2013 Society of Economic Geologists, Inc.

Economic Geology, v. 108, pp. 4558

Sulfide Saturation in Mafic Magmas: Is External Sulfur Required for

Magmatic Ni-Cu-(PGE) Ore Genesis?
EDWARD M. RIPLEY AND CHUSI LI
Department of Geological Sciences, Indiana University, Bloomington, Indiana 47405

Abstract
The importance of externally derived sulfur in the genesis of sulfide-rich, magmatic Ni-Cu-(platinum group
element [PGE]) deposits remains a key, yet unresolved, issue. Calculations utilizing a variety of mafic magma
types indicate that, in general, sulfide saturation by fractional crystallization occurs after Ni has been depleted
due to olivine crystallization. Cu- and PGE-rich layers may form during relatively later stages of closed-system
crystallization, but unless the collection of the cotectic proportion of immiscible sulfide is extremely efficient,
the mass of sulfide is too small to produce an economic deposit. We show that there are numerous processes
that may lead to early sulfide saturation in mafic/ultramafic magmas. Contamination of mantle-derived mag-
mas by siliceous country rocks or their partial melts will lower the sulfur content needed to induce sulfide liq-
uid saturation, typically by amounts ranging from 200 to 700 ppm. The mixing of magmas, particularly if the
result is to lower the liquidus temperature of the mixed magma, may also lower the sulfur content needed to
attain sulfide saturation by similar amounts. An increase in magma fO2 related to the addition of volatiles such
as H2O and CO2 is less effective in decreasing the sulfur concentration needed to achieve sulfide liquid satu-
ration. Contamination processes that lead to an increase in the activity of SiO2 in the melt, and hence may pro-
mote orthopyroxene rather than olivine crystallization, aid in generating relative Ni enrichment in remaining
liquid as a result of the lower DNi (mineral melt) value of orthopyroxene relative to olivine. Although conta-
mination and magma mixing may produce early sulfide saturation without the addition of externally derived
sulfur, Ni-rich sulfide deposits can form in such cases only from large-volume, open systems, where the effi-
ciency of sulfide collection is high. No matter what the liquidus minerals may be, without the addition of coun-
try rock-derived sulfur, the mass of sulfide necessary to generate economic Ni-Cu-(PGE) concentrations
requires efficient sulfide collection from large, but not necessarily unrealistic, volumes of magma. Small
deposits (430 Mt of sulfide) may form from the collection of cotectic proportions of sulfide from less than 50
km3 of magma. Larger deposits such as those at Norilsk could involve more than 200 km3 of magma; this vol-
ume of magma is not unreasonable, particularly in rift/plume-related settings. Despite such possibilities, sulfur
isotope data clearly indicate that externally derived sulfur has been involved in the formation of many large de-
posits, and that collection of mantle-derived sulfide in sufficient quantities to produce orebodies is a rare
process. We propose that magmatic Ni-Cu-(PGE) sulfide ore formation normally requires significant sulfide
supersaturation, and that the addition of sulfur derived from xenoliths is the most viable mechanism for pro-
ducing sulfide well above the cotectic proportion in mafic/ultramafic magmas.

Introduction crystallization or the introduction of external sulfur must

THE ATTAINMENT of sulfide saturation of a magma and the
separation of immiscible sulfide liquid, together with the
presence of an appropriate physical environment for the col-
lection and concentration of the metal-rich sulfide liquid, are
key ingredients for the formation of magmatic Ni-Cu-(plat-
inum group element [PGE]) deposits. Experimental studies
of the solubility of sulfur in mafic magmas (e.g., Haughton et
al., 1974; Shima and Naldrett, 1975; Buchanan and Nolan,
1979; Danckworth et al., 1979; Wendlandt, 1982; Mavrogenes
and ONeill, 1999; Holzheid and Grove, 2002; ONeill and
Mavrogenes, 2002; Clemente et al., 2004; Jugo et al., 2005;
Scaillet and Macdonald, 2006; Liu et al., 2007; Moune et al.,
2009) have shown that several processes can lead to the at-
tainment of sulfide saturation in a mafic magma. Mavrogenes
and ONeill (1999) showed that the solubility of sulfide in a
mafic magma increases with decreasing pressure, and there-
fore concluded that mafic magmas that are emplaced at rela-
tively shallow depths are unlikely to be sulfide saturated.
They noted that, in order for sulfide saturation to be achieved
in the low-pressure environment, either extensive fractional
Corresponding author: e-mail, ripley@indiana.edu
0361-0128/13/4082/45-14
occur. The experimental data on sulfide and sulfate solubility
in mafic magmas that are now available can be used to illus-
trate that several processes of contamination of mafic magma,
or mixing of magma types, can lead to the saturation of a
mafic magma with sulfide or sulfate liquid. However, one of
the major questions still posed with respect to magmatic Ni-
Cu-(PGE) ore deposition is if externally derived sulfur is an es-
sential ingredient in the process. Recent articles illustrate the
strongly contrasting points of view. For example, Keays and
Lightfoot (2010) state that crustal sulfur must be added to
mantle-derived magmas to promote ore genesis, whereas Seat
et al. (2009) provide evidence from the Nebo-Babel deposit
that mantle S can be sufficient to promote ore formation.
The only process for the attainment of sulfide saturation in
a mafic magma that does not require some type of interaction
with country rocks, or mixing of magmas, is fractional crystal-
lization. The increase in S concentration of a melt as a result
of crystallization of olivine, pyroxene, feldspar, etc., may lead
to the separation of what is known as cotectic proportions of
sulfide liquid. Sulfide-poor but PGE-enriched layers such as
that found in the Sonju Lake intrusion (e.g., Miller, 1999; Park
et al., 2004) are thought to result from fractional crystallization
Submitted: February 27, 2012
45 Accepted: May 16, 2012
46 RIPLEY AND LI
of a tholeiitic melt. Tao et al. (2007) proposed that sulfide sat-
uration was achieved due to olivine and chromite crystalliza-
tion in magma that gave rise to PGE-enriched disseminated
sulfide zones in the Jinbaoshan intrusion in China.
Most, but perhaps not all, of the worlds large Cu-Ni-(PGE)
deposits exhibit evidence for mixing processes in their evolu-
tion, either involving a compositionally distinct mantle-derived
magma or a crustal contaminant. Chemical mixing of magmas
has been proposed as an integral process in the formation of
high-grade PGE deposits, such as those of the Merensky Reef
in the Bushveld Complex and the JM Reef in the Stillwater
Complex (e.g., Campbell et al., 1983; Naldrett and von
Gruenewaldt, 1989), and has recently been proposed by
Godel et al. (2011) as a key process in the generation of the
Nebo-Babel deposits. Li and Ripley (2009) showed that mix-
ing of two magmas could lead to the attainment of sulfide sat-
uration in the mixed magma. However, the degree of sulfide
saturation is critically dependent on the sulfur contents of the
end-member magmas, and the amount of sulfide liquid that
may form will typically be small. Sulfide-poor but PGE-rich
deposits could result, but it is not clear whether sulfide-rich,
large tonnage Cu-Ni-(PGE) deposits could be produced.
Another type of mixing involves mafic magmas and partial
melt derived from assimilated, siliceous wall rocks. Irvine
(1975) deduced that the addition of SiO2 to a mafic magma
could decrease sulfide solubility and promote the formation
of an immiscible sulfide liquid. Magma silicification or felsifi-
cation has been proposed as a mechanism to induce sulfide
saturation in mafic magmas by Irvine (1975), Li and Naldrett
(2000), Lightfoot and Hawkesworth (1997), and Seat et al.
(2009). Li and Ripley (2005) also showed that silicification of
a mafic magma can lead to sulfide saturation, but questions
similar to those for magma mixing remained, in terms of the
mass of sulfide that could be sequestered as a result of such a
process.
Lehmann et al. (2007) and Holwell et al. (2007) have called
upon the assimilation of carbonate to increase magma fO2 and
decrease the sulfide solubility of magmas related to the Ni-
Curich ores of the Jinchuan deposit and the PGE-rich occur-
rences in the Platreef of the Bushveld Complex, respectively.
Changes in magma fO2 are related to changes in key composi-
tional values such as the FeO/Fe2O3 ratio. Experimental
results suggest that at relatively low fO2 conditions, where sul-
fide is the predominant sulfur species, small increases in fO2
of a magma coupled with decreased amounts of FeO will de-
crease the S concentration needed to attain sulfide saturation.
However, whether this process can account for the large
amounts of sulfide in systems such as Jinchuan remains un-
clear. Perhaps the most obvious contaminant to induce sulfide
saturation in magmas is S itself. This was pointed out by
Mavrogenes and ONeill (1999) as a result of their experi-
mental work on the effect of pressure on sulfide solubility,
and has been proposed by many researchers based on physi-
cal parameters such as the spatial proximity of S-rich country
rocks or S isotope results. Sulfur isotope results have provided
strong evidence for the introduction of S in komatiite-related
deposits (e.g., Naldrett, 1966; Lesher and Groves, 1986; Bekker
et al., 2009), the deposits at Norilsk (Grinenko, 1985; Gor-
bachev and Grinenko, 1973; Li et al., 2003b), sulfide deposits
in the Duluth Complex (e.g., Ripley, 1981; Ripley and Al-Jassar,
0361-0128/98/000/000-00 $6.00
1987; Lee and Ripley, 1995; Arcuri et al., 1998; Ripley et al.,
2007), Voiseys Bay (Ripley et al., 1999, 2002), Uitkomst (Li et
al., 2002), and the recent discoveries of Kabanga in southern
Africa (Maier et al., 2011). In fact, it is often the absence of
compelling S isotope evidence that has led researchers to
seek alternatives to external S addition to promote sulfide sat-
uration in mafic magmas (e.g., Seat et al., 2009). There are at
least two reasons why 34S values of sulfide ores that fall in
the range generally considered to be normal for S in the upper
mantle (~0 to 2, Ripley, 1999, and references therein)
should not necessarily be taken as an indicator that S conta-
mination has not occurred. One is the situation discussed by
Ripley et al. (2002) for the Voiseys Bay deposit. At Voiseys
Bay, only the ores in the Reid Brook zone show 34S values
that are different from those of mantle S. However, when
examined in detail, it is clear that the Proterozoic Tasiuyak
Gneiss, a locally sulfidic/graphitic pelite that is a likely source
of externally derived S, is characterized by a wide range of
34S values (~ 17 to +18; Ripley et al., 2002). These val-
ues are not unusual for sulfide found in sedimentary rocks,
and of particular significance is the fact that the weighted av-
erage 34S of the Tasiuyak Gneiss is 1.4. Clearly, if ho-
mogenization occurred in the magma, anomalous 34S values
should not be expected. This example holds for any situation
where the 34S value of country-rock S is near 0. A second
possibility is that if sulfides accumulated in a chamber where
additional uncontaminated magma passed, isotopic exchange
reactions (Ripley and Li, 2003) could result in 34S values that
approach those of mantle values. This process would be
expected in conduit systems where magma passage may have
led to increases in the tenor of accumulated sulfide. In
Archean systems, where sedimentary sulfides may be charac-
terized by 34S values that depart only slightly from those of
mantle values, the use of multiple S isotopes (e.g., 33S as well
as 34S) has proven valuable in the determination of externally
derived S from Archean rocks. Archean sedimentary sulfides
often record anomalous 33S values (deviation from the ter-
restrial fractionation line), thought to be linked to low oxygen
values in the Archean atmosphere (e.g., Farquhar and Wing,
2003). Anomalous 33S values have been utilized by Bekker
et al. (2009) to confirm the importance of crustal S in komatiite-
related deposits. For magmatic deposits where Archean rocks
are unlikely to have supplied S to the magma, 33S will be of
little value as an indicator of external S addition where 34S
values are near 0. There can be little argument that crustal
S has played a significant role in ore genesis in those deposits
where S isotope values are strongly anomalous. In their
recent paper, Keays and Lightfoot (2010) used chalcophile
element geochemistry of basaltic rocks at Norilsk compared
to those of the Deccan to arrive at their conclusion that crustal
S is essential for the genesis of magmatic Ni-Cu deposits.
Others (e.g., Lehmann et al., 2007; Seat et al., 2009) refute
such a claim, and the question remains if other mechanisms
which can promote sulfide saturation in a magma can lead to
the generation of magmatic Ni-Cu sulfide orebodies. In this
paper we review many of the processes that can lead to the at-
tainment of sulfide saturation in a mafic magma and attempt
to constrain some key parameters that are of importance in
situations where the addition of externally derived S to a
magma is not indicated, and hence may not be of significance
46
 SULFIDE SATURATION IN MAFIC MAGMAS: IS EXTERNAL SULFUR REQUIRED FOR MAGMATIC Ni-Cu-(PGE) ORE GENESIS? 47

