University of Florida Trace Element Vanadium Research Discussion

I am writing a paper on the trace metal vanadium and am needing help with research and writing what the trace metal is used for. The following is a guideline for the paper:

Synthesis of the findings and discussion (5-6 pages single spaced including tables and equations)

1. Divide into different topics based on your literature search and your interest

2. Discuss relevant topics for a given metal and make sure there are logic connections

3. While writing the paper, try to synthesize information based on different sources.

4. You want to show “1+1=3”, i.e., your contribution via your interpretation, comparison, and hypotheses. Make sure to cite references where you get the information.

5. Please place figures and tables within the text, and not at the end of the paper. Make them small but readable.

I have found quite a bit of references regarding vanadium but need help with explaining the chemistry behind its used in redox batteries and other forms of technology.

Chemical Geology 417 (2015) 68–89
Contents lists available at ScienceDirect
Chemical Geology
journal homepage: www.elsevier.com/locate/chemgeo
Vanadium: Global (bio)geochemistry
Jen-How Huang a,⁎, Fang Huang b, Les Evans c, Susan Glasauer c
a
b
c
Environmental Geosciences, University of Basel, CH4056 Basel, Switzerland
CAS Key Laboratory of Crust-Mantle Materials and Environments, School of Earth and Space Sciences, USTC, Hefei, Anhui 230026, China
School of Environmental Sciences, University of Guelph, Guelph, ON N1G 2W1, Canada
a r t i c l e
i n f o
Article history:
Received 18 May 2015
Received in revised form 22 September 2015
Accepted 23 September 2015
Available online 28 September 2015
Keywords:
Vanadium
Geochemistry
Biogeochemical cycling
Microorganism
Isotope
Mineralogy
Soil
a b s t r a c t
Redox-sensitive transition group elements are involved in almost all fundamental geochemical processes. Of
these elements, vanadium (V) contributes a particularly powerful tool to decipher the Earth’s history and its
link to extraterrestrial bodies. A comprehensive view of V includes the formation and interaction between the
Earth’s interior layers, the evolution of the Earth’s surface to a habitable zone, biogeochemical cycling, and anthropogenic impacts on the environment. Tracing the geochemical behavior of V through the Earth’s compartments reveals critical connections between almost all disciplines of Earth sciences. Vanadium has a history of
application as a redox tracer to address the early accretion history of the Earth, to identify connections between
the mantle and crust by subduction and melting, and to interpret past surface environments. The geochemical
cycling of V from the deep Earth to the surface occurs through magmatism, weathering and digenesis, reflecting
variations of fO2 and V species in different Earth compartments. Minerals form a link between deep Earth reservoirs of vanadium and surface environments, and the study of V in minerals has increased the understanding of V
cycling. Finally, the exploitation of V has been increasing since the Industrial Revolution, and significant amounts
of V have been released as a consequence into natural systems. Environmental concerns are promoting new areas
of research to focus on V cycling between water, air, soil and sediment compartments. An increased understanding of V in all compartments, and knowledge of the processes that connect the compartments, is vital to tracing
the fate of this intriguing element in natural systems.
© 2015 Elsevier B.V. All rights reserved.
Contents
1.
2.
3.
4.
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Vanadium geochemistry in terrestrial earth and planetary systems . . . . . . . . . . .
2.1.
Vanadium distribution in the Earth and in extraterrestrial materials . . . . . . .
2.1.1.
Meteorite, bulk earth, and moon . . . . . . . . . . . . . . . . . . .
2.1.2.
Mantle and crust . . . . . . . . . . . . . . . . . . . . . . . . . .
2.2.
Partitioning of vanadium during magmatism . . . . . . . . . . . . . . . . . .
2.2.1.
Valences and coordination of vanadium in melts and minerals . . . . . .
2.2.2.
Partitioning of vanadium between metallic and silicate melts . . . . . .
2.2.3.
Partitioning of vanadium between minerals and silicate melts . . . . . .
2.3.
Temporal and spatial variation of mantle fO2 constrained by vanadium . . . . . .
2.4.
Vanadium isotopes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Occurrence of vanadium in minerals . . . . . . . . . . . . . . . . . . . . . . . . .
3.1.
Vanadium associated with primary rocks and minerals . . . . . . . . . . . . .
3.2.
Weathering and diagenesis of secondary vanadium in minerals and fossil fuels . .
Cycling and transport of vanadium in surface environments . . . . . . . . . . . . . .
4.1.
Aqueous speciation chemistry of vanadium . . . . . . . . . . . . . . . . . . .
4.1.1.
4.1.2. The role of complexation in V mobilization . . . . . . . . . . . .
4.1.2.
Vanadium redox transformation in surface environments . . . . . . . .
4.2.
Microbial controls on V geochemistry: redox, complexation, and sorption reactions
⁎ Corresponding author.
E-mail address: jen-how.huang@unibas.ch (J.-H. Huang).
http://dx.doi.org/10.1016/j.chemgeo.2015.09.019
0009-2541/© 2015 Elsevier B.V. All rights reserved.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
69
69
69
69
69
71
71
72
72
72
73
73
73
73
75
75
76
76
77
J.-H. Huang et al. / Chemical Geology 417 (2015) 68–89
4.3.
Vanadium in surface environments: water, soil, and sediment
4.3.1.
Freshwater environments . . . . . . . . . . . .
4.3.2.
Marine environments . . . . . . . . . . . . . .
4.3.3.
Soils and sediments . . . . . . . . . . . . . . .
4.3.4.
Adsorption to metal oxide minerals . . . . . . . .
4.4.
Vanadium in the atmosphere . . . . . . . . . . . . . . .
4.5.
Anthropogenic sources . . . . . . . . . . . . . . . . .
5.
Conclusions and outlook . . . . . . . . . . . . . . . . . . . .
Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . .
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
1. Introduction
Vanadium is increasingly applied to studies of almost all fundamental geochemistry disciplines. Research on V geochemistry has, however,
lagged behind that for other transition metals. This is the first comprehensive review that addresses the geochemical behavior of V under dramatically different conditions, from the terrestrial Earth to other rocky
planets, from magmatic to environmental studies, and from partitioning
at high temperature to speciation in hydrous systems. Such a holistic approach is vital to link chemical reservoirs to the evolution of the bulk
Earth. Vanadium is widely distributed in igneous and sedimentary
rocks and minerals as a mildly incompatible, refractory, lithophilic element. The average crustal abundance of V is similar to that of Zn and
Ni (Reimann and Caritat, 1998) although it is more dispersed in the
crust than either element; concentrated mineral deposits are consequently rare. Much of the historical interest in V has derived from the
occurrence of the common redox states in Earth surface environments
(+ 3, + 4, + 5) as a consequence of V’s geochemical evolution. The
chemical speciation and the solubility of V species are strong functions
of pH and Eh conditions (Taylor and Van Staden, 1994), making it a
key redox indicator. More recently, interest in V partitioning during
high temperature – high pressure magma differentiation processes
has increased, and new possibilities to exploit V stable isotope chemistry are being explored (Nielsen et al., 2011; Nielsen et al., 2014; Prytulak
et al., 2011; Prytulak et al., 2013). Studies on the partitioning of V among
the possible reservoirs in the Earth’s crust have provided critical insights
into the evolution of these reservoirs (Fig. 1).
The lithophilic nature of V explains its predominance in Earth surface environments, where it plays a role in a wide range of biological
processes. The function of V in biology likely evolved with the chemical
differentiation of the Earth’s surface environments. This leads to the intriguing possibility that V played a major role in biological electron
transfer early in Earth’s history (Rehder, 2008a; Rehder, 2008b).
The complex redox chemistry of V and its particular application to
renewable energy technologies will likely add to the demand for this element in the future. The demand for V used in construction materials,
by far the greatest consumer of mined V, can be expected to further increase the volume of this element that is cycled through terrestrial,
aquatic and atmospheric systems.
In the following, we explore the chemical behavior of V in the Earth’s
major compartments, including magma, rocks, sediments and organisms as well as in extraterrestrial matter.
2. Vanadium geochemistry in terrestrial earth and planetary systems
2.1. Vanadium distribution in the Earth and in extraterrestrial materials
Knowing the V composition of the major geochemical reservoirs is
critical for understanding the geochemical behavior of V in terrestrial
magmatism, and for applying V to constrain fundamental processes of
the Earth. Such processes include, but are not limited to, core segregation, mantle and atmospheric evolution, and the development of ore
deposits.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
69
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
79
79
79
81
82
84
84
85
85
85
2.1.1. Meteorite, bulk earth, and moon
The starting point for V on Earth is accretion and core formation. A
comparison of the V content between CI chondrites and the Earth provides important constraints on both processes. CI chondrites are considered to represent the primitive undifferentiated materials in the solar
system and are likely the most important building blocks from which
the Earth was formed (McDonough and Sun, 1995; Palme and O’Neill,
2003; Sun and McDonough, 1989). The average V contents of CI chondrites estimated in a few studies are quite consistent, ranging from 55
to 56.5 mg kg− 1 (Anders and Grevessc, 1989; McDonough and Sun,
1995; Palme, 1988; Wasson and Kallemeyn, 1988). These values are
lower than the average V contents of the silicate earth (82 mg kg−1)
estimated in McDonough and Sun (1995) because V is more depleted
in Ca-Al-rich inclusions than are Al and rare earth elements. The bulk
V in the CI chondrite is consequently diluted.
The bulk Earth has a V content of 95 mg kg−1, lower than that in the
metallic core (120 mg kg−1) (McDonough and Sun, 1995). The ratio of
the V content between silicate Earth (or primitive mantle) and CI chondrite normalized to Mg content is 0.62 (McDonough and Sun, 1995),
showing a terrestrial depletion. Numerous studies have demonstrated
that V depletion is best explained by the preferential partitioning of V
into the metallic core at high pressure and low fO2 during the “deep
magma ocean” process (Gessmann and Rubie, 1998; McDonough and
Sun, 1995; Palme and O’Neill, 2003; Ringwood et al., 1991; Wänke
and Dreibus, 1986). Simple mass-balance calculations show that the
core could contain half of the total V budget of the bulk earth
(McDonough, 2003). The Moon mantle is depleted in V, as well as in Cr
and Mn, similar to the Earth’s mantle (Ringwood, 1966; Ringwood
et al., 1991). This implies that the Moon was most likely derived from
the mantle of the Earth or from an impactor larger than Mars that
experienced V depletion due to core forming processes (Drake et al.,
1989; Gessmann and Rubie, 2000; Ringwood et al., 1991).
2.1.2. Mantle and crust
Post-accretion processes have led to distinct signatures of V in the
mantle and crust. Vanadium contents in peridotites range from a few
mg kg−1 to around 100 mg kg−1, substantially lower than the concentration in mid-ocean ridge basalts (MORB), ocean island basalts (OIB),
and island arc basalts (IAB) (Fig. 2). This indicates that V is mildly incompatible during mantle partial melting. Clinopyroxene, garnet, and
spinel are the main hosts for V in mantle peridotite (Johnson et al.,
1990). Because the valence of V is sensitive to fO2, peridotites from different tectonic settings show variable V contents and correlations with
other major trace elements, providing an important tool to constrain
fO2 of the mantle through the Earth’s history (Canil, 2002; Canil, 2004).
A summary of 6590 basalt samples from spreading centers shows large
variations, from less than 100 mg kg−1 to 800 mg kg−1, and gives an
average V composition of ~275 mg kg−1 (data from http://www.petdb.
org/). Similar ranges are also observed in OIB and IAB (see (Mallmann
and O’Neill, 2009) for a recent summary) and reflect the effect of source
composition, melting degree, and melting style.
Trace element compositions for continental crust can be estimated
via weighted averages of a large number of representative rock units
70
J.-H. Huang et al. / Chemical Geology 417 (2015) 68–89
Fig. 1. Vanadium storage on earth.
exposed in the crustal surface, or measured on fine-grained sediments
or sedimentary rocks such as shale, loess, and tillite (Clarke, 1889;
Condie, 1993; Gao et al., 1998a; Gao et al., 1998b; Rudnick and Gao,
2003; Taylor and McLennan, 1995). As summarized in Rudnick and
Gao (2003), V contents in the bulk continental crust estimated in previous works vary substantially from 96 to 230 mg kg−1 (Condie, 1993;
Gao et al., 1998b; Rudnick and Gao, 2003; Taylor and McLennan,
1981; Taylor and McLennan, 1985; Taylor and McLennan, 1995). Such
discrepancy could reflect regional chemical heterogeneity in the continental crust, inaccuracy of the estimating methods involving data quality and unrepresentative samples, or proportions of mafic lower crust
versus felsic upper crust applied to calculate the bulk value. Nonetheless, it is generally agreed that the lower crust (196 mg kg−1) has a V
content within the range of mantle-derived basalts, but substantially
higher than the primitive mantle (82 mg kg−1), the middle crust
(107 mg kg− 1) and the upper crust (97 mg kg−1) (McDonough and
Sun, 1995; Rudnick and Gao, 2003) (Fig. 2). If the upper continental
crust was essentially formed by extracting felsic melt from mafic
sources (lower continental crust or subducted oceanic crust), or by
intra-crustal differentiation, then V as well as Sc and Cr are compatible
during the igneous processes forming the upper crust.
