48
(2003) 985.
11. Otmaèiæ, H., Telegdi, J., Papp, K., Stupnišek-Lisac, E., J.
Appl. Electrochem. 34 (2004) 545.
12. Muresan, L., Varvara, S., Stupnišek-Lisac, E., Otmaèiæ, H.,
Marušiæ, K., Horvat-Kurbegoviæ, Š., Robbiola, L., Rahmou-
ni, K., Takenouti, H., Electrochimica Acta, in press.
13. Serghini-Idrissi, M., Bernard, M. C., Harrif, F. Z., Joiret,
S., Rahmouni, K., Srhiri, A., Takenouti, H., Vivier, V.,
Ziani, M., Electrochimica Acta, 50 (2005) 4699.
14. Robbiola, L., Pereira, N., Thaury, K., Fiaud, C., Labbé,
J.-P., Decuprification phenomenon of Cu-Sn alloys in
aqueous solution at nearly neutral pH condition, METAL
98, Ed. W. Mourey & L. Robbiola, James & James Sci-
ence Publ., London (1998) 136–144.
15. Joiret, S., Rahmouni, K., Srhiri, A., Takenouti, H., Vivier,
V., Eurocorr 2004, Nice September 2004, CD No 287.
16. Rahmouni, K., Robbiola, L., Srhiri, A., Takenouti, H.,
Ventimiglia, E., 56th Annual Meeting of the International
Society of Electrochemistry, 25–30 September 2005,
Busan (Korea).
17. Rahmouni, K., Ph. D. thesis, “Corrosion and protection of
bronzes covered with patina: Electrochemical and spectro-
scopic study of the surface of archaelogical artefacts and
synthesis of equivalent patina on a commercial bronze”,
University Paris 6 and University Ibn Tofail (2005).
76 K. MARUŠIÆ et al., Corrosion Protection of Synthetic Bronze Patina, Chem. Biochem. Eng. Q. 21 (1) 71–76 (2007)