for Ni-Cu-(PGE) ore formation. We utilize the equation of Li
and Ripley (2009) to estimate the concentration of sulfur at
sulfide saturation (SCSS) in a mafic magma; the equation is
based on experimental results from a variety of sulfide-satu-
rated and compositionally variable systems (see above) and
gives results that are similar to the equation of Liu et al.
(2007). We use the equation in conjunction with the MELTS
routine of Ghiroso and Sack (1995) to evaluate how processes
such as fractional crystallization, compositional variations re-
lated to contamination, and changes in intensive parameters
such as fO2 affect the attainment of sulfide saturation in mag-
mas. We also evaluate the importance of the mass of magma
involved in potential ore-forming systems where the intro-
duction of externally derived S may not have occurred.
It should be noted that the equation for SCSS by Li and
Ripley (2009) does not take into account chalcophile ele-
ments such as Ni and Cu. Ariskin et al. (2010) and Evans et
al. (2008) have shown that chalcophile elements, in addition
to Fe, may exert a significant control on the solubility of S in
mafic to ultramafic magmas. The potential influence of Ni
and Cu on sulfur solubility is particularly important when
assessing the role of mantle-derived S in ore genesis. As S is
held primarily within Ni, Cu, and Fe sulfide minerals in the
mantle, complexing with these metals is to be expected. This
will be true for high-degree or second-stage mantle melts,
where immiscible sulfide liquid may ultimately dissolve in
silicate liquids. There is no reason for the breaking of S-Cu
or S-Ni bonds in favor of S-Fe, and both Ni and Cu must
exert a control on the capacity of a mafic magma to dissolve
sulfide. If S is introduced into a mafic/ultramafic melt in a
form that does not involve metal complexing, such as H2S,
the mass action effect will strongly favor the formation of Fe-
S complexes. ONeill and Mavrogenes (2002) have stated
that the sulfide capacity of a melt with ~10% FeO is con-
trolled mostly by the Fe content. Until additional experi-
mental data that treat the effects of chalcophile element
concentration on sulfide solubility over a broad range of
compositions are available, we are unable to update our re-
gression equation.
Fractional Crystallization and Sulfide Saturation
In Figure 1 we illustrate the sulfide saturation curves for a
high-Al olivine tholeiite (HAOTTable 1) liquid thought to
be a reasonable parental magma for troctolitic rocks (e.g.,
Duluth Complex and, potentially, Voiseys Bay; Ripley et al.,
2007; Scoates and Mitchell, 2000), a high-FeO picritic liquid
thought to be a potential parental magma for rocks that host
the Eagle deposit (Ding et al., 2010), and a high-FeO,
siliceous high-MgO basalt liquid similar to the B1 magma
type proposed by Barnes at al. (2010) as a potential parental
magma to the Lower and lower Critical zones of the Bushveld
Complex. In this discussion the magmas are assumed to be
anhydrous; the effect of H2O on the sulfide saturation curve
is discussed below. For the high-Al olivine tholeiite (Fig. 1a)
the sulfide saturation value rises slightly from that of the 1-kb
liquidus due largely to iron enrichment in the liquid related
to crystallization of plagioclase and olivine in a ratio of 3:1
(Table 2). It is not until later in the sequence, and particularly
when spinel becomes a liquidus mineral, that the sulfide sat-
uration value begins to decrease. In the case of a picritic
parental magma, the S content required for saturation at the
liquidus is considerably higher than that required to saturate
the HAOT liquid, but the saturation value decreases as
olivine initially crystallizes, followed by olivine plus clinopy-
roxene (Fig. 1b; Table 2). Where clinopyroxene and plagio-
clase begin to crystallize, the FeO content of the residual
magma rises, and the S content required for sulfide liquid
saturation increases as well. In the case of the FeO-rich,
siliceous high-MgO basalt, the SCSS progressively declines,
as initial olivine crystallization is quickly followed by orthopy-
roxene (~3.5% crystallization), and then clinopyroxene and
plagioclase. These three examples illustrate the types of vari-
ations in the SCSS trajectories that may characterize mafic
magmas.
Treating S as a perfectly incompatible element during sili-
cate crystallization, we indicate when fractional crystallization
should induce sulfide saturation in the HAOT and picritic
magmas. A value of 1,000 ppm S is representative of a melt

TABLE 1. Composition of Magmas Used to Assess SCSS (values in wt %)

1 2 3 4 5 6 7 8

SiO2 47.77 46.56 48.95 52.30 55.37 48.00 47.26 71.42

TiO2 1.29 1.25 2.84 1.17 0.38 1.81 0.27
Al2O3 19.90 19.40 12.82 12.70 11.53 11.69 6.55 13.70
Fe2O3 0.71 1.62 2.05 1.17 1.32 1.84 1.27
FeO 10.37 10.74 10.06 5.87 8.16 10.91 8.83 1.20
MgO 7.10 6.92 7.73 13.08 13.79 13.67 30.23 0.28
CaO 9.69 9.44 12.52 9.56 6.58 9.83 5.04
Na2O 2.67 2.60 2.51 3.73 1.85 1.75 0.45 5.40
K2O 0.50 0.49 0.51 0.43 1.02 0.48 0.10 4.03
H2O 0.00 1.00 0.00 0.00 0.00 0.00 0.00 4.00

Columns: 1) High-Al olivine tholeiite; Ripley et al., 2007; liquidus T = 1,229C

2) High-Al olivine tholeiite, 1 wt % H2O; Ripley et al., 2007; liquidus T = 1,169C
3) Ocean island tholeiite; Kent et al., 1999; liquidus T = 1,205C
4) Siliceous high-MgO basalt; e.g., Barnes, 1989; liquidus T = 1,347C
5) High-FeO, siliceous high-MgO basalt; e.g., Barnes et al., 2010; liquidus T = 1,382C
6) High-FeO picrite; Ding et al., 2009; liquidus T = 1,337C
7) Komatiite; Barnes, 2007; liquidus T = 1,617C
8) Partial melt from pelite; Hoffer and Grant, 1980

0361-0128/98/000/000-00 $6.00 47
48 RIPLEY AND LI

FIG. 1. Sulfur content at sulfide liquid saturation versus wt % of magma crystallization. The sulfur axis refers either to the
sulfur content at sulfide liquid saturation or, in the case of fractional crystallization, to the sulfur content of the remaining
liquid. Fractionating assemblages are given in Table 2. All calculations were done using a pressure of 1 kbar. Note the rela-
tively small differences between the HAOT at QFM and QFM-2 (a), and a similar relationship for the high-FeO picrite (b).
Trends for the contaminated magmas, and the hydrous HAOT, are similar, with much lower SCSS values than their anhy-
drous and uncontaminated counterparts. Sulfur enrichment due to fractional crystallization is illustrated for initial concen-
trations of 800, 1,000, and 1,500 ppm. Sulfide saturation is attained at less than ~20% crystallization in contaminated mag-
mas with 1,000 ppm S. Of particular note is the fact that orthopyroxene is the fractionating phase rather than olivine and
plagioclase in the HAOT magma contaminated with 15% SiO2 (Table 2). Although the S content of the magma contaminated
with SiO2 only would drop to 850 ppm, saturation would still be attained at ~10% crystallization. Orthopyroxene is the frac-
tionating phase at ~3.5% crystallization for the high-FeO, siliceous high-MgO basalt (c). Because DNi(mineral-melt) values
are lower for orthopyroxene relative to olivine, the crystallization of orthopyroxene may be relevant for the formation of Ni-
rich sulfide orebodies where sulfide saturation was attained after significant magma crystallization.

