Fe Isotope Fraction at Ion During Hydro Thermal Ore Deposition

Geochimica et Cosmochimica Acta 70 (2006) 3011–3030
www.elsevier.com/locate/gca
Iron isotope fractionation during hydrothermal ore deposition

and alteration
a,*
Gregor Markl , Friedhelm von Blanckenburg b, Thomas Wagner a
a
Institut für Geowissenschaften, Wilhelmstr 56, D-72074 Tübingen, Germany
b
Institut für Mineralogie, Callinstrasse 3, D-30167 Hannover, Germany
Received 25 July 2005; accepted in revised form 3 February 2006
Abstract
Iron isotopes fractionate during hydrothermal processes. Therefore, the Fe isotope composition of ore-forming minerals characterizes
either iron sources or fluid histories. The former potentially serves to distinguish between sedimentary, magmatic or metamorphic iron
sources, and the latter allows the reconstruction of precipitation and redox processes. These processes take place during ore formation or
alteration. The aim of this contribution is to investigate the suitability of this new isotope method as a probe of ore-related processes. For
this purpose 51 samples of iron ores and iron mineral separates from the Schwarzwald region, southwest Germany, were analyzed for
their iron isotope composition using multicollector ICP-MS. Further, the ore-forming and ore-altering processes were quantitatively
modeled using reaction path calculations. The Schwarzwald mining district hosts mineralizations that formed discontinuously over
almost 300 Ma of hydrothermal activity. Primary hematite, siderite and sulfides formed from mixing of meteoric fluids with deeper crust-
al brines. Later, these minerals were partly dissolved and oxidized, and secondary hematite, goethite and iron arsenates were precipitated.
Two types of alteration products formed: (1) primary and high-temperature secondary Fe minerals formed between 120 and 300 °C, and
(2) low-temperature secondary Fe minerals formed under supergene conditions (<100 °C). Measured iron isotope compositions are var-
iable and cover a range in d56Fe between 2.3& and +1.3&. Primary hematite (d56Fe: 0.5& to +0.5&) precipitated by mixing oxi-
dizing surface waters with a hydrothermal fluid that contained moderately light Fe (d56Fe: 0.5&) leached from the crystalline
basement. Occasional input of CO2-rich waters resulted in precipitation of isotopically light siderite (d56Fe: 1.4 to 0.7&). The differ-
ence between hematite and siderite is compatible with published Fe isotope fractionation factors. The observed range in isotopic com-
positions can be accounted for by variable fractions of Fe precipitating from the fluid. Therefore, both fluid processes and mass balance
can be inferred from Fe isotopes. Supergene weathering of siderite by oxidizing surface waters led to replacement of isotopically light
primary siderite by similarly light secondary hematite and goethite, respectively. Because this replacement entails quantitative transfer
of iron from precursor mineral to product, no significant isotope fractionation is produced. Hence, Fe isotopes potentially serve to iden-
tify precursors in ore alteration products. Goethites from oolitic sedimentary iron ores were also analyzed. Their compositional range
appears to indicate oxidative precipitation from relatively uniform Fe dissolved in coastal water. This comprehensive iron isotope study
illustrates the potential of the new technique in deciphering ore formation and alteration processes. Isotope ratios are strongly dependent
on and highly characteristic of fluid and precipitation histories. Therefore, they are less suitable to provide information on Fe sources.
However, it will be possible to unravel the physico-chemical processes leading to the formation, dissolution and redeposition of ores in
great detail.
Ó 2006 Elsevier Inc. All rights reserved.
1. Introduction sible tools for deciphering geochemical processes (see the

