Sunteți pe pagina 1din 20

Marine and Petroleum Geology xxx (2017) xxx-xxx

Contents lists available at ScienceDirect

Marine and Petroleum Geology


journal homepage: www.elsevier.com

F
OO
Review article

Large salt accumulations as a consequence of hydrothermal processes associated with


‘Wilson cycles’: A review Part 1: Towards a new understanding
Martin Hovlanda⁠ ,⁠ ∗⁠ , Håkon Rueslåttenb⁠ , Hans Konrad Johnsenb⁠

PR
a
Tech Team Solutions, Stavanger, Norway
b
Independent Consultant, Trondheim, Norway

ARTICLE INFO ABSTRACT

The formation of large salt deposits is observed especially in areas with a geological history of high tectonic activ-

D
Keywords:
Hydrothermal salt ity. Over the last decade it has become a well-established fact that heavy brines form and solid salts precipitate,
Wilson cycles due to the thermodynamic and physico-chemical properties of seawater at high temperatures and pressures en-
Physico-chemical reactions countered within hydrothermal systems. This article reviews the modern theoretical and experimental research
Thermodynamics
behind these findings, and also describes geological settings that most likely cause brine- and salt-forming hy-
TE
Large salt accumulations
drothermal processes to occur. This analysis has led to the identification of a set of specific conditions, properties,
Subduction
Rifting and processes (referred to as Conceptual elements) that are used to explain the often complex processes of brine
Ocean spreading behavior that leads to hydrothermal formation of solid salt.
Hidden salts The main objective of this review is to present hydrothermal conditions known to occur during Wilson cycles:
subduction, collision, and rifting, e.g., zones of repeated tectonic unrest, where brines (commonly derived from
seawater) are concentrated into heavy brines and precipitate solid salts. The internal heat of the Earth and its
EC

interaction with deeply-circulating seawater in hydrothermal systems and also the immense recycling of crustal
materials, including porous oceanic crust and serpentinite (hydrated) rocks via mantle processes may lead to the
formation of salt accumulations. It is also acknowledged that such brines and solid salts may often be stored
sub-surface for long time periods, extending from one Wilson cycle to another. Thus, on the basis of this analysis,
it is cautiously suggested that large amounts of salts ‘hidden’ inside subduction zones may appear on the surface
during subsequent rifting and oceanization phases.
In Part 2 of this review, the Conceptual elements, which are described and discussed herein (Part 1) are applied
RR

to selected cases, including the Andean Mountains, the East African Rift, the Red Sea Rift, and other locations.

tonic processes, the only conceivable process for producing solid salts
1. Introduction was solar evaporation of seawater (e.g., Alling, 1928).
As years went by, all the work done on the solar evaporation of sea-
CO

Over the last two centuries large volumes of salts (mostly chlorides water and the formation of salts (evaporites) has formed a large and
and sulphates) have been discovered both on and below Earth's sur- seemingly consistent knowledge base for the explanation and explo-
face. (The term “salt” is here used to indicate a seawater-derived ionic ration of giant salt bodies (Hsü et al., 1973a,b; Warren, 1999, 2006;
compound in its solid state). The discovery of these large volumes, 2010, 2016). Thus, the solar evaporite theory has become so dominant
however, came much earlier than the realization of how the planet's in the last century that it is more-or-less considered ‘forbidden’ to tam-
crust was dynamically and thermally structured. Therefore, the theories per with it.
for salt accumulations were established prior to the understanding of Nevertheless, the large salt accumulations in deep marine basins
UN

plate tectonics. Being unaware of the forces and heat involved in tec are difficult to explain with a surface seawater evaporation model
(Scribano et al., 2017, and references therein; Christeleit et al., 2015;

∗ Corresponding author.
Email address: mthovland@gmail.com (M. Hovland)

https://doi.org/10.1016/j.marpetgeo.2017.12.029
Received 30 September 2017; Received in revised form 17 December 2017; Accepted 21 December 2017
Available online xxx
0264-8172/ © 2017.
M. Hovland et al. Marine and Petroleum Geology xxx (2017) xxx-xxx

Lugli et al., 2015). The internal heat of the Earth and its interaction
with deeply-circulating seawater in hydrothermal systems, and the im- 1.2. The origin of sea salts
mense recycling of crustal materials via mantle processes (Wilson cy-
cles with subduction, rifting, and collision; e.g., Liou et al., 2014) have, Salt accumulations formed by hydrothermal activity, may have two
seemingly, been neglected as a potential explanatory model for the sources of origin; (a) precipitated salt from seawater; and (b) salts de-
formation of large salt accumulations, either above or below ground rived directly from crust and mantle masses. Subgroup (b) may also in-

F
(Sozansky, 1973; Momenzadeh, 1990; Hovland et al., 2006a,b, c; 2014, clude previously formed and hidden salts, either in the form of solids, or
2015, 2016; Scribano et al., 2017; Hardie, 1990). brines. Solids and brines encountered by the hydrothermal systems may
It has long been recognized that salts (evaporites) play economically be refined relative to seawater composition.