0361-0128/98/000/000-00 $6.00 48
 SULFIDE SATURATION IN MAFIC MAGMAS: IS EXTERNAL SULFUR REQUIRED FOR MAGMATIC Ni-Cu-(PGE) ORE GENESIS? 49

TABLE 2. Fractionating Minerals from the Magmas Illustrated in Figure 1 troctolitic rocks from the Duluth Complex that show no evi-
dence for crustal contamination contain from 500 to 600 ppm
TC
Magma fO2 (first appearance) Minerals1 S and suggest that the parental HAOT contained far less than
1,000 ppm S (Ripley et al., 2007). For comparative purposes,
HAOT QFM-2 1,233 pl we show S trajectories as a result of fractional crystallization
1,198 pl,ol for initial concentrations of 800, 1,000, and 1,500 ppm. In the
1,118 pl,ol,sp
1,078 pl,cpx,sp
case of the picritic magma, saturation at 1 kb is reached at
~40% crystallization for a starting magma with 1,000 ppm S.
HAOT QFM 1,229 ol For the HAOT, ~40% crystallization is also required to in-
1,209 pl,ol duce sulfide liquid saturation. Of additional note is that, due
1,140 pl,ol,opx,cpx to the shallow trend of the SCSS curve where plagioclase and
1,120 pl,ol,cpx
1,100 pl,cpx,sp clinopyroxene are liquidus minerals (Table 2), and the posi-
tive slope where Fe enrichment accompanies the cessation of
HAOT, 1% H2O QFM 1,169 ol olivine crystallization, only small amounts of immiscible sulfide
1,161 pl,ol liquid can be generated as a result of fractional crystallization
1,079 pl,ol,sp
1,069 pl,ol,sp,cpx
until late stages of the process. In strong contrast, sulfide sat-
uration is attained in the high-FeO, siliceous high-MgO basalt
HAOT + 10% pelite melt QFM-2 1,187 pl,ol at ~17% fractional crystallization for a magma with an initial
1,060 pl,ol,sp S concentration of 1,000 ppm.
1,040 pl,cpx,sp Concentrations of platinum group elements and Cu are
HAOT + 30% SiO2 melt QFM-2 1,207 opx expected to increase during fractional crystallization due to
1,180 pl,opx their high degree of incompatibility. This is not the case for
1,060 pl,opx,sp Ni, which is readily accepted into the olivine structure. Dis-
tribution coefficients for Ni between olivine and liquid in sul-
High-FeO picrite QFM-2 1,337 ol
1,175 ol,pl,cpx
fide-undersaturated conditions are a strong function of MgO
1,150 pl,cpx contents and range between 4 and 20 at MgO concentrations
from 22 to 4 wt % (see summary in Li et al., 2003a). For the
High-FeO picrite QFM 1,337 ol picritic parental magma (assuming an initial concentration of
1,200 ol,cpx Ni = 300 ppm), where the ratio of olivine crystallized to pla-
1,175 cpx,pl
gioclase plus clinopyroxene at the point of sulfide saturation
High-FeO picrite + 10% QFM 1,321 ol is ~1.7, the concentration of Ni in the residual liquid drops to
pelite melt 1,170 cpx 127 ppm using an olivine/liquid D value of 7. The drop in Ni
1,140 cpx,pl concentration for the HAOT at the 40% crystallized point,
1,000 cpx,pl,sp
where sulfide saturation is predicted, is similar (139 ppm) due
High-FeO, siliceous QFM 1,382 ol in part to an olivine/liquid D value of 15. The drop in Ni con-
high-MgO basalt 1,356 opx centration is far less extreme for the high-FeO, siliceous high-
1,191 cpx MgO basalt, where orthopyroxene replaces olivine as a liq-
1,183 cpx,pl uidus phase before sulfide saturation is attained. Because the
High-FeO, siliceous QFM 1,361 ol DiNi (orthopyroxene-liquid) value for orthopyroxene (~13,
high-MgO basalt, 1% H2O 1,240 opx Hashizume and Hariya, 1992; Frei et al., 2009) is consider-
1,110 cpx ably lower than the DiNi (olivine-liquid) value, higher con-
1,090 cpx,pl centrations of Ni will be available for partitioning into sulfide
990 cpx,pl,sp
liquid. Of obvious concern is whether sulfide-saturated mag-
High-FeO, siliceous QFM 1,363 ol mas where prior crystallization of olivine is indicated and that
high-MgO basalt + 1,322 opx contain less than 140 ppm Ni can produce Ni sulfide-rich ore-
10% pelite melt 1,132 cpx,pl bodies. We will return to this point below, but end the current
1,032 cpx,pl,sp discussion with an example where fractional crystallization
1 cpx = clinopyroxene, ol = olivine, opx = orthopyroxene, pl = plagioclase,
appears to have played a strong role in the distribution of S
sp = spinel
and metals in an intrusion.
The Sonju Lake intrusion is a relatively small (~1,200 m
thick) layered body in the Beaver Bay Complex associated
with the ~1.1 Ga Duluth Complex (Miller and Ripley, 1996).
derived from 20% melting of a mantle with 200 ppm S (Palme Miller (1999) and Park et al. (2004) have described the pres-
and ONeill, 2004). Lower degrees of melting may produce ence of Cu- and PGE-enriched zones in the layered sequence
melts with higher S concentrations, but, if sulfide saturation at the level of Fe oxide-rich gabbro (Fig. 2). Miller (1999)
is attained, immiscible sulfide may be retained in the mantle suggested that the reef represents the level at which sulfide
(e.g., Keays, 1995; Naldrett, 2011). In the case of a magma saturation was achieved in the crystallizing magma. Li and
such as a HAOT, which may be itself derived from fractional Ripley (2005) confirmed that sulfide saturation should have
crystallization of a more primitive liquid, the S concentration been achieved at the time that spinel became a liquidus min-
of the derivative liquid may be more elevated. However, eral using an equation which is the predecessor to that used

0361-0128/98/000/000-00 $6.00 49
50 RIPLEY AND LI

FIG. 2. Stratigraphic column from the Sonju Lake intrusion, showing the depletion of Ni upward, and the enrichment in
Pt and Pd in the ferrogabbro (modified from Miller, 1999, and Park et al., 2004). Abbreviations: Units in the Sonju Lake
intrusion are preceded by the abbreviation sl: Sonju Lake apatite diorite (slad), dunite (sld), ferrogabbro (slfg), gabbro
(slg), monzodiorite (slmd), melatroctolite (slmt), and troctolite (slt). Country rocks are the Finland quartz monzodiorite
(fqmd) and Finland granite (flg).

in this manuscript (Li and Ripley, 2009). What is particularly
important is the fact that at the level of the Cu and PGE
enrichment, the concentration of Ni has dropped to values of
less than 20 ppm. The field and geochemical evidence
strongly suggests that Ni was sequestered by olivine prior to
the attainment of sulfide saturation in the magma. The Sonju
Lake intrusion perhaps represents an end-member case
where extensive Ni depletion occurred prior to the arrival of
sulfide saturation in the magma by fractional crystallization.
Contamination and Sulfide Saturation
Assimilation of volatiles
The addition of volatile species such as H2O, CO2, CH4,
and H2S to a magma may occur without melting or dissolution
of major minerals. We will treat the potential for the addition
of volatile S species to a magma below. The role of H2O in the
solubility of S in mafic magmas has received only minor
attention (e.g., Liu et al., 2007; Gorbachev and Bezman, 2011;
Moune et al., 2009), and further experimental study is re-
quired to assess the importance of species such as H2S and
HS in sulfide dissolution within mafic magmas. The addition
of H2O to a magma is frequently linked to an increase in fO2
and decrease in FeO content (and hence the capacity of a
magma to dissolve sulfide) via a reaction such as the following:
H2O + 2 FeO = Fe2O3 + H2. However, Carmichael (1991) and
Righter et al. (2008) have discussed the uncertainty of H2O
addition to a magma resulting in an increase of magma fO2.
This is particularly true if H2O dissolution is via molecular
H2O rather than an OH complex. Even without a substantial
change in fO2, the SCSS of an H2O-bearing mafic magma may
be much lower relative to the anhydrous counterpart. In Fig-
ure 1a the SCSS curve for an HAOT with 1 wt % H2O at
QFM is shown to be lower than that of the anhydrous variety
by ~280 ppm at liquidus temperatures. Ripley et al. (2007)
have shown that a value of 1 wt % H2O is reasonable for many
of the magmas that produced sulfide-bearing intrusions in the
Duluth Complex. A large part of the H2O may have been de-
rived via assimilation of fluid produced as country rocks were
dehydrated. In the case of the high-FeO, siliceous high-MgO
basalt, the addition of 1 wt % H2O lowers the SCSS by ~135
ppm at the liquidus, and by greater values throughout the
course of crystallization. The effect is very similar to that pro-
duced by the assimilation of a pelite-derived partial melt (see
below). Although additional study of the effect of H2O addi-
tion on the solubility of S in mafic magmas is warranted, it ap-
pears that hydrous magmas may attain sulfide liquid satura-
tion much sooner than low-H2O magmas.
A similar uncertainty in terms of increases of magma fO2 is
associated with assimilation of CO2. Dissolution mechanisms