review by Johnson et al., 2004a). The observed natural
The stable isotopes of transition metals such as Cr, Fe, variations of their isotope ratios have been attributed to
Cu, or Mo have recently attracted much attention as pos- a number of processes such as fluid–solid reactions (Barling
et al., 2001; Rouxel et al., 2003; Albarède, 2004), redox
*
Corresponding author. Fax: +49 7071 29 3060.
reactions in aqeuous fluids (Matthews et al., 2001;
E-mail address: markl@uni-tuebingen.de (G. Markl). Schauble et al., 2001; Ellis et al., 2002; Welch et al., 2003;
0016-7037/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.gca.2006.02.028
3012 G. Markl et al. 70 (2006) 3011–3030
Anbar et al., 2005), inter-mineral equilibrium fractionation (2) to relate the observed isotope fractionation trends to
(Polyakov, 1997; Polyakov and Mineev, 2000; Williams a physico-chemical model of iron ore formation, and
et al., 2004), the involvement of micro-organisms (Johnson (3) to investigate whether Fe isotopes can be used as pre-
et al., 2004b), the uptake of metals by marine organisms dictive tools for ore formation and hydrothermal
(Maréchal et al., 2000) or by plants and humans (Walczyk alteration.
and von Blanckenburg, 2002; Walczyk and von Blancken-
burg, 2005). All studies agree that at ambient temperatures
both biological and abiotic processes can cause shifts in tran- 2. Geological context
sition metal isotope ratios of a few permil per mass unit. A
demonstration of such fractionations at higher temperatures The Variscan Schwarzwald gneissic and granitic base-
is still a matter of debate (Zhu et al., 2000, 2002; Beard and ment and the overlying Triassic Buntsandstein host a large
Johnson, 2004; Williams et al., 2004). One of the reasons number of hydrothermal vein-type deposits (Fig. 1). Many
for the absence of clear high-temperature fractionations is of them contain various Fe-bearing phases, mainly hema-
that equilibrium isotope fractionation is predicted to de- tite, goethite and siderite. Within an area of 120 by
crease with increasing temperature, and the expected shifts 40 km, more than 400 individual veins are known. Most
are in the range of instrumental detection limits (Polyakov, of these veins are sub-economic, but they have been mined
1997; Schauble, 2004). Metal isotope fractionation is expect- for Ag, Pb, Zn, Co, U or Cu since Roman times. One
ed, and has been observed, during hydrothermal ore forma- deposit (the Clara mine near Wolfach) is still active and ex-
tion which takes place within an intermediate temperature ploits a barite–fluorite vein. Based on their mineralogy,
range and also involves a large range of inorganic aqueous several hydrothermal mineralization styles can be distin-
geochemical reactions that would entail isotope fraction- guished, which are, for example, Sb–Ag-bearing quartz
ation. For Cu isotopes, Zhu et al. (2000) could show that veins (occurring throughout the entire district), Co–Ni–
samples of one mineral type from the same deposit exhibits Ag–Bi–U-bearing barite–fluorite veins (in the Wittichen
large variations of copper isotope ratios, but it is not clear area), Fe–Mn-bearing quartz–barite veins (in the Eisen-
by which process these variations are caused. Graham bach area), Cu–Bi-dominated quartz–barite veins in the
et al. (2004) suggested that large shifts in Fe isotope compo- area between Freudenstadt and Neubulach in the northern
sition (over 4& in the 56Fe/54Fe ratio) can occur in iron Schwarzwald, or the Pb–Zn–(Ag)-bearing quartz–fluorite
deposits associated with magmatic processes, i.e., in a Cu– assemblages in the Southern Schwarzwald (Metz et al.,
Au porphyry–skarn complex. There, the variability was ex- 1957; Bliedtner and Martin, 1988).
plained by mixtures of iron from different, specifically igne- The hydrothermal vein systems are believed to have
ous and sedimentary, sources. The data of Rouxel et al. formed by mixing of ascending deep-seated, relatively re-
(2003) indicate that during alteration of oceanic basalts, iso- duced formation waters with infiltrating meteoric, oxidiz-
topically light iron is preferentially leached, transported in ing surface waters (Fig. 2; Werner and Dennert, 2004).
the fluid and then deposited in hydrothermal Fe–Si deposits, Based on C, S, H and O isotope studies, the deep-seated
whereas the isotopically heavy iron remains mainly in the fluid was chemically surprisingly uniform in space and
basaltic alteration products such as celadonite. This was time, had a temperature of around 300–350 °C and lea-
attributed to a kinetic control during leaching, or to prefer- ched metals (including the iron) from the basement gran-
ential incorporation of the heavier isotope into celadonite. ites and gneisses of the Schwarzwald in 7–8 km depth
All these early observations suggest that stable transition (Schwinn and Markl, 2005; Schwinn et al., 2006). The
metal isotopes do indeed have potential as tools in petroge- surface waters passed through various sedimentary units
netic studies, but a conceptual framework for their interpre- of Lower Triassic to Upper Jurassic age; some of these
tation is still not in place. strata contain significant amounts of Fe as hematite, goe-
The topic of this contribution is to provide a concept for thite or chlorite (Fig. 2). The mixing of the two fluids oc-
Fe isotope ratio variations in hydrothermal and supergene curred at a depth of about 1–2 km and the temperature
iron ores. For this purpose, we report a large new dataset of vein formation depended on the mixing ratio of the
comprising 51 samples of various iron minerals from two fluids (Schwinn et al., 2006). At this stage, hematite
hydrothermal and sedimentary deposits in a well-investi- and/or siderite were precipitated as primary Fe-bearing
gated ore district of Central Europe, the Schwarzwald in phases, reflecting the specific fluid compositions involved
southwest Germany. The purpose of this study is (primary deposits ‘‘pD’’ on Fig. 2). Support for this
model comes from the observation that the crystalline
(1) to investigate the iron isotope variability of primary rocks show abundant secondary alteration, with chloriti-
and secondary iron minerals in a large, geologically zation of biotite and sericitization and albitization of
diverse ore district formed over a period of 300 million feldspars being the most notable alteration reactions.
years which has been well described (e.g., Metz et al., Oxygen isotope investigations of Simon and Hoefs
1957; Werner and Franzke, 2001; Werner et al., (1987) and Hoefs and Emmermann (1983) indicate the
2002; Werner and Dennert, 2004; Markl, 2004; Schw- interaction of a meteoric fluid with the basement rocks
inn and Markl, 2005; Schwinn et al., 2006), at temperatures well below 500 °C.
Hydrothermal iron isotope fractionation 3013
Fig. 1. Simplified geological map of the Schwarzwald region in southwest Germany with sample locations. The numbers refer to Table 1.
Based on observed structures, fluid inclusion studies and which is of Tertiary age. For the other veins, the isotopic
a comparison with similar veins in the Rhenish Massif (e.g., dating of pitchblende (U–Pb and U–Xe, Xe–Xe), primary
Wagner and Cook, 2000; Wagner and Boyce, 2003), the hematite (U–He), and K-bearing minerals (K–Ar) revealed
commonly Sb-bearing quartz veins appear to be related three mineralization events. The first was dated at 310–
to late stages of the Variscan orogeny (310–280 Ma). 280 Ma, at the end of the Variscan orogeny (Hofmann
Hence, the hydrothermal deposits of the Schwarzwald area and Eikenberg, 1991; Segev et al., 1991; Meshik et al.,
have been classified into Variscan Sb–(±Ag ± Bi ± Au)– 2000; Wittichen area, Menzenschwand deposit), a second
quartz veins and post-Variscan fluorite–barite–quartz one at 150–110 Ma (Segev et al., 1991; Wernicke and Lip-
veins. Most of the post-Variscan veins do not host minerals polt, 1993; Wernicke and Lippolt, 1997; Hohberg and
suitable for radiometric dating. For some of these veins Eisenbach area), and a third one at 50–30 Ma related to
such as Badenweiler, however, a Tertiary age is proven the formation of the Rhine Graben structure (Hofmann
by their occurrence on the Rhine Graben boundary fault, and Eikenberg, 1991; Menzenschwand deposit). K–Ar
3014 G. Markl et al. 70 (2006) 3011–3030
Fig. 2. Iron reservoirs and their respective iron isotope ratios relevant for ore-forming and alteration processes in the Schwarzwald region. Primary Fe
deposits (pD): fluid mixing and precipitation of siderite (intermediate Eh, Dsid–Fe(II)aq = 1.5& to 0.5&) or hematite (high Eh, Dhem–Fe(II)aq = +1.5&),
T = 100–200 °C. Secondary Fe deposits (sD): weathering and dissolution (high Eh, DFe(II)aq–silicate = 1.2& to 0.1&), reprecipitation within deposit
(high Eh, Dgoethite or hematite–Fe(II)aq = +1.5&), T = 100–200 °C. Note that D is the difference between product and precursor.
and Ar–Ar dates of sericitized feldspars from the basement rocks (Fig. 3a–c). The primary siderite was partly oxidized
and the overlying Triassic Buntsandstein sandstone are in at high temperatures and a secondary high-temperature
the range 150–110 Ma (Lippolt and Kirsch, 1994; Zuther hematite formed. Hence, the primary iron minerals were
and Brockamp, 1988; Meyer et al., 2000), which correlates redeposited either locally or into a more distant deposit
well with the second hydrothermal event. by secondary processes. Temperatures were high
It is clear from the geological and metallogenic frame- (>100 °C) during this alteration. Later supergene (i.e.,
work that the Schwarzwald iron ore deposits formed dis- low temperature) weathering after partial or complete ero-
continuously over a large age range. As a consequence, sion of the sedimentary cover led to decomposition and
they contain both primary and secondary iron minerals. oxidation of the primary siderite. During this process, bot-
These are distinguished from each other based on their tex- ryoidal goethite or hematite (‘‘glaskopf’’, Fig. 3d–f) or very
tures. Primary iron minerals (hematite, siderite) are be- fine-grained hematite as thin, reddish leaf-style crystals
tween 1 and 20 mm in size. They typically consist of large were formed at pressures and temperatures close to those
euhedral crystals, intergrown with primary gangue miner- prevailing at the surface (sD on Fig. 2). The complete
als such as barite or quartz or grown directly onto the host replacement of siderite is indicated by a residue comprising
a b
FLK12 HEB7
c d
Fe12 Fe16
e f
Fe9 Fe7
Fig. 3. Photographs of some typical iron ore samples. (a) Fresh, euhedral, primary hematite crystals on a granite breccia from Lenzkirch (sample FLK
12). (b) Fresh hematite crystals in a quartz vug from Erlets near Hausach (sample HEB7). (c) Dark brown, partly oxidized siderite crystals from
Neuenbürg (sample Fe12). (d) Fibrous secondary hematite (‘‘red glaskopf’’) from the Rappenloch mine (sample Fe16). (e) Botryoidal secondary goethite
(‘‘brown glaskopf’’), Neuenbürg (sample Fe9). (f) Fibrous, glossy secondary goethite (‘‘brown glaskopf’’), Clara mine (sample Fe7). Scale bar is 1 cm.
either hollow pseudomorphs or dark brown crystal relicts, (1) Primary iron minerals (crystalline hematite (FeIII;
which now consist of Fe(III)-oxides and -hydroxides Fig. 3a and b), chalcopyrite (FeIII), siderite (FeII;
(mainly goethite). In the following, we will use the terms Fig. 3c) and pyrite (FeII)) from hydrothermal veins.
primary (high-temperature hematite and siderite of first The hydrothermal veins occur exclusively in the
generation), secondary (high-temperature hematite of sec- Schwarzwald basement and in the Buntsandstein cov-
ond generation) and alteration or weathering assemblage er sequence at the eastern border of the Schwarzwald.
(low-temperature mineralization). They do not penetrate the other overlying sedimenta-
ry strata.
3. Sampling strategy and sample description (2) On a few samples, partial oxidation processes of the
primary siderite at high temperatures led to the for-
This study combines samples from the entire area of the mation of crystalline, secondary hematite. Based on
Schwarzwald district and its surroundings, from various textures such as in sample M590, in which microcris-
types of iron mineral-bearing veins and from Fe-bearing talline hematite is ‘‘sandwiched’’ between two gener-
Mesozoic sediments. The iron ores were formed in a period ations of high-temperature siderite, this secondary
spanning more than 200 Ma of formation history. The hematite definitely formed at higher temperatures.
sample locations are displayed on Fig. 1 and the samples (3) Supergene iron minerals include the Fe(III)-arsenates
are briefly described in Table 1. The following types of Ba-pharmacosiderite and scorodite and in particular
samples were investigated: the Fe(III)-oxides and -hydroxides goethite and
3016 G. Markl et al. 70 (2006) 3011–3030
Table 1
Samples, sample descriptions, location code for Fig. 1a, and measured iron isotope values from Schwarzwald iron minerals and ores
Sample No. No. on Fig. 1 Mineral p/s Locality d56Fe d57Fe Sample description
Mass-spectrometric duplicates;
bold: sample average
FE2 6 Siderite p/s Mine Clara, Wolfach 0.80 1.17 Dark brown, lensoid
0.80 1.18 crystals, oxidized, in vug
0.80 1.18 on barite
FE3 6 Siderite p Mine Clara, Wolfach 0.99 1.52 Light brown, radially
1.00 1.53 aggregated, fresh crystals
0.99 1.52 up to 5 mm, on barite,
overgrown by calcite and
hematite
M420 7 Siderite p/s Mine Sophia, Wittichen 0.86 1.27 Dark brown, oxidized,
0.85 1.27 rhombohedral crystals in
0.85 1.27 vug on barite crystals
M590 7 Siderite I p Mine Sophia, Wittichen 1.36 2.01 First generation, coarse-
grained, anhedral
siderite on barite
M590 7 Siderite II p Mine Sophia, Wittichen 0.74 1.07 Fresh rhombohedral
0.75 1.17 crystals on small
0.74 1.12 hematite crystals which
overgrown an earlier
generation of siderite
FE10 1 Siderite p/s Neuenbürg 0.86 1.24 Dark brown oxidized
0.80 1.21 crystals on barite
0.83 1.23
FE12 1 Siderite p/s Neuenbürg 0.73 1.05 Oxidized, dark brown
0.72 1.03 rhombohedral crystals
0.72 1.04 on sandstone
WEIN2202 6 Chalkopyrite p Mine Clara, Wolfach 0.34 0.45 Fresh, golden, anhedral
0.32 0.48 masses in barite
0.33 0.47
Repeat dissolution 0.25 0.43
0.32 0.47
0.28 0.45
WEIN2202 6 Pyrite p Mine Clara, Wolfach 0.29 0.46 Fresh, golden crystals in
0.26 0.41 and with chalcopyrite
0.28 0.44
FE14 5 Chalcopyrite p Mine Friedrich- 0.30 0.40 Fresh, golden mass in
Christian, 0.33 0.44 quartz
Wildschapbach 0.32 0.42
FE14 5 Pyrite p Mine Friedrich- 0.01 0.02 Fresh, golden crystals in
Christian, 0.03 0.01 and with chalcopyrite in
Wildschapbach 0.01 0.02 quartz