Average V content of the juvenile upper continental crust was
estimated by the map model by which the proportions of rocks were
determined by geological maps and stratigraphic successions (Condie,
1993). It increases with time from 70 to 73 mg kg−1 (pre-Archean) to
91–106 mg kg− 1 (post-Archean), and shows negative correlations
J.-H. Huang et al. / Chemical Geology 417 (2015) 68–89
71
Fig. 2. Correlations of V contents with SiO2 and Sc in geochemical endmembers. PM, primitive mantle; UCC, upper continental crust; MCC, middle continental crust; LCC, lower continental
crust; BCC, bulk continental crust; N-MORB, normal mid-ocean ridge basalts. Data source: crust, data Rudnick and Gao (2003); PM, Sun and McDonough (1989); N-MORB, http://www.
petdb.org/.
with Cr and Gd/Yb (Fig. 3). In contrast to V, the MgO, Cr, and Ni contents
decrease dramatically from Archean upper continental crust to the
Phanerozoic crust. Such variations may reflect the fundamental change
with time in growth mechanism of the felsic upper continental crust. It
has been proposed that the subducted hot oceanic crust (MORB composition) could be partially melted because of the larger geothermal gradient of the Earth during the Archean (2.5–3.8 Ga) (Boehler, 2000). The
melts could interact with the mantle wedge, producing tonalite–
tronhjemite–granodiorite (TTG) suites with high Mg and Cr contents;
in contrast, post-Archean juvenile upper continental crust grew mainly
via intra-crustal differentiation of basaltic and andesitic rocks, followed
by extraction of granitic melts without significant interaction with
the upper mantle (e.g. (Taylor and McLennan, 1981; Taylor and
McLennan, 1985; Taylor and McLennan, 1995)). Because fO2 is not
well constrained during crustal differentiation, it is not clear whether variations of residual mineral phases in magma played a dominant
role in producing the higher V content in the post-Archean upper
crust. In essence, the question remains: is the temporal increase in
V content a consequence of increasing fO2 during magma evolution?
Or does it reveal changes in temperature-pressure conditions for
continental growth from Archean to Phanerozoic?
2.2. Partitioning of vanadium during magmatism
A fO2-sensitive element can be exploited to estimate the oxidation
states in magmatic processes such as partial melting, magma differentiation, and core formation. Vanadium is an important element for this
application because it has more valence states in natural materials
than many other multi-valence elements. In the following, the chemical
partitioning of V during deep processes is summarized. The mineralogy
of V in further discussed in Section 3.
2.2.1. Valences and coordination of vanadium in melts and minerals
In oxides, silicate minerals, and melts, V exists as +2, +3, +4, and
+ 5 (Shearer et al., 2006a; Sutton et al., 2005), and a large amount of
V enters the metallic core as V0 (McDonough, 2003; McDonough and
Sun, 1995). Using V K-edge X-ray absorption near edge spectroscopy
(XANES), Sutton et al. (2005) determined the valence of V in natural
and synthetic basalt glasses and developed a microscale oxybarometer
to study the variation of fO2 in the melting and evolution of the Earth
and its moon, and Mars. Application of this method revealed that fO2
varies from logIW-1.6 in lunar glasses to logIW + 4.0 in terrestrial
glasses by at least 6 log units (Karner, 2006; Sutton et al., 2005).
Fig. 3. Variations of V, Cr, and Gd/Yb of continental crust with time showing decreasing contribution of mantle materials to the continental crust. Data are from Condie (1993).
72
J.-H. Huang et al. / Chemical Geology 417 (2015) 68–89
Vanadium is dominantly V(IV) in terrestrial melts, V(III) in lunar melts,
and a mixture of V(III) and V(IV) in Martian melts; glasses synthesized
in air contain V(V) (Karner, 2006). In minerals, V also exists as a mixture
of all valence states with their proportions varying as a function of fO2.
For instance, a XANES study on natural titanomagnetite from a layered
mafic intrusion show that V in magnetite is mostly V(III) with minor
V(IV) occupying octahedral sites in the spinel structure (Balan et al.,
2006).
2.2.2. Partitioning of vanadium between metallic and silicate melts
In order to more effectively use V to constrain redox variation in
high-temperature planetary processes, it is critical to understand the
partitioning of V between different phases at varying fO2. Several experimental studies have been performed to study the partitioning of V
between silicate melt and metallic melt (Metallic/silicateD) to constrain
the fO2 variation in the mantle during core formation (Drake et al.,
1989; Frost et al., 2008; Mann et al., 2009; Wade and Wood, 2005;
Wood, 2008; Wood et al., 2008). Metallic/silicateDV and Metallic/silicateDCr
may not be sensitive to silicate melt composition, but decrease dramatically with increasing fO2 and Si content in the metallic melt (Tuff
et al., 2011; Wood, 2008). Wade and Wood (2005) estimated that
core/mantle
DV ranges from 1.5–2.2, similar to the values reported in
McDonough (2003). To reconcile the depletion of V in the silicate
earth mentioned earlier with the lithophile tendency of V, high
temperature with a fixed fO2 (and iron oxide in the mantle) or a continuously increasing fO2 is required during core segregation (Frost
et al., 2008; Wade and Wood, 2005; Wood, 2008), while a recent
experiment study in the effect of oxygen content in the core on
core/mantle
DV also suggests accretion of the core under oxidizing conditions (Frost et al., 2008; Wade and Wood, 2005). The model of high
temperature with a fixed fO2 can be excluded because the temperature
required for an appropriate metallic/silicateDV is too high, and is unrealistic
for the base of a magma ocean (Frost et al., 2008; Wade and Wood,
2005). However, it is still an open question whether the terrestrial accretion occurred under reducing conditions, where fO2 in the silicate
materials was coupled with iron oxide content due to oxidation in the
late stage of the accretion process (Frost et al., 2008; Tuff et al., 2011;
Wade and Wood, 2005; Wood, 2008), or under oxidizing conditions
where the core has higher oxygen content than previously believed
(Siebert et al., 2013).
2.2.3. Partitioning of vanadium between minerals and silicate melts
Mineral/melt
DV has been intensively investigated, with a special focus
on the role of fO2 in changing the valence and thereby the partitioning
of V at high temperature. Numerous experimental studies clearly
show that Mineral/meltDV are dominantly controlled by fO2. Oxygen fugacity controls valence states, coordination numbers, and incorporation
mechanism of V in both melts and minerals (Karner et al., 2006;
Mallmann and O’Neill, 2009; Righter et al., 2006; Righter et al., 2011;
Shearer et al., 2006a; Shearer et al., 2006b). Furthermore, the addition
of other high valence cations (such as Fe3+ and Ti4+) into the silicate
melt can also change the coordination environment of V, and thus its
partitioning between minerals and melt (Giuli et al., 2004). There is
still, however, a lack of studies that quantify the effect of Fe3 +, Ti4+,
and H2O on mineral/meltDV and on other single valent elements (such as
Sc and Ga). This may add uncertainty to interpreting the similarity of
V/Sc and V/Ga among the arc lavas, OIB, and MORB.
Under the P-T-fO2 conditions of the upper mantle, mineral/meltDV generally decreases in the order of spinel, amphibole, clinopyroxene, garnet, orthopyroxene, and olivine (Adam and Green, 2006; Canil, 1999;
Canil, 2004; Karner et al., 2006; Karner et al., 2008; Mallmann and
O’Neill, 2009; Righter et al., 2006; Righter et al., 2011). Olivine/meltDV
and orthopyroxene/meltDV decrease dramatically from N N 1 to b 0.01 with
increasing valence state from +2 to +5, while V(III) is more compatible
than V(II) and V(V) in clinopyroxene (Karner et al., 2008; Mallmann
and O’Neill, 2009). Because most minerals are solid solutions with
significant variation in chemical composition, crystal chemistry could
play an important role in controlling the valence state and thus the partition coefficients of V. Experimental studies on Martian basalt systems
show that augite/meltDV is greater than pigeonite/meltDV at the same fO2 because the higher Al and Na content in augite facilitate the coupled substitution of V(III) and V(IV) into the augite structure (Karner et al.,
2008). Nonetheless, it is increasingly accepted that V is compatible during mantle melting only in extremely reduced environments
(ΔlogfO2(FMQ)-3), while it is incompatible or lithophilic (rock-loving)
during magmatism at the crustal conditions, consistent with the lower
V content in mantle peridotites than in basalts (e.g., (Canil, 1999;
Canil, 2002; Lee, 2005; Mallmann and O’Neill, 2009). It is notable that
V is compatible during generation of the felsic melt, assuming a mafic
derivative of the upper continental crust as discussed earlier. This
could be a sequence of V partitioning between the dominant residual
phases (including garnet, clinopyroxene, and amphibole) and highly
polymerized melt.
2.3. Temporal and spatial variation of mantle fO2 constrained by vanadium
Temporal variation of the mantle fO2 from Archean to Cenozoic provides important constraints on Earth’s evolution and interactions
among the mantle, crust, hydrosphere, and atmosphere. The V content
of peridotites is commonly correlated with major elements such as
MgO and Al2O3, likely reflecting the influence of partial melting (Canil,
2002; Canil, 2004). The content of V in igneous rocks is generally correlated with magma differentiation indicators. Vanadium may be affected
by the fractional crystallization of mafic minerals because it is compatible in clinopyroxene with fO2 near the fayalite-magnetite-quartz buffer
(Canil, 2004; Karner et al., 2006). The ratios of V to single valent elements (such Sc, Y, Ga, and Ti) are, however, less sensitive to magmatic
differentiation or to the source heterogeneity of basalts. Such ratios
could, therefore, be used to filter out the effect of differentiation and
source to address the fO2 of the mantle source of mid-ocean ridge
basalts (MORBs), ocean island basalts (OIBs), and island arc basalts
(IABs) (Canil, 2002; Lee, 2005; Li and Lee, 2004; Mallmann and
O’Neill, 2009). A careful examination of different trends between V
and Al2O3 in peridotites from variable tectonic settings reveals that
abyssal peridotites were formed under a similar fO2 to the MORB, but
slightly lower than that of the spinel-facies Archean cratonic lithosphere
(Canil, 2002). Furthermore, Canil (2002) argued that the Archean mantle may have had similar fO2 to the modern mantle, and thus gradual oxygenation of the Earth’s atmosphere does not hinge on mantle melting.
The lack of temporal variation in fO2 for the mantle since the Archean is
also supported by the similar V/Sc between Archean basalts (up to
3.5 Ga) and modern MORB (6.34 ± 0.62 vs. 6.74 ± 1.11, 1σ) (Li and
Lee, 2004).
Comparing V/Sc between mantle sources from different tectonic settings also provides a tool to address spatial variation of the mantle fO2
(Lee, 2005; Li and Lee, 2004; Mallmann and O’Neill, 2009). Based on
the observation that arc basalts have V/Sc that is indistinguishable
from MORB (Lee, 2005; Li and Lee, 2004), some authors suggest that
the fO2 of the mantle wedge is at the same level as the MORB source
(~FMQ). This conclusion seems to contradict the observation that the
primitive basalts and spinel peridotites from variable arcs have slightly
higher Fe3 +/Fe2 + and thus higher fO2 than MORB and its mantle
sources, respectively (Ballhaus, 1993; Kelley and Cottrell, 2009;
Macdonald et al., 2000; Parkinson et al., 2007). Studies on the V/Sc in
the ophiolite peridotites from Alaska, Yukon and British Columbia also
suggest a narrow fO2 range between NNO and NNO-1 during melting,
slightly higher than the MORB mantle (Canil et al., 2006). More complications are brought by almost identical Zn/Fe between MORB and primitive arc lavas (Lee et al., 2010), and slightly lower δ56Fe in arc lavas than
in MORB (Dauphas et al., 2009); this does not support the higher fO2 in
the source of arc lavas relative to the MORB source if both have similar
initial Zn/Fe and δ56Fe. Reconciling such discrepancies is challenging.