0361-0128/98/000/000-00 $6.00 50
 SULFIDE SATURATION IN MAFIC MAGMAS: IS EXTERNAL SULFUR REQUIRED FOR MAGMATIC Ni-Cu-(PGE) ORE GENESIS?
of carbonic species in a magma are important in evaluating
potential magma fO2 increases. At relatively high pressure,
CO2 dissolves as CO2 3 (e.g., Blank and Brooker, 1994) with
relatively little variation in magma fO2. Stanley et al. (2011)
found no correlation between CO2 concentration and the
ratio of Fe3+ to total Fe in a basaltic liquid. In contrast, Wen-
zel et al. (2002) proposed that CO2-rich fluids released during
the decomposition of dolomitic xenoliths in the Ioko-Dovyren
intrusion resulted in locally elevated magma fO2 values. Mollo
et al. (2010) suggested the oppositethat carbonate assimila-
tion involving magma-H2O-CO2 interaction may lead to a
decrease in magma fO2 values. Methane may be a component
of fluids produced by dehydration of carbonaceous pelites
(e.g., Ohmoto and Kerrick, 1977; Andrews and Ripley, 1989).
The assimilation of CH4 is expected to reduce the fO2 of the
magma and may be of significance in the reduction of sulfate
in magmas with oxidation states that permit sulfate plus sul-
fide coexistence or the presence of sulfate only (e.g., Jugo et
al., 2005; Jugo, 2009). The assimilation of graphite, perhaps
accompanied by CH4, by relatively oxidized, sulfate-bearing
magmas has also been proposed as an important process for
decreasing magma fO2 and causing the reduction of sulfate in
the magma to sulfide (e.g., Nixon, 1998; Jugo and Lesher,
2005; Thakurta et al., 2008a, b). Tomkins et al. (2012) have
also suggested the assimilation of graphite-bearing felsic mag-
mas by oxidized arc basalts as a mechanism to promote
magma reduction and the formation of immiscible sulfide
liquid.
The assessment of sulfur-free volatile assimilation on sul-
fide solubility in mafic magmas remains difficult because fO2
changes in natural magmas are often linked with assimilation
processes that alter magma compositions other than Fe2+/Fe3+
ratios. Li and Ripley (2005) showed that fO2 /fS2 ratios in many
of the experiments listed above can be expressed as a function
of melt composition. The effect of crystallization at QFM
rather than QFM-2 on sulfide saturation in the HAOT and
high-Fe picritic liquids is shown in Figure 1a and b. Saturation
values at QFM are lower than those at QFM-2, as expected,
but the effect is small compared to variations in sulfide satu-
ration values related to assimilation of nonvolatile elements
(see below). Jugo (2010) also noted that changes in fO2 below
~QFM resulted in little change in the magma Fe2+/Fe ratio,
and hence little change in the SCSS. Increases in the magma
fO2 above QFM result in decreases in the Fe2+/Fe ratio, as
well as oxidation of sulfide to sulfate. Due to the increase in S
solubility within oxidized magmas, the concentration of S
needed to obtain sulfide liquid saturation also rises (e.g.,
Jugo, 2009).
Assimilation of siliceous country rocks
We utilize two compositions of contaminants to illustrate
the effect of the assimilation of siliceous country rocks on the
sulfide saturation state of a mafic magma. One composition is
the granitic partial melt produced from a pelite in the exper-
iments of Hoffer and Grant (1980; Table 1). The second is
simply addition of SiO2 only, as may result from magma inter-
action with an SiO2-rich sedimentary or metamorphic rock, or
as may result from disequilibrium and polystage melting of
pelitic rocks (e.g., Ripley and Alawi, 1988; Mariga et al.,
2006). The Hoffer and Grant (1980) melt contains 4% H2O,
0361-0128/98/000/000-00 $6.00
51
so the potential effect of H2O on sulfide saturation is included
in the assessment.
In Figure 3, we have compared the sulfide saturation value
at the liquidus temperature (1 kb) for a variety of mafic mag-
mas and the corresponding values at the liquidus for the same
magmas after 10% assimilation of the Hoffer and Grant (1980)
partial melt. All sulfide saturation values decreased from a
high (in terms of differences between contaminated and un-
contaminated liquids) of 681 ppm S in a komatiitic liquid to a
low of 246 ppm in a siliceous, high-MgO basalt liquid. As ex-
pected, the more siliceous the initial magma, the less decrease
recorded by the addition of the partial melt. Starting basaltic
compositions with 47 to 48 wt % SiO2 showed decreases in
sulfide saturation values from 365 to 488 ppm. The computa-
tions illustrate the exceptionally strong effect that assimilation
may have on komatiitic and picritic magmas, but, in all cases,
the assimilation of a granitic contaminant by a mafic magma
lowers the SCSS and may lead to early generation of a sulfide
liquid.
We utilize the SiO2-only contaminant with the HAOT to
illustrate a potentially important change in the mineralogy of
fractionating phases in the contaminated magma (Table 2).
With 10% contamination by the Hoffer and Grant (1980)
melt, all resultant magma compositions retained olivine as an
early crystallizing mineral. With the addition of 15% SiO2 to
the HAOT composition, not only is the sulfide saturation
value decreased by 569 ppm, but the liquidus phase changes
to orthopyroxene. In the case of the high-FeO, siliceous high-
MgO basalt, the elevated initial SiO2 concentration drives the
early crystallization of orthopyroxene, without assimilation of
a siliceous contaminant (although such assimilation may have
been responsible for the elevated SiO2 content initially). As
noted above, because of the lower DiNi (orthopyroxene-liquid)
value for orthopyroxene relative to olivine (Hashizume and
Hariya, 1992; Frei et al., 2009), increased concentrations of
Ni will be available to sulfide liquids where similar percent-
ages of orthopyroxene crystallize rather than olivine. We sug-
gest that such a process may be significant for norite-hosted,
Ni-rich orebodies. In this case, not only does assimilation
reduce the capacity of the magma to dissolve sulfide, but Ni
availability is also enhanced relative to magmas from which
olivine is the primary early crystallizing ferro-magnesian min-
eral. Figure 3 also illustrates that very little, if any, fractional
crystallization is required to promote sulfide saturation in
contaminated siliceous magma. To summarize, analysis of sul-
fide solubility systematics of mafic magmas clearly shows that
assimilation of siliceous contaminants will cause a significant
drop in the S concentration needed for the attainment of sul-
fide saturation. In most cases, little, if any, fractional crystal-
lization of the contaminated magma is required to achieve
sulfide saturation. Where the aSiO2 is sufficiently elevated, ortho-
pyroxene, rather than olivine, is an early crystallizing mineral
and, because of the lower DNi value of orthopyroxene relative
to olivine, additional Ni is available for incorporation into a
sulfide liquid. Fractional crystallization of most low-SiO2 mafic
magmas will not promote sulfide saturation until 20 to 40%
crystallization. Although PGE and Cu contents may continue
to rise with crystallization, significant Ni will be sequestered
in olivine. Assimilation of volatiles such as H2O and CO2 may
or may not result in fO2 increases of the crystallizing magma,
51
52 RIPLEY AND LI

FIG. 3. Comparison of the SCSS from uncontaminated magmas and those contaminated with 10% of the partial melt de-
rived from a pelite (Table 1; Hoffer and Grant, 1980). Number 7 is a mixture of 30% HAOT with 1 wt % H2O and 70% high-
FeO picrite. The SCSS of the mixed magma is somewhat less than that caused by contamination of the high-FeO picrite by
10% of the pelite-derived melt. Calculations performed using a pressure of 1 kbar.

but, in any case, the drop in the sulfide saturation value
caused by fO2 increase (up to ~QFM) alone is minor com-
pared to that caused by assimilation of Si-rich contaminants.
Mixing of magmas
Li and Ripley (2005) confirmed that mixing of two magmas
may induce sulfide saturation of the resultant magma. We will
not dwell on this point as the systematics are much the same
as contamination involving country rock. We illustrate that
the SCSS for a mixture of 70% high-Fe picrite with 30% of
HAOT containing 1 wt % H2O is lower than that of the un-
contaminated picrite by ~300 ppm. This amount is slightly
less than that produced by assimilation by the HAOT of 10%
of the melt produced by pelite partial melting (Fig. 3), but
illustrates that mixing of chemically distinct magmas is a viable
mechanism to reduce the sulfur concentration necessary to
saturate a resultant magma in sulfide liquid. In the case of the
picrite-HAOT mixture, the FeO content of the mixed magma
differs only slightly from either of the end-members; the prin-
cipal control on the lowering of the saturation value is the
drop in the liquidus temperature of the mixed magma.
Although a lower SCSS value may result relative to that of
fractional crystallization by any of the processes mentioned
above, the mixing of mafic magmas alone will be unlikely to
produce sulfide saturation unless both end-members are
close to their saturation values. This feature has been pointed
out by Naldrett (2011) and is illustrated in what is essentially
an end-member case for the high-FeO picrite-HAOT mixture
shown in Figure 4. The degree of sulfide supersaturation (de-
fined as the difference in the sulfur concentration and that of
the SCSS, which is the minimum necessary to promote the
formation of sulfide liquid droplets) that may be attained as a
result of mixing is a strong function of the shape (concavity)
of the SCSS curve. Figure 4a illustrates that the concavity of
the high-FeO picrite-HAOT mixture is so slight that only mix-
tures of magmas very near the end-member saturation values
will attain sulfide saturation upon mixing. In contrast, where
variations in SCSS values of mixtures are more strongly con-
cave (e.g., Li and Ripley, 2005) or vary as shown in Figure 4b
for an ocean island tholeiite-HAOT mixture, mixing of mafic
magmas can be a mechanism to produce sulfide supersatura-
tion. Figure 4b illustrates that even in this case the mixing
process will be effective only if the end-members are within
~100 ppm of their sulfide saturation values. Tomkins et al.
(2012) have shown how the mixing of an S-bearing oxidized
arc basalt with a reduced graphite-bearing felsic melt could
result in reduction of sulfate and sulfide supersaturation in
the mixed magma.
Addition of externally derived sulfur
Perhaps the most efficient means of causing prolonged sul-
fide supersaturation in a mafic magma is through the addition
of S derived from country rocks. As outlined above, in many
localities the combination of isotopic (e.g., Ripley and Li,
2003) and geochemical (e.g., Keays and Lightfoot, 2010) data
strongly points to the involvement of externally derived S in