Fe5 6 Ba-pharmacosiderite s Mine Clara, Wolfach 0.20 0.27 2 mm large brown cubic
0.13 0.18 crystals on barite
0.16 0.22
FE6 6 Scorodite s Mine Clara, Wolfach 0.21 0.33 3 mm large bluish
0.26 0.40 crystals on
0.24 0.37 pseudomorphed
chalcopyrite
Fe1-1 17 Hematite p Mine Rappenloch, 0.29 0.43 Relatively large, fresh
Eisenbach 0.27 0.43 crystals
0.28 0.43
Fe4-1 8 Hematite sh Mine Anton, Heubach 0.83 1.23 Small, free, euhedral
near Wittichen 0.91 1.35 crystals, very thin, red, in
0.87 1.29 barite vugs
Repeat dissolution 0.83 1.19
0.79 1.17
0.81 1.17
Fe5-1 1 Hematite s Neuenbürg 0.44 0.66 Layer of red, feathery
0.44 0.66 crystals
0.44 0.66
(continued on next page)
Table 1 (continued)
FE3 6 Hematite s Mine Clara, Wolfach 0.62 0.89 Small, red, thin crystals
0.70 1.03 in vugs on siderite,
0.66 0.96 together with calcite
FLK12 18 Hematite p Lenzkirch 0.40 0.61 Large, fresh, euhedral
crystals up to 1 cm on
granite porphyry
FE1 19 Hematite p Sirnitz, Münstertal 0.49 0.74 Fresh, euhedral crystals
up to 2 mm on quartz-
cemented granite breccia
FLK20 18 Hematite p Lenzkirch 0.08 0.10 Large euhedral crystals
0.06 0.10 on granite porphyry,
0.07 0.10 very slightly oxidized
FE4 6 Hematite sh Mine Clara, Wolfach 0.89 1.33 2 mm large tabular
0.91 1.28 crystals on quartz
0.90 1.30
M590 7 Hematite sh Mine Sophia, Wittichen 0.85 1.16 Very tiny, red, free
0.77 1.09 crystals, on earlier and
0.81 1.13 below later generation of
siderite
FE8 3 Hematite s Kienberg near 1.06 1.65 Layer of red feathery
Freudenstadt 1.10 1.63 crystals on fibrous
1.08 1.64 goethite
PHB4 9 Hematite s Hohberg, St. Roman 1.48 2.19 Outermost layer of 3 cm
1.52 2.31 thick fibrous, radial
1.50 2.25 hematite (‘‘red
glaskopf’’)
PHB4 9 Hematite s Hohberg, St. Roman 0.58 0.83 Innermost core layer of
0.58 0.81 3 cm thick fibrous, radial
0.58 0.82 hematite (‘‘red
glaskopf’’)
METZ557 13 Hematite p Mine Eisenwand, Zell a. 0.08 0.16 Layer of up to 6 mm
H. 0.08 0.09 large crystals in quartz
0.08 0.12
HEB7 11 Hematite p Erlets near Hausach 0.42 0.55 Up to 5 mm large single
0.41 0.59 crystals in quartz vug
0.41 0.57
FE9 1 Hematite s Neuenbürg 0.24 0.34 Layer of red, feathery
0.25 0.35 crystals
0.24 0.35
FE11 12 Hematite p Mine Otto, 0.31 0.46 Up to 5 mm large, fresh,
Schottenhöfe, Zell a. H. 0.30 0.44 euhedral crystals on
0.30 0.45 microcrystalline
hematite
FE11 12 Hematite s Mine Otto, 0.98 1.47 Microcrystalline, dense
Schottenhöfe, Zell a. H. 0.97 1.43 hematite on barite
0.97 1.45
FE13 10 Hematite s Streckfeld near Wolfach 1.15 1.68 Layer of red, feathery
1.17 1.75 crystals on goethite
1.16 1.72
GMS04 20 Hematite p Menzenschwand 0.04 0.07 Large, fresh, euhedral
crystals in vug on granite
FE15 17 Hematite p Mine Rappenloch, 0.06 0.09 Layer of up to 5 mm
Eisenbach 0.03 0.02 large, fresh crystals
0.04 0.03
FE15 17 Hematite p Mine Rappenloch, 0.10 0.17 Fresh, microcrystalline
Eisenbach 0.13 0.18 layer
0.11 0.18
FE16 17 Hematite s Mine Rappenloch, 0.80 1.15 Fibrous crystals (‘‘red
Eisenbach 0.76 1.10 glaskopf’’), layered
0.78 1.12
FE17 16 Hematite p Farrenberg near 0.03 0.04 Layer of up to 3 cm
Eisenbach 0.00 0.01 large, fresh crystals on
0.01 0.01 granite
(continued on next page)
3018 G. Markl et al. 70 (2006) 3011–3030
Table 1 (continued)
FE18 14 Hematite p Lägerfelsen, Triberg 0.30 0.43 Layer of crystals up to
0.28 0.43 5 mm on rhyolite
0.29 0.43
FE19 15 Hematite p Fahlenbach near 0.56 0.81 Feathery crystals in
Eisenbach 0.50 0.78 barite on granite
0.53 0.79
FE20 2 Goethite s Mine Schrotloch, 0.80 1.18 Fibrous crystals (‘‘brown
Hornisgrinde 0.81 1.20 glaskopf’’)
0.81 1.19
FE21 1 Goethite s Neuenbürg 0.64 0.93 Fibrous crystals (‘‘brown
0.64 0.96 glaskopf’’)
0.64 0.94
Fe7 6 Goethite s Mine Clara, Wolfach 0.80 1.22 Fibrous mass (‘‘brown
0.80 1.21 glaskopf’’), shiny black
0.80 1.21 on the surface, fibres up
to 5 mm long
QFS26 4 Goethite s Mine Dorothea, 0.51 0.70 Fibrous mass (‘‘brown
Freudenstadt 0.56 0.76 glaskopf’’), fibres up to
0.53 0.73 2.5 cm long
Repeat 0.50 0.76
dissolution 0.48 0.74
0.49 0.75
WEIN2321 12 Goethite s Mine Otto, 1.13 1.68 Fibrous goethite
Schottenhöfe, Zell a. H. 1.11 1.63 (‘‘brown glaskopf’’) on
1.12 1.66 barite
FE8 3 Goethite s Kienberg near 0.52 0.82 Fibrous, brown
Freudenstadt 0.53 0.78 glaskopf, below hematite
0.52 0.80
Fe2-1 24 Goethite Donaueschingen, E of 0.00 0.02 1 mm large oolites from
Schwarzwald 0.04 0.07 sedimentary iron ore
0.02 0.02
Fe3-1 23 Goethite Kandern, W of 0.01 0.01 1 mm large oolites from
0.00 0.01
FE22 26 Goethite Geisingen, Baar, E of 0.02 0.05 1 mm large oolites from
0.05 0.06
0.11 0.17
0.05 0.08
FE23 25 Goethite Achdorf, Wutach E of 0.44 0.63 1 mm large oolites from
0.42 0.62
FE24 21 Goethite Ringsheim near Lahr, W 0.06 0.03 1 mm large oolites from
of Schwarzwald 0.06 0.06 sedimentary iron ore
0.06 0.04
FE25 22 Goethite Schliengen S Freiburg, 0.28 0.38 1 cm large oolites from
W of Schwarzwald 0.31 0.43 sedimentary iron ore
0.30 0.41
FE26 27 Goethite Thalheim near 0.27 0.35 1 mm large oolites from
Tuttlingen, E of 0.31 0.44 sedimentary iron ore
Schwarzwald 0.29 0.39
FE27 23 Goethite Kandern S Freiburg, E 0.09 0.07 1 cm large oolites from
of Schwarzwald 0.11 0.14 sedimentary iron ore
0.10 0.10
FE28 Sandstone Schramberg 0.22 0.35 Whole-rock analysis,
brick-red, fine-grained
sandstone
FE29 Shale Dotternhausen 0.21 0.28 Whole-rock analysis,
Lias e, black shale
FE32 Shale Holzmaden 0.03 0.04 Whole-rock analysis,
Lias e, black shale
p denotes primary minerals, s secondary minerals formed under supergene conditions and sh denotes secondary minerals formed at higher than supergene
temperatures.
hematite, the latter two in their fibrous form (‘‘glas- only briefly repeated here. Mineral separates were first
kopf’’, Fig. 3d–f). All these phases formed by weath- decomposed in 6 M HCl. Iron was separated from the min-
ering and oxidation of the primary iron ores in the eral matrix by anion exchange chromatography and
hydrothermal veins. In some cases, the secondary checked for purity by ICP-OES. Isotope measurements
minerals are pseudomorphs that formed by in situ were performed on a ThermoFinnigan Neptune instrument
replacement of the primary minerals. In other cases, using the sample-standard-sample bracketing technique. A
the iron was transported over some millimeters, cen- detailed description of the chemical and mass-spectromet-
timeters or meters before it was re-precipitated. ric procedures is presented by Schoenberg and von Blanck-
Hence, the secondary minerals occur both directly enburg (2005). They found that the overall 95% confidence
on the relicts or close to the relicts of the primary interval reproducibility is ±0.049& (d56Fe) and 0.071&
ones. In Table 1, minerals separated from the same (d57Fe). The isotope data shown in Table 1 contain both
hand specimen are indicated by a common sample mass-spectrometric duplicates (repeat analyses of the same
number. They were typically sampled millimeters to purified Fe aliquot) and real sample duplicates (repeat
a few centimeters from each other. decompositions of separate sample powders). The mass-
(4) Iron-bearing sediments from the Buntsandstein, the spectrometric duplicates reproduce within the ranges given
Lias and the Dogger were analyzed to constrain the above, while repeat decompositions might exhibit sample
variations in Fe isotope composition of the sedimen- heterogeneity. This issue will be addressed elsewhere, but
tary cover sequence, which may have acted as source samples from Hohberg (PHB4) and Rappenloch (Fe15
of Fe for some of the post-Variscan veins in the base- and Fe16) give an indication of heterogeneity within and
ment (Fig. 2). The Dogger sediments are coastal between samples. All isotope ratios are reported relative
oolithic limestones which host large amounts of to the IRMM-14 Fe isotope standard. The 56Fe/54Fe ratio
iron-oxides and -hydroxides. Their depositional age has been converted into d56Fe according to
overlaps with the radiometric ages of some of the mas-
56
sive hydrothermal hematite veins in the Schwarzwald. 56 Fe=54 Fesample
d Fesample ¼ 56 Fe=56 Fe
1 100
IRMM-14
4. Analytical procedures and results and the d57Fe likewise.
4.1. Analytical procedures 4.2. Results
From all samples except the sedimentary oolitic iron The iron isotope compositions are listed in Table 1 and
ores, of which whole-rock powders were used, hand-picked displayed in Fig. 4. The d56Fe values vary between about
mineral separates were used for analysis. All measurements 1.5& and +0.9&, hence over almost 2.5 delta units.
were performed at the geochemistry laboratory of the Uni- The isotope ratios are distinct between and within mineral
versity of Hannover. The analytical method is explained in groups. They are also different and much more variable
detail by Schoenberg and von Blanckenburg (2005), and is than the +0.04% to +0.38% range (relative to IRMM-14)
Fig. 4. Iron isotopic compositions of primary and secondary iron minerals from hydrothermal veins, and of sediments from the Schwarzwald area.
3020 G. Markl et al. 70 (2006) 3011–3030
found for igneous rocks (Beard et al., 2002; Poitrasson and atively large distances (the samples were deposited 50 km
Freydier, 2005) or the 0.2& to +0.3& range found for apart from each other).
clastic sediments and river suspended loads (Beard and In addition to the oolitic Dogger ores, three whole-rock
Johnson, 2004). samples of iron-bearing Mesozoic sediments from the
Schwarzwald area have been analyzed. While a Buntsand-
4.2.1. Primary and secondary oxidic iron minerals from stein (sandstone, sample FE28) has a d56Fe value of
hydrothermal veins +0.22&, two black shales from the Schwäbische Alb (sam-
Large variations of their d56Fe values were measured in ples FE29 and FE32) have values of 0.21 and 0.03,
both primary and secondary Fe(III)-oxides and -hydrox- respectively. Hence, the shales show similar isotopic varia-
ides (goethite and hematite). Primary hematite crystals tions—but shifted to slightly lighter d56Fe values—as the
have d56Fe values between 0.49& and +0.53&, whereas oolitic Dogger ores. Overall, the d56Fe range measured in
secondary hematite displays two sample groups: hematite sediments is within the range observed for modern clastic
formed during supergene alteration (group ‘‘s’’ in Table marine sediments (0.2& to +0.3&, Beard and Johnson,
1) contains light iron throughout with d56Fe values be- 2004). Basement granites and gneisses from the study area
tween 1.5& and 0.58&, while secondary hematite have not been analyzed for their Fe isotopic composition,
formed at high temperatures (group ‘‘sh’’ in Table 1) shows but in general, crystalline basement rocks have d56Fe values
d56 Fe between +0.81& and +0.9&. Within one-layered fi- around +0.1& to 0.38& (Beard et al., 2002; Poitrasson
brous hematite sample (Fig. 3d), d56Fe values vary over a and Freydier, 2005).
distance of about 3 cm from core to rim between 0.58&
and 1.5& (sample PHB 4 in Table 1). Fibrous goethite 5. Thermodynamic modelling of iron leaching, precipitation
ranges between 1.12& and 0.51&, very similar to and alteration
low-T hematites.
The data indicate that various processes contribute to
the change in iron isotope compositions. In principle, three
4.2.2. Primary and secondary non-oxidic iron minerals from stages can be distinguished, during which fractionation can
hydrothermal veins occur, which are leaching, precipitation and supergene
Pyrite, chalcopyrite and siderite show relatively small alteration. Considering that redox reactions play an impor-
variations in their iron isotopic compositions, but exhibit tant role during these processes, particularly for iron iso-
large differences to other minerals from the same vein. tope fractionation, we have thermodynamically modeled
Coexisting pyrite and chalcopyrite from the Clara deposit various reaction parameters and iron speciation during
have indistinguishable iron isotopic compositions of about (1) leaching of iron from crystalline basement rocks
d56Fe = 0.30&, whereas samples from the Friedrich- through fluid–rock interaction (Figs. 5 and 6), (2) precipi-
Christian deposit show a difference between pyrite tation of Fe minerals through fluid mixing in the vein
(d56Fe = 0.01&) and chalcopyrite (0.32&). The chalco- deposits (Fig. 7), and (3) supergene weathering of siderite
pyrite value is almost identical in both deposits. The side- through interaction with oxidizing surface-derived waters
rites from three different deposits (including the Clara (Fig. 8). This involved calculations of phase diagrams
deposit) have d56Fe values in the range between 1.36& and various reaction path simulations in the model system
and 0.72&; the Clara siderites contain iron with d56Fe Na–K–Ca–Fe–Al–Si–C–Cl–O–H and several subsystems.
of 1.0& (fresh siderite) and 0.8& (slightly oxidized to We have not included sulfur into the model system for
goethite). A similar pattern is observed at Neuenbürg, two main reasons. First of all the solubility of Fe in hydro-
where the d56Fe value of fresh siderite is 0.83 while that thermal fluids with high salinities is mostly controlled by Cl
of the oxidized sample is 0.72&. Three samples from complexes. The concentration of sulfur is several orders of
the Sophia mine near Wittichen show d56Fe of 1.36& magnitude smaller, resulting in only a subordinate control
(fresh siderite I), 0.72& (fresh siderite II) and 0.85& on Fe solubility. Second, no reliable thermodynamic data
(slightly oxidized siderite, probably of a second genera- for reduced Fe–S species such as Fe(HS)+ or Fe(HS)2 are
tion). Small variations between 0.16& and 0.24& are available and there are only data for one oxidized Fe–S
displayed by the secondary Fe(III)-arsenates Ba-pharma- species, the FeSO4 0 complex.
cosiderite and scorodite, of which, however, only two sam-
ples were analyzed. 5.1. Setup of the thermodynamic model
4.2.3. Sediments The calculations were carried out with the HCh software
Relatively small variations are observed in goethite from package (Shvarov and Bastrakov, 1999), which models het-
the Dogger oolitic iron ores, with d56Fe values being be- erogeneous equilibria and reaction progress by minimiza-
tween +0.02& and +0.42&. The range indicates that the tion of the Gibbs free energy of the total system
iron source of these sediments was rather homogeneous (Shvarov, 1978, 1981). This package contains an extensive
and that the isotope fractionation during low-temperature data management facility, which allows efficient calculation
precipitation was similar in coastal environments over rel- of the thermodynamic properties of mineral-fluid reactions.
Fig. 6. Predominance diagram of aqueous Fe species in pH—log aCl

