1
The isotope geochemistry of zinc and copper
1
2
Frédéric Moynier
1,2
, Derek Vance
3
, Toshiyuki Fujii
4
, Paul Savage
5
3
4
1
Institut de Physique du Globe de Paris, Université Paris Diderot, Université Sorbonne
5
Paris Cité, CNRS UMR 7154, 1 rue Jussieu, 75238 Paris Cedex 05
6
2
Institut Universitaire de France, Paris, France
7
3
ETH Zürich, Institute for Geochemistry and Petrology, Department of Earth Sciences,
8
Clausiusstrasse 25, 8092 Zürich, Switzerland.
9
4
Research Reactor Institute, Kyoto University, 2-1010 Asashiro Nishi, Kumatori, Sennan,
10
Osaka 590-0494, Japan
11
5
Department of Earth and Environmental Sciences, University of St Andrews, Irvine
12
Building, Fife KY16 9AL, United Kingdom
13
14
Corresponding author: moynier@ipgp.fr
15
16
INTRODUCTION
17
18
Copper, a native metal found in ores, is the principal metal in bronze and brass. It
19
is a reddish metal with a density of 8920 kg m
-3
. All of copper's compounds tend to be
20
brightly colored: for example, copper in hemocyanin imparts a blue color to blood of
21
mollusks and crustaceans. Copper has three oxidation states, with electronic configurations
22
of Cu
0
([Ar]3d
10
4s
1
), Cu
+
([Ar]3d
10
), and Cu
2+
([Ar]3d
9
). Cu
0
does not react with aqueous
23
2
hydrochloric or sulfuric acids, but is soluble in concentrated nitric acid due to its lesser
24
tendency to be oxidized. Cu(I) exists as the colorless cuprous ion, Cu
+
. Cu(II) is found as
25
the sky-blue cupric ion, Cu
2+
. The Cu
+
ion is unstable, and tends to disproportionate to Cu
0
26
and Cu
2+
. Nevertheless, Cu(I) forms compounds such as Cu
2
O. Cu(I) bonds more readily
27
to carbon than Cu(II), hence Cu(I) has an extensive chemistry with organic compounds.
28
In aqueous solutions, Cu
2+
ion occurs as an aquacomplex. There is no clearly
29
predominant structure among the four-, five-, and six- fold coordinated Cu(II) species
30
(Chaboy et al. 2006). Hydrated Cu(II) ion has been represented as the hexaaqua complex
31
Cu(H
2
O)
6
2+
, which shows the Jahn-Teller distortion effect (Sherman 2001; Bersuker 2006),
32
whereby the two Cu-O distances of the vertical axial bond (Cu-O
ax
) are longer than four
33
Cu-O distances in the equatorial plane (Cu-O
eq
). The Jahn-Teller effect lowers the
34
symmetry of Cu(H
2
O)
6
2+
from octahedral T
h
to D
2h
. The sixfold coordination of hydrated
35
Cu(II) species is questioned by a finding of fivefold coordination (Pasquarello et al. 2001;
36
Chaboy et al. 2006; Little et al. 2014b; Sherman et al. 2015). The bond distance related to
37
Cu(H
2
O)
6
2+
is considered to reflect a rapid switch between the square pyramid and trigonal
38
bipyramid configurations (Pasquarello et al. 2001; de Almeida et al. 2009). The fivefold
39
coordination is supported by computational (Amira et al. 2005) and spectroscopic
40
(Benfatto et al. 2002) studies.
41
In aqueous media at elevated temperatures, Cu(I) is thermodynamically more stable
42
than Cu(II). The structures of Cu(I) species are thought to be due to the splitting of
43
degenerate 4p
x,y,z
orbitals by a ligand field (Kau et al. 1987). Cu(I) complexes possess
44
simple linear structures (Fulton et al. 2004) due to 4p
z
and 4p
x,y
orbitals. The splitting of
45
4p
x,y
orbital and/or the formation of degenerate 4p
y,z
orbitals give the Cu(I) species
46
3
threefold coordination structures (T-shaped or trigonal planar coordination). For the
47
fourfold tetrahedral coordination (T
d
) structure, the p
x,y,z
orbitals may be close to
48
degenerate.
49
Zinc is an element of Group 2B, the last column of the d block. Zinc is not a
50
transition metal by definition because it has a d subshell that is only partially occupied.
51
Zinc has two oxidation states, with electronic configurations of Zn
0
([Ar]3d
10
4s
2
) and
52
Zn
2+
([Ar]3d
10
), where Zn(II) has 3d
10
with two electrons per orbital. Zinc is sometimes
53
included with the transition metals because its properties are more similar to these than to
54
the post-transition metals, whose properties are determined by partially filled p subshells.