0361-0128/98/000/000-00 $6.00 52
 SULFIDE SATURATION IN MAFIC MAGMAS: IS EXTERNAL SULFUR REQUIRED FOR MAGMATIC Ni-Cu-(PGE) ORE GENESIS?
a matic, sulfide-rich Ni-Cu-PGE deposits. However, as we have
b digestion or thermomechanical erosion, the amount of sulfide
FIG. 4. Comparison of SCSS trajectories for mixed magmas: (a) high-FeO
picrite and HAOT, and (b) ocean island tholeiite and HAOT. Mixing of mafic
magmas is potentially important for the attainment of sulfide supersaturation
where end-member S concentrations are close to those required for sulfide
saturation and the shape of the SCSS curve for the mixture is strongly con-
cave or has a pronounced inflection.
ore genesis. The importance of external S in magmatic Ni-
Cu-(PGE) ore genesis has been highlighted, and debated, by
many authors, and it is not the purpose of this manuscript to
further review the premise. However, two points regarding
external sulfur addition to mafic magmas deserve further
mention. One is that the addition of H2S- or HS-bearing flu-
ids from country rocks in a contact aureole is unlikely to pro-
duce sulfide in excess of the cotectic proportion. This is
largely due to the fact that external fluids will be added to
magmas only up to their solubility limits. In many cases, fluid
in country rocks will be channeled either by layer-parallel
flow or away from the magmatic heat source. Where sulfide is
derived from xenoliths, the concentration of S in the contam-
inated magma may far exceed saturation values for the pre-
vailing temperature. In the case of komatiites, the process of
thermomechanical erosion described by Lesher and cowork-
ers (e.g., Lesher and Keays, 2002; Houl et al., 2012) also pro-
vides a mechanism for the incorporation of sulfide (whether
derived via devolatilization, sulfide dissolution, or sulfide melt-
ing) that greatly exceeds saturation values. Processes such as
these that promote sulfide supersaturation in a mafic magma
0361-0128/98/000/000-00 $6.00
53
are clearly of potential significance in the generation of mag-
illustrated above, many processes that do not involve the in-
corporation of external S may promote sulfide saturation of a
mafic magma. Whether an economic accumulation of sulfide
may result is a function of the size of the magmatic system,
rate and duration of magma emplacement, and efficiency of
sulfide collection.
The Importance of Magma Volume and Chamber Shape
Given that mafic magma may reach sulfide saturation as a
result of fractional crystallization, and much sooner via con-
tamination involving a siliceous end-member or by magma
mixing, the question remains whether an economic sulfide
accumulation may be produced without the addition of exter-
nal S. In all cases described above, except those of xenolith
liquid produced is that of, or very near, the cotectic propor-
tion (Barnes, 2007; Li and Ripley, 2009). Both Barnes (2007)
and Li and Ripley (2009) describe methods to calculate the
cotectic proportion of sulfide. The amount of sulfide liquid
produced per crystallization increment is small (generally
<1.5 wt %), and very efficient collection is required to pro-
duce horizons with more than ~2 vol % sulfide. In the case of
a magma with a sulfide saturation value that has been decreased
due to contamination and mixing, the mass of collectible sul-
fide is constrained by the initial concentration of sulfide in the
system. This is an obvious point, but one of extreme impor-
tance when considering ore-forming systems. For example,
for the Sonju Lake intrusion the parental magma is estimated
to have contained no more than 1,000 ppm S. Miller and Rip-
ley (1996) estimate that the intrusion is considerably larger
than its exposed area of 9 km2, with a strike length of ~20 km.
We estimate a potential volume of ~30 km3 of magma, but of
particular significance is the fact that, even with the improb-
able collection of all available sulfide, only a ~2-m-thick mas-
sive sulfide horizon could be produced in the intrusion. As
stated above, Ni concentration dropped to extremely low val-
ues prior to sulfide saturation. When sulfide saturation was
attained in the intrusion as a result of fractional crystalliza-
tion, the R-factor would have been ~260, assuming sulfide
saturation affected all portions of the remaining 60% liquid.
A 2-m-thick massive sulfide would grade ~3 wt % Cu (100
ppm Cu in the parent magma) and 1.3 ppm Pt (3.5 ppb Pt in
initial liquid). A 2-m-thick massive sulfide would be an un-
likely candidate for a Cu ore, but a PGE reef could be pro-
duced. In the Sonju Lake intrusion, the zone of PGE and Cu
enrichment is dispersed over ~100 m rather than 1 or 2 m,
with grades of ~0.06% Cu and 30 ppb Pt, but less than 70
ppm Ni. These are expected values for dispersion of the
immiscible liquid with the composition described above.
Without the addition of external S, most magmas that crys-
tallize as closed systems (whether uncontaminated or after
contamination involving S-poor country rocks) are unlikely to
produce sufficient quantities of sulfide to form economic
deposits, unless the volume of magma from which sulfide is
derived is large. Barnes (2007) evaluated whether cotectic
amounts of sulfide were sufficient to account for the large,
disseminated sulfide orebody at Mt. Keith. Large, dissemi-
nated sulfide-bearing intrusions are logical candidates where
53
54 RIPLEY AND LI

crystallization of cotectic proportions of sulfide may have
been responsible for ore generation. Barnes (2007) con-
cluded that sulfide in excess of cotectic amounts was required
to produce the Mt. Keith orebody and proposed the mechan-
ical deposition of transported sulfide, formed in response to
the introduction of sulfur from country rocks.
In a very large volume system such as the impact melt sheet
at Sudbury, concentrations of 600 to 1,000 ppm S (as esti-
mated by Keays and Lightfoot, 1999; Lightfoot et al., 2001)
are consistent with the estimated mass of sulfide in the Sud-
bury ores (Lesher and Thurston, 2002; Keays and Lightfoot,
2004; Fig. 5). For a system such as Norilsk (sulfide content
between 275 and 750 Mt; Lightfoot and Hawkesworth, 1997;
Naldrett, 2011), the amount of sulfide present far exceeds
that which could be collected from closed-system crystalliza-
tion of magma that produced tabular, ~1,000-m-thick (or less)
sills. If contamination not involving S addition is a key process
for the attainment of saturation, then open-system processes
are likely to have been operative to result in the concentration
of sufficient quantities of sulfide. The mass and grade of sul-
fide at Norilsk have been the focus of many studies, virtually
all of which focus on sulfide collection from passing magmas
in a conduit system, followed by some type of metal upgrad-
ing (e.g., Brgmann et al., 1993; Naldrett and Lightfoot,
1999; Arndt et al., 2003; Kerr and Leitch, 2005; Lightfoot and
Keays, 2005; Li et al., 2009). Figure 5 illustrates that for the
Norilsk ores, the estimated sulfide mass, if collected from
magma with ~1,000 ppm S available as immiscible sulfide,
would require between ~34 and 100 km3 of magma. For com-
parison, the 30-Mt Ovoid orebody at Voiseys Bay (Lightfoot
et al., 2012) would have required sulfide collection from ~3.7
km3 of magma. For both of these systems the required
amounts of magma may not be unreasonable; the involve-
ment of 100 km3 in an ore-forming system represents only a
small fraction of that which generated the Siberian Traps
(e.g., Reichow et al., 2009). For the small but high-grade
(~3.5% Ni) Eagle deposit, the 4-Mt sulfide orebody could
have been generated from as little as 0.5 km3 of magma con-
taining 1,000 ppm S. A more realistic estimate for the Eagle
deposit also serves to illustrate that Ni contents in parental
magmas of ~150 ppm may give rise to sulfide assemblages
with tenors similar to those found in many deposits. Figure 1b
shows that sulfide saturation of a picritic magma contami-
nated with a partial melt from a pelite may occur at ~15%
crystallization. The Ni content of the liquid at this time would
be ~155 ppm. Crystallization of an additional 10% involves
clinopyroxene and results in the Ni content of the liquid
falling to no less than 140 ppm. The computed mass of
immiscible sulfide available for collection from the residual
liquid over this interval is ~0.3 wt %, resulting in an R-factor
of ~335. With Ni contents in the residual liquid of 140 to 150
ppm, R-factors of 200 to 350 will produce Ni tenors in the sul-
fide liquid of 2 to 3 wt %. Hence, collection of this sulfide
could generate an orebody similar in grade to many of the
worlds large deposits. For the small, 4-Mt orebody at Eagle,
the amount of silicate magma needed based on the above cal-
culation is ~2 km3 .These values would appear to be very rea-
sonable quantities of magma to be delivered in a developing
rift system. The example from the Sonju Lake intrusion also
serves to emphasize that if the amount of mantle-derived sul-
fide present in a moderate-sized, sheet-type intrusion were
to have been collected in a restricted magmatic conduit, an

FIG. 5. Plot of tonnes of sulfide in ore versus S concentration in a source magma. Cubic kilometers of magma are shown
assuming that all the sulfur in magma is collected to generate the ore. For example, if only 200 ppm S is available as immis-
cible sulfide during magma crystallization, deposits similar in size to those at Norilsk (between 275 and 750 Mt) can be gen-
erated from between 200 and 300 km3 of magma. Small orebodies such as Eagle can be formed by collection of immiscible
sulfide from less than 10 km3 of magma.