space, constructed at 300 °C, 2.5 kbar and an H2O activity of 0.66. The
log f O2 was fixed at 31, which conforms to the lower endpoint of the
alteration trend (see Fig. 5). Increasing the log f O2 to 30 (upper endpoint
Fig. 5. The log f O2 —pH diagram constructed at 300 °C, 2.5 kbar and an of the alteration trend) essentially results in the same topology, with only
H2O activity of 0.66, showing mineral stability relationships in the system the field of FeO disappearing. At the conditions of hydrothermal fluid–
K–Fe–Si–Al–O–H. The thin solid line shows the magnetite–hematite rock interaction in the basement of the Schwarzwald district, FeCl2 is by
transition, whereas the thick solid and thick dotted lines delineate stability far the predominant aqueous Fe species. Note that the stoichiometries of
fields of silicate minerals at activities of K+ of 0.1 and 0.01, respectively. aqueous species are given in the notation used by Shock et al. (1997).
The alteration trend (grey arrow) indicates the concurrent change in pH
and oxidation state via progressive hydrothermal alteration (transition metric and LA-ICP-MS data; Markl, unpubl. data) with
from rock-buffered to fluid-buffered conditions) of the basement granites/ this average granite at 300 °C and 2.5 kbar. Several differ-
gneisses. Note that at 300 °C and 2.5 kbar the neutral point corresponds to
ent meteoric end-member fluid compositions were modeled
a pH of 4.95.
by equilibrating pure water at 50 °C and 1 bar with (1) cal-
cite, at pO2 and pCO2 constrained at atmospheric values,
Thermodynamic data for aqueous species were essentially (2) calcite, at atmospheric pCO2, and (3) calcite only. Sim-
taken from the SUPCRT92 database and subsequent up- ulations of the alteration of siderite assumed that the oxi-
dates (Johnson et al., 1992; Shock et al., 1997; Sverjensky dizing surface-derived water was pure water at 30 °C and
et al., 1997), with additional data for several aluminum spe- 1 bar, with the dissolved oxygen content constrained from
cies from Tagirov and Schott (2001). Thermodynamic data equilibration with O2 gas at atmospheric partial pressure.
for silicate, oxide, hydroxide and carbonate minerals were
taken from the internally consistent dataset of Holland 5.2. Results of the calculations
and Powell (1998). All calculations of individual activity
coefficients of aqueous species applied an extended De- 5.2.1. Leaching of iron from crystalline basement rocks
bye-Hückel model using the b-gamma equation for NaCl Leaching of iron from the basement was modeled using
as the background electrolyte (Oelkers and Helgeson, a typical granitic assemblage of K-feldspar, quartz and two
1990; Shock et al., 1992). micas. At 2.5 kbar and 300 °C, K-feldspar is stable with
For the calculation of fluid–rock equilibria, whole-rock either muscovite or biotite (annite) only (Fig. 5), but chlo-
analyses of typical granites from the Schwarzwald area rite (daphnite) is a typical stable phase. At these conditions,
(Emmermann, 1977) were used to constrain the bulk rock a more-or-less neutral fluid (i.e., pH around 4.9) with real-
composition of the crystalline basement. This composition istic K+ activities between 0.1 and 0.01 would be in the sta-
was then recalculated to the system Na–K–Ca–Fe–Al–Si– bility field of muscovite. Rock-buffering would move it to
Cl–O–H, which represents a simplified model of the rock the invariant point defined by K-feldspar, chlorite (daph-
composition (excluding Ti and Mn). The starting composi- nite) and muscovite (plus quartz), where magnetite would
tion of the brine end-member used in the fluid mixing cal- be the stable iron oxide phase (Fig. 5). Through progressive
culations was obtained by equilibrating a fluid having total fluid–rock interaction, slightly to strongly acidic and/or
solute concentrations and major cation ratios (K/Na and oxidized fluid batches penetrating the rock along previous-
Ca/Na) typical of the post-Variscan hydrothermal system ly reacted fractures would drive the fluid composition into
in the Schwarzwald district (as derived from microthermo- the hematite stability field and would finally arrive at the
3022 G. Markl et al. 70 (2006) 3011–3030
Fig. 7. Representative results of reaction path modeling for mixing of a deep saline brine (fluid1) with surface-derived meteoric water (fluid2). The starting
composition of the brine end-member was obtained by equilibrating a fluid having total solute concentrations and major cation ratios (K/Na and Ca/Na)
typical of the post-Variscan hydrothermal system with an average Schwarzwald granite at 300 °C and 2.5 kbar. The meteoric end-member chemistries were
modeled by equilibrating pure water at 50 °C and 1 bar with calcite, at pO2 and pCO2 constrained at atmospheric values (model 1, left column), and
calcite, at atmospheric pCO2 (model 2, right column). The composition of the mixed fluid was then calculated as a function of the mass fraction of fluid2,
defined as mfluid2/(mfluid1 + mfluid2), at a fixed pressure of 500 bar. It can be seen that in model 1 hematite is the stable iron phase formed throughout most
of the simulation, whereas in model 2 magnetite is formed up to a mass fraction of fluid2 of about 0.75, where it is subsequently replaced by siderite and
ankerite. The results of the simulations represent two end-member scenarios, which correspond to the two typical vein assemblages found in the
Schwarzwald district.
invariant point defined by pyrophyllite, muscovite and