55
Fresh zinc has a shiny metallic luster, but it tarnishes easily. It is hard and brittle,
56
becomes malleable with increasing temperature, and melts at 419.53°C. Metallic zinc is
57
easily oxidized and hence it is used as a reducing agent. Reduction of acids like HCl to
58
H
2
by Zn
0
is well known.
59
In compounds or complex ions, Zn is present only as Zn(II). Hydrated Zn
2+
is
60
generally thought to be present as the octahedral Zn(H
2
O)
6
2+
, this being the most stable
61
structure (Mhin et al. 1992). Besides the marked preference for sixfold coordination, Zn(II)
62
can easily be fourfold or fivefold coordinated. The coordination number is attributable to
63
a balance between bonding energies and repulsions among the ligands.
64
Zinc and Cu are both moderately volatile elements, with 50% condensation
65
temperatures (T
c
) of 726K and 1037K, respectively (Lodders 2003). It was long thought
66
that Zn behaved as a lithophile element during planetary (and especially, Earth’s)
67
differentiation, hence there is negligible Zn in Earth’s core (e.g. McDonough 2003). This
68
4
assumption was used to place broad bounds on the amount of S (which has a similar T
c
to
69
Zn) in Earth’s core (around 1.7wt% Dreibus and Palme 1996). However, more recent work
70
indicates that Zn behaves as a moderately siderophile element, with potentially ~30% of
71
terrestrial Zn stored in Earth’s core (Siebert et al. 2011), significantly affecting the
72
conclusions of Dreibus and Palme (1996). Zinc is the most abundant lithophile element
73
with a T
c
<750K, 100 times more abundant than the second-most abundant (Br, Tc=546K).
74
Its high abundance relative to other moderately volatile elements (due to the relatively high
75
binding energies per nucleon of its isotopes) makes Zn a good tracer of volatility in rocks
76
and a major application of its isotopes has been related to understanding volatility
77
processes.
78
Copper is a siderophile and highly chalcophile element (Siebert et al. 2011), with
79
~2/3 of the terrestrial Cu thought to be stored in Earth’s core (Palme and O'Neill 2003).
80
Copper is also moderately volatile, but is the most refractory of the chalcophile elements,
81
meaning that Cu may be a good tracer of the role of sulphides during differentiation and
82
igneous processes.
83
Zn is comprised of five natural stable isotopes,
64
Zn (49.2%),
66
Zn (27.8%),
67
Zn
84
(4.0%),
68
Zn (18.4%) and
70
Zn (0.6%) and Cu of two stable isotopes,
63
Cu (69.2%), and
85
65
Cu (30.8%) (Shields et al. 1964). Due to their relatively high first ionization potentials
86
(9.4 eV for Zn and 7.7 eV for Cu), the measurement of Zn and Cu isotope ratios by
87
Thermal-Ionization Mass-Spectrometry (TIMS) is very difficult. This explains the very
88
limited amount of Zn and Cu isotopic data produced before the advent of Multiple-
89
Collector Inductively-Coupled-Plasma Mass-Spectrometry (MC-ICP-MS). In addition,
90
since Cu has only two stable isotopes it is not possible to use a double spike technique to
91
5
correct for instrumental bias on TIMS. Since the first commercialized MC-ICP-MS in the
92
late 90s and the first ‘high precision’ Zn and Cu isotope ratio measurements (Maréchal et
93
al. 1999), more than 500 papers have been published (source: ISI Web of Science) on
94
various geochemical topics associated with Zn and Cu isotopes (e.g. oceanography,
95
cosmochemistry, environmental sciences, medical sciences). With the exception of
96
medical sciences, for which there is a dedicated chapter in this volume, here we review
97
these varied applications and discuss the potential of these isotope systems for future
98
studies.
99
100
METHODS
101
102
Measurement of Zn and Cu isotope ratio was originally made using TIMS (Shields
103
et al. 1964; Shields et al. 1965; Rosman 1972). As for any element with only two isotopes,
104
it was not possible to properly assess the instrumental isotopic fractionation for Cu and the
105
analytical uncertainty was therefore poor (no better than 2 ‰/amu; Shields et al. 1964;
106
Shields et al. 1965). With five stable isotopes, for Zn it is possible to correct for
107
instrumental bias and TIMS was originally used with double spike methods to measure Zn
108
isotopic compositions. The earliest measurements, on the older generation of TIMS were
109
associated with analytical precisions of around 1 ‰/amu (Rosman 1972; Loss et al. 1990),
110
but modern generation TIMS can reach precisions of 0.1-0.2 ‰/amu (Ghidan and Loss
111
2011).
112
The vast majority of recent Cu and Zn isotopic data have been acquired by MC-
113
ICP-MS, either by standard-sample bracketing (e.g. Maréchal et al. 1999; Mason et al.
114