OO
important roles in petroleum prospecting and extraction. Thus, accord- Serpentinite will contain both types of salts. The contribution from
ing to Warren (2006, 2010) and Mohriak et al. (2012), they play a ma- the two sources is difficult to distinguish from each other, but Braitsch
jor role in i) facilitating hydrocarbon migration; ii) providing tectonic (1971), who performed a thorough study of salt compositions in the gi-
and sealing control for hydrocarbon trap formation, and, thus, iii) play ant Zechstein salt accumulations in Europe, concludes that the salt ac-
a role in hydrocarbon reservoir distribution. However, from this current cumulations actually do not have the same composition as that of the
review, it becomes evident that the roles of salt formation and accumu- oceans. Thus, he found that the salt accumulations could not originate
lation may deserve more attention in global tectonic (crustal) research, directly from the simple evaporation of seawater.
as salts may produce areas of weakness in and around suture zones that A study aiming at determining the size of different sources of chlo-

PR
are later prone to rifting. This is evidenced by the presence of salts in rine (Graedel and Keene, 1996) concluded that 99.6% of the Earth's
early rifts, also before the sea has invaded the rift, e.g., the East African chlorine is to be found in the mantle, while 0.3% is found in the crust
Rift. and only a mere 0,1% resides in the oceans. However, a more modern
assessment, by Kendrick et al. (2017), suggests that the budget is much
1.1. Objectives and organization more complicated.
According to Wallace and Anderson (2000), the “Yearly amounts of
The objectives of this review article are to present hydrothermal con- reactive volatiles H2⁠ O and Cl that are subducted are in the range of what

D
ditions known to occur during Wilson cycles, where salts may concen- is returned surfaceward by arc magmatism. If there were no surfaceward
trate into heavy brines and solid salts, starting from seawater. We set return of subducted H2⁠ O and Cl, then the formation, alteration, and sub-
forth to identify various stages of the Wilson cycles where such condi- duction of oceanic crust would comprise a net drain on the water and
tions are likely to be present. Hydrodynamics, thermodynamics and cer- salt in the oceans. The geologic record and understanding of tectonism
TE
tain properties of brines, rocks and salts, lead to predictive processes reveal that continental crust and ocean water have existed for most of
and results. A number of ‘Conceptual elements’ are used to explain the of- Earth's history and subduction has not caused the oceans to dry up. Thus
ten complex processes of brine behavior and the formation of hydrother- water lost to altered oceanic crust is probably largely returned to the
mal solid salts (Table 1). oceans via dewatering of subducted crust and subduction-related mag-
These ‘Conceptual elements’ are applied in the following chapters, matism.” This clearly shows that tectonic cycling of elements is active
including Part 2 of the article, to explain the new model for hydrother- on Earth, and indicates how important it is to understand the full im-
EC

mal salt formation. In addition, a number of geological cases are dis- plications of Wilson cycles that cause seawater to interact and exchange
cussed in order to substantiate and visualize the model (mainly found in elements with hot, pristine mantle.
Part 2: ‘Application on selected cases’).
This review is divided into the current Part 1: 'Towards a new under-
standing'; and Part 2: 'Application on selected cases'. Part 1 thus, concen- 1.3. The current evaporite theory
trates on the theoretical basis for the application of the ‘Conceptual ele-
RR

ments’ in Table 1 on the selected geological examples, mainly found in The conventional (solar) ‘evaporite theory’ is, generally, resting on
Part 2. three main pillars: a) dry climate, b) precipitation in shallow water, and
c) a relatively continuous supply of seawater that renews the content of
Table 1 salt in the evaporation basin (without diluting the brine more than the
‘Conceptual elements’ applied to explain hydrothermal salt formation. The items refer to rate of evaporation). To explain the observed salt thicknesses of several
the thermodynamics of physical and chemical reactions, including items that characterize
kilometers often found at great water depths, the solar evaporation the-
observable phenomena.
CO

ory alone does not provide a sound explanation (Selley, 2005; Scribano
Conceptual elements et al., 2017). The discussion about the possibility of accumulating such
giant deposits of ‘evaporites’ with solar evaporation alone, really made
‘Thermo- and fluid-dynamics’ (When seawater encounters heat at high pressures in
the headlines subsequent to the DSDP (Deep Sea Drilling Project), Leg
the Earth's crust, phase separation may occur; e.g. the formation of heavy brines,
solid salt and low-saline vapor in the supercritical and boiling domains of seawater
13 drilling in the western part of the Mediterranean. How was it possi-
and derived brines). ble to explain the formation of thick ‘evaporite’ deposits at over 3 km
‘Solubility and precipitation of salts’ (Salts may precipitate over large temperature water depth?
UN