0361-0128/98/000/000-00 $6.00 54
 SULFIDE SATURATION IN MAFIC MAGMAS: IS EXTERNAL SULFUR REQUIRED FOR MAGMATIC Ni-Cu-(PGE) ORE GENESIS?
economic ore accumulation could have resulted. The conduit
environment provides a physical setting where small amounts
of sulfide from large volumes of magma may collect; in addi-
tion, subsequent upgrading may occur as sulfide-undersatu-
rated magmas utilize the conduit and partially dissolve the
accumulated sulfide (e.g., Kerr and Leitch, 2005; Li et al.,
2009). Ni enrichment of a low-Ni sulfide that initially formed
from a system where Ni was sequestered by olivine could also
occur. Maier et al. (2001) pointed out the importance of con-
duits in the formation of magmatic Ni-Cu-PGE sulfide
deposits, as well as the lack of such deposits in large layered
intrusions. The point here is that geochemical data are con-
sistent with the mineralization that is found in intrusions such
as Sonju Lake being related to processes that did not involve
the introduction of external S (or even significant contamina-
tion of any type). Had this sulfide been able to accumulate in
a conduit setting, there appears to be little reason to suspect
that an orebody could not have resulted. Barnes et al. (2011)
have recently described an Ni-rich disseminated sulfide ore-
body that occurs as a stratiform layer in the Fazenda Mirabela
intrusion of Brazil. The authors note that this is a very unusual
occurrence produced as a result of mixing of two composi-
tionally distinct magmas, one of which contained suspended
sulfide droplets. 34S values of sulfide minerals (1.00.5)
are consistent with the premise that sulfur is of mantle origin
(Almeida Lazarin, 2011).
Isotopic data from Norilsk (e.g., Gorbachev and Gri-
nenko, 1973; Grinenko, 1985; Arndt et al., 2003; Li et al.,
2003b), Voiseys Bay (Ripley et al., 1999, 2002), and Eagle
(Ding et al., 2012) indicate that externally derived S has
been involved in ore genesis in each of these systems. Sulfur
isotope data (e.g., Ripley et al., 2010) suggest that as much
as 50% of the S contained in the Norilsk deposits is of coun-
try-rock origin; this permits the speculation that, consider-
ing the size of the Norilsk deposits, economic deposits
could have formed even without the addition of external S.
Recent evaluations suggest that this is an unlikely scenario
(Keays and Lightfoot, 2010). However, it is clear that ex-
tremely efficient collection of mantle-derived sulfide from
appropriate volumes of magma will reduce the importance
of externally derived S in the formation of magmatic ore de-
posits. The fact that many large Ni-Cu-(PGE) deposits show
evidence for external S addition suggests that such efficient
collection of immiscible sulfide derived solely from mantle
S is a rare process. Early attainment of sulfide saturation
may be pivotal not only for Ni availability but also for poten-
tial sulfide collection in conduits prior to dissemination
within sheet- or sill-like intrusions. The external addition of
S to a mafic/ultramafic magma may not only supply addi-
tional sulfide mass, but also trigger early sulfide saturation
and, potentially, more rapid collection of immiscible sulfide
liquid. Given that early attainment of sulfide saturation is
possible in contaminated magma as we have shown above,
the mechanisms of collection of immiscible sulfide liquid re-
main as important topics for researchers concerned with the
genesis of magmatic Ni-Cu-(PGE) deposits. Mungall and Su
(2005), Chung and Mungall (2009), and de Bremond dArs
et al. (2001) have provided constraints on physical proper-
ties related to the collection of immiscible sulfide liquid
droplets. Additional research of this type is vital to enhance
0361-0128/98/000/000-00 $6.00
55
our understanding of all processes that may lead to the for-
mation of magmatic Ni-Cu-(PGE) deposits.
Conclusions
An assessment of mechanisms that can promote sulfide sat-
uration in mafic magmas indicates that externally derived S
may not be required for Ni-Cu-(PGE) ore formation in open
systems where the volume of magma involved is large and an
efficient process of sulfide collection is operative. Conduits
provide a physical environment where sulfide collection and
upgrading may occur, related in part to the passage of multi-
ple pulses of magma. The fact that many large Ni-Cu-(PGE)
sulfide deposits show strong evidence for the incorporation of
crustally derived S suggests that efficient collection of man-
tle-derived sulfide is a rare process, and that sulfide supersat-
uration is often required. Sulfide supersaturation in magmas
is accomplished via the liberation of sulfur from xenoliths in
intrusions or via similar transport due to thermomechanical
erosion near the base of komatiite lava flows.
Closed-system fractional crystallization of mantle-derived
magmas will be unlikely to produce sulfide accumulations of
sufficient size for economic exploitation. Sulfide saturation in
most uncontaminated mantle-derived magmas will not occur
via fractional crystallization until ~20 to 40% crystallization.
Sulfide accumulations that may form will be Ni poor because
of sequestration of Ni by early crystallizing olivine. Systems
with sufficient volume have the potential to generate Cu-rich
PGE reefs. The contamination of mafic magmas by siliceous
country rocks, or mixing of magma types, will cause a reduc-
tion in the S concentration necessary for the attainment of
early sulfide saturation of between ~ 200 and 700 ppm. This
facilitates sulfide saturation but does not necessarily trigger it.
Where the aSiO2 is sufficiently elevated to cause orthopyrox-
ene rather than olivine crystallization, more Ni is available for
incorporation into sulfide. Ni sulfide-rich ores could form
preferentially in orthopyroxene-bearing rocks, but only from
large-volume systems.
Addition of volatiles such as H2O and CO2 to mafic magmas
is frequently linked to an increase in magma fO2 and a lower-
ing of the SCSS. Assimilation of oxygen-bearing volatiles does
not necessarily result in an increase of magma fO2; in addition,
at QFM and below an increase in fO2 results in only small
changes in the SCSS of the magma. Assimilation of siliceous,
and potentially hydrous, partial melts derived from country
rocks and mixing of chemically distinct magmas are more
viable processes for decreasing the SCSS of a mafic magma,
relative to increased fO2 alone. Methane, as well as graphite,
may serve as a reductant in high fO2 magmas where sulfate is
predominant and the solubility of S is considerably higher
than in reduced magmas.
Acknowledgments
Discussions with A.J. Naldrett, C.M. Lesher, R.R. Keays,
Jon Hronsky, P.C. Lightfoot, W.D. Maier, S.-J. Barnes, M.
Fiorentini, and S.J. Barnes regarding S saturation in mafic
magmas are appreciated and helped to stimulate the writing
of this manuscript. Reviews by Wolf Maier and Alexey Ariskin
improved the manuscript presentation. We thank Steve Barnes,
who served as associate editor for the manuscript, and Editor
Larry Meinert, for their comments and supervision. Research
55
56 RIPLEY AND LI
on magmatic ore deposits at Indiana University has been sup-
ported by NSF Grant EAR 1016031 to E.M. Ripley and C.
Li. A gift to the Indiana University Economic Geology pro-
gram from Rio Tinto is gratefully acknowledged.
REFERENCES
Almeida Lazarin, F., 2011, Geologic, petrologic and isotopic studies of the
Santa Rita and Peri-Peri nickel-copper sulfide deposits, northeastern
Brazil: Unpublished thesis, Brasilia, University of Brasilia, 69 p.
Andrews, M., and Ripley, E.M., 1989, Mass transfer and sulfur fixation in the
contact aurole of the Duluth Complex, Dunka Road Cu-Ni deposit, Min-
nesota: Canadian Mineralogist, v. 27, p. 293310.
Arcuri, T., Ripley, E.M., and Hauck, S.A., 1998, Sulfur and oxygen isotope
studies of the interaction between pelitic xenoliths and basaltic magma at
the Babbitt and Serpentine Cu-Ni deposits, Duluth Complex, Minnesota:
ECONOMIC GEOLOGY, v. 93, p. 10631075.
Ariskin, A.A., Danyyushevsky, L.V., Bychkov, K.A., and Barmina, G.S., 2010,
Modeling sulphide solubility in MORB glasses and Ol cumulate piles: The
opposite effect of NiO vs FeO in the melt: Ontario Geological Survey, 11th
International Platinum Symposium, Sudbury, Ontario, June 2010, Abstract
volume, Miscellaneous Release, 4 p.
Arndt, N.T., Czamanske, G.K., Walker, R.J., Chauvel, C., and Fedorenko,
V.A., 2003, Geochemistry and origin of the intrusive hosts of the Norilsk-
Talnakh Cu-Ni-PGE sulfide deposits: ECONOMIC GEOLOGY, v. 98, p. 495515.
Barnes, S.-J., Maier, W.D., and Curl, E.A., 2010, Composition of the mar-
ginal rocks and sills of the Rustenburg Layered Suite, Bushveld Complex,
South Africa: Implications for formation of the platinum group element
deposits: ECONOMIC GEOLOGY, v. 105, p. 14911511.
Barnes, S.J., 1989, Are Bushveld U-type parent magmas boninites or conta-
minated komatiites: Contributions to Mineralogy and Petrology, v. 101, p.
447457.
2007, Cotectic precipitation of olivine and sulfide liquid from komatiite
magma and the origin of komatiite-hosted disseminated nickel sulfide min-
eralization at Mount Keith and Yakabindie, Western Australia: ECONOMIC
GEOLOGY, v. 102, p. 299304.
Barnes, S.J., Osborne, G.A., Cook, D., Barnes, L., Maier, W.D., and Godel,
B., 2011, The Santa Rita nickel sulfide deposit in the Fazenda Mirabela in-
trusion, Bahia, Brazil: Geology, sulfide geochemistry, and genesis: ECO-
NOMIC GEOLOGY, v. 106, p. 10831110.
Bekker, A., Barley, M.E., Fiorentini, M.L, Rouxel, O.J., Rumble, D., and
Beresford, S.W., 2009, Atmospheric sulfur in Archean komatiite-hosted
nickel deposits: Science, v. 326, p. 10801089.
Blank, J.G., and Brooker, R.A., 1994, Experimental studies of carbon dioxide
in silicate melts: Solubility, speciation and stable carbon isotope behavior:
Reviews in Mineralogy and Geochemistry, v. 30, p. 157186.
Brgmann, G.E., Naldrett, A.J., Asif, M., Lightfoot, P.C., Gorbachev, N.S.,