daphnite (always plus excess quartz). This means that the
system would evolve from rock- to fluid-buffered and dur-
ing this evolution, where hematite would become the stable
iron oxide phase in the rocks. This is indeed what is ob-
served within alteration zones in the hydrothermally over-
printed Schwarzwald granites.
A fluid having a composition between these two invari-
ant points, i.e., along this alteration path, would have an
iron speciation displayed in Fig. 6. At an oxygen fugacity
of 1031 bar and at acid to neutral pH, the dissolved iron
is almost exclusively present in its divalent state. This pic-
ture does not change at slightly elevated fO2 values (e.g.,
at an fO2 of 1030 bar only the stability field of the FeO
aqueous species disappears, but the rest of the topology re-
mains identical). Depending on chlorinity, Fe2+, FeCl+ or
FeCl2 are the dominant aqueous iron species. For the
hydrothermal solutions responsible for granite alteration
in the Schwarzwald, having high salinities between 20
and 40 wt% NaCl equivalent, FeCl2 is the predominant
aqueous iron species outweighing other species by about
two orders of magnitude. Biotite or chlorite would be the
main source of iron in the fluid and this, in turn, means
that iron leached from the basement in 7–8 km depth and
at 300 °C would be present as divalent iron in solution
and no redox process would be involved during leaching.
5.2.2. Fluid mixing with surface water