intervals due to their wide solubility range at elevated temperatures and pressures). The most cited papers from this scientific drilling were those by Ken-
‘Fresh water formation’ (Condensation of vapor in hydrothermal systems produces
neth Hsü et al. (1973a, 1977), and Ryan et al. (1973), about the history
fresh (‘distilled’) water that can re-dissolve solid salt along its ascending path).
‘Formation of leaky salt stocks (domes)’ (Solid salt is permeable at depths,
of the Mediterranean Salinity Crisis. They also provide a rare glimpse of
allowing brine to flow through. Solid salt may precipitate upon cooling during the the bitter controversy that arose during the interpretation and publica-
ascent, and thus, build the salt stock continuously higher). tion of the drilling results: “The significance of the DSDP discovery was,
‘Refining of salts’ (Marine salts have different solubilities and the more soluble salts however, shadowed by controversies. That the Mediterranean evapor-
achieve higher concentrations than others, before precipitating. Some salts have
ite is Messinian (Late Miocene) in age seemed to be just about the only
crystal water attached to their crystal structure (e.g. tachyhydrite). These salts may
melt in their own crystal water upon heating and leave the solid salt as flowing consensus after the Leg 13 drilling. Even the shipboard scientific staff
brines). could not reach an agreement on the genesis of this unusual formation
‘Salt Preservation’ (Solid salts may be preserved by dry climate, dense brines, or …” “The desiccated deep-basin model (e.g., Ch. 43, of the Leg 13 cruise
sheaths of anhydrite, gypsum, carbonates, or fine grained clastic or biogenic report) was authored by Hsü, Cita, and Ryan because they were the only
sediments).
‘Hidden salts’ (Salts and brines may form at certain thermodynamic conditions members of the shipboard staff totally convinced on its plausibility.”
albeit not always observable; e.g., in subduction zones, within- or beneath oceanic- (Hsü et al., 1977).
and continental crust and within deep seated slab-remnants. This salt may become
visible at later stages in Wilson cycles).
M. Hovland et al. Marine and Petroleum Geology xxx (2017) xxx-xxx

For salt thicknesses of some tens of metres, the ‘evaporite’ explana- When seawater encounters hot mantle, the water-rock interactions
tion may be realistic. But, according to Warren (2006): “… the scale of are causing dissolution and hydrolysis of minerals in the rock, resulting
lacustrine evaporite development, past and present, pales in comparison in uptake of a significant amount of water in the form of OH-groups in
to some ancient marine-fed saline systems. There are no modern coun- new formed minerals. Simultaneously, the brine becomes denser due to
terparts to such ancient marine-associated systems.” This is because, ac- its uptake of salts and other dissolved elements from the mantle rocks.
cording to Warren (2006, 2010), there are no places on Earth where These reactions are exothermic in nature, and extra heat is produced

F
seawater is evaporating at a high enough rate to satisfy the climatic during the reactions, especially if serpentinization is involved, which
and shallow-water requirements of the conventional ‘evaporite’ theory: is often the case. In addition, during serpentinization, hydrogen is also
“Our inability to compare all aspects of evaporite deposition, past and produced, which represents a very strong reduction agent in chemical

OO
reactions involving iron-bearing minerals.
present, reflects a present-day lack of both longterm shallow epeiric seas
These reactions form brines that are much denser than the ini-
and of large drawdown basins fed by seawater seeps.” (Warren, 2006).
tial brine (seawater), without consuming large amounts of heat from
These disagreements and difficulties to give proper explanations to
the surroundings. Also, in contrast to the static example above, the
the processes of evaporite formations show all too clearly that a revi-
hydrothermal processes are extremely dynamic. The initially formed
sion of the traditional ‘evaporite’ theory is required. For example, the
brines may indeed be subjected to heat from other parts of the system as
term lacustrine evaporites, used above, easily brings ambiguities into the
they move around, or by dynamic hot magma bodies (volcanism). This
discussion. This is because the evaporation of fresh water (from rivers,
is likely to happen in both rifting and subduction systems, and provides

PR
etc.) gives negligible amounts of salts upon evaporation, but hydrother-
the necessary energy to produce accumulations of salts.
mal meteoric water may give significant amounts of salts upon evapora-
If we accept the concept that salts and brines may survive over ge-
tion like we see in the East African Rift. In most cases it seems this term
ological time spans in the mantle/deep crust, they may also experience
is applied to evaporation of hydrothermal brines which reaches the sur-
energy supply from several stages of the Wilson cycle before appear-
face; for example, in Danakil (Afar), Ethiopia, where dense brines em-
ing as observable salt accumulations. The Earth has undergone several
anate onto the surface, dry out and become preserved in the dry and hot
periods during which supercontinents have assembled and rifted apart
climate (Talbot, 2007).
again. Therefore, the supply of energy and the availability of seawater