J.-H. Huang et al. / Chemical Geology 417 (2015) 68–89
Nonetheless, future experimental studies on the partitioning of V between mantle minerals and hydrous melt under variable fO2 and P-T
conditions similar to the arc systems would be helpful to address the
implications of V and other elemental data in MORB and arc lavas.
2.4. Vanadium isotopes
50
V and 51V are the two most common stable isotopes of vanadium,
with atomic abundances of 0.25% and 99.75%, respectively. Pioneering
studies on vanadium isotopic composition in meteorites and Earth samples, using an Atlas CH-4 mass spectrometer, indicate that meteoritic
and terrestrial materials have similar 50V/51V, with an error of 1%; this
suggests that there is no significant difference in irradiation histories
during early accretion and evolution of the solar system (Balsiger
et al., 1969; Balsiger et al., 1976; Pelly et al., 1970). Such an error does
not allow V isotopes to be regularly applied in geochemical studies.
The main obstacles to measuring V isotopes are the weak signal of 50V
relative to 51V in geological samples, and chemical separation of V
from matrix elements such as Cr and Ti. The use of multi-collector inductively coupled plasma-mass spectrometry and optimized chemical
separation procedure dramatically improved the external precision of
V isotope analysis of terrestrial samples by two orders of magnitude,
i.e. ~ 0.10‰ (2σ) (Nielsen et al., 2011; Prytulak et al., 2011). Taking
VISSOX (Vanadium Isotope Standard Solution Oxford) as the international V isotope standard, the Oxford Group observed substantial isotope fractionation in meteorites and in terrestrial materials. For
example, δ51V of the Allende chondrite is − 1.78‰, whereas USGS
whole rock standards range from −0.55 (AGV-2) to −1.04‰ (BHVO2) (Nielsen et al., 2011; Prytulak et al., 2011). Further studies revealed
V isotope fractionation in more terrestrial and meteorite samples
(Nielsen et al., 2014; Prytulak et al., 2011; Prytulak et al., 2013).
Recently, first-principles calculations predicted that V isotopes can be
substantially fractionated among V species with different valences in
aqueous systems and during adsorption of V(V) to goethite (Wu et al.,
2015). These studies indicate a promising future for the application of
V isotopes to a variety of fundamental questions, from high temperature magmatism to low temperature environmental sciences.
3. Occurrence of vanadium in minerals
Minerals link V circulating in the Earth’s interior to weathering and
biogeochemical processes at the surface. Vanadium typically occurs as
an accessory component of minerals. This can be attributed to the relatively low concentration of V in the Earth’s crust, and the tendency of V
to disperse due to its sensitive response to acidity and oxygen. The size
and charge of V species enable it to substitute for common transition
elements, e.g., Fe and Al in primary and secondary minerals. Vanadium
mineral classes, encompassing primary and secondary minerals, include
oxides and hydroxides, silicates, and relatively rare sulfides; the coordination of V in minerals can be tetrahedral, pyramidal, or octahedral.
There is relatively little information available on the diverse V minerals,
which likely reflects their rarity, in spite of the ubiquitous presence of V
in geological materials. The review by Evans and White (1987) gives an
excellent introduction to “the colorful vanadium minerals”. In the
following section, we focus on minerals that contain V as a significant
constituent, including important primary and secondary ore minerals.
3.1. Vanadium associated with primary rocks and minerals
Because V readily substitutes for Fe in minerals, it is more abundant
in mafic than in felsic rocks. Trivalent cationic V is a common lattice
component of primary minerals (Hurlbut and Klein, 1977). Typical concentrations in basalts and gabbros range from 200 to 300 mg kg−1, compared to concentrations in granites from 5 to 80 mg kg− 1 (Nriagu,
1998). Titaniferous magnetite is the principle source of mined V,
which is produced primarily as a byproduct of iron and titanium mining.
73
As mentioned, V occurs in magnetite mostly as V(III), with minor V(IV)
occupying octahedral sites (Balan et al., 2006). Titaniferous magnetite is
commonly associated with layered mafic intrusions such as the
Bushveld Igneous Complex in South Africa, one of the largest mafic intrusions in the world. The Panzhihua vanadium-titanium-iron ore
deposit in Southwest China is hosted as layers or lenses in a gabbroic
layered intrusion in the Emeishan Large Igneous Province (Zhong and
Zhu, 2006; Zhou, 2005), and large deposits of V occur in the titaniferous
magnetite deposits of Quebec (Kish, 1972). There is still debate about
the formation of the ore deposits associated with mafic intrusions. It is
unclear whether the V-rich magnetite was formed from immiscible
Fe-Ti oxide liquid at the late-stage of tholeiitic magma differentiation
(e.g., (Reynolds, 1985; Ripley et al., 1998; Zhou, 2005) or whether the
titano-magnetite accumulation was formed via crystallization as
a liquidus phase (Pang et al., 2008). The predominance of V(III)
in vanadiferous titanomagnetite, with up to 10% V(IV), has been
established using X-ray absorption spectroscopy (XANES) (Table 1).
Montroseite [VO(OH)] is a primary V(III) ore mineral first observed in
the western U.S. (Weeks et al., 1950), where it is the likely source of V
for associated oxidized minerals (Evans and Garrels, 1958; Weeks,
1961), discussed in the next section.
Vanadate [V(V)] ore minerals are associated with low temperature, non-sulfidic mineralization in parts of southern Africa, including Namibia (Boni et al., 2007), Zambia (Pelletier, 1930) and
Angola (Millman, 1960). The vanadate ore deposits located in the
Otavi Mountainland, Namibia occur in fractures within host dolomite and were formed by precipitation from vanadate-bearing solutions circulating through karst bodies (Boni et al., 2007). Minerals
include mottramite [PbCu(VO4)(OH)], descloizite [(Pb, Zn)2VO4(OH)]
and vanadinite [Pb5(VO4)3Cl]. These relatively rare minerals are
thought to have formed during the mid-Miocene, when the climate became drier and chemical weathering was subsequently limited in
southern Africa (Boni et al., 2007). The deposits, now mined out, once
contained several million tons of extractable V ore.
3.2. Weathering and diagenesis of secondary vanadium in minerals and
fossil fuels
Weathering, transport, sedimentation, and diagenesis processes
mobilize V from high temperature magmatic and metamorphic rocks
to form low temperature sedimentary rocks. The diverse mineralogy
and wide range of environments where V-bearing secondary minerals
occur reflect the complexity of V chemistry (Clark, 1993; Fleischer,
1987).
The redox weathering of primary reduced V and U minerals
(e.g., montroseite, uraninite) has led to extensive tabular and roll
front deposits of uranyl vanadate minerals such as carnotite
(K2(UO2)2(VO4)2·3H2O) and related carnotite group minerals such
as tyuyamunite (Ca-variant) in the Colorado Plateau (Weeks,
1961). Cycles of reduction and oxidation fueled by organic matter
likely transformed the source minerals over time into the diverse mineral assemblages that currently exist (Hansley and Spirakis, 1992;
Spirakis, 1996). Minerals in the Colorado Plateau encompass V in +3,
+ 4 and + 5 valence states and reflect a wide range of reduction
potentials and pH values during formation (Evans and Garrels, 1958).
Where oxygen exposure was low and the pH was neutral, partly
oxidized minerals such as duttonite [VO(OH)2 ] and simplotite
[Ca(VO)2 (VO4 ) 2 ∙ 5H2 O] resulted (Roach and Thompson, 1959).
More intensive weathering that leads to higher acidity and oxidizing
conditions resulted in minerals such as carnotite and hewettite
[Ca(V6O16)∙6H2O] (Elston and Botinelly, 1959; Roach and Thompson,
1959). The role of microorganisms in forming these V mineral deposits
is discussed in Section 4.2.
Vanadium incorporates in the structure of secondary minerals such
as clays (micas, smectites, illite, kaolinite, and chlorite), Fe-Mn hydrous
oxides, and organic rich facies during diagenesis, metamorphism, and
74
J.-H. Huang et al. / Chemical Geology 417 (2015) 68–89
Table 1
Speciation of vanadium in different compartments of the surface environment.
Compartment
Vanadium speciation
Remarks
References
Coastal waters, USA
V(IV): up to 0.56 µg L−1; V(V): up to 1.68 µg L−1,
Wang and Sañudo
Wilhelmy (2009b)
Lake Sediment, Venezuela
Fish muscle
Mussel
Coal, Kentucky, USA
Red mud leachate, Hungary
Vanadiferous titanomagnetite
Magnetite, South Africa
Hematite, India
Soil extracts, Belgium,
Denmark, Sweden
Oil shale
Sedimentary limestone,
Coke waters;
Peat pond; Alberta, Canada
Mineral water,
Lake water, Yangzhou, China
V(IV): 0.4–25.8 mg kg−1; V(V): 1.4–9.2 mg kg−1,
V(IV): 0.92 mg kg−1
V(IV): 1.52 mg kg−1
V(III) and V(IV) in oxygen coordination
V(V) detected
V(III) predominance with up to 10% of V(IV)
V(III) predominance with up to 9–16% of V(IV)
V(III) and V(IV) (41–72%)
V(IV): 0.06–0.14 mg L−1; V(V): 0.6–3.0 mg L−1 (n = 3)
High V(IV) and V(V) concentrations in summer;
low in spring
Separation with Chelex 100 resin
Oil production with 10,000 platforms
Speciation with LC-ICP-MS
Soils
V(IV), 46.8 μg g−1; V(V) 26 mg kg−1 (Pintung, Taiwan)
V(IV), 49.2 mg kg−1; V(V) 51.8 mg kg−1 (SRM 2709)
V(IV): 12.2 mg L−1; V(V), 16.5 mg L−1 (n = 1)
V(IV): bDL; V(V): 2.53 μg L−1 (n = 1)
V(IV): 520–7120 mg kg−1; V(V): 11–42 mg kg−1 (n = 4)
Oil refining leachate, Taiwan
Lake water, Wuhan, China
Soils, South Africa
Sea water
River water
Mineral water
Lake water, Ji Lin, China
Particulates from diesel
engine
Urban aerosols, USA
Seawater, Galician, Spain
Seawater, China
River water,
Lakewater,
Well water,
Rain water,
Prophyrin associated V(IV)
9% and 19% of V as prophyrin associated V(IV) (n = 2)
V(IV): 48–1090 μg L−1; V(V): 926–5137 μg L−1 (n = 2)
V(IV): 105 μg L−1; V(V): 30 μg L−1 (n = 1)
V(IV): bDL-3.19 μg L−1; V(V): 5.07–45 μg L−1 (n = 10)
V(IV): undetectable; V(V) 4.36 μg L−1 (n = 1)
V(IV): 0.61–0.90 μg L−1; V(V) 0.50–0.86 μg L−1 (NASS-4
and CASS-3) (n = 2)
V(IV): 0.38 μg L−1; V(V): bDL (SLRS-3) (n = 1)
V(IV): 0.38 μg L−1; V(V): 0.60 μg L−1 (n = 1)
V(IV): 0.25 μg L−1; V(V): 1.0 μg L−1 (n = 1)
V(IV): 10.2–20.1 ng mile−1; V(V):
2.3–109 ng mile−1 (n = 4)
V(II), V(III), V(IV) and V(V) identified with XANES
V(IV): ca. 0.15–4.5 mg kg−1; V(V):
ca. 0.4–7.7 mg kg−1 (n = 10)
V(IV): bDL; V(V): up to 0.25 μg L−1
V(IV): 0.5–1.6 μg L−1; V(V): 0.5–1.8 μg L−1 (n = 3)
V(IV): 1.0 μg L−1; V(V): 1.2 μg L−1 (n = 1)
V(IV): 1.7–3.4 μg L−1; V(V): 1.2–1.3 μg L−1 (n = 2)
V(IV): 0.4–11.7 μg L−1; V(V): 1.4–4.2 μg L−1 (n = 2)
V(IV): 0.2–0.4 μg L−1; V(V): 1.5–1.6 μg L−1 (n = 2)
Colina et al. (2005)
Speciation by EXAFS
Speciation by EXAFS
Speciation by XANES
Speciation by HERFD-XAS
Maylotte et al. (1981)
Burke et al. (2012)
Balan et al. (2006)
Bordage et al. (2011)
Extraction with 10 mM CaCl2
Speciation by HPLC-ICP-MS
Determined by XRD
Determined by ESR
Speciation by HPLC-ICP-MS
Baken et al. (2012)
Speciation by HPLC-ICP-MS
Speciation by GFASS after ion-exchange resin
separation
Speciation by HPLC-ICP-MS
Aureli et al. (2008)
Zhao et al. (2006)
UV detection after LC separation
Speciation by FETV-ICP-OES
AAS determination after NaCO3 extraction
Speciation by HPLC-ICP-MS
ICP-MS detection after anion-exchange membrane
separation
Acetate soluble fraction
Ekstrom et al. (1983)
Premović et al. (1986)
Li et al. (2007)
Kuo et al. (2007)
Jen and Yang (1994)
Wu et al. (2005)
Mandiwana and Panichev
(2004)
Wann and Jiang (1997)
Jia et al. (2012)
Shafer et al. (2012)
Acetate soluble fraction
ICP-MS detection after Chelex resin separation and
XANES
AAS determination after Chelex 100 resin separation Alcalde-Isorna et al. (2011)
ICP-OES determination after CTAB- modified alkyl
Xiong et al. (2010)
silica micro-column
N: number of samples measured; bDL: below detection limit.