This process is supported by a line of evidence presented
by Schwinn et al. (2006). The hot, high-salinity hydrother-
mal fluid rises quickly from 7 to 8 km depth to a depth of
about 1.5 km, where the actual precipitation of the hydro-
thermal minerals occurs by mixing with an oxidized sur-
face-derived meteoric water of low temperature. The
solubility of Fe(II) strongly depends on chlorinity of the
fluid and on temperature (Seward and Barnes, 1997; Wood
and Samson, 1998; Yardley, 2005). Hence, dilution of the
highly saline brine by low-salinity surface water would pro-
vide a plausible mechanism for iron ore precipitation.
However, redox effects may be involved in such precipita-
tion processes. Therefore, the mixing process is modeled
for two end-member scenarios, an oxidized one (left col-
umn in Fig. 7; surface water in equilibrium with atmo-
spheric oxygen) and a reduced one (right column in
Fig. 7; surface water not in equilibrium with atmospheric
oxygen; fO2 is buffered by H2O dissociation only). It is
obvious that a fluid in equilibrium with calcite (i.e., reacted
Fig. 8. Results of the simulation of weathering of siderite by surface- with the Mesozoic sediments covering the Schwarzwald)
derived water at an increasing water/rock ratio. The calculations were and atmospheric O2 and CO2 will precipitate mostly hema-
performed assuming that the oxidizing surface-derived water was pure tite during this mixing process. The Fe(II)/Fe(III) ratio
H2O at 30 °C and 1 bar, with the dissolved oxygen content constrained by changes from Fe(II)-dominated in the deep, hot fluid to
equilibration with O2 gas at atmospheric partial pressure. The amount of
goethite formed at the expense of siderite increases with increasing water/
Fe(III)-dominated in solutions with a significant propor-
rock ratio until all siderite is consumed; this results in a sharp decrease of tion of meteoric water. The total iron content drastically
the total dissolved iron coupled with an increase in the oxidation state in drops at a mixing ratio of about 0.7 (brine/meteoric). Inter-
the water. estingly, the fluid mixture has a temperature of 120–150 °C
at this point, which conforms to the mode of fluid inclusion
homogenization temperatures in the Schwarzwald veins.
3024 G. Markl et al. 70 (2006) 3011–3030
The precipitation process of the relatively reduced fluid perature leaching of iron from the basement. We will now
is completely different. At high-brine/meteoric fluid ratios, proceed to summarize the isotope fractionation effects
magnetite would precipitate, while at fluid compositions relevant to these processes. A prerequisite to this effort is
dominated by the meteoric fluid, siderite and ankerite form an inventory of the essential iron isotope fractionation fac-
instead. Their formation is not predicted for the oxidized tors relevant to these processes that have been published to
case, which indicates that the commonly observed siderite date.
in the later stages of the Schwarzwald hydrothermal veins
was formed from more reduced and/or CO2-enriched flu- 6. Iron isotope systematics
ids. This mixing process is much less effective in precipitat-
ing the iron and the Fe(II)/Fe(III) ratio is always 6.1. Iron isotope fractionation factors
dominated by Fe(II). The pH curves for both mixing mod-
els are very similar. The models show that in the hematite- The determination of fractionation factors for iron is, as
bearing veins significant oxidation occurs during the pre- for all other transition metal isotopes, still in its infancy.
cipitation process, whereas no oxidation occurs during pre- The available data are patchy and mostly not applicable
cipitation of siderite. While a homogeneous source for the to natural systems in a straightforward manner. The cur-
deep fluid can be inferred, the meteoric fluid is most prob- rent status has recently been reviewed by Anbar (2004),
ably of slightly to strongly variable composition. Its oxida- and Johnson et al. (2004a). Basically, three independent
tion state may be dependent on possible interactions with approaches exist, which are (1) theoretical molecular parti-
organic-rich black shale sediments in the Mesozoic cover tion functions, calculated from molecular and condensed-
sequence of the Schwarzwald and the depth of the equili- phase vibrational frequencies reviewed by Schauble
bration with the host rocks (at oxic or anoxic conditions). (2004), (2) laboratory experiments under conditions that
are as close as possible to those in nature, and (3) inferring
5.2.3. Supergene weathering of siderite deposits empirical fractionation factors from natural observations.
Low-temperature alteration of iron-bearing ore deposits One of the principle obstacles in the experiments is estab-
is a final process which can result in the fractionation of lishing whether chemical equilibrium has been fully
iron isotopes. Specifically, we have modeled supergene achieved, or whether measured isotope differences are due
alteration of a siderite deposit through reaction with oxi- to kinetic effects (see review by Beard and Johnson,
dizing surface-derived water at increasing water/rock (w/ 2004). The empirical approach makes use of known mass
r) mass ratios (Fig. 8), i.e., at each step a fixed amount of fluxes, and known isotope ratios of influxes, effluxes and
fresh siderite was equilibrated with an increasing amount the compartments involved. Fractionation factors are cal-
of the aqueous solution. This appears applicable to quite culated using Rayleigh-type mass balance equations. Often,
a number of veins in the northern and central Schwarzwald natural isotope fractionation involves a series of fraction-
(e.g., Neuenbürg, Freudenstadt, Otto mine). In most of ation steps, and only a cumulative fractionation factor is
these deposits, earlier siderite is almost completely replaced obtained that includes the sum of all steps. Here, the cur-
by botryoidal hematite or goethite. With increasing water/ rent status of fractionation factors relevant to this study
rock ratio, the amount of goethite formed at the expense of is reviewed in detail. Rather than endorsing a particular
siderite steadily increases until all siderite is consumed; this fractionation factor it is deliberately left to the reader to as-
occurs at a log (w/r) of about 3.9 (Fig. 8). As long as both sess the uncertainty introduced into the interpretation.
siderite and goethite are stable, the total dissolved Fe con-
centration, the Fe(II)/Fe(III) ratio, the oxygen fugacity and 6.1.1. Fractionation among aqueous species
pH are buffered by this two-phase assemblage and remain The most prominent and best-documented equilibrium
constant. Total consumption of the siderite results in a very fractionation factor is that of Fe(II) oxidation in aqueous
sharp transition, which is characterized by a decrease of the Fe-containing solutions. Schauble et al. (2001) have deter-
total dissolved Fe concentration by about 6 orders of mag- mined a theoretical DFe(III)aq–Fe(II)aq value of +6& for the
nitude, an increase in oxygen fugacity coupled with a de- equilibrium reaction between the Fe(II)-hexaquo complex
crease in the Fe(II)/Fe(III) ratio, and a decrease of the and the Fe(III)-hexaquo complex at 25 °C, and of +1.5&
pH from about 5.8 to 4.7 (Fig. 8). With further increase at 250 °C value (Dproduct–reactand; here and in the following,
of the water/rock ratio, the total dissolved Fe concentra- the isotope fractionation DB–A refer to the 56Fe/54Fe iso-
tion decreases slightly, reflecting the effect of progressive tope ratio with A being the reactand compartment and B
dilution of the weathering solution. If similar calculations being the product compartment). These authors used
are performed at temperatures above 55 °C at atmospheric vibrational spectroscopic data for a number of Fe complex-
pressure, hematite substitutes goethite as the reaction es and an empirical force field model (Modified Urey-Brad-
product. ley Force Field ‘‘MUBFF’’). Anbar et al. (2005) have used
In summary, the results of the thermodynamic calcula- Density Functional Theory (‘‘DFT’’) and obtained
tions demonstrate that iron isotope fractionation governed DFe(III)aq–Fe(II)aq of +3.2& for the equilibrium reaction be-
by redox reactions is likely to occur during the precipita- tween the Fe(II)-hexaquo complex and the Fe(III)-hexaquo
tion and the alteration processes, but not during high-tem- complex at 25 °C, and of +1.2& at 250 °C (Anbar et al.,
2005). The isotope fractionation between various Fe(III)- iron-reducing bacteria to induce the reduction. The relative
chloro complexes and the Fe(II)-hexaquo complex is pre- direction of this reductive fractionation is opposite to that
dicted to be of the same direction, but smaller. For exam- determined for iron oxidation, and is also similar in
ple, a d value of 4& was predicted for the equilibrium magnitude (Johnson et al., 2004b). These results are entire-
reaction between FeðIIIÞðH2 OÞ4 Cl2 þ complex and the ly consistent with the general observation of redox
Fe(II)-hexaquo complex at 25 °C, and of ca. 1& at reactions producing the largest, and possibly reversible,
250 °C, respectively (Schauble et al., 2001). A fractionation iron isotope effects. A non-redox-related fractionation
of about half that magnitude was predicted for the reaction factor is that of Fe(II)aq adsorption onto a solid, for which
between FeðIIIÞCl4 and the Fe(II)-hexaquo complex an experimental DFe(II)sorbed–Fe(II)aq value of 2& was
(Schauble et al., 2001). Iron in the FeCl2 complex is ca. determined (Johnson et al., 2005). This, again, is interpret-
2& lighter than the Fe(II)-hexaquo complex at 25 °C, ed to reflect a kinetic effect. A very different result was
and ca. 1& lighter at 250 °C (Schauble et al., 2001). The obtained for adsorption of Fe(II) onto goethite, which
oxidation of ferrous iron has been also calibrated in labo- resulted in DFe(II)sorbed–Fe(II)aq of +1.5& to +2.5& (Icopini
ratory experiments, and the probably best-calibrated Fe et al., 2004). The disagreement between these two estimates
fractionation factor resulted (reviewed by Beard and John- is still a matter of intense debate (Johnson et al., 2005).
son, 2004). The equilibrium reaction between the reactand Finally, the precipitation of siderite from Fe(II)aq was
Fe(II)aq and the product Fe(III)aq led to a DFe(III)aq–Fe(II)aq predicted from spectroscopic data to result in siderite that
value of +2.9& at 25 °C (Welch et al., 2003). This fraction- is approximately 1.5& lighter than the iron dissolved in the
ation factor is in excellent agreement with the DFT-derived fluid (Polyakov and Mineev, 2000; Schauble et al., 2001). A
theoretical value (Anbar et al., 2005), but it is half of that preliminary experimental determination of this isotope
predicted by MUBFF (Schauble et al., 2001). This reaction fractionation resulted in Dsiderite–Fe(II)aq of 0.5& (Wiesli
has not yet been calibrated at higher temperatures. . . et al., 2004).
The fractionation upon weathering was simulated exper-
6.1.2. Fluid–solid fractionation imentally by exposing hornblende and goethite to a variety
Theoretical equilibrium fractionation factors using a of leaching reagents (acetate, oxalate, citrate, and the sider-
combination of DFT (Anbar et al., 2005) and Möss- ophore desferrioxamine mesylate; Brantley et al., 2001;
bauer-derived (Polyakov, 1997) fractionation factors have Brantley et al., 2004). Leaching of hornblende released flu-
been determined for the formation of hematite and goethite ids that were lighter by 0.1–1.2&, while no fractionation
from the Fe(III)-hexaquo complex (Anbar et al., 2005). occurred upon leaching of goethite (Brantley et al., 2004).
The Dhematite–Fe(III)OOH is 1.8& at 25 °C, and 0.6& at
250 °C, whereas the Dgoethite–Fe(III)OOH is 3.5& at 25 °C, 6.1.3. Solid–solid fractionation
and ca. 1.3& at 250 °C, respectively. The light Fe isotope These fractionation factors are entirely based on predic-
composition of hematite was experimentally predicted by tions from spectroscopic data (Polyakov, 1997). For the
Skulan et al. (2002). However, fractionation between minerals measured in this study the following equilibrium
Fe(III)aq and hematite is strongly reaction rate-dependent. fractionation factors are predicted at 25 °C presented rela-
At the fastest reaction rates the hematite was up to 1& tive to Fe(III)aq = 0& using the reduced partition function
lighter than the fluid, but when extrapolated to infinitely of Anbar et al. (2005): siderite = 4.9&; goe-
slow precipitation rates (assuming that this would present thite = 3.4&; hematite = 1.9&; pyrite = +1.8&. At
the equilibrium fractionation), the fractionation was zero. 250 °C, the sequence is siderite = 1.8&; goe-
This was explained by a rate-dependent kinetic effect thite = 1.3&; hematite = 0.6&; pyrite = +0.7&.
(Skulan et al., 2002). The disagreement between the theo-
retical equilibrium effects, predicting light Fe in hematite, 6.2. Iron isotope fractionation pathways
and the experimental results, predicting light Fe only
during a kinetic fractionation, might be explained by an We can now proceed to evaluate the measured data in
inaccuracy in the Mössbauer-derived mineral fractionation view of these fractionation factors. We do this by first
factors, or in the experiments (Anbar et al., 2005). Bullen establishing possible iron isotopic compositions of the
et al. (2001) have measured a bulk fractionation factor hydrothermal fluid. Then, we proceed to discuss how
DFe(III)OOH–Fe(II)aq of +1& for the oxidation of Fe(II)aq chemical reactions in the fluids are involved in the modifi-
to Fe(III)OOH in both laboratory experiments and a nat- cation of these compositions.
ural stream situation. This bulk fractionation is compatible
with the combination of fractionation effects by (1) forma- 6.2.1. Leaching of iron from the basement
tion of a dissolved oxidation product, i.e., Fe(III) species in Silicate rocks usually contain unfractionated iron
solution, enriched in 56Fe, followed by (2) kinetically (d56FeIRMM14 = +0.09&; Beard et al., 2002). Recently,
controlled precipitation of a solid Fe(III)OOH depleted some heavily differentiated granites have been measured
in 56Fe (Beard and Johnson, 2004). The equilibrium of which d56Fe ranges between +0.1& and +0.38& (Poi-
reduction of solid ferric iron cannot be simulated easily, trasson and Freydier, 2005). This is the range we assume
and preliminary experiments have involved dissimilatory for the composition of leached basement rocks. The exper-
3026 G. Markl et al. 70 (2006) 3011–3030
imental fractionation factors outlined above appear to occurs if the fraction of the oxidizing surface fluid (fluid2
indicate that leaching of silicates would produce fluids that on Fig. 7) is larger than 70%, and most of the Fe dissolved
are slightly depleted in 56Fe over 54Fe by 1& at the most, in the fluid would be transferred into hematite.
but that leaching of oxides would not entail an isotope We illustrate the composition of a solid formed during a
fractionation. For the case of hydrothermal leaching of given mineralization event at various degrees of Fe(II)aq
Fe from rock-forming basement minerals such as amphi- remaining in solution after oxidation in a Rayleigh-type
bole, biotite, Fe-oxides or pyroxenes appears to result in calculation (Fig. 9). The grey curve gives an example for
Fe(II)aq of slightly lighter compositions (Rouxel et al., bulk hematite (Dhematite–Fe(II)aq = +1.5) precipitating from
2003). Quantitative hydrothermal dissolution of silicates, an oxidized Fe(II) fluid with an initial d56Fe of 0.5&.
however, would result in an iron-bearing fluid of unfrac- This resulting hematite has a composition of between +1
tionated compositions. Overall, fluid–rock interaction (the fraction of oxidized and precipitated Fe(II)aq is small)
involving a relatively reduced fluid would be expected to re- and 0.5& (all Fe(II)aq oxidized and precipitated). This is
sult in a Fe isotope composition of around 0& to 0.5& exactly the range observed for primary hematite. In this
(maximum 1.0&). simple scenario, the Fe isotope compositions would serve
The thermodynamic modeling has shown that all lea- as proxy for the degree of fluid oxidation and Fe precipita-
ched Fe is in the ferrous state, and that the vast majority tion. However, more complex scenarios are also possible in
of leached fluids is within the FeCl2 predominance field which the ferrous Fe fluid is oxidized in separate stages.
(Fig. 6). FeCl2 would fractionate Fe by ca. 1& relative Then, the later generations of hematite would precipitate
to the Fe(II)-hexaquo complex. However, if the transfer from residual fluids depleted in heavy Fe during previous
of iron from basement to fluid entails immediate and oxidation events. The bulk hematites formed would inte-
quantitative formation of the chloro complex, the fluids grate over a series of compositional increments (indicated
would preserve the isotopic composition of the leached by the progressively evolving instantaneous precipitate).
iron without further modification. The high salinities of Based on this model, it is not surprising that an individual
20–40 wt% NaCl equivalent further ensure a high degree mineral from one sample may show highly variable iron
of mobility of these fluids (Yardley, 2005). We would, isotopic compositions, even in the primary, high-tempera-
therefore, expect a high availability of iron having a rela- ture minerals. For example, primary, euhedral hematite
tively homogenous isotopic composition in the fluid. The from the Rappenloch deposit shows surprisingly variable
large range of observed inter-mineral and within-mineral iron isotopic compositions. In a single hand-sized layered
variations in the hydrothermal veins, however (Fig. 4), ore specimen larger crystals in a vuggy layer have a d56Fe
which are furthermore not in the sequence as suggested of +0.04&, whereas a microcrystalline layer has a value
from inter-mineral fractionation factors, seem to suggest of 0.11&. Another typical coarse-grained primary hema-
that these fluids are strongly modified by secondary pro- tite from the same location has a value of 0.28&.
cesses. Minerals precipitated from these fluids trace these
fluids’ composition.
6.2.2. Fluid mixing with surface water