D
during billions of years in the Earth's history make deep salt and brine
2. Formation of hydrothermal salts
production unavoidable. In short, the salt accumulations we observe to-
day may be the result of processes that date much further back in time
2.1. Salts forming in hydrothermal systems
than the latest stage in the Wilson cycle of that area.
TE
Bearing in mind all the Wilson cycles encountered during the Earth's
Besides plate tectonics, two of the last century's perhaps most impor-
history the question whether the mantle has provided the necessary
tant discoveries within marine geology, were indeed related to salt and
amounts of energy to produce the observed salt is not an issue. It should
hydrothermal processes: 1) The discovery of hot brines in the Atlantis
also be noted that Taylor and McLennan (1985) estimated that in total
II Deep of the Red Sea (e.g., Charnock, 1964); and 2) The discovery at
2% of the continental crust consists of salts (in sedimentary and pale-
the East Pacific Rise of hot vents, so-called ‘black smokers’ (Lonsdale,
osedimentary units). This is an enormous amount, which is undergoing
EC

1977). Both discoveries are considered ‘game-changers’ in the quest for transitions during rifting, subduction, and continental collisions (e.g.,
understanding seawater circulation in the oceanic crust. It is now well during Wilson cycles).
understood that the driving force for this circulation is heat from the un-
derlying mantle; so-called ‘forced convection’, with auto-circulation of
seawater when the temperature gradient between a hot (>1100 °C) ris- 2.3. Serpentinization of mantle rocks
ing pluton (igneous intrusion) and the ocean floor (at typically 2–4 °C)
RR

is adequately steep. Hydrothermal circulation of seawater in the Earth's Over the last ~40 years, there has been a revolution in our under-
crust at temperatures above 400 °C and >3 km below sea surface, will standing of how the oceanic lithosphere has evolved through interac-
inevitably lead to the formation of concentrated brines and the precipi- tion with seawater. Along slow-spreading ridges there are areas where
tation of solid salt. It is suggested that hydrothermal processes may ex- seismic profiling shows more than 15 km thick ‘amagmatic’ zones repre-
plain both the location and the amount of salt deposits in a better way senting relatively cool lithosphere, where seawater has reacted with the
than solar evaporation. Even so, this does not prevent solar evaporation upper mantle to form serpentinite rocks (Pearce, 2002; Dick et al., 2003;
from being an active contributor to brine densification and precipitation Mével, 2003; Snow and Edmonds, 2007; Ildefonse et al., 2007; Miranda
CO

of solid salt in many cases (Schreiber et al., 2007). and Dilek, 2010 ). This zone overlies almost unaltered ultramafic mantle
rocks. On the seafloor above such areas the serpentinization processes
2.2. Energy considerations are recognized by the occurrences of hydrothermal venting, both hot
(∼360 °C) black smokers and low-temperature (40–90 °C) white smokers.
Attempts to explain observed evaporites based on mantle heat act- The black color of black smokers is due to the emittance of abundant
ing on seawater in a static setting, will not produce the accumulations sulfide particles, while the ‘white smokers’ are characterized by having
observed in nature. The concept of producing evaporites from seawa- high pH-values (9–10), abundant CO2⁠ and methane contents (e.g. Mével,
UN

ter based on the heat from stationary magma/mantle bodies is not a 2003), and are precipitating carbonate and magnesium hydroxide min-
straightforward evaporation process. This is demonstrated by a simple erals onto the seafloor. The water that comes up in the hot springs is
calculation: A 1 km cube of hot mantle contains approximately the heat commonly found to have significantly lower salinity than seawater.
necessary to evaporate a 1 km cube of seawater, which leads to a 1 km It is well known that the upper mantle consists of dry magne-
square of salt of ∼15 m thickness; i.e., far from being comparable to the sium-rich silicate rocks (‘peridotite’), containing abundant iron-bear-
vast amounts of salt observed in some geological settings. However, the ing olivine ((Mg,Fe)2⁠ SiO4⁠ ). These minerals always coexist with orthopy-
premises for this calculation are wrong. A more correct calculation must roxene (MgFeSiO3⁠ ) ± Ca-rich clinopyroxene ± Al-Cr spinel and ± Ni-Fe
take into account several other factors than the heat-transfer between sulfides. When seawater encounters peridotite through deep cracks in
two static objects; e.g., water-rock interactions. the hot oceanic crust, spontaneous water-rock reactions start. These re-
actions are referred to as serpentinization, because peridotite is trans