Average dissolved ocean V = 1.8–2.3 μg L−1 (reviewed in (Emerson and Huested, 1991)).
hydrous thermal processes (Algeo and Maynard, 2008; Brumsack, 2006;
Condie, 1993; Morford et al., 2005). In clay minerals, V(IV) substitutes
for Al(III) in the octahedral sheet as VO2+ (Gehring et al., 1993). Structural similarities between Fe and V oxides have suggested the substitution of V(III) for Fe(III) in Fe-oxides (Taylor and Giles, 1970), and it was
shown subsequently that V can replace octahedrally coordinated Fe(III)
in the diaspore structure of goethite (Schwertmann and Pfab, 1994;
Schwertmann and Pfab, 1996). The incorporation of reduced V species
(III, IV) in clays and other secondary minerals provides a useful redox
indicator for past anoxic conditions (Gehring et al., 1994; Gehring
et al., 1999; Lebedel et al., 2013).
Crude oil can contain high concentrations of V due to the sorption or
uptake of V by algae, plankton and plants and incorporation into porphyrin groups (e.g., chlorophyll). The organic matter settles in basins
where degradation takes place, causing anoxic conditions to develop.
Decomposition of organic matter and diagenetic processes tend to concentrate trace elements in these materials, leading to an enrichment of
V that can be diagnostic for the depositional environment and facilitates
identification of the source rock (Barwise, 1990).
In oil, V tends to concentrate in the heavy crude fractions, an association that is enhanced in the presence of reduced sulfur species
(Punanova, 2014). Vanadium is commonly one of the most abundant
metal elements in crude oil deposits (up to around 1200 mg L−1)
(López and Lo Mónaco, 2004). Asphaltene in crude oil contains up to
5000 mg L− 1 V (López et al., 1995). Barwise (1990) developed a
classification scheme which relates the Ni/V ratio to the depositional
environment; Ni is also an abundant metal in petroleum. The environments identified by Barwise (1990) for organic matter accumulation
leading to oil genesis are marine (Ni/V b 0.5) and lacustrine or terrestrial
(Ni/V 1–10); however, the typically poor preservation of the organic
groups that contain Ni and V in terrestrial deposits generally excludes
these materials from the classification (Barwise, 1990). Classification
by Ni/V ratios has been used widely to distinguish sources of crude oil
between marine and terrestrial environments, most recently in crude
deposits in Nigeria, Russia, Venezuela and South Africa (Akinlua et al.,
2015; Lopez et al., 2015; Ogunlaja et al., 2014; Yasnygina et al., 2015).
Vanadium in coal is generally considered a trace element, due to
concentrations that are typically less than 100 mg kg− 1 (Sia and
Abdullah, 2011). Coal may contain up to 70% (by volume) of noncarbonaceous materials, which are often aluminosilicate minerals such
as clays, or carbonates. Vanadium appears to be located preferentially
in association with clay minerals (Chen et al., 2011; Huggins et al.,
2009; Sia and Abdullah, 2011; Wang et al., 2008). For example, in one
study of coal (Illinois #6), the estimated V content of associated illite
was 420 mg kg− 1, whereas the organic fraction contained around
20 mg kg−1 V (Huggins et al., 2009).
Shales containing high amounts of organic matter, known as black or
carbonaceous shales, are typically enriched in V relative to “average”
J.-H. Huang et al. / Chemical Geology 417 (2015) 68–89
shale as well as containing relatively high total sulfur (Lipinski et al.,
2003). Carbonaceous shales can contain up to 16,000 mg kg-1
(Michibata et al., 1991; Vine and Tourtelo, 1970) compared to noncarbonaceous shales that average 130 mg kg -1 V (Turekian and
Wedepohl, 1961). As observed for coal, V in shale associates preferentially with the clay minerals (e.g., Peacor et al. (2000) and Lipinski
et al. (2003)). In one recent study, soils developed on V-enriched
black shale were shown to retain the signature of high V concentration,
and V in the soil was distributed more homogeneously than in the parent rock; the fresh shale had V concentrations ranging from 225 to
5162 mg kg− 1, and the soils from 187 to 442 mg kg− 1 (Xu et al.,
2013). The difference between the unweathered and weathered material was attributed to a two-stage differentiation process: 1) chemical
differentiation, largely by leaching of more mobile chemical components, and 2) chemical homogenization, which includes hydrolysis of
resistant silicate minerals as well as continued preferential leaching of
some elements (Xu et al., 2013).
4. Cycling and transport of vanadium in surface environments
Although many minerals can be enriched in V, the processes that
control the release and mobilization of V in weathering environments
have been much less studied as compared to other trace element
e.g. As and Hg (Huang, 2014; Selin, 2009; Smedley and Kinniburgh,
2002). The diverse pools of V found in Earth surface environments
reflect conditions of redox, pH, concentration and complexation that
shed light on geochemical and biogeochemical processes that link the
reservoirs of V (Fig. 4, Table 1). The following sections address V chemistry and biochemistry in the context of Earth surface environments.
4.1. Aqueous speciation chemistry of vanadium
The chemical behavior of V in water controls the biogeochemical
cycling of V in surface environments. Vanadium speciation is a complex
75
function of pH, Eh, concentration, solution chemistry and, in many
surface environments, biology. Vanadium exists in the + 3, + 4 and
+5 oxidation states in natural waters. Vanadate species [V(V)] are thermodynamically stable in oxic conditions, while V(IV) is stable under
suboxic conditions, and V(III) is found in anoxic environments (Sadiq,
1988); redox chemistry is discussed in more detail in Section 4.4.3.
Aqueous V species in natural systems occur mainly as V(V) and V(IV),
as shown using different chemical and spectroscopic speciation methods
(Table 1; Alcalde-Isorna et al., 2011; Aureli et al., 2008; Colina et al., 2005;
Jia et al., 2012; Li et al., 2007; Wang and Sañudo Wilhelmy, 2009b; Wann
and Jiang, 1997; Wu et al., 2005; Xiong et al., 2010; Zhao et al., 2006).
Vanadium speciation is a strong function of concentration and pH.
Five V(V) species have been reported: the pervanadyl ion, VO+
2 , vanadic
2−
acid, H3VO04 and its three conjugate bases, H2VO−
and VO34 −
4 , HVO4
(Table 2). Unless otherwise indicated, all the thermodynamic data
used in the following Tables are derived from the NIST data base
(Martell and Smith, 2004) and recalculated, if necessary, to zero ionic
strength (Verweij, 2013). The mass action and mass balance equations
used to generate the Figures from these Tables were solved using the
computer program MICROQL (Westall, 1979), rewritten in Visual Basic
Version 6.1 by one of the authors (L.E.).
In dilute solutions at neutral pH, the dominant species of V(V) are
the phosphate-like mononuclear vanadate oxyanions Hn VO n4 − 3
(Wanty and Goldhaber, 1992; Wehrli and Stumm, 1989). The species
3−
VO+
occurs in extremely
2 dominates under acidic conditions, and VO4
alkaline conditions (Fig. 5).
Vanadic acid, H3VO04, is a minor constituent within the pH range 2–5.
Vanadium has a notable tendency to polymerize, forming species containing up to 10 atoms of V (Wanty and Goldhaber, 1992). Polynuclear
species have enhanced stability, as shown using reaction enthalpy for V4
and V5 cyclic species (McCann et al., 2013). Eleven soluble poly-nuclear
species have been reported by Petterson et al. (1983) and fourteen are
listed in Martell and Smith (2004) containing 2, 3, 4, 5, 6 or 10 atoms
of vanadium (Table 2). The speciation of V(V) at 0.5 mM concentration
Fig. 4. Schematic presentation of vanadium biogeochemical cycling in the surface environment.
76
J.-H. Huang et al. / Chemical Geology 417 (2015) 68–89
Table 2
Dissociation constants, log K, for mono-nuclear vanadium(V) species (calculated from the
Gibb’s energy values given in Wanty and Goldhaber (1992), 25 °C, 1 atm. Pressure, zero
ionic strength (I).
log K
VO+
2
+ 2H2O ⇋ H3VOo4 +
+
H3VOo4 ⇋ H2VO−
4 +H
H
+
−3.67
−3.40
log K
H2VO−
4
HVO2−
4
⇋
⇋
HVO2−
+ H+
4
VO3−
+ H+
4
−8.06
−13.28
is illustrated using the formation constants given in Elvingson et al.
(1996) (Table 3, Fig. 6).
Vanadium(IV) is prevalent in reduced groundwater (Bosque-Sendra
et al., 1998; Hirayama et al., 1992; Nakano et al., 1990) and in water
influenced by volcanic (Minelli et al., 2000) and industrial activities
(Banerjee et al., 2003). Under moderately reducing and acidic conditions, V(IV) occurs as the vanadyl cation, VO2 + or VO(OH)+ (Table 4,
see also Fig.7). The aqueous chemistry of V(IV) is strongly influenced
by its associated complexes with inorganic and organic cations
(Wanty and Goldhaber, 1992) (Table 4), which is further discussed
in Section 4.1.2.
Vanadium(III) is thermodynamically stable over a wide range of
pH values in anoxic environments (Wright and Belitz, 2010). It
hydrolyzes rapidly in aqueous solution as VOH2 +, V(OH)+
2 and
V2(OH)4+
(Pajdowski, 1966; Pajdowski and Jeżowska-Trzebiatowska,
2
1966) and readily precipitates as insoluble V(III) oxyhydroxides
(Wanty and Goldhaber, 1992). Because of its sensitivity to oxygen,
V(III) species are relatively rare in surface and near-surface waters
(Aureli et al., 2008; Wällstedt et al., 2010; Wang and Sañudo Wilhelmy,
2009b) (Table 1).
Table 3
Poly-nuclear soluble vanadium(V) species.
H2V2O2−
7
HV2O3−
7
HV2O4−
7
V3O5−
10
V4O4−
12
HV4O5−
13
V4O6−
13
V5O5−
15
V6O6−
18
H3V10O3−
28
H2V10O4−
28
HV10O5−
28
V10O6−
28
been suggested as the essential intermediates for the reduction of
V(V) (Kustin et al., 1974; Wells and Kuritsyn, 1970), and V is known
to be stabilized against reduction by binding to humic substance at
circumneutral pH values (Lu et al., 1998). The strong tendency of
V(IV) to form stable complexes helps explain the presence of V(IV) in
natural waters under toxic conditions (Wehrli, 1987). Aluminum and titanium oxide surfaces have been shown to catalyze V(IV) oxidation
(Wang and Sañudo Wilhelmy, 2009b; Wehrli and Stumm, 1988;
Wehrli and Stumm, 1989), and complexation with organic molecules
may serve to inhibit oxidation.
4.1.1. 4.1.2. The role of complexation in V mobilization
The complexation of vanadium by organic and inorganic ligands can
elevate its mobility at solid-water interfaces (Wanty and Goldhaber,
1992). Vanadium(IV) forms complexes more readily than does
V(V) and with a wider range of organic and inorganic ligands. Common
complexes formed by V with inorganic ligands include H4VO4(C2O4)3−
2
2−
+
+
and H4VO4C2O−
VOF2, VOF−
4 for V(V); VOCl , VOF
3 , VOF4 , VOSO4,
VO(C2O4)22 −, VO(OH)C2O−
4 , VO(CH3COO)2, VO(CH3COO), VOCO3,
+
VO(OH)CO−
3 for V(IV); and VSO4 for V(III) (Wanty and Goldhaber,
1992).