and Fredorenko, V.A., 1993, Siderophile and chalcophile metals as tracers
of the evolution of the Siberian Trap in the Norilsk region, Russia:
Geochimica et Cosmochimica Acta, v. 57, p. 20012018.
Buchanan, D.I., and Nolan, J., 1979, Solubility of sulfur and sulfide immisci-
bility in synthetic tholeiitic melts and their relevance to Bushveld Complex
rocks: Canadian Mineralogist, v. 17, p. 483494.
Campbell, I.H., Naldrett, A.J., and Barnes, S.J., 1983, A model for the origin
of the platinum-rich sulfide horizons in the Bushveld and Stillwater Com-
plexes: Journal of Petrology, v. 24, p. 133165.
Carmichael, I.S.E., 1991, The redox state of basic silicic magmas: A reflec-
tion of their source regions?: Contributions to Mineralogy and Petrology, v.
106, p. 129141.
Chung, H.-Y., and Mungall, J.E., 2009, Physical constraints on the migration
of immiscible fluids through partially molten silicates, with special refer-
ence to magmatic sulfide ores: Earth and Planetary Science Letters, v. 286,
p. 1422.
Clemente, B., Scaillet, B., and Pichavent, M., 2004, The solubility of sulfur
in hydrous rhyolitic melts: Journal of Petrology, v. 45, p. 21712196.
Danckworth, P.A., Hess, P.C., and Rutherford, M.J., 1979, The solubility of
sulfur in high-TiO2 mare basalts, in Proceedings of the Lunar and Planetary
Science Conference, 10th: New York, Pergamon Press, p. 517530.
de Bremond dArs, J., Arndt, N.T., and Hallot, E., 2001, Analog experimen-
tal insights into the formation of magmatic sulfide deposits: Earth and Plan-
etary Science Letters, v. 186, p. 371381.
Ding, X., Li, C., Ripley, E.M., Rossell, D., and Kamo, S., 2010, The Eagle
and East Eagle Sulfide ore-bearing mafic-ultramafic intrusions in the
0361-0128/98/000/000-00 $6.00
Midcontinent Rift System, upper Michigan: Geochronology and petrologic
evolution: Geochemistry Geophysics Geosystems, v. 11, 22 p., doi: 10.1029/
2009 GC002546.
Ding, X., Ripley, E.M., Shirey, S.B., and Li, C., 2012, Os, Nd, O and S iso-
tope constraints on country rock contamination in the conduit-related
Eagle Cu-Ni-(PGE) deposit, Midcontinent rift system, upper Michigan:
Geochimica et Cosmochimica Acta, v. 89, p. 1030.
Evans, K.A., ONeill, H.St.C, and Mavrogenes, J.A., 2008, Sulphur solubility
and sulphide immiscibility as a function of the concentration of manganese,
nickel, tungsten and copper at 1 atm amd 1400C: Chemical Geology, v.
255, p. 236249.
Farquhar, J., and Wing, B.A., 2003, Multiple sulfur isotopes and the evolu-
tion of the atmosphere: Earth and Planetary Science Letters, v. 213, p.
113.
Frei, D., Liebscher, A., Franz, G., Wunder, B., Klemme, S., and Blundy, J.,
2009, Trace element partitioning between orthopyroxene and anhydrous
silicate melt on the lherzolite solidus from 1.1 to 3.2 GPa and 1,230 to
1,535C in the model system Na2O-CaO-MgO-Al2O3-SiO2: Contributions
to Mineralogy and Petrology, v. 157, p. 473490.
Ghiroso, M.S., and Sack, R.O., 1995, Chemical and mass transfer in mag-
matic processes IV. A revised and internally consistent thermodynamic
model for the interpolation and extrapolation of liquid-solid equilibria in
magmatic systems at elevated temperatures and pressures: Contributions
to Mineralogy and Petrology, v. 119, p. 197213.
Godel, B., Seat, Z., Maier, W.D., and Barnes, S.-J., 2011, The Nebo-Babel
Ni-Cu-PGE sulfide deposit (West Musgrave block, Australia): Pt. 2. Con-
straints on parental magma and processes, with implications for mineral ex-
ploration: ECONOMIC GEOLOGY, v. 106, p. 557584.
Gorbachev, P.N., and Bezmen, N.I., 2011, The influence of the iron and
ox/red conditions on the solubility of sulfur in water saturated silicate melts:
Vestnik Otdelenia nauk o Zemle Ran, v. 3, 3 p., doi: 10.2205/2011
NZ000153.
Gorbachev, N.S., and Grinenko, L.N., 1973, The sulfur isotope ratios of sul-
fides and sulfates of the Oktyabrsky sulfide deposit, Norilsk region, and the
problem of its origin: Geokhimiya, v. 8, p. 11271136. (in Russian)
Grinenko, L.N., 1985, Sources of sulfur of the nickeliferous and barren gab-
bro-dolerite intrusions of the northwest Siberian platform: International
Geology Review, v. 28, p. 695708.
Hashizume, H., and Hariya, Y., 1992, Partitioning of the transition metal Mn,
Fe, Co, Ni and Zn between orthopyroxene and silicate melt: Ionic size ef-
fect to the partitioning behavior: Journal of the Faculty of Science,
Hokkaido University, v. 23, p. 211226.
Haughton, D.R., Roeder, P.L., and Skinner, B.J., 1974, Solubility of sulfur in
mafic magmas: ECONOMIC GEOLOGY, v. 69, p. 451467.
Hoffer, E., and Grant, J.A., 1980, Experimental investigation of the forma-
tion of cordierite-orthopyroxene parageneses in pelitic rocks: Contributions
to Mineralogy and Petrology, v. 73, p. 1522.
Holwell, D.A., Boyce, A.J., and McDonald, I., 2007, Sulfur isotope variations
within the Platreef Ni-Cu-PGE deposit: Genetic implications for the origin
of sulfide mineralization: ECONOMIC GEOLOGY, v. 102, p. 10911110.
Holzheid, A., and Grove, T.L., 2002, Sulfur saturation limits in silicate melts
and their implications for core formation scenarios for terrestrial planets:
American Mineralogist, v. 87, p. 227237.
Houl, M.G., Lesher, C.M., and Davis, 2012, Thermomechanical erosion at
the Alexo mine, Abitibi greenstone belt, Ontario: Implications for the
genesis of komatiite-associated Ni-Cu-(PGE) mineralization: Mineralium
Deposita, v. 47, p. 105128.
Irvine, T.N., 1975, Crystallization sequences of the Muskox intrusion and
other layered intrusions: II. Origin of the chromite layers and similar
deposits of other magmatic ores: Geochimica et Cosmochimica Acta, v. 39,
p. 9911020.
Jugo, P.J., 2009, Sulfur content at sulfide saturation in oxidized magmas:
Geology, v. 37, p. 415418.
2010, Exponential increase of sulfur content at sulfide saturation with
increasing oxygen fugacity: Implications for Cu-Ni-PGE mineralization
[abs]: Ontario Geological Survey, 11th International Platinum Symposium,
Sudbury, Ontario, June 2010, Abstract volume, Miscellaneous Release, 4 p.
Jugo, P.J., and Lesher, C.M., 2005, Redox changes caused by evaporate and
carbon assimilation at Norilsk and their role in sulfide precipitation [abs.]:
Geological Society of America Abstracts with Programs, v. 39, p. 360.
Jugo, P.J., Luth, R.W., and Richards, J.P., 2005, An experimental study of the
sulfur content is basaltic melts saturated with immiscible sulfide or sulfate
liquids at 1300C and 1.0 GPa: Journal of Petrology, v. 46, p. 783798.
56
 SULFIDE SATURATION IN MAFIC MAGMAS: IS EXTERNAL SULFUR REQUIRED FOR MAGMATIC Ni-Cu-(PGE) ORE GENESIS?
Keays, R.R., 1995, The role of komatiitic and picritic magmatism and S-sat-
uration in the formation of ore deposits: Lithos, v. 34, p. 118.
Keays, R.R., and Lightfoot, P.C., 1999, The role of meteorite impact, source
rocks, protores and mafic magmas in the genesis of the Sudbury Ni-Cu-
PGE sulfide ore deposits: Geological Association of Canada Short Course
Notes 13, p. 329366.
2004, Formation of Ni-Cu-platinum group element sulfide mineraliza-
tion in the Sudbury impact melt sheet: Mineralogy and Petrology, v. 82, p.
217258.
2010, Crustal sulfur is required to form magmatic Ni-Cu sulfide de-
posits: Evidence from chalcophile element signatures of Siberian and Dec-
can Trap basalts: Mineralium Deposita, v. 45, p. 241257.
Kent, A.J.R., Clague, D.A., Masahikko, H., Stolper, E.M., Hutcheon, I.D.,
and Norman, M.D., 1999, Widespread assimilation of a seawater-derived
component at Loihi seamount, Hawaii: Geochimica et Cosmochimica Acta,
v. 63, p. 27492761.
Kerr, A., and Leitch, A.M., 2005, Self-destructive sulfide segregation systems
and the formation of high-grade magmatic ore deposits: ECONOMIC GEOL-
OGY, v. 100, p. 311332.
Lee, I., and Ripley, E.M., 1995, Genesis of Cu-Ni sulfide mineralization in
the South Kawishiwi intrusion, Spruce Road area, Duluth Complex, Min-
nesota: Canadian Mineralogist, v. 33, p. 723743.
Lehmann, J., Arndt, N., Windley, B., Zhou, M.-F., Wang, C.Y., and Harris,
C., 2007, Field relationships and geochemical constraints on the emplace-
ment of the Jinchuan intrusion and its Ni-Cu-PGE sulfide deposit, Gansu,
China: ECONOMIC GEOLOGY, v. 102, p. 7594.
Lesher, C.M., and Groves, D.I., 1986, Controls on the formation of komati-
ite-associated nickel-copper sulfide deposits, in Friedrich, G.H., Genkin,
A., Naldrett, A.J., Ridge, J.D., Sillitoe, R.H., and Vokes, F.M., eds., Geol-
ogy and metallogeny of copper deposits, Proceedings of the Twenty-Sev-
enth International Geological Congress, Moscow: Berlin, Springer-Verlag,
p. 4362.
Lesher, C.M., and Keays, R.R., 2002, Komatiite-associated Ni-Cu-(PGE) de-
posits: Mineralogy, geochemistry and genesis: Canadian Institute of Mining
Special Volume 54, p. 579617.
Lesher, C.M., and Thurston, P.C., 2002, A special issue devoted to the min-
eral deposits of the Sudbury BasinPreface: ECONOMIC GEOLOGY, v. 97, p.
13731375.
Li, C., and Naldrett, A.J., 2000, Melting reactions of gneissic inclusions with
enclosing magma at Voiseys Bay, Labrador, Canada: Implications with re-
spect to ore genesis: ECONOMIC GEOLOGY, v. 95, p. 801814.
Li, C., and Ripley, E.M., 2005, Empirical equations to predict the sulfur con-
tent of mafic magmas at sulfide saturation and applications to magmatic
sulfide deposits: Mineralium Deposita, v. 40, p. 11731180.
2009, Sulfur contents at sulfide-liquid or anhydrite saturation in silicate
melts: Empirical equations and example applications: ECONOMIC GEOL-
OGY, v. 104, p. 405412.
Li, C., Ripley, E.M., Maier, W.D., and Gomwey, T.E.S., 2002, Olivine and
sulfur isotopic compositions of the Uitkomst Ni-Cu sulfide ore-bearing
complex, South Africa: Evidence for sulfur contamination and multiple
magma emplacements: Chemical Geology, v. 188, p. 149159.
Li, C., Ripley, E.M., and Mathez, E.A., 2003a, The effect of S on the parti-
tioning of Ni between olivine and silicate melt in MORB: Chemical Geol-
ogy, v. 201, p. 295306.
Li, C., Ripley, E.M., and Naldrett, A.J., 2003b, Compositional variation of