3
M. Hovland et al. Marine and Petroleum Geology xxx (2017) xxx-xxx

formed into serpentinite by pervasive hydrolysis reactions (Scribano et is 10.5 kg: “Since the serpentinization process requires pure H2⁠ O, the
al., 2017). involved seawater-derived serpentinization fluid will undergo total
Experimental reactions show that olivine is preferentially dissolved out-salting.” However, the difficulty is to document this, in nature.
at temperatures around 250–300 °C, while orthopyroxene is rapidly dis- A common feature of serpentinite rocks are their patterns of perva-
solved at temperatures above 400 °C (Martin and Fyfe, 1970). Sev- sive fracturing (Malvoisin et al., 2017). The swelling of the rock exerts
eral accessory minerals may also form in the serpentinization process, strong mechanical forces on the rocks surrounding the reaction zones,

F
such as brucite or talc ± magnetite ± FeNi sulfides ± FeNi alloys ± chlo- causing fracturing on several scales. Even fractures on the micrometre
rite ± tremolite, depending on the physico-chemical conditions of the scale will immediately be filled with brine due to the high capillary
system (Früh-Green et al., 2004). In particular, the formation of NiFe al- forces involved. These fine fractures provide large interaction surfaces

OO
loys (taenite and awaruite) demand strongly reduced conditions. These between the brine and the rock and chemical reactions will cause fast
conditions are best achieved where serpentinization is at a non-equi- consumption of the pore water. Thus, the salinity of the brine may reach
librium stage; i.e. at the reaction front where olivine is present. Once high saturations and eventually also precipitate solid salts (Sharp and
olivine is completely altered, the oxygen fugacity associated with the Barnes, 2004).
rock transformation will increase markedly . The mechanisms for the migration of ‘hidden salt’ from serpentiniza-
The formation of NiFe-alloys in addition to the formation of hydro- tion zones were discussed by Scribano et al. (2017). Two mechanisms
gen, demonstrates that the serpentinization process also includes reduc- were suggested; (a) the hydrothermal salt plume; and (b) the buoyant

PR
tion processes that are balanced out by the oxidation of iron (Fe2⁠ + →Fe diapir. The first model depends heavily on a continued flow of hot hy-
3
⁠ +
) to form magnetite and also serpentinite containing some ferric iron. drothermal water from the serpentinization zone that has sufficient en-
A general formula, including both olivine and pyroxene with their con- ergy to move such heavy brines and even slurries of solid salt parti-
tent of magnesium and iron, is suggested by Evans (2004). This equa- cles in heavy brines. Dewatering of serpentinite at higher temperatures
tion also illustrates the redox reactions in the serpentinization process. (>400° C) may constitute sufficient energy to move such salt plumes.
The valences of iron in the minerals are indicated by (II) for ferrous and The second model refers to the formation of salt stocks by hot brines
(III) for ferric (oxidized) iron in the equation: migrating in the middle of the salt body; provided that the salt stock
is situated within the ‘Holness zone’, where salt is permeable (see Ch.

D
1.2 Mg1⁠ .8Fe(⁠ II)0⁠ .2SiO4⁠ + 0.76 Mg0⁠ .9Fe(⁠ II)0⁠ .1SiO3⁠ + 2.088
2.4.7 in the current paper). This assumes subcritical conditions and flow
H2⁠ O → Mg2⁠ .85Fe(⁠ II)0⁠ .11Fe(⁠ III)0⁠ .08Si1⁠ .96O5⁠ (OH)4⁠ + 0.042 of hot brines that reach saturation upon cooling at the upper part of
Fe(⁠ II)O*Fe(⁠ III)2⁠ O3⁠ + 0.176 H2⁠ the salt body, where solid salts are precipitating according to their spe-
cific solubility at each particular temperature and pressure interval. This
TE
Olivine + Pyroxene + Water → Serpentine + Magnetite + Hydrogen model thus, includes a refining process of the salt, where e.g. halite pre-
(+Heat) cipitates long before calcium and magnesium chlorides, upon cooling.
Sharp and Barnes (2004) conclude that subduction of serpentinite
The reaction is strongly exothermic, causing the serpentinization
may be the most important contributor to bring Cl down to mantle
process to increase its longevity due to its own heat production.
depths, thereby being a main contributor to the global cycle of Cl. A
EC

high-salinity fluid plume is suggested to evolve in the subduction zone


2.3.1. Serpentinization and salt
during serpentinite dehydration of a subducting slab that will affect and
The average density of a fresh peridotite is approximately 3.3 g/cm3⁠ ,
intensify water-rock interactions and melting properties above the slab.
while serpentine minerals have much lower density (∼2.5 g/cm3⁠ ), due to
the uptake of OH-groups in the new-formed crystal lattices. This causes
up to 40% swelling of the rock at complete alteration. Consequently, 2.4. The hydrothermal salt model
any brine (seawater at sub-critical conditions) will increase its salinity in
RR