A large number of organic compounds can complex V, including
fulvic and humic acids, via oxygen groups (Cheshire et al., 1977;
Wehrli, 1987). Table 5 provides an overview of some studies that have
investigated the complexation of V(IV) and V(V) by natural organic
matter. It is known that vanadate oxyanions form relatively weak complexes with organic compounds, as compared to the V(IV) oxycation
VO2 + (Wehrli, 1987). Nevertheless, V(V)-organic complexes have
4.1.2. Vanadium redox transformation in surface environments
V(IV) and V(III) are generated via biotic and abiotic reduction of
V(V) (Bautista and Alexande, 1972; Bredberg et al., 2004; Carpentier
et al., 2003; Li et al., 2009; Wanty and Goldhaber, 1992). Microbial reduction is discussed in the next section.
An Eh–pH diagram illustrates the distribution of vanadium’s three
common oxidation states, V(III), V(IV) and V(V), constructed using the
thermodynamic data given in Wanty and Goldhaber (1992) (Fig. 7).
The oxidation–reduction reactions and the calculated standard electrode potential, Eo, are given in Table 6.
It is apparent that V(V) species predominate in oxic environments at
low pH values, and in both oxic and anoxic environments at pH values
greater than about 8 (Fig. 7). Francavilla and Chasteen (1975) observed
that VO+
2 concentrations in solution decreased with increasing pH from
2 to 6 because of the formation of (VOOH)2+
2 and VO(OH)2 species. The
6−
increased pH may, however, cause condensation of VO+
2 to V10O28 ,
which is not directly involved in the reduction reaction (Lu et al.,
1998). The vanadyl ion, VO2+, and its hydrolyzed species, VOOH+, do
not exist at pH values above about 7 if not complexed with inorganic
and organic ligands; they are stable below this value in environments
that are oxic to anoxic. Vanadium(III) species in solution exist only
under anoxic conditions and at pH values up to about 10, where they become unstable relative to V(V) species.
The oxyhydroxide of V(IV), VO(OH)2(s) and the hydroxide of V(III),
V(OH)3(s), can form at relatively low concentrations at 25 °C (Fig. 8).
The oxidation–reduction reactions for V which include solids is shown
in Table 7. The precipitation of V in soils and sediments is discussed in
Section 4.3.3.
Fig. 5. Speciation of vanadium(V) at low concentrations using the thermodynamic data
given in Table 2.
Fig. 6. Speciation of vanadium(V) at 0.5 mM concentration (thermodynamic constants are
from Elvingson et al. (1996), 25 °C, 1 atm. Pressure, I = 0.15 M NaCl).
J.-H. Huang et al. / Chemical Geology 417 (2015) 68–89
77
Table 4
Formation constants, log β, for vanadium(IV) with some inorganic ligands (calculated from the Gibb’s energy values provided in Wanty and Goldhaber (1992), 25 °C, 1 atm. Pressure, zero
ionic strength (I).
Vanadium(IV)
log β
VO2+ + H2O ⇋ VO(OH)+ + H+
2VO2+ + 2H2O ⇋ (VO(OH))2+
+ 2 H+
2
VO2+ + Cl− ⇋ VOCl+
−5.7
−6.7
−0.1
Organic compounds such as humic, fulvic, and ascorbic acids can
reduce V(V) to V(IV), and dissolved sulfide will reduce V(IV) to V(III)
under conditions of pH and ionic strength common in natural systems
(Table 8).
Because V solubility depends strongly on oxygen presence, the redox
potential can control V mobility in surface environments (Fig. 7), which
has been demonstrated in studies of soil, saturated sediments and rock
diagenesis (Breit and Wanty, 1991; Francois, 1988; Yang et al., 2014).
The V(IV) and V(III) species generated by low reduction potential will
associate rapidly with the particulate fraction via complexation or
adsorption, formation of insoluble hydroxides, and incorporation into
minerals (Francois, 1988; Szalay and Szilágyi, 1967). By extension,
the oxidation of V(III) and V(IV) to V(V) during weathering of Vcontaining minerals is a major cause of V mobilization in soil and sediment (Yang et al., 2014). There is evidence that reduced V is subsequently complexed by naturally occurring organic substances (Ferrer
and Baran, 2001); this will decrease V solubility if the organic matter
is insoluble. For example, complexation of V(IV) by high molecular
weight fulvic acids can lead to aggregation (Templeton and Chasteen,
1980), and the reduction of V(V) by humic substances with consequent
complexation by organic matter is a probable source of V associated
with fossil plant detris and bioliths (Szalay and Szilágyi, 1967).
Reduced V(IV) and V(III) can hydrolyze and form insoluble hydroxides
(Carpentier et al., 2003; Li et al., 2007; Ortiz-Bernad et al., 2004), which
could also contribute to V that is preserved in fossils. Reduced sediments are generally considered sinks for V. Suboxic conditions may,
however, result in the release of V into the solution phase due to the
reductive dissolution of adsorbing phases, e.g. Mn and Fe (hydr)oxides
(Fox and Doner, 2003). For example, the mobilization of V associated
with saline sediments in evaporation ponds was interpreted as a complex function of the reductive dissolution of Mn and V oxides in the
presence of organic matter (Amrhein et al., 1993).
4.2. Microbial controls on V geochemistry: redox, complexation, and
sorption reactions
The flexible redox chemistry of V and its role as an essential trace
element suggest that it may have once played a larger role in the chemistry of life (Rehder, 2008a; Rehder, 2008b). Some microorganisms can
Fig. 7. Eh–pH predominance diagram for soluble species of vanadium.
log β
VO2+ + SO2−
⇋ VOSOo4
4
VO2+ + CO2−
⇋ VOCOo3
3
+
VO2+ + CO2−
+ H2O ⇋ VO(OH)CO−
3
3 +H
2.46
3.47
1.52
respire using V(V) and even V(IV) as electron acceptors (Carpentier
et al., 2003; Li et al., 2009; Li et al., 2007; Lyalikova and Yurkova, 1992;
Ortiz-Bernad et al., 2004; Zhang et al., 2014). The relatively low toxicity
of vanadium and the high reduction potential for the V(V)/V(IV) redox
couple suggests that V will be used in preference to other electron
acceptors, such as Fe(III). The association of V with microbes via sorption and uptake processes may cause vanadium to concentrate in sediments, which is speculated to be a major source of this element in
association with fossil fuels (Breit, 1988; Lewan, 1984). In the following,
we highlight several studies to provide background on V-microbe interactions. For excellent discussions of V bioinorganic chemistry, see
Chasteen (1983) and Rehder (2008a).
Many microorganisms are highly resistant to V. For example, Pseudomonas and Thiobacilli were shown to tolerate V(V) at 5000 mg L−1
(Lyalikova and Yurkova, 1992). A high tolerance to V is surprising
given the several common valence states of V, which could be expected
to disrupt redox reactions (Bell et al., 2004). Soils that have been exposed to crude oil and refined oil products harbor microorganisms
that are particularly tolerant to vanadium, which is perhaps not surprising given reported concentrations of around 1200 mg kg−1 V in crude
oil (Crans et al., 1998). For example, bacteria isolated from soils at an
oil refinery tolerated up to 510 mg kg−1 vanadyl sulfate, and accumulated around 35,200 mg kg− 1 V (bacterial dry weight) (Hernández
et al., 1998). A similar high tolerance was observed for bacteria isolated
from soils contaminated with crude oil in Russia and Saudi Arabia (Bell
et al., 2004). Tolerance to V can be widespread in a microbial community. For example, bacteria present in a water sample taken from a South
African mine containing around 420 mg L−1 V(V) at pH 3 encompassed
17 phylogentic orders (Kamika and Momba, 2014); one isolate was found
to tolerate up to around 700 mg L−1 V(V) at pH 7; tolerance was slightly
lower at pH 3.
The relatively low toxicity of vanadium and its favorable redox
chemistry make it a predictable electron acceptor for bacterial respiration. A number of microorganisms, including Pseudomonas vanadiumreductans T-1, Pseudomonas isachenkovii A-1, Geobacter metallireducens,
Enterobacter cloacea and species of Shewanellae have demonstrated respiration of soluble V(V) to produce V(IV) in pH-neutral conditions
(Carpentier et al., 2003; Li et al., 2007; Lyalikova and Yurkova, 1992;
Ortiz-Bernad et al., 2004; van Marwijk et al., 2009; Zhang et al., 2014).
Vanadium(III) has been detected during the dissimilatory reduction of
V(V) by Shewanella putrefaciens CN-32, indicating highly anoxic
conditions (Li et al., 2009; Li et al., 2007). Because of the relationship
between V oxidation state and solubility, dissimilatory V respiration
has been suggested as a strategy to remove V from contaminated
water (Carpentier et al., 2003; Ortiz-Bernad et al., 2004). For example,
acetate injected into an aquifer contaminated with U and V enhanced
the growth and activity of Geobacteraceae, and V was effectively
removed from the groundwater (Ortiz-Bernad et al., 2004). Removal
of V(V) from groundwater by bioreduction was suggested in another
study, which demonstrated stimulation of bacteria that reduced V(V) in
columns of aquifer sediment that were exposed to V-containing water
(Yelton et al., 2013). These studies were focused on bacteria, but archaea
were recently shown to also be capable of V reduction. Two methanogenic archaea, one mesophilic and one thermophilic, reduced V(V) to
V(IV) under growth conditions only, at concentrations up to 10 mM.
Vanadium reduction occurred at the expense of methanogenesis and
resulted in extracellular precipitates of V(IV) (Zhang et al., 2014). In
78
J.-H. Huang et al. / Chemical Geology 417 (2015) 68–89
Table 5
Vanadium complexation with natural organic matter.
Organic
Substance
Vanadium
Species
Interaction
Comments
References
DOC
V(V)
Humic substance-V(V) complexes less than 1% of total V
Canada (2010)
DOM
V(V)
V complexation with DOM is not important
Humic substance
V(V)
Fulvic acid
Peat humic acid
V(IV)
V(IV)
V(IV)
Soil fulvic acid
V(IV)
Soil organic
matter
V(IV)
V(V) monomer complexation with humic substance at pH 7.2 with
formation equilibrium constant of 108 M−1.
V(IV) bound to carboxylic groups of humic substance
Partially complexed at pH 2, predominantly complexed at pH 4 and 6
Complexed by oxygen or mixed oxygen-nitrogen donor groups in the
humic acid.
Complexation
Aggregation as (VO)2(heavier fractioned fluvic acid)6
Association of 4.2% of total soil V with humic and fluvic acids
DOC: N10 mg L−1, pH: 8, Wabamun
Lake, Canada
DOC: N10 mg L−1, pH: 7, Allard river,
Canada
DOC: N10 mg L−1, pH: 6.5,
Colombiere river, Canada
DOC: b1 mg L−1, pH: 8, St. Lawrence
Gulf, Canada
V: 2.0–2.5 μg kg−1; pH: 7.72
Mississippi river, USA
V(V): 12.8–166 mg L−1, Humic
substance: 0.025–0.58 mM, 25 °C
–
V: 0.51 g L−1
V:12.8–28 µg L−1; pH 5; fulvic acid:
2.5 × 10−3 M
Soil obtained from Scotland
Shiller and Boyle (1987)
Lu et al. (1998)
Wilson and Weber (1979)
Goodman and Cheshire
(1975)
Templeton and Chasteen
(1980)
Cheshire et al. (1977)
-: no information
summary, V reduction has been demonstrated for a wide phylogenetic
range of microorganisms, typically results in extracellular precipitates
of V(IV), and may or may not require growth conditions.
Although the bioreduction of V appears to lead to V immobilization,
the activity of microorganisms may also mobilize V from solid phases.
Sulfur oxidizing Thiobacillus sp. facilitated the leaching of V from soils
and minerals via oxidation of sulfide minerals (Briand et al., 1999;
Gomez and Bosecker, 1999). Neutrophilic bacteria, including species
of Pseudomonas and its relative Shewanella, mobilized V from shale at
a former uranium mine, which was attributed to the production of
chelating agents (Kalinowski et al., 2004). Shewanella putrefaciens
CN-32 exposed to carnotite under reducing conditions respired
mineral-bound V(V) in preference to U(VI), resulting in higher solution
concentrations of V (Glasauer, 2015; Li et al., 2009). It is an interesting
speculation that the dissimilatory reduction of V and U by microbes
may have contributed to forming the roll front and tabular ore-grade
deposits of carnotite and associated minerals in the Colorado Plateau.