Partial oxidation of fluids by oxidized surface waters
would form Fe(III)aq that should be around 3& heavier
at 25 °C, or 1.5& heavier at 250 °C. The model for O2-sat-
urated fluids (left column in Fig. 7) show that hematite
forms at this stage. Fractionation factors for hematite for-
mation suggest slightly lighter compositions for hematite
than the Fe(III)aq. Therefore, an overall maximum frac-
tionation of +1.5& would be expected for the solids pre-
cipitating from the Fe(II)-bearing solutions. Given that
primary fluids are predicted to have compositions of
0.5& to 0&, hematite precipitating at equilibrium is pre-
dicted to contain iron with d56Fe of ca. +0.5& to +1.0&.
In our samples, the range of measured d56Fe in primary
Fig. 9. Rayleigh-type calculation showing the composition of a solid
hydrothermal hematite is larger with 0.5& to +0.5&. formed at various degrees of Fe(II)aq remaining in solution after
These in part lighter compositions can be explained by oxidation. The grey curve gives an example for bulk hematite
mass balance effects. Only if the fraction of Fe(II) oxidized (Dhematite–Fe(II)aq = +1.5) precipitating from an oxidized Fe(II) fluid with
and precipitated is small, compositions that are close to the an initial d56Fe of 0.5&. This hematite obtains an integrated composition
of between +1 (fraction Fe(II)aq oxidized and precipitated is small) and
predicted fractionation factors would result. If the fraction
0.5& (all Fe(II)aq oxidized and precipitated). The later generations of
of Fe oxidized and precipitated is large, the hematite hematite would precipitate from residual fluids depleted in heavy Fe
formed will obtain a composition close to that of the fluid. during previous oxidation events (instantaneous precipitate). The remain-
Indeed it has been shown that a significant drop in Fe(II) ing fluid will always have the composition of the lower, dashed curve.
The setting for carbonate and sulfide precipitation is duced, CO2-bearing fluid. The rather high d56Fe value of
quite different. These do not involve an oxidation step. the hematite is confirmed by texturally identical hematite
The right column in Fig. 7 shows that siderite precipitates from the mineralogically similar Anton mine (d56Fe:
from reduced surface fluids that have dissolved carbonate +0.84&) which is situated close to the Sophia mine and
from overlying sediments and are enriched in HCO3 and by a texturally similar hematite from the Clara mine
CO3 2 . Equilibrium fractionation factors suggest that sid- (d56Fe: +0.9&). The isotopic difference between siderite
erite incorporates light iron relative to a Fe(II)-bearing flu- and high-temperature secondary hematite of +1.6& to
id (Dsiderite–Fe(II)aq = 4.9& to 0.5&). Indeed, measured +2.3& lies exactly in the range predicted from the above-
siderite contains light iron throughout (d56Fe = 1.4& to mentioned siderite–hematite fractionation factors. This
0.7&), but the heavy Fe of this range is not as light as indicates that siderite and hematite achieved thermody-
would be expected if siderite precipitated from an initial namic equilibrium during this fluid-assisted high-tempera-
fluid of 0.5& with a fractionation factor of 60.5&. ture alteration process. At thermodynamic equilibrium
One reason might be that the measured siderites formed and exchange of Fe between both minerals involved, min-
from fluids that always transferred a substantial fraction eral isotope compositions would reflect the relative equilib-
of their Fe(II)aq into the siderite, thereby not resulting in rium fractionation factors of these minerals.
any net isotope fractionation due to mass balance con-
straints. In any case, the composition of the siderites, 6.2.4. Low-temperature alteration
namely, the absence of very light compositions that might The situation is entirely different during low-tempera-
potentially have formed by precipitating siderite from an ture alteration, where no chemical equilibrium is estab-
oxidizing fluid that has experienced depletion in heavy Fe lished. The thermodynamic model (Fig. 8) has shown
(the curve labelled remaining solution in Fig. 9) is entirely that goethite replaces siderite when exposed to oxidizing
compatible with the inferred formation process. Unlike fluids, and that no Fe is released in the process. In that case
hematite formation that involves oxidizing surficial fluids we would expect that goethite exactly inherits the Fe iso-
(left column in Fig. 7), siderite precipitates from HCO3 - tope composition of the siderite precursor, despite the frac-
enriched fluids. The siderite compositions confirm that tionation factor that would predict goethite that is 1.5&
the siderites precipitated from non-oxidizing environments, heavier than siderite at 25 °C. This is exactly the pattern
as predicted by the thermodynamic model. that is observed. Secondary goethite yields a range in
It can be assumed that sulfides (chalcopyrite, pyrite) pre- d56Fe of 1.1& to 0.5&, very similar to primary siderites
cipitated from reduced, sulfide-rich hydrothermal solu- that range from 1.36& to 0.7&. Therefore, these goe-
tions. Chalcopyrite and pyrite both contain iron with thite compositions demonstrate the feasibility of the sug-
d56Fe of 0.3& to 0&. Based on the fractionation factors gested alteration processes. Furthermore, it appears that
discussed above it is expected that sulfides would incorpo- replacement products serve to identify the Fe isotope com-
rate iron that is much heavier than found in the Fe(II) fluid position of their precursor (provided that no additional Fe
from which they precipitate and that is also heavier than is brought into the system by the oxidizing alteration fluid).
hematite. Considering that this is not observed, these iso- Within the Clara mine, alteration of primary sulfides
tope compositions are compatible with near-quantitative (d56Fe around 0.3&) led to precipitation of the second-
precipitation from a fluid of light initial composition. ary, supergene Fe(III)-arsenates Ba-pharmacosiderite and
scorodite (d56Fe: 0.16& and 0.24&), partly on pseudo-
6.2.3. High-temperature alteration morphed chalcopyrite, partly some cm to dm away from it.
Three hematite samples from Table 1 were interpreted to Given that the isotope compositions of the alteration prod-
show a high-temperature alteration of primary Fe miner- ucts are so similar to their precursors, we can assume that a
als. They cover a range in d56Fe between +0.5& and complete iron transfer has taken place between the two. A
+0.9&. The texturally most convincing evidence for detailed description of iron isotope systematics within sin-
high-temperature alteration is shown sample M590 from gle deposits can be found in Appendix A.
the Sophia mine near Wittichen. In this sample, very thin
leaf-like hematite (d56Fe: +0.81&) is found as oxidation 6.2.5. Sedimentary Fe deposits
product on an earlier generation of siderite (d56Fe: Oolithic iron ores of the Dogger (Middle Jurassic) con-
1.36) and is also overgrown by a younger generation of tain goethites which have d56Fe values between +0.02&
siderite (d56Fe: 0.74&). Both siderites are interpreted as and +0.42&. This range is close to crystalline rocks
being of high-temperature origin (>150 °C). This is based (d56Fe: +0.04& to +0.38&; Beard et al., 2002; Poitrasson
on textural observations and on fluid inclusion studies. and Freydier, 2005), and also close to the range of modern
The older siderite generation shows dark brown patches clastic marine sediments (d56Fe: 0.2& to +0.3&; Beard
resulting from partial oxidation and thus we suggest that and Johnson, 2004). We suggest the following evolution
the hematite formed via oxidation of the first generation to explain these features: (1) dissolved iron is homogenized
siderite during a relatively oxidizing event of the high-tem- during transport, or is derived from isotopically uniform
perature, vein-filling processes. The second generation of sources, resulting in relatively uniform pre-oxidation iso-
siderite was then precipitated during a new influx of re- tope ratios, (2) Fe(II)aq is oxidized, with Fe(III)aq being
3028 G. Markl et al. 70 (2006) 3011–3030
about 3& heavier, and (3) goethite is precipitated from of the precursor (siderite) by the predicted fractionation
Fe(III)aq, with d56Fe being about 3.4& lighter than factor.
Fe(III)aq at equilibrium. Steps (2) and (3) compensate each Finally, the range of compositions of goethite from
other isotopically (depending on the actual mass balances oolitic iron ores seems to indicate oxidative precipitation
and potential further kinetic fractionation effects), such from a relatively uniform coastal water Fe reservoir. Un-
that goethite is deposited with d56Fe close to that of the ori- like Fe isotopes that equilibrate under metamorphic or
ginal seawater. Another, equally speculative explanation is magmatic conditions, the iron isotope systematics found
that the dissolved (e.g., coastal) Fe(II)aq was always imme- when hydrothermal fluids are involved shows remarkable
diately and quantitatively oxidized and precipitated into variations. Importantly, these variations are consistent
goethite, such that the original dissolved Fe compositions with both thermodynamic formation models and predicted
are now preserved by the goethite. Finally, it is possible Fe isotope fractionation factors. Therefore, Fe isotopes
that oolite formation involves diagenetic reactions, which have considerable potential as tracers of ore-forming
might fractionate iron towards light compositions on processes.
reduction, and towards heavy compositions on re-oxida-
tion and precipitation into oolitic Fe(III). Acknowledgments
7. Summary and conclusion We are grateful to Marc Mamberti and Ronny Schoen-
berg for their support of the Fe isotope analyses presented
Iron isotopes in hydrothermal mineralizations encom- here. We acknowledge the constructive reviews by S. Kes-
pass a range of 2.5& in d56Fe values. Given that such var- ler, B. Bergquist, C. Smith, O. Rouxel, and an anonymous
iability occurs within single veins, sometimes even within reviewer.
an individual layered hematite sample, we can discount
the possibility that these isotope ratios provide information Associate editor: Martin B. Goldhaber
on the iron source (Graham et al., 2004). Rather, we sug-
gest a scenario in which the iron isotopes trace the fluid his- Appendix A. Detailed description of iron isotope systematics
tory. This first entails leaching of Fe from basement. The in single deposits
resulting fluids are rich in Cl to which Fe(II)aq is strongly
bound (d56Fe = 0.5&). These hydrothermal fluids can In the following section, we describe iron isotopic varia-
precipitate Fe mineral through two different depositional tions within single deposits. Samples from the Clara mine
mechanisms. One is mixing with oxygen-rich surface show the largest variability of all deposits investigated.
waters, resulting in the precipitation of isotopically heavy Weathering of primary sulfides (d56Fe around 0.3&) led
hematite. The second is input of CO2-rich fluids (produced to precipitation of secondary, supergene Fe(III)-arsenates
by surface carbonate sediment dissolution) which leads to Ba-pharmacosiderite and scorodite (d56Fe: 0.16& and
the precipitation of light siderite. Given that the two pre- 0.24&) partly on pseudomorphed chalcopyrite, partly
cipitation products are isotopically distinct (heavy hematite some cm to dm away from it. Alteration and oxidation
and light siderite, respectively), their composition serves to of primary siderite in the Clara mine (d56Fe = 0.99&) re-
trace the responsible depositional mechanisms. The residu- sults in significant shifts of the Fe isotopic composition.
al fluids can be further oxidized, or they can be recharged Siderite, which is dark brown and hence partially oxidized
by fresh iron-rich hydrothermal fluids of unfractionated to goethite, exhibits a d56Fe of 0.8&, which is also the
composition. Therefore, a large range of mixing ratios value of secondary goethite. Secondary fibrous hematite
can be produced. Such mixing histories are recorded at has an even heavier value of 0.66&.
the meter scale by isotope differences between minerals of Four samples from the Sophia mine near Wittichen
the same generation, on the cm scale by differences between comprise three siderites and one hematite. Of those, two
minerals within one hand specimen, and on the millimeter siderites and the hematite come from the same hand spec-
scale by intimately zoned hematite and goethite aggregates. imen (M590). The hematite (d56Fe = +0.81&) is found as
Magnetite is never formed as this would require either low oxidation product on an earlier generation of siderite
fO2 (carbonate-free fluids) or high fO2 (carbonate-bearing (d56Fe = 1.36) directly overgrown by a younger genera-
fluids). tion of siderite (d56Fe = 0.74&). The very heavy value
Secondary alteration products are goethite and hema- of the hematite is confirmed by texturally identical hema-
tite, which form by replacement of siderite at both high tite from the Anton mine (d56Fe = +0.84&) which is situ-
and low temperatures. The low-T secondary minerals ap- ated close to the Sophia mine. Both mined almost identical
pear to inherit the composition of their precursors Co–Ag-barite veins. A partly oxidized siderite from a dif-
throughout. Therefore, Fe isotopes can serve to identify ferent sample in the Sophia mine is slightly isotopically
precursors in such replacement reactions. In contrast, lighter (d56Fe = 0.85&), which, however, might also be
high-T alteration produces replacement products at ther- a primary feature not related to the oxidation. The sider-
modynamic equilibrium. The Fe isotope composition of ite–hematite combination implies that there is a very signif-
these products (hematite, goethite) is different from that icant isotopic fractionation during common precipitation
of minerals containing iron in different valence states which Beard, B.L., Johnson, C.M., 2004. Fe isotope variations in modern and
qualitatively agrees with the results from the Clara mine ancient Earth and other planetary bodies. In: Johnson, C.M., Beard,
B.L., Albarède, F. (Eds.), Geochemistry of Non-Traditional Stable
detailed above. Isotopes, vol. 55. Mineralogical Society of America, Washington, DC,
At the barite–goethite deposit of Neuenbürg, partially pp. 319–357.
oxidized siderite (0.83& to 0.72&), secondary fibrous Beard, B.L., Johnson, C.M., Skulan, J.L., Nealson, K.H., Cox, L., Sun,
crystalline hematite (0.24& and 0.44&) and secondary H., 2002. Application of Fe isotopes to tracing the geochemical and
fibrous goethite (0.64&) have values in the same range, biological cycling of Fe. Chem. Geol. 195, 87–117.