the reaction zone. However, some substitution of chloride for hydroxyl This chapter provides some more background information pertinent
is found in the new-formed minerals (e.g. iowaite, brucite, and serpen- to further explain the ‘Conceptual elements’ in Table 1.
tine) in addition to other sea-salt elements incorporated in the crystal
lattices of the serpentinite minerals (Sharp and Barnes, 2004; Manuella 2.4.1. Model description
et al., 2016; Vogel et al., 2014). Based on previous work in rifted salt basins, mainly the Red Sea
According to Sharp and Barnes (2004) a significant amount of Mg2⁠ + (Hovland et al., 2015), a total of five relevant steps or phases have been
and Na+ ⁠ water-soluble chlorides are hosted in fresh serpentinite. By
CO

identified, whereby salt is naturally processed in one way or another to


leaching of serpentinites with distilled water they extracted an amount eventually end up as salt accumulations. Different types of salts are in-
of salts corresponding to more than 0.5 wt% Cl. They suggest that cluded in various structures and accumulations, including 'diapirs' (salt
Cl-bearing minerals are incorporated in the serpentine both as inter- stocks) and layered deposits.
stitial water soluble solid salts (e.g. halite and bischofite) and also These steps are:
as insoluble salt components incorporated in the crystal structure of
fine-grained serpentinite (Sharp and Barnes, 2004). Because Cl consti- 1) Seawater encounters hot rocks at moderate depths in the fault zones.
tutes 55% by weight of sea salt, 0.5 wt% of Cl in the serpentinite indi-
UN

Salt precipitation takes place due to boiling. Upon cooling of this sys-
cates a total salt content of about 0.9 wt%. tem, salts are dissolved and most of them are returned to the sea.
Fluid flux rates for water that enters into serpentinization reaction 2) As fracture volume expands and hydrothermal brines move further
zones in the Earth's crust are calculated to be in the range of 60–600 mol down towards the hot igneous body, high-pressure ‘supercritical’
m−⁠ 2 a−⁠ 1 . This can be translated into water volumes of up to 10,000 m3 ⁠
phase separation occurs, resulting in salt precipitation and the for-
per km− ⁠ 2 a−
⁠ 1. If it is assumed that this is mainly seawater, a corre-
mation of low density (0.3 g/cc at the critical point) water vapor and
sponding amount of 300 tons per km− ⁠ 2 a−
⁠ 1 of sea salt is enriched in
heavy brines that migrate further down and concentrate beyond sat-
these reaction zones, either as heavy brines or as precipitated solid uration.
salt (mainly magnesium and sodium chlorides), as shown by Sharp and
Barnes (2004).
According to Scribano and Viccaro (2014), the theoretical amount
of salt left or “produced” by changing one m3⁠ peridotite to serpentinite,

4
M. Hovland et al. Marine and Petroleum Geology xxx (2017) xxx-xxx

3) The circulating fluids cool the system further, and the upwards mi-
grating low salinity vapor is condensing, and starts to dissolve pre- 2.4.2. Solubility and precipitation in P/T-space
viously deposited salts. The different solubilities of sea salts lead to The common sea salts are polar ionic compounds. When precipitated
a refining of the salt types, whereby the most soluble ones are pre- as solid salts the ions are mainly bound together by electrostatic forces.
dominantly dissolved and displaced first. When these brines migrate When these salts are immersed in water, which has a very large dielec-
upwards due to fluid and/or tectonic pressures, they are cooled and tric constant, the water molecules are weakening the electrostatic bond-

F
some salts are re-precipitated before the brine reaches the seafloor. ing. This is causing the ions, both cations and anions, to lose contact
4) Upon reaching the seafloor, the brines are cooled further, and will with the solid salt and brought into solution. For this reason, the dielec-
easily become supersaturated with respect to some of the salt types. tric constant of water has been regarded as a proxy for the solubility of

OO
The remaing brine creates ponds of heavy brines along the seafloor polar ionic salts, while non-polar compounds (like oil) are less soluble
when topography and sea currents allow. Salts will then precipitate in water.
from these heavy brines on the seafloor (ponding/mini-basin forma- However, while most of the common sea salts increase their solubil-
tion) and form stratified deposits. This process includes another re- ity at higher temperature (with the exception of calcium sulfate, anhy-
fining of the salts. drite), the dielectric constant of water decreases with increasing temper-
5) In the continued subsurface processes, brines will migrate through ature; e.g., from 88 units at 0 °C and down to 73 units at 100 °C (Pan et
salt deposits as they build up. This brine migration also includes al., 2013). Furthermore, the dielectric constant of water is reduced right
a) dissolution and re-precipitation, b) pressure build-up due to down to only 2 units, at the critical point (CP) (Hayashi and Hakuta,