The mobility of co-occurring V and U is highest under weakly reducing
conditions as V(IV) and U(VI). Fully oxidized V and U will co-precipitate
as carnotite minerals, whereas reduced species will precipitate in separate mineral phases. Organic matter in association with V and U species
may have created cycles of oxidizing and reducing conditions that eventually formed the deposits, as suggested by several researchers (Evans
and Garrels, 1958; Hansley and Spirakis, 1992; Hostetler and Garrels,
1962; Spirakis, 1996).
In addition to the control on V geochemistry that microbes exert
through redox reactions, adsorption to microbial surfaces can also play
a role in V cycling. The microbial surface contains different organic functional groups that contribute negative and positive charges, e.g. carboxyl,
phosphoryl, hydroxyl and amide groups. Although the net surface charge
is negative at pH values where most life exists, both anionic and cationic
metal and metalloid species will be retained within the cell envelope
(French et al., 2013a; French et al., 2013b). In addition, the generally
high surface-to-volume ratio of microbial cells facilitates contact
between the surface and sorbates, which enables sorption and uptake.
Because of the high density of functional groups in the microbial cell
wall that can react with cations and anions, it is feasible that strong
metal binding will impact the structural integrity of the wall. This was
observed for cultures of Shewanella putrefaciens CN-32 exposed to
V(IV). Under oxic conditions, V(IV) stabilized membrane fluidity,
which is controlled by lipid packing (French et al., 2013a; French
et al., 2013b). Exposure to V(IV) did not, however, impact membrane
fluidity under anoxic conditions although more V (as VO2+) was associated with the cell wall, which could not be explained definitively
and reflects the complex biological chemistry of V.
The positive correlation between organic matter and V in fossil fuels
is likely due in part to the strong interaction of V with microbial surfaces. An investigation of V(V) biosorption by Halomonas sp. GT-83
showed an uptake capacity of 52,700 mg kg−1 dry weight at pH 3
(Ghazvini and Mashkani, 2009). The biosorption efficiency is likely
highest at pH 3 (relative to more alkaline values tested) because most
V(V) exists in cationic form at low pH (Fig. 8). Electrostatic attraction
was important for the adsorption of V(V) to Halomonas GT-83, as indicated by an observed decrease in adsorption with increasing salinity.
Temperature appears to have little impact on V(V) bisorption, whereas
increasing microbial cell density may reduce V(V) adsorption capacity
due to cell aggregation (Ghazvini and Mashkani, 2009).
Sorption/desorption, reduction-oxidation, complexation and precipitation reactions all play important roles in the geochemical cycling of V
in surface and near-surface environments (Fig. 9). As discussed, the
Table 6
Oxidation–reduction reactions for the oxidation states of vanadium (calculated from the Gibb’s energy values given in Wanty and Goldhaber (1992), 25 °C, 1 atm. Pressure, zero ionic
strength (I).
Vanadium(V)–Vanadium(IV)
VO+
2
+
−
2+
+ 2 H + e ⇋ VO + H2O
+
−
2+
H2VO−
+ 3H2O
4 + 4 H + e ⇋ VO
+
−
+
H2VO−
4 + 3 H + e ⇋ VOOH + 2H2O
Vanadium(IV)–Vanadium(III)
+
−
+
H2VO−
4 + 4 H + 2e ⇋ V(OH)2 + 2H2O
+
−
+
HVO2−
+
5
H
+
2e
⇋
V(OH)
4
2 + 2H2O
Eo
1.000
1.419
1.084
Eo
0.691
0.029
Vanadium(IV)–Vanadium(III)
2+
+
−
Eo
3+
VO + 2 H + e ⇋ V + H2O
VO2+ + H+ + e− ⇋ VOH2+
VO2+ + H2O + e− ⇋ V(OH)+
2
VOOH+ + H+ + e− ⇋ V(OH)+
2
0.191
−0.038
0.297
J.-H. Huang et al. / Chemical Geology 417 (2015) 68–89
79
conditions, which can fluctuate seasonally. Light can be a contributing
factor; in one study, diel variations in V concentrations that were observed in an acidic stream could be attributed to photoreduction (Kay
et al., 2011).
Fig. 8. Eh–pH predominance diagram for soluble and solid species of vanadium.
cycling of V between major V-containing minerals by weathering
processes and processes controlled by microorganisms, including
complexation and redox-driven reactions, are key to V’s mobility. In
the following sections, we expand the discussion of these processes by
focusing on particular environments for which vanadium research has
been active.
4.3. Vanadium in surface environments: water, soil, and sediment
4.3.1. Freshwater environments
As discussed in previous sections, the presence of soluble V is a complex function of the solution chemistry, the presence of sorbing colloids,
and the source material. The relationship between V in solution and
environmental factors can vary as a function of the environmental
setting; for example, a strong association with colloidal or particulate
Fe explained low dissolved V content in streams (Wällstedt et al.,
2010), whereas the concentration of dissolved V in major rivers was
related more strongly to weathering and source rock than to the association with colloids or organic matter (Shiller and Boyle, 1987). At sites
where V is detected in water at elevated concentrations, commonly in
association with mining activities, dissolved V concentrations may
reach several hundred mg L− 1, e.g., 30–220 mg L− 1 in Wyoming
River, U.S.A. (Schroeder, 1970; Shiller and Boyle, 1987). A study of
groundwater in California showed a similar correlation between V and
the source rock, in particular basaltic and andesitic rocks (Schroeder,
1970; Wright and Belitz, 2010). The concentrations of V in that study
were highest in oxic and alkaline groundwater, where V is present as
dissolved V(V).
Vanadium concentrations in water can be seasonally dependent.
Higher concentrations of dissolved V during warm (summer) conditions have been documented in freshwater and marine systems,
e.g., in Long Island Sound (Wang and Sañudo Wilhelmy, 2009a),
Toyko Bay and Shimizu Port, Biwa Lake (Sato and Okabe, 1978; Sato
and Okabe, 1982) and Florida Bay (Caccia and Millero, 2003). Elevated
concentrations of V were observed during the seasonal transition from
winter to spring for water draining from a wetland (Pourret et al.,
2012). Several factors can contribute to seasonal V mobility. The association of V with particulates in water is a strong function of pH and Eh
Table 7
Oxidation–reduction reactions for vanadium which include solid species for V(IV) and
V(III). (calculated from the Gibb’s energy values given in Wanty and Goldhaber (1992),
25 °C, 1 atm. Pressure, zero ionic strength (I).
Eo
Reduction reaction
H2VO−
4
HVO2−
4
+
−
+ 2 H + e ⇋ VO(OH)2(s) + H2O
+ 4 H+ + 2e− ⇋ V(OH)3(s) + H2O
VO(OH)2(s) + 2 H+ + e− ⇋ V(OH)+
2 + H2O
VO(OH)2(s) + H+ + e− ⇋ V(OH)3(s)
1.057
0.892
0.324
0.248
4.3.2. Marine environments
In seawater, V is one of the two most abundant trace elements
(molybdenum is the other) due to rock weathering. Concentrations of
dissolved V in the open ocean average 35–45 nmol L−1 (as reviewed
in Emerson and Huested (1991)) and higher concentrations occur
locally (Table 1). V in the ocean exhibits conservative behavior,
although non-conservative behavior is common in coastal waters
where concentrations are typically lower (reviewed in Wang and
Sañudo Wilhelmy (2009b)). Most V in open ocean seawater is dissolved rather than particulate (N 0.45 μm), as shown by studies in the
North Atlantic (Hoffman et al., 1974; Riley and Taylor, 1972; Riley and
Saxby, 1982). These studies concur with the much higher reservoir of
dissolved V (2.7 × 1015 g) as compared to suspended particulate V
(1.8 × 1010 g) estimated for the global ocean (Hope, 2008). Ocean sediment contains, however, the largest reservoir of V (1.7 × 1020 g) on the
Earth’s surface (Hope, 2008), mainly in marine shale and carbonaceous
sediments (Ekstrom et al., 1983). This reflects that the ocean floor is the
ultimate sink of V in the global circulation (Bengtsson and Tyler, 1976;
Miramand and Fowler, 1998). Submarine volcanoes are important
sources of V in marine sediments; V concentrations in active ridge
sediments may reach N 400 mg kg−1 (Boström and Fisher, 1971). Atmospheric deposition is another important source of V input to global
oceans and is discussed in Section 4.4.
The abundance of V in seawater, its occurrence in several oxidation
states that control solubility, and its role as an essential trace element
to life have made it a valuable proxy for establishing paleoredox and
paleoproductivity conditions (reviewed in Tribovillard et al. (2006)).
In oxic seawater, the main species of V(V) are HVO24 − and H2VO−
4 .
Common V(IV) species are VO2 +, VOOH+, and insoluble VO(OH)2
(Rehder, 2008a; Rehder, 2008b). Vanadyl species can be further
reduced under anoxic conditions to V(III), which readily forms solid
oxides of V2O3 and V(OH)3 that can accumulate in sediments. The relative accumulation rates of V, as well as those of other redox sensitive
elements such as Mo and U, increase dramatically with decreasing fO2
in the ocean (Shaw et al., 1990; Tribovillard et al., 2006; Wanty and
Goldhaber, 1992). The distinct chemical signatures of V and other multivalent elements that are preserved in sedimentary marine rocks can
be excellent paleo-oxygen barometers, an indice for paleo fO2 variation,
for recording the history of fO2 variation in the atmosphere (Algeo and
Maynard, 2004; Algeo and Maynard, 2008; Anbar and Rouxel, 2007;
Tribovillard et al., 2006). Vanadium supplied to oceans from weathered
detrital inputs from the continental crust has been used with other
redox tracers to track oceanic evolution from oxic-suboxic to anoxic
(non-sulfidic, anoxygenic) and euxinic (sulfidic, oxygenic) conditions.
Precipitation with organic carbon occurs at redox boundary zones,
producing a strong correlation between organic carbon and V, as well
as other trace elements (Tribovillard et al., 2006). In euxinic environments, V is mainly hosted in authigenic phases and is less associated
with organic carbon. For example, there is a weak correlation between
V and total organic carbon content in oceanic black shale, which distinguishes V from other transition metals with strong euxinic affinity
(Algeo and Maynard, 2004; Algeo and Maynard, 2008). To relate trace
element signatures in sediments and sedimentary rock, concentrations
are typically normalized to aluminum content; aluminum represents
the aluminosilicate fraction, which has low mobility during diagenesis, in contrast to carbonate and opal mineral fractions (reviewed in
Tribovillard et al. (2006)). The reference shales may not, however
accurately represent the sediments being studied (Van der Weijden,
2002). Concentrations of V in shales span a wide range globally and
can vary widely across local geographic areas (see Xu et al. (2013)
and Section 3.2).
80
J.-H. Huang et al. / Chemical Geology 417 (2015) 68–89
Table 8
Potential pathways of vanadium redox transformation in the surface environment.
Redox reaction
Redox couple/reagents
Comments
Reference
V(IV) to V(III)
H2S
Wanty and Goldhaber (1992)
V(V) to V(IV)
V(V) to V(IV)
Humic substance
Fulvic acid
V(V) to V(IV)
benzene-1,4-diol
(Quinol)
V(V) to V(IV)
Insoluble peat humic acid
V(V) to V(IV)
Hydroxylamine
V(V) to V(IV)
V(V) to V(IV)
Benzene-1,2-diol (catechol) and
its deriviatives
L-ascorbic acid
V(IV) to V(V)
O2
V(III) to V(IV)
O2
Slow reduction kinetics; V: 0.5–12.2 mg L−1,
pH 3.6–6.8, H2S: 8.1 × 10−4 to 0.471 atm;
ionic strength 0.05–0.1
pH b6; V(V): 25.5 mg L−1
Higher reduction rate at pH 2 than pH 6
Semiquinone radical was not involved
Reduction rates increase with increasing acidity
via formation of VO2QH2+ and VO(OH)QH22+ V:
25.5 and 51 mg L−1; quinol: 27–324 g L−1;
HClO4: 5.03–503 g L−1; at 16, 25, 32 and 40 °C
pH: 7.2, 1.17 g L−1, 1 g humic acid.
No V(III) was identified
+
+
3+
Via formation of VO+
2 ⋅NH3OH and V(OH)2 ⋅NH3OH V:
51 mg L−1; Hydroxylamine: 0.33–6.6 g L−1; 25-60 °C,
HClO4: 100–503 g L−1.