Bliedtner, M., Martin, M., 1988. Erz—und Minerallagerstätten des
but hematite is heavier than the siderite as was also ob- Mittleren Schwarzwaldes. Geologisches Landesamt Baden-Wuerttem-
served in the Clara and Sophia mine samples. This shows berg, Freiburg i.Br.
that the oxidation processes operate in a similar way in Brantley, S.L., Liermann, L.J., Bullen, T.D., 2001. Fractionation of Fe
very different types of deposits. Obviously, mass balance isotopes by soil microbes and organic acids. Geology 29, 535–538.
dependent equilibrium fractionation leads to successively Brantley, S.L., Liermann, L.J., Guynn, R.L., Anbar, A.D., Icopini, G.A.,
Barling, J., 2004. Fe isotope fractionation during mineral dissolution
heavier Fe isotopic compositions during oxidation of the with and without bacteria. Geochim. Cosmochim. Acta 68, 3189–3204.
primary iron-bearing minerals. Bullen, T.D., White, A.F., Childs, C.W., Vivit, D.V., Schulz, M.S., 2001.
On the other hand, two localities appear to tell a differ- Demonstration of significant abiotic iron isotope fractionation in
ent story. At the Otto mine, large euhedral, primary hema- nature. Geology 29, 699–702.
tite crystals have a d56Fe value of +0.3&, whereas Ellis, A.S., Johnson, T.M., Bullen, T.D., 2002. Chromium isotopes and
the fate of hexavalent chromium in the environment. Science 295
microcrystalline, most probably secondary hematite from (5562), 2060–2062.
the same sample has a d56Fe of 0.97&. Fibrous second- Emmermann, R., 1977. A petrogenetic model for the origin and evolution
ary goethite from another sample has a value of 1.12&. of the Hercynian granite series of the Schwarzwald. N. Jahrb. Mineral.
Similarly, the Kienberg sample FE8 shows large varia- Abh. 128, 219–253.
tions in the same direction towards lighter values within Graham, S., Pearson, N., Jackson, S., Griffin, W., O’Reilly, S., 2004.
Tracing Cu and Fe from source to porphyry: in situ determination of
millimeters; older fibrous goethite (d56Fe = 0.52&) is Cu and Fe isotope ratios in sulfides from the Grasberg Cu–Au deposit.
overgrown by feathery hematite (d56Fe = 1.08&). Chem. Geol. 207, 147–169.
Primary, euhedral hematite from the Rappenloch mine Hoefs, J., Emmermann, R., 1983. The oxygen isotope composition of
shows surprisingly variable iron isotopic compositions. In Hercynian granites and pre-Hercynian gneisses from the Schwarzwald,
a single hand-sized layered ore specimen larger crystals in SW Germany. Contrib. Mineral. Petrol. 83, 320–329.
Hofmann, B., Eikenberg, J., 1991. The Krunkelbach uranium deposit,
a vuggy layer have a d56Fe of +0.04&, whereas a micro- Schwarzwald, Germany. Correlation of radiometric ages, U–Pb, U–
crystalline layer has a value of 0.11&. Another typical Xe–Kr, K–Ar, Th–U with mineralogical stages and fluid inclusions.
coarse-grained primary hematite from the same location Econ. Geol. 86, 1031–1049.
has a value of 0.28&. A sample of layered, fibrous hema- Holland, T.J.B., Powell, R., 1998. An internally consistent thermodynamic
tite of secondary origin from the Rappenloch mine has a data set for phases of petrological interest. J. Metam. Geol. 16, 309–
343.
value of 0.78&, whereas all primary hematites vary only Icopini, G.A., Anbar, A.D., Ruebush, S.S., Tien, M., Brantley, S.L., 2004.
between 0.28& and +0.04&. Centimeter-sized crystals of Iron isotope fractionation during microbial reduction of iron: the
a different sample from Farrenberg near Eisenbach (same importance of adsorption. Geology 32, 205–208.
mineralization type as Rappenloch, 500 m apart from each Johnson, C.M., Beard, B.L., Albarède, F., 2004a. Geochemistry of non-
other) have a d56Fe value of +0.01&, i.e., identical to the traditional stable isotopes. In: Rosso, J.J. (Ed.), Reviews in Mineralogy
and Geochemistry, vol. 55. Mineralogical Society of America, Wash-
crystalline layer in the aforementioned sample. Two further ington, DC.
samples of layered, botroidal hematite (‘‘red glaskopf’’) Johnson, C.M., Beard, B.L., Roden, E.E., Newman, D.K., Nealson, K.H.,
have d56Fe values of 0.28& and 0.78&, respectively. 2004b. Isotopic constraints on biogeochemical cycling of Fe. In:
Hence, one mineral not only from the same mining district, Johnson, C.M., Beard, B.L., Albarède, F. (Eds.), Geochemistry of Non-
but also from the same mine and even from the same sam- Traditional Stable Isotopes, vol. 55. Mineralogical Society of America,
Washington, DC, pp. 359–408.
ple show highly variable iron isotopic compositions. Johnson, C.M., Beard, B.L., Welch, S.A., Roden, E.E., Croal, L.R.,
Newman, D.K., Nealson, K.H., 2005. Experimental constraints on Fe
References isotope fractionation during magnetite and Fe carbonate formation
coupled to dissimilatory hydrous ferric oxide reduction. Geochim.
Albarède, F., 2004. The stable isotope geochemistry of Copper and Zinc. Cosmochim. Acta 69, 963–993.
In: Johnson, C.M., Beard, B.L., Albarède, F. (Eds.), Geochemistry of Johnson, J.W., Oelkers, E.H., Helgeson, H.C., 1992. SUPCRT92: A
Non-Traditional Stable Isotopes, vol. 55. Mineralogical Society of software package for calculating the standard molal thermodynamic
America, Washington, DC, pp. 409–427. properties of minerals, gases, aqueous species, and reactions from 1 to
Anbar, A.D., 2004. Iron stable isotopes: beyond biosignatures. Earth 5000 bar and 0 to 1000 °C. Comp. Geosci. 18, 899–947.
Planet. Sci. Lett. 217, 223–226. Lippolt, H.J., Kirsch, H., 1994. Isotopic Investigation of Post-Variscan
Anbar, A.D., Jarzecki, A.A., Spiro, T.G., 2005. Theoretical investigation Plagioclase Sericitization in the Schwarzwald Gneiss Massif. Chem.
of iron isotope fractionation between Fe(H2O)3+6 and Fe(H2O)2+6: Erde 54, 179–198.
implications for iron stable isotope geochemistry. Gechim. Cosmochim. Maréchal, C.N., Nicolas, E., Douchet, C., Albarède, F., 2000. Abundance
Acta 69, 825–837. of zinc isotopes as a marine biogeochemical tracer. Geochem. Geophys.
Barling, J., Arnold, G.L., Anbar, A.D., 2001. Natural mass-dependent Geosyst. 1, 1999GC000029.
variations in the isotopic composition of molybdenum. Earth Planet. Markl, G., 2004. Wie kommt das Silber ins Gestein? Die Bildung der
Sci. Lett. 193, 447–457. Schwarzwälder Erzgänge und ihrer Mineralien. In: Markl, G., Lorenz,
3030 G. Markl et al. 70 (2006) 3011–3030
S. (Eds.), Silber Kupfer Kobalt—Bergbau im Schwarzwald. Markstein Simon, K., Hoefs, J., 1987. Effects of meteoric water interaction on
Verlag, Filderstadt. Hercinian granites from the Südschwarzwald, Southwest Germany.
Matthews, A., Zhu, X.K., O’Nions, R.K., 2001. Kinetic iron stable Chem. Geol. 61, 253–261.
isotope fractionation between iron (-II) and (-III) complexes in Skulan, J.L., Beard, B.L., Johnson, C.J., 2002. Kinetic and equilibrium Fe
solution. Earth Planet. Sci. Lett. 192, 81–92. isotope fractionation between aqueous Fe(III) and hematite. Geochim.
Meshik, A.P., Lippolt, H.J., Dymkov, Y.M., 2000. Xenon geochronology Cosmochim. Acta 66, 2995–3015.
of Schwarzwald pitchblendes. Mineral. Deposita 35, 190–205. Sverjensky, D.A., Shock, E.L., Helgeson, H.C., 1997. Prediction of the
Metz, R., Richter, M., Schürenberg, H., 1957. Die Blei- Zink- Erzgänge thermodynamic properties of aqueous metal complexes to 1000 °C and
des Schwarzwaldes. Beih. Geol. Jahrb. 29, 1–277. 5 kb. Geochim. Cosmochim. Acta 61, 1359–1412.
Meyer, M., Brockamp, O., Clauer, N., Renk, A., Zuther, M., 2000. Tagirov, B., Schott, J., 2001. Aluminum speciation in crustal fluids
Further evidence for a Jurassic mineralizing event in central Europe. revisited. Geochim. Cosmochim. Acta 65, 3965–3992.
K–Ar dating of geothermal alteration and fluid inclusion systematics Wagner, T., Boyce, A.J., 2003. Sulphur isotope geochemistry of black
in wall rocks of the Käferteige fluorite vein deposit in the northern shale-hosted antimony mineralization, Arnsberg, northern Rhenish
Black Forest, Germany. Mineral. Deposita 35, 754–761. Massif, Germany: implications for late-stage fluid flow during the
Oelkers, E.H., Helgeson, H.C., 1990. Triple-ion anions and polynuclear Variscan orogeny. J. Geol. Soc. London 160, 299–308.
complexing in supercritical electrolyte solutions. Geochim. Cosmochim. Wagner, T., Cook, N.J., 2000. Late-Variscan antimony mineralization in
Acta 54, 727–738. the Rheinisches Schiefergebirge, NW Germany: Evidence for stibnite
Poitrasson, F., Freydier, R., 2005. Heavy iron isotope composition of precipitation by drastic cooling of high-temperature fluid systems.
granites determined by high resolution MC-ICP-MS. Chem. Geol. 222, Mineral. Deposita 35, 206–222.
132–147. Walczyk, T., von Blanckenburg, F., 2002. Natural iron isotope variations
Polyakov, V.B., 1997. Equilibrium fractionation of the iron isotopes: in human blood. Science 295, 2065–2066.
estimation from Mössbauer spectroscopy data. Geochim. Cosmochim. Walczyk, T., von Blanckenburg, F., 2005. Deciphering the iron isotope
Acta 61, 4213–4217. message of the human body. Intern. J. Mass Spectrom. 242, 117–
Polyakov, V.B., Mineev, S.D., 2000. The use of Mössbauer spectroscopy in 134.
stable isotope geochemistry. Geochim. Cosmochim. Acta 64, 849–865. Welch, S.A., Beard, B.L., Johnson, C.M., Braterman, P.S., 2003. Kinetic
Rouxel, O., Dobbek, N., Ludden, J., Fouquet, Y., 2003. Iron isotope and equilibrium Fe isotope fractionation between aqueous Fe(II) and
fractionation during oceanic crust alteration. Chem. Geol. 202, 155–182. Fe(III). Geochim. Cosmochim. Acta 67, 4231–4250.
Schauble, E.A., 2004. Applying stable isotope fractionation theory to new Werner, W., Dennert, V., 2004. Lagerstätten und Bergbau im Schwarzwald.
systems. In: Johnson, C.M., Beard, B.L., Albarède, F. (Eds.), Landesamt für Geologie, Rohstoffe und Bergbau Baden-Württemberg,
Geochemistry of Non-Traditional Stable Isotopes, vol. 55. Mineralog- Freiburg, Germany, 334 p.
ical Society of America, Washington, DC, pp. 65–111. Werner, W., Franzke, H.J., 2001. Postvariskische bis neogene Bruchtekto-
Schauble, E.A., Rossmann, G.R., Taylor, H.P., 2001. Theoretical nik und Mineralisation im südlichen Schwarzwald. Z. Dt. Geol. Ges.
estimates of equilibrium Fe-isotope fractionations from vibrational 152, 405–437.
spectroscopy. Geochim. Cosmochim. Acta 65, 2487–2497. Werner, W., Franzke, H.J., Wirsing, G., Jochum, J., Lüders, V.,
Schoenberg, R., von Blanckenburg, F., 2005. An assessment of the Wittenbrink, J., 2002. Die Erzlagerstätte Schauinsland bei Freiburg
accuracy of stable Fe isotope ratio measurements on samples with im Breisgau. Aedificatio Verlag, Freiburg.
organic and inorganic matrices by high-resolution multicollector ICP- Wernicke, R.S., Lippolt, H.J., 1993. Botryoidal hematite from the
MS. Intern. J. Mass Spectrom. 242, 257–272. Schwarzwald, Germany. Heterogenous U distributions and their
Schwinn, G., Markl, G., 2005. REE systematics in hydrothermal fluorite. bearing on the helium dating method. Earth Planet. Sci. Lett. 114,
Chem. Geol. 216, 225–248. 287–300.
Schwinn, G., Wagner, T., Baldorj, B., Markl, G., 2006. Quantification of Wernicke, R.S., Lippolt, H.J., 1997. UTh-He evidence of a Jurassic
mixing processes in ore-forming hydrothermal systems by combination continuous hydrothermal activity in the Schwarzwald basement,
of stable isotope and fluid inclusion analyses. Geochim. Cosmochim. Germany. Chem. Geol. 138, 273–285.
Acta 70, 965–982. Wiesli, R.A., Beard, B.L., Johnson, C.M., 2004. Experimental determi-
Segev, A., Halicz, L., Lang, B., Steinitz, G., 1991. K–Ar dating of nation of Fe isotope fractionation between aqueous Fe(II), siderite and
manganese minerals from the Eisenbach region, Black Forest, south- ‘‘green rust’’ in abiotic systems. Chem. Geol. 211, 343–362.
west Germany. Schweiz. Mineral. Petrogr. Mitt. 71, 101–114. Williams, H.M., McCammon, C.A., Peslier, A.H., Halliday, A.N.,
Seward, T.M., Barnes, H.L., 1997. Metal transport in hydrothermal ore Teutsch, N., Levasseur, S., Burg, J.P., 2004. Iron isotope fractionation
fluids. In: Barnes, H.L. (Ed.), Geochemistry of Hydrothermal Ore and the oxygen fugacity of the mantle. Science 204, 1656–1659.
Deposits, third ed. Wiley, New York, pp. 435–486. Wood, S.A., Samson, I.M., 1998. Solubility of ore minerals and
Shock, E.L., Oelkers, E.H., Johnson, J.W., Sverjensky, D.A., Helgeson, complexation of ore minerals in hydrothermal solutions. Rev. Econ.
H.C., 1992. Calculation of the thermodynamic properties of aqueous Geol. 10, 33–80.
species at high pressures and temperatures. J. Chem. Soc. Faraday Yardley, B., 2005. Metal concentrations in crustal fluids and their
Transact. 88, 803–826. relationship to ore formation. Econ. Geol. 100, 613–632.
Shock, E.L., Sassani, D.C., Willis, M., Sverjensky, D.A., 1997. Inorganic Zhu, X., Guo, Y., Williams, R., O’Nions, K., Matthews, A., Belshaw, N.,
species in geological fluids: correlations among standard molal Canters, G., de Waal, E., Weser, U., Burgess, B., Salvato, B., 2002.
thermodynamic properties of aqueous ions and hydroxide complexes. Mass fractionation processes of transition metal isotopes. Earth
Geochim. Cosmochim. Acta 61, 907–950. Planet. Sci. Lett. 200, 47–62.
Shvarov, Y.V., 1978. Minimization of the thermodynamic potential of an Zhu, X., O’Nions, K., Guo, Y., Belshaw, N., Rickard, D., 2000.
open chemical system. Geochem. Internat. 15, 200–203. Determination of natural Cu-isotope variation by plasma-source mass
Shvarov, Y.V., 1981. A general equilibrium criterion for an isobaric- spectrometry: implications for use as geochemical tracers. Chem. Geol.
isothermal model of a chemical system. Geochem. Internat. 18, 38–45. 163, 139–149.
Shvarov, Y., Bastrakov, E., 1999. HCh: a software package for Zuther, M., Brockamp, O., 1988. The fossil geothermal system of the
geochemical equilibrium modeling. User’s guide. Australian Geolog- Baden–Baden trough, northern Black Forest, Germany. Chem. Geol.
ical Survey Organisation, Record 1999/25, 61 p. 71, 337–353.

Fe Isotope Fraction at Ion During Hydro Thermal Ore Deposition

Uploaded by

Copyright:

Available Formats

Fe Isotope Fraction at Ion During Hydro Thermal Ore Deposition

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Fe Isotope Fraction at Ion During Hydro Thermal Ore Deposition

Uploaded by

Copyright:

Available Formats

Geochimica et Cosmochimica Acta 70 (2006) 3011–3030

Iron isotope fractionation during hydrothermal ore deposition

Received 25 July 2005; accepted in revised form 3 February 2006

1. Introduction sible tools for deciphering geochemical processes (see the

4. Analytical procedures and results and the d57Fe likewise.

4.1. Analytical procedures 4.2. Results

Fig. 6. Predominance diagram of aqueous Fe species in pH—log aCl

invariant point deﬁned by pyrophyllite, muscovite and

5.2.2. Fluid mixing with surface water

6.2.2. Fluid mixing with surface water

You might also like