PR
salt-clogging and release, and c) extrusion of salts/salt slurries and 2010) and, thus, water is losing most of its solubility for the common
brines (e.g., the formation of salt glaciers, salt walls, salt injectites, sea salts at the transition to the superctitical domain (Liebscher and
and ‘diapirs’). Heinrich, 2007).
Each mineral has its own thermodynamic behavior according to the
These steps are not to be understood as completely isolated or se- availability of water, and according to temperature, and pressure vari-
quential, as several of the steps may occur simultaneously, although one ations. This is normally described by a phase diagram for each type of
step may at a certain time be more dominant than the other. Further- salt. But, because the sea salts interact with each other in solution, the

D
more, it must be noted that the accumulation of salt in specific regions, phase diagrams will be dependent on the total ionic composition, mak-
is fully dependent on the local conditions of ‘salt preservation’, as we ing the phase diagrams even more complex than shown in Fig. 1, which
discuss in Ch 2.4.12. It should therefore not come as a surprise that al- is only valid for one type of salt, NaCl (halite).
though large quantities of salt are produced in a particular subsurface or The reverse process, - precipitation of salts, is a consequence of
TE
near-surface area, the salt may not accumulate due to poor local preser- their solubility (Table 2). Upwards directed brine flow inside salt struc-
vation conditions. tures effectively maintains the salt concentration at the saturation level
as it flows and cools at the same time. This leads to precipitation of
solid salts all along the flow path; e.g., saturated NaCl brine at ∼100 °⁠ C
and ambient pressure dissolves up to 390 g of NaCl per liter of water.
EC
RR
CO
UN

Fig. 1. Phase diagram for the system H2⁠ O-NaCl showing the distribution of the three components liquid solution (L), vapor (V) and solid halite (H) in the P/T-domain. It illustrates where
the various components (phases: L, V and H) occur at different pressures and temperatures typically encountered in hydrothermal systems associated both with rifting and subduction.
(Based on Driesner and Heinrich, 2007). The blue-colored points, A, B, and C, refer to Example 1 in the text. (For interpretation of the references to color in this figure legend, the reader
is referred to the Web version of this article.)

5
M. Hovland et al. Marine and Petroleum Geology xxx (2017) xxx-xxx

Table 2 anhydrite and finely disseminated halite developed around the heating
Solubility at 20 and 100 °⁠ C for some pertinent salt minerals (CRC Handbook of Chemistry
elements. The ‘salt stock’ made its own accommodation space by push-
and Physics: Lide, 1990).
ing away the gravel (Hovland et al., 2006a). The experiment demon-
Solubility at 20 °C Solubility at 100 °C strates that anhydrite and halite precipitates even in wet environments,
Salts (g/100 g) (g/100 g) where conditions are favorable, e.g., close to strong natural heat-sources
located where seawater can freely circulate (‘wet evaporation’).
NaCl (halite) 35.7 39.1

F
Na2⁠ SO4⁠ *10H2⁠ O 11.0 92.7
(mirabilite) 2.4.4. The discovery of high-pressure phase separation
KCl (sylvite) 34.7 56.7 When seawater is heated and pressurized to near the CP a ‘two-phase

OO
KMgCl3⁠ *6H2⁠ O 64.5 decomposes mode’ occurs, i.e., most of the volume becomes a vapor (about 95 vol%)
(carnallite)
K2⁠ SO4⁠ *MgSO4⁠ *6H2⁠ O 19.2 59.8
with a density of 0.3 g/cm3⁠ , and the remaining liquid becomes a dense
(schoenite) brine as shown in the phase diagrams (Figs. 1 and 2) to be discussed in
MgSO4⁠ *7H2⁠ O (epsomite) 71 91 the next chapter.
2K2⁠ CO3⁠ *3H2⁠ O 129.4 268.3 The first to study phase separation of seawater at high pressures
MgCl2⁠ *6H2⁠ O (bischofite) 167 367
CaCl2⁠ *6H2⁠ O 279 536
and temperatures, e.g., at supercritical conditions (T = ∼407 °C,
CaSO4⁠ (anhydrite) 0.21 0.16 p = ∼300 bar), were Bischoff and Rosenbauer (1989). However, because