Formation of V(V)-reductant complex as intermediate
HClO4: 20.1–100 g L−1; 25 °C,
Formation of VO2+ complexes with the derivates of
ascobic acid
Three order of magnitude faster oxidation kinetics
for V(IV) adsorbed to Al2O3 and TiO2 than solution V(IV)
V: N 0.51 mg L−1, pH: 4–9
–
Lu et al. (1998)
Wilson and Weber (1979)
Wells and Kuritsyn (1970)
Szalay and Szilágyi (1967)
Nazer and Wells (1980)
Kustin et al. (1974)
Ferrer and Baran (2001)
Wehrli and Stumm (1988), Wehrli and
Stumm (1989)
Fallab (1967)
-: no information
New approaches to apply trace metal data to paleoceanographic
systems extend conventional approaches based on differences in trace
metal accumulation between oxygenated and non-oxygenated environments. (Algeo and Rowe, 2012) present a method which applies
trace-metal/TOC ratios to assess watermass restriction and deepwater
renewal times in restricted anoxic marine systems. This can be effective
because the onset of ocean anoxia results in the drawdown of trace
metals. Although their investigation uses Mo as the example, V may
also be suitable given its similar redox properties, as described above.
As for Mo, the uptake of V in anoxic marine environments depends on
both the concentration of V and on the concentration of the sedimentary organic matter which hosts the element.
Vanadium isotope fractionation was presented in Section 2.4. The
feasibility of applying V isotopes – 50V (0.24%) and 51V (99.76%) – as a
new paleoredox proxy for ocean sediments is promising (Wu et al.,
2015). The proposed approach is based on the balance of V fluxes into
and out of the ocean at steady state, with rivers as the sole input and
marine sediments and hydrothermal deposits as the outputs. The
Fig. 9. Vanadium mobilization and immobilization processes at the water-soil/sediment interface.
J.-H. Huang et al. / Chemical Geology 417 (2015) 68–89
model remains to be tested; there are currently no V isotope measurement for seawater or ocean sediments (Wu et al., 2015).
4.3.3. Soils and sediments
Vanadium in soils may originate from source rock, atmospheric
deposition and anthropogenic sources. Vanadium in soils and sediments
appears to be generally immobile as sorbed and precipitated phases. Its
mobilization is often decoupled from other metals and metalloids, and
is governed by weathering and redox processes (Shiller and Mao,
2000; Yang et al., 2014). The speciation of V in soils is strongly impacted
by the formation of complexes with anions present in the soil solution.
This is demonstrated for V(IV) in a simulated soil solution (Fig. 10).
Because sulfate ions are stronger complexant ligands for VO2+ than
chloride ions, the VOSO04 complex is much more important than the
VOCl+ complex, even though the SO24 − ion concentration is lower
than Cl− (Fig. 10).
The aqueous chemistry of V(III) is similar to Fe(III) in that the
vanadic, V3+, ion hydrolyzes in water to form the VOH2+ and V(OH)+
2
ions (Table 9). By analogy with the Fe(III) system, the V3+ ion can be
expected to form relatively weak complexes with the Cl− ion, and
strong complexes with SO2−
ions although the VOSO04 complex is not
4
the major species. A formation constant, log β, of 3.28 has been calculated for the VSO+
4 ion (Wanty and Goldhaber, 1992). This compares
with values of 3.89, 3.81 and 2.25 for Al3+, Cr3+ and Fe3+ respectively.
A dinuclear, tetravalent species, V2(OH)42 +, is only important at
relatively high V3 + concentrations. Complexation with sulfate, at a
typical soil sulfate concentration, occurs at very acidic pH values,
which are not expected for the majority of soils (Fig. 11).
Vanadium forms oxide minerals at elevated concentrations. These
include vanadium pentoxide, V2O5(s), vanadium oxyhydroxide,
VO(OH)2(s), vanadium tetroxide, V2O4(s), vanadium trioxide, V2O3(s)
and vanadium hydroxide, V(OH)3(s) (Table 10). The oxides V2O5(s),
V2O4(s) and V2O3(s) are generally prepared at high temperatures
(N300 °C) (Cook, 1947), conditions unlikely to occur in soils or sediments and are therefore unlikely to precipitate under ambient conditions. However, VO(OH)2(s) and V(OH)3(s) can form at 25 °C (Garrels,
1953) and so are possible vanadium oxide precipitates in soils and
sediments.
Vanadium pentaoxide, V2O5(s), is the most important synthesized V
compound, and is used in a number of industrial processes. It may enter
soils through atmospheric deposition or the disposal of wastes in the
vicinity of industrial establishments. It is, however, relatively soluble
in water, with a minimum solubility at about pH 4 (Fig. 5). The solubilities of VO(OH)2(s) and V(OH)3(s), on the other hand, both decrease
Fig. 10. Speciation of vanadium(IV) in a simulated soil solution using the thermodynamic
data given in Table 4.
81
Table 9
Formation constants, log β, for vanadium(III) with some inorganic ligands (calculated
from the Gibb’s energy values given in Wanty and Goldhaber (1992), 25 °C, 1 atm.
Pressure, zero ionic strength (I).
log β
Vanadium(III)
3+
2+
+
V + H2O ⇋ VOH + H
−2.30
V3+ + 2H2O ⇋ V(OH)o2 + 2 H+ −6.33
log β
3+
2V + 2H2O ⇋ V2(OH)4+
2
V3+ + SO2−
⇋ VSO+
4
4
+
+H
−3.79
3.28
as pH increases (Fig. 12), supporting the strong tendency toward
decreased solubility as a function of valence.
Another likely means to immobilize V is through association with
soil organic matter and retention by clay minerals and Fe oxides
(Mikkonen and Tummavuori, 1994). The interaction of V with soil
minerals has not, however, been well studied to date, and can be difficult to predict because of the complexities of soil mineralogy, the
chemical complexity of the soil solution, and the diverse chemistry
of V. As a consequence, estimates of the mobility and bioavailability
of V in soils are contradictory. This is also due in part to the methods
that are commonly used to assess mobility. Selective sequential extraction (SSE) techniques are widely used to characterize the reactivity of V and other elements associated with different soil
components. The quality of the results obtained by SSE depends on
the assurance that a given treatment or reagent can selectively and
specifically extract the intended phases (Hass and Fine, 2010). A
lack of standardization in SSE protocols can complicate the comparison of results between studies (Hass and Fine, 2010). We present a
few examples of results from studies that applied SSE techniques to
assess V fractions in soil in order to illustrate the wide range of
outcomes.
Sequential leaching of V in a variety of agricultural and industrial
soils from Poland indicated the predominance of V in the soluble
fraction (Poledniok and Buhl, 2003). An investigation of Egyptian
and Greek soils differing in origin and properties showed that the
bicarbonate-DTPA extractable V (related to soluble V) was positively
correlated with total soil V, soil pH, clay, silt and cation exchange
capacity, but negatively with organic matter (Tsadilas and Shaheen,
2010). An investigation of German soils highlighted a large range in V
extractability for different soils: EDTA extraction recovered 0.2–35%
of V soluble in aqua regia, which is considered to represent total V
(Gäbler et al., 2009). Other studies have documented the majority
of V in the soil residual fraction, which may comprise up to 93% of
total V in some soils (Cappuyns and Swennen, 2014; Ovari et al.,
2001; Teng et al., 2011). In general, most V in soils appears to have
low mobility except under strongly acidic conditions (Cappuyns and
Swennen, 2014). The mobile V in soils occurs mainly as V(V) and only
Fig. 11. Speciation of V(III) in the presence of SO2−
ions.
4
82
J.-H. Huang et al. / Chemical Geology 417 (2015) 68–89
Table 10
Some oxide, oxyhydroxide and hydroxide precipitates of vanadium.
Reaction
log Kso
Reaction
log Kso
Vanadium(V)
V2O5(s) + 2 H+ ⇋
2VO+
2 + H2O
−1.36
Vanadium(IV)
VO(OH)2(s) + 2 H+ ⇋
VO2+ + 2H2O
6.11
Vanadium(III)
V(OH)3(s) + 3 H+ ⇋
V3+ + 3H2O
7.59
minor amounts are present as V(IV) (Baken et al., 2012; Burke et al.,
2012). Vanadate is also the most available form for plants (Tian et al.,
2015). Although red mud, the fine sediment byproduct of bauxite processing, is not soil, a recent study provides insight into V mobilization
from mineral solids. This study was motivated by the failure of a red
mud dam at Ajka, Hungary, which released sediment with elevated concentrations of potentially toxic elements, including V (Burke et al.,
2012). All of the V was present in the red mud as V(V) at pH 13, determined using X-ray absorption spectroscopy (XANES), and was relatively immobile. Neutralization of the sediment to circumneutral pH values
resulted in immobilization of Cr and As, but mobilized V. It was surmised that dissolution of V-bearing minerals during neutralization,
anion exchange processes, and the formation of polymeric V species
could have contributed to enhancing V mobility at lower pH (Burke
et al., 2012). In comparison, pH was shown to influence V mobility indirectly in floodplain and paddy soils, and redox status controls V mobility
(Frohne et al., 2015; Shaheen et al., 2014a; Shaheen et al., 2014b). These
microcosm experiments with controlled redox variations further highlight that the length of the redox period changes substantially with
the concentration of V in porewater (Shaheen et al., 2014b). Higher
porewater V concentrations were detected in the experiment with
short-term overflooding, probably due to kinetic limitation of anoxic V
immobilization.
Other environmental factors may also impact V mobility. In one study,
aging caused V(V) added to soil to become less soluble, bioavailable and
toxic (Burke et al., 2012). Vanadium(IV) appears to predominate in soils
over other valence states (Kuo et al., 2007; Mandiwana and Panichev,
2004) (Table 1), possibly reflecting the incorporation of VO2+ into the
lattice of soil minerals (Gehring et al., 1993) as well as favorable conditions for dissimilatory reduction by microorganisms. Drought and
rewetting greatly increase the leaching of V, as observed in soil column
studies (Yang et al., 2014). The observation that the leaching behavior of
V did not correlate with its sorption affinity suggests that weathering
and oxidation of V-containing minerals is responsible for an increase
in mobile V during the drought period, which was flushed during
rewetting.
Atmospheric deposition is an important input of V to soils located far
from a point source, such as peat bogs (Cloy et al., 2011). Because the
matrix in peat bogs is almost exclusively organic, the migration of V is
likely retarded by the formation of organic complexes. Vanadium in
wetland soils is found mainly in the acid-digestible fraction (N 70%)
(Fox and Doner, 2003). Lead is well known to be immobile in organic
soils; the general similarity between the vertical distribution of V and
Pb suggests that V is similarly immobile in peat bogs (Cloy et al.,
2011). Seasonal effects can also play a role (Pourret et al., 2012).
Few countries have established standards for V contamination in soil.
In Russia, a maximum of 150 mg kg−1 is allowed in agricultural soils
(GOST-17.4.4.02-84, 1985). The general soil quality guideline for V in
Canada is 130 mg kg−1 (Canadian_Environmental_Quality_Guidelines,
1999).
4.3.4. Adsorption to metal oxide minerals
The water-soil partitioning of V in soils and sediments reveals the
generally low mobility of soil V, with adsorption coefficients of up to
750 000 L kg−1 (Table 1). Most research on the sorption of V by minerals has focused on sorption to metal oxides. In addition to the surface
chemistry of the sorbent, V adsorption and desorption depend strongly
on the concentration, chemical valence and chemical species (Table 11)
(Peacock and Sherman, 2004; Wehrli and Stumm, 1988; Wehrli and
Stumm, 1989).
Adsorption of both V(IV) and V(V) onto metal oxide minerals has
been interpreted in terms of the formation of inner-sphere bidentate
Fig. 12. Solubility of some vanadium oxides as a function of pH.
J.-H. Huang et al. / Chemical Geology 417 (2015) 68–89
83
Table 11
Solid-solution partition coefficients (Kds) calculated from experimental data for a range of metal oxides, clays and environmental materials.