olivine and sulfur isotopes in the Norilsk and Talnakh intrusions, Siberia:
Implications for ore-forming processes in dynamic magma conduits: ECO-
NOMIC GEOLOGY, v. 98, p. 6986.
Li, C., Ripley, E.M., and Naldrett, A.J., 2009, A new genetic model for the
giant Ni-Cu-PGE sulfide deposits associated with Siberian flood basalts:
ECONOMIC GEOLOGY, v. 104, p. 291301.
Lightfoot, P.C., and Hawkesworth, C.J., 1997, Flood basalts and magmatic
Ni, Cu and PGE sulphide mineralization: Comparative geochemistry of the
Norilsk (Siberian Trap) and West Greenland sequences: American Geo-
physical Union Monograph 100, p. 357380.
Lightfoot, P.C., and Keays, R.R., 2005, Siderophile and chalcophile metal
variations in flood basalts from the Siberian Trap, Norilsk region: Implica-
tions for the origin of the Ni-Cu-PGE sulfide ores: ECONOMIC GEOLOGY, v.
100, p. 439462.
Lightfoot, P.C., Keays, R.R., and Doherty, W., 2001, Chemical evolution and
origin of nickel sulfide mineralization in the Sudbury Igneous Complex,
Ontario, Canada: ECONOMIC GEOLOGY, v. 96, p. 18551875.
0361-0128/98/000/000-00 $6.00
57
Lightfoot, P.C., Keays, R.R., Evans-Lambswood, D., and Wheeler, R., 2012,
S saturation history of Nain Plutonic Suite mafic intrusions: Origin of the
Voiseys Bay Ni-Cu-Co sulfide deposit, Labrador, Canada: Mineralium
Deposita, v. 47, p. 2350.
Liu, Y., Samaha, N.T., and Baker, D.R., 2007, Sulfur concentration at sulfide
saturation (SCSS) in magmatic silicate melts: Geochimica et Cosmochim-
ica Acta, v. 71, p. 17831799.
Maier, W.D., Li, C., and De Waal, S.A., 2001, Why are there no major Ni-Cu
sulfide deposits in large layered mafic-ultramafic intrusions: Canadian Min-
eralogist, v. 39, p. 547556.
Maier, W.D., Barnes, S.-J., and Ripley, E.M., 2011, The Kabanga Ni sulfide
deposits, Tanzania: A review of ore-forming processes: Reviews in Eco-
nomic Geology, v. 17, p. 217234.
Mariga, J., Ripley, E.M., and Li, C., 2006, Petrologic evolution of gneissic
xenoliths in the Voiseys Bay intrusion, Labrador, Canada: Mineralogy, re-
actions, partial melting, and mechanisms of mass transfer: Geochemistry
Geophysics Geosystems, v. 7, 25 p., doi: 10.1029/2005GC 001184.
Mavrogenes, J.A., and ONeill, H.St.C., 1999, The relative effects of pres-
sure, temperature, and oxygen fugacity on the solubility of sulfide in mafic
magmas: Geochimica et Cosmochimica Acta, v. 63, p. 11731180.
Miller, Jr., J.D., 1999, Geochemical evaluation of platinum group element
(PGE) mineralization in the Sonju Lake intrusion, Finland, Minnesota:
Minnesota Geological Survey Circular 44, 31 p.
Miller, Jr., J.D., and Ripley, E.M., 1996, Layered intrusions of the Duluth
Complex, USA, in Cawthorn, R.G., ed., Layered intrusions: Amsterdam,
Elsevier, p. 257301.
Mollo, S., Gaeta, M., Freda, C., Di Rocco, T., Misiti, V., and Scarlato, P.,
2010, Carbonate assimilation in magmas: A reappraisal based on experi-
mental petrology: Lithos, v. 114, p. 503514.
Moune, S., Holtz, F., and Botcharnikov, R.E., 2009, Sulfur solubility in an-
desitic to basaltic melts: Implications for Hekla volcano: Contributions to
Mineralogy and Petrology, v. 157, p. 691707.
Mungall, J.E., and Su, S., 2005, Interfacial tension between magmatic sulfide
and silicate liquids: Constraints on kinetics of sulfide liquation and sulfide
migration through silicate rocks: Earth and Planetary Science Letters, v.
234, p. 135149.
Naldrett, A.J., 1966, The role of sulphurization in the genesis of iron-nickel
sulfide deposits of the Porcupine district, Ontario: Canadian Institute of
Mining and Metallurgy, Transactions, v. 69, p. 147155.
2011, Fundamentals of magmatic sulfide deposits: Reviews in Economic
Geology, v. 17, p. 150.
Naldrett, A.J., and Lightfoot, P.C., 1999, Ni-Cu-PGE deposits of the Norilsk
region, Siberia: Their formation in conduits for flood basalt volcanism:
Geological Association of Canada Short Course Notes 13, p. 195250.
Naldrett, A.J., and von Gruenewaldt, G., 1989, Association of platinum-
group elements with chromite in layered intrusions and ophiolite com-
plexes: ECONOMIC GEOLOGY, v. 84, p. 180187.
Nixon, G.T., 1998, Ni-Cu sulfide mineralization in the Turnagain Alaskan-
type complex: A unique magmatic environment: British Columbia Ministry
of Employment and Investment, Geological Fieldwork Report 1998-1, p.
18.118.11.
Ohmoto, H., and Kerrick, D., 1977, Devolatilization equilibria in graphitic
systems: American Journal of Science, v. 277, p. 10131044.
ONeill, H.St.C., and Mavrogenes, J.A., 2002, The sulfide capacity and the
sulfur content at sulfide saturation of silicate melts at 1400C and 1 bar:
Journal of Petrology, v. 43, p. 10481087.
Palme, H., and ONeill, H.St.C., 2004, Cosmochemical estimates of mantle
composition: Treatise on Geochemistry, v. 2, p. 138.
Park, Y.-R., Ripley, E.M., Miller, Jr., J.D., Li, C., Mariga, J., and Shafer, P.,
2004, Stable isotopic constraints on fluid-rock interaction and Cu-PGE-S
redistribution in the Sonju Lake intrusion, Minnesota: ECONOMIC GEOL-
OGY, v. 99, p. 325338.
Reichow, M.K., Pringle, M.S., AlMukhamedov, A.I., Allen, M.B., Andre-
ichev, V.I., Buslov, M.M., Davies, C.E., Fedoseev, G.S., Fitton, J.G., Inger,
S., Medvedev, A.Ya., Mitchell, C., Puchkov, V.N., Safonova, I.Y., Scott,
R.A., and Saunders, A.P., 2009, The timing and extent of the eruption of the
Siberian Traps large igneous province: Implications for the end-Permian
environmental crisis: Earth and Planetary Science Letters, v. 277, p. 920.
Righter, K., Yang, H., Costin, G., and Downs, R.T., 2008, Oxygen fugacity in
the Martian mantle controlled by carbon: New constraints form the nakhlite
MIL 03346: Meteoritics and Planetary Science, v. 43, p. 17091723.
Ripley, E.M., 1981, Sulfur isotopic studies of the Dunka Road Cu-Ni deposit,
Duluth Complex, Minnesota: ECONOMIC GEOLOGY, v. 76, p. 610620.
57
58 RIPLEY AND LI
1999, Systematics of sulfur and oxygen isotopes in mafic igneous rocks
and related Cu-Ni-PGE mineralization: Geological Association of Canada
Short Course Notes, v. 13, p. 133158.
Ripley, E.M., and Alawi, J.A., 1988, Petrogenesis of pelitic xenoliths at the
Babbitt Cu-Ni deposit, Duluth Complex, Minnesota, U.S.A.: Lithos, v. 21,
p. 143159.
Ripley, E.M., and Al-Jassar, T., 1987, Sulfur and oxygen isotope studies of
melt-country rock interaction, Babbitt Cu-Ni deposit, Duluth Complex,
Minnesota: ECONOMIC GEOLOGY, v. 82, p. 87107.
Ripley, E.M., and Li, C., 2003, Sulfur isotope exchange and metal enrich-
ment in the formation of magmatic Cu-Ni-(PGE) deposits: ECONOMIC
GEOLOGY, v. 98, p. 635641.
Ripley, E.M., Park, Y.R., Li, C., and Naldrett, A.J, 1999, Sulfur and oxygen
isotopic evidence of country rock contamination in the Voiseys Bay Ni-Cu-
Co deposit, Labrador, Canada: Lithos, v. 47, p. 5368.
Ripley, E.M., Li, C., and Shin, D., 2002, Paragneiss assimilation in the gene-
sis of magmatic Ni-Cu-Co sulfide mineralization at Voiseys Bay, Labrador:
34S, 13C, and Se/S evidence: ECONOMIC GEOLOGY, v. 97, p. 13071318.
Ripley, E.M., Taib, N.I., Li, C., and Moore, C.H., 2007, Chemical and min-
eralogical heterogeneity in the basal zone of the Partridge River intrusion:
Implications for the origin of Cu-Ni sulfide mineralization in the Duluth
Complex, Midcontinent Rift System: Contributions to Mineralogy and
Petrology, v. 154, p. 3554.
Ripley, E.M., Li, C., Moore, C.H., and Schmitt, A.K., 2010, Micro-scale S
isotope studies of the Kharaelakh intrusion, Norilsk region, Siberia: Con-
straints on the genesis of coexisting anhydrite and sulfide minerals:
Geochimica et Cosmochimica Acta, v. 74, p. 634644.
Scaillet, B., and Macdonald, R., 2006, Experimental and thermodynamic
constraints on the sulfur yield of peralkaline and metaluminous silica flood
eruptions: Journal of Petrology, v. 47, p. 14131437.
Scoates, J.S., and Mitchell, J.N., 2000, The evolution of troctolitic and high
Al basaltic magmas in Proterozoic anorthosite plutonic suites and implica-
tions for the Voiseys Bay massive Ni-Cu sulfide deposit: ECONOMIC GEOL-
OGY, v. 95, p. 677702.
0361-0128/98/000/000-00 $6.00
Seat, Z., Beresford, S.W., Grguric, B.A., Gee, M.A.M., and Grassineau, N.V.,
2009, Reevaluation of the role of external sulfur addition in the genesis of
Ni-Cu-PGE deposits: Evidence from the Nebo-Babel Ni-Cu-PGE deposit,
West Musgrave, Western Australia: ECONOMIC GEOLOGY, v. 104, p.
521538.
Shima, H., and Naldrett, A.J., 1975, Solubility of sulfur in ultramafic melt and
the relevance of the system Fe-S-O: ECONOMIC GEOLOGY, v. 70, p.
960967.
Stanley, B.D., Hirschmann, M.M., and Withers, C., 2011, CO2 solubility in
Martian basalts and Martian atmospheric evolution: Geochimica et Cos-
mochimica Acta, v. 75, p. 59876003.
Tao, Y., Li, C., Hu, R., Ripley, E.M., Du, A., and Zhong, H., 2007, Petroge-
nesis of the Pt-Pd mineralized Jinbaoshan ultramafic intrusion in the Per-
mian Emeishan large igneous province: Contributions to Mineralogy and
Petrology, v. 153, p. 321337.
Thakurta, J., Ripley, E.M., and Li, C., 2008a, Geochemical constraints on the
origin of sulfide mineralization in the Duke Island Complex, southeastern
Alaska: Geochemistry Geophysics Geosystems, v. 9, 34 p., doi: 10.1029/
2008GC 001982.
2008b, Prerequisites for sulfide-poor and sulfide-rich Cu-Ni-PGE min-
eralization in Alaskan-type complexes: Journal of the Geological Society of
India, v. 72, p. 611622.
Tomkins, A.G., Rebryna, K.C., Weinberg, R.F., and Schaefer, B.F., 2012,
Magmatic sulfide formation by reduction of oxidized arc basalt: Journal of
Petrology, v. 53, p. 15371567, doi: 10.1093/petrology/egs025.
Wendlandt, R.F., 1982, Sulfide saturation of basalt and andesite melts at high
pressures and temperatures: American Mineralogist, v. 67, p. 877885.
Wenzel, T., Baumgartner, L.P., Brugmann, G.E., Konnikov, E.G., and Kislov,
E.V., 2002, Partial melting and assimilation of dolomitic xenoliths by mafic
magma: The Ioko-Dovyren intrusion (North Baikal region, Russia): Journal
of Petrology, v. 43, p. 20492074.
58