Material
Components
Kd (L kg−1)
pH
Comments
Reference
Agricultural soil, China
Mine tailing
Soils, South Africa
Dust
Marine sediment, (MESS-3)
Clay soil, USA
Sandy soil
Floodplain soils, Belgian
Ferrihydritea
Not specified
1′400–10′950
5′170–70′530
9.37–526
23.4
48.6
~150
~1.71
~21′800–35′500
~90′000
~60′000
~2′000
4.3
8.1
–
H2O leaching
Yang et al. (2014)
H2O extraction
Mandiwana et al. (2005)
5.6
5.9
–
8
9
8
c = 5–100 mg L−1
Wang et al. (2003)
10 mM CaCl2 extraction
c = 5.1 mg V L−1
I = 0.01
c = 10.2 mg V L−1
I = 0.7
Ca(NO3)2 extraction
Cappuyns and Salbbinck (2012)
Brinza et al. (2008)
–
32 °C
40 °C
50 °C
60 °C
c = 10–50 mg L−1
c = 0.41–12.8 mg L−1, I = 0.1
c = 2.55 mg L−1, I = 0.1
Gobeil et al. (2005)
Prathap and Namasivayam (2010)
–
Szalay and Szilágyi (1967)
V(V)
V(V)
Not specified
Not specified
Ferrihydrite
Not specified
Sandy soils,
Silty soils,
Loamy soils,
Clay soils,
River suspended particulate matter
Fe(III)/Cr(III) hydroxide
Not specified
TiO2
Al2O3
TiO2
Al2O3
Insoluble peat humic acid
V(IV)b
Not specified
V(V)
V(V)
V(IV)
247–34′100
7′560–323′600
4′200–755′000
4′540–489′167
29,600
~170–270
~220–530
~260–430
~280–900
~2500–225′000
~1′500–9′500
~5000–49,500
~2500–47,500
~6–60
4.2–7.3
–
7.9–8.2
1–.8–3.5
2.75–4
0.5–11.5
1–10
2.1
Trefry and Metz (1989)
Gäbler et al. (2009)
Wehrli (1987)
-: no information.
a
calculation infeasible at pH 4–7.
b
Kd from pH 3.5–5 can not be estimated, however, higher values are expected.
complexes for the vanadyl ion, VO2+, and mono-dentate complexes for
vanadate ions, HnVOn-3
(Motschi and Rudin, 1984; Peacock and
4
Sherman, 2004; Wehrli and Stumm, 1989). Recent work on the adsorption of V(V) ions onto goethite (Peacock and Sherman, 2004) has demonstrated the formation of two bidentate complexes, ≡Fe2OVO(OH)+
4
and ≡ Fe2OVO2OHo; their intrinsic formation constants were determined by batch adsorption experiments (Table 12). The calculated adsorption of V(V) ions onto goethite as a function of pH using the
Diffuse Layer Model and the constants given in Peacock and Sherman
(2004) shows very high affinity across environmentally relevant pH
values (Fig. 13). Using EXAFS, the two complexes were shown to be
inner-sphere bidentate corner-sharing complexes with the edges
hydroxyls of goethite.
The solution pH can be expected to control V sorption, given the
strong dependence of V speciation on pH. Adsorption involving
ligand exchange can, however, offer exceptions to what might be
expected based on simple electrostatic repulsion. For example, adsorption of V(V) at low pH to positively charged iron oxide mineral surfaces
can be expected to occur via VO+
2 cations (Figs. 6, 7). At higher pH,
V(V) occurs as conjugate acids of VO3−
that can be expected to adsorb
4
to positively charged oxide mineral surfaces by ligand exchange, analogous to the sorption behavior of phosphate. Charge dependence for V
sorption has been observed in several studies. For example, the adsorption of vanadate to α-FeOOH and δ-Al2O3 was suppressed below pH
values b~3 and N ~ 8, and was suppressed at pH N ~8 for Fe(III)/Cr(III)
hydroxide at 2.5 mg L−1; sorption decreased around pH 9 for TiO2 at
the same concentration. Adsorption of vanadyl cations to TiO2 and
Al2O3 was suppressed below pH ~ 3 and ~ 4, respectively (Wehrli,
1987). The adsorption behavior was explained by the predominance
of VO2 + at pH values b~3, resulting in repulsion by the positively
charged TiO2 and Al2O3 surface (Wehrli and Stumm, 1989).
The adsorption of V to mineral surfaces depends on concentration.
Concentration is particularly important to consider for V(V) because of
the formation of polynuclear V(V), i.e., metavanadate (e.g. V3O3−
9 ), species, which can result in decreased sorption (Fig. 6); for example, less
V(V) was sorbed to goethite between pH 6 and 9 at 25 mg L−1 V than
at 2.5 mg L−1 V in one study (Peacock and Sherman, 2004). Competition
with other ions also impacts sorption. Vanadate competes with other
oxyanions such as phosphate, arsenate, selenate and molybdate for
sites on positively charged mineral surfaces (Blackmore et al., 1996;
Brinza et al., 2008; Wang et al., 2003). Prathap and Namasivayam
(2010) proposed an order of adsorption among common oxyanions
and chloride to positively charged metal oxide surfaces of: vanadate N
phosphate N selenate N molybdate; sulfate, nitrate and chloride showed
no effect on V(V) adsorption. In another study, vanadate outcompeted
Table 12
Surface reactions and intrinsic complexation constants for V(V) onto goethite calculated
using the Diffuse Layer Model (Specific surface =37.2 m2 g−1; [SOH]T = 6.02 sites nm−2;
log KH1 = 6.78; log KH2 = − 10.10).
log Kint
Adsorption reaction
o
2 ≡ Fe-OH +
2 ≡ Fe-OHo +
VO3−
4
VO3−
4
+
+4H ⇋≡
+ 3 H+ ⇋ ≡
Fe2-OVO(OH)+
4 + 2H2O
Fe2-OVO2OHo + 2H2O
42.87
41.26
Fig. 13. Adsorption of vanadium(V) onto goethite.
84
J.-H. Huang et al. / Chemical Geology 417 (2015) 68–89
arsenate for adsorption to Fe2O3, but had an equal affinity for Al2O3
(Jeong et al., 2007). The adsorption of negatively charged V oxyanion
species will increase the adsorption of cationic metals to positively
charged surfaces due to charge screening, formation of cation-anion
complexes at the surface and surface precipitation (Zhao et al., 2004).
4.4. Vanadium in the atmosphere
Atmospheric V is a key component of the V cycle. The major natural
sources of atmospheric V globally are marine aerosols produced when
bubbles burst at the sea surface, continental dust from crustal
weathering, and volcanic debris (Zoller et al., 1973). Volcanic activity
contributes V associated with tephra and ashes, and in hydrothermal
springs (Bortnikova et al., 2009; Mizuno et al., 2008; Smichowski
et al., 2003). The total V added to the atmosphere from natural sources
has been estimated at 37 × 109 g yr.−1 and is comparable in magnitude
to V contributed by anthropogenic sources (Duce and Hoffman, 1976).
Duce and Hoffman (1976) estimated that around 10% of anthropogenic
atmospheric V is deposited in the global oceans.
The abundance of V transported in the atmosphere from continental pollution sources may make it a useful tracer for atmospheric
pollutant transport (Duce and Hoffman, 1976). For example, the
accumulation of anthropogenic V, as determined for a peat bog core in
Scotland, could be correlated with industrial activities begun in the
mid-to-late 19th century (Cloy et al., 2011). Deposition rates peaked
(~1.3 to 2.0 mg m−2 yr.−1) in the mid-20th century before decreasing
to 0.1–0.3 mg m− 2 y− 1 in the early years of the 21st century. This
was attributed to more stringent emission standards in the United
Kingdom (Cloy et al., 2011). Studies on a Swiss peat bog revealed
an increased deposition of anthropogenic V since 1936, reflecting
the growing impact of industrialization on atmospheric inputs (Krachler
et al., 2003). Although fossil fuel combustion has increased since some
early studies on V atmospheric deposition (e.g., (Duce and Hoffman,
1976; Zoller et al., 1973)) and controls on emissions in some countries
have reduced the inputs of V to the atmosphere, which is reflected by
lower rates of land deposition (Cloy et al., 2011; Harmens et al.,
2007). Elevated concentrations of V in Sphagnum moss collected from
rain-fed peat bogs near the Athabasca oil sands in Alberta, Canada
support, however, that emissions from mining and processing bitumen
do still contribute to increased land deposition of V locally (Shotyk et al.,
2014).
Climate cycles have a strong impact on the atmospheric deposition
of V, which is true in general for many trace metals. For example, V
accumulation in the Swiss peat bog discussed above was highest during
the Younger Dryas cold climate event (centered at 10,590 14C yr. B.P.),
with corresponding atmospheric fluxes of approximately 14 mg m−2
yr.−1; this value is ~40 times background (Krachler et al., 2003). Vanadium concentrations in sections of the 3623 m Vostok deep Antarctic ice
core, dated from 4600 to 410,000 years BP, were low during interglacial
periods and warm interstadials, and much higher during the coldest
periods of the last four ice ages (Gabrielli et al., 2005). Higher historical
V concentrations preserved in peat have been attributed to increased
atmospheric soil dust, which is a consequence of climate as well as
changes in land use, e.g., forest clearing for agriculture (Gabrielli
et al., 2005; Krachler et al., 2003). Because particulates, especially
geologically-derived dusts, are a major contributor to atmospheric
V (Sakata and Asakura, 2011; Song and Gao, 2009; Theodosi et al.,
2010), changes in climate that increase dust loads in the atmosphere
should increase the transport of V.
Vanadium is found within a range of particle sizes in atmospheric
compartments. The speciation of V by size fraction in Japanese rainwater showed that 45–100% of V was present as labile forms with molecular weight b 1000 Da (Kawakubo et al., 1997). Non-labile V was mainly
found in the higher molecular weight (≤10,000 Da) or insoluble fractions. Another study in Japan found that ~50% of suspended atmospheric V occurred in particles smaller than 2 μm in size (Adachi et al., 1997).
The association of V with mineral dusts can be expected to mirror the
association of V with soil minerals, although this has been given little
attention. In one study, V in dry atmospheric deposition was strongly
correlated with Al, indicating clay mineral association (Adachi et al.,
1997; Sakata and Asakura, 2011).
Vanadium that reaches the atmosphere from volcanic emission and
fossil fuel combustion is generally expected to occur as V(V) due to the
ubiquity of oxygen (Challenge, 2010). A recent study detected mainly
V(IV) species in diesel vehicle emissions, with higher emissions of
V(V) from vehicles equipped with exhaust after-treatment; V(V) was
the major component in the finest size fraction (Table 1) (Shafer et al.,
2012). In another study, V was detected in association with organic carbon and sulfate in air-born emission plumes released from ships in port.
It was speculated that V(V) may catalyze the oxidation of reduced sulfur
species to form sulfate and sulfuric acid. This supports that V redox
chemistry, likely driven by SO2 and photochemistry, may occur in atmospheric particles and is a health concern because of synergistic effects of
V particulates and sulfates (Ault et al., 2010).
4.5. Anthropogenic sources
Around 85% of global V produced is used as ferrovanadium in the
steel industry, while another 10% is used in the aerospace industry in
the form of V-Al-Ti alloys (Moskalyk and Alfantazi, 2003). Although
the other industrial applications of V (including catalysts, glass coating
and rechargeable batteries; Table 13) have less impact, some uses, in
Table 13
Industrial uses of vanadium.
Chemical forms
Application
References
Ferrovanadium (FeV)
Chrome-vanadium steels
Vanadium high-carbon steel alloys, high speed tool steels (surgical instruments and tools)
Springs and transmission gears
are fabricated from
Jet engines, high-speed airframes and dental implants, missile cases, nuclear reactor
components, jet engine housings and associated airframes
Cladding titanium to steel
Manufacturing sulfuric acid
maleic anhydride production
ceramics production
Glass coatings
Electrochemical conversion coating
Rechargeable batteries
Vehicle exhaust catalyst
Superconductor
Color modifiers in mercury-vapor lamp, photographic developer, drying agent in various
paints and varnishes, and reducing agent, production of pesticides, and the black dyes
(e.g., mordants), inks, and pigments, (Vanadium Production, 2003).
Moskalyk and Alfantazi (2003)
Moskalyk and Alfantazi (2003)
V-Al-Ti alloys
Vanadium foil
Vanadium pentoxide (V2O5)
Vanadium dioxide (VO2)
Vanadate
Vanadium redox couple
V2O5-WO3 (MoO3)/TiO2
Vanadium and V alloys
Other applications
Moskalyk and Alfantazi (2003)
Lositskii et al. (1966)
Abon and Volta (1997)
Manning et al. (2004)
Guan and Buchheit (2004)
Joerissen et al. (2004)
Liu et al. (2012), Shafer et al. (2012)
Ponyatovskii et al. (2009)
Moskalyk and Alfantazi (2003)
J.-H. Huang et al. / Chemical Geology 417 (2015) 68–89
particular those related to energy generation and storage, are likely to
increase in the future.
The waste V generated by the ferrous metallurgy industry is
mainly in liquid and particulate form (WHO, 1988). The recovery of
V from these wastes is around 60–70%; the remaining 30–40% of V is
discharged into th…