PR
they used pressure cells without visual access, they were unable to vi-
sually observe the phase transitions. Such experiments were later car-
Upon cooling the brine to 20 °⁠ C, solid salt will precipitate, as the solubil-
ried out by Tester et al. in the early 1990's: “When a temperature of
ity is reduced to 357 g per liter (Table 2). Thus, salt structures, like salt
around 405 °C was attained, a ‘cloud’ formed in the chamber and we
stocks (domes) or salt walls, may build up.
witnessed a sudden transition from one-phase flow to a two-phase flow,
Gypsum and anhydrite have retrograde solubilities, i.e., their solubil-
as solid salt ‘shock-crystallized’ when the solubility of NaCl dropped to
ities decrease with increasing temperature (Table 2). Furthermore, an-
near-zero over a temperature range of only a few degrees C at the crit-
hydrite has a lower solubility than gypsum, especially at higher temper-

D
ical point of brine.” (J. Tester, pers. com., 2002). The resulting NaCl
atures, and will precipitate from seawater at temperatures above 130 °C.
and Na2⁠ SO4⁠ solids inside the reaction chamber were found to consist
It is also pertinent to note that gypsum, which is essentially the same
of 10–100 μm ‘highly amorphous kernel-shaped particles’ (Tester et al.,
mineral as anhydrite, but with some additional crystal water, is rarely
1993).
observed in salt accumulations. Gypsum is the low-temperature mineral
TE
The reason why salts are practically insoluble in supercritical va-
of the two, and should be abundant in salts formed at low tempera-
por is the extreme lowering of the dielectric constant, from ∼80 to ∼2
ture, provided it does not lose its crystal water rapidly after deposition
(Braitsch, 1971). units. This causes a reduction of the ionic dissociation constant for the
Halite is actually the least soluble chloride and does not have the dissolved salts to such a degree that the salts precipitate. Other impor
ability to bind crystal water. Chlorides of magnesium, calcium and
EC

potassium, as well as more exotic salts, such as tachyhydrite, have the


ability to bind crystal water and, therefore, tend to stay in solution long
after halite saturation is reached. Thus, in hot water-deprived systems,
brines of magnesium and calcium may escape the system as very hot
liquids. Upon cooling, the water turns into crystal water and the salts
become solid. This is a viable explanation for the formation of tachyhy-
drite, a salt so hygroscopic that it does not survive as a solid when ex-
RR

posed to normal air humidity. Thus, the source for tachyhydrite would
most probably be from a mixed salt accumulation that is occasionally
exposed to more heat, which enables intergranular magnesium and cal-
cium chlorides in halite to liquefy in their crystal water (e.g., at temper-
atures above 120 °⁠ C) and escape out of the salt masses in hot liquid form.
To be preserved and solidify upon cooling, these brines must be some-
CO

how protected by equally saturated brines, a situation to be discussed


later.

2.4.3. Salt-formation by boiling – ‘wet evaporation’


A prerequisite for the formation of solid salt and observable salt
accumulations associated with hydrothermal systems is the supply of
seawater. At ocean depths of less than 2800 m (i.e., pressures below
UN

300 bars), seawater will not reach its CP. However, at sufficiently high
temperatures, seawater will boil, even when confined in sediments or
in fractures of the crust. Although it is well known that the boiling of
seawater at surface may produce large quantities of salts (e.g., Talbot,
2007), it is not readily clear what happens when seawater boils under
Fig. 2. Images showing the salt formation, from Simulation case #6 of Coumou et al.
water.
(2009). The figure shows the results from a simulation where sea water meets hot mantle,
This was, however, demonstrated by a boiling experiment in a after a period of 4000 years. Color coding reflects salinity. The grey areas indicate where
7 m × 7 m and 3 m deep pool filled with seawater, where the boiling sys- solid salts are formed and fresh vapor leaves the system. Depth is indicated in metres be-
tem included an open 200 L steel container with a 5 kW heating ele- low sea level. a) Shows the temperature distribution. b) Shows the established water flow
lines. (For interpretation of the references to color in this figure legend, the reader is re-
ment buried in gravel. The boiling was controlled by three thermome-
ferred to the Web version of this article.)
ters and continued for 11 days. An artificial ‘salt stock’ consisting of

6
M. Hovland et al. Marine and Petroleum Geology xxx (2017) xxx-xxx

tant properties of the supercritical vapor are its low viscosity and high sary to use phase diagrams (Fig. 1). With such diagrams it is possible
diffusivity (Tester et al., 1993; Bellissent-Funel, 2001), which facilitates to state where and when (in the T/P space) salt will become saturated
fast diffusion into porous rocks (‘stockwork’ veins/fissures) and sedi- and precipitate (to solid halite) and when vapor (low salinity to fresh
ments. ‘steam’) will form.
Butterfield (2000) shows a scanning electron microscopy (SEM) im- The phase diagram in Fig. 1 shows the regions in P/T-space where
age of halite-coated basalt. His interpretation of this image is that a the various phases occur, including the concentration of salt (% NaCl).

F
sub-seafloor volcanic eruption (in deep water) “… caused immediate In the figure, occurrence of the liquid phase is marked with (L), the va-
phase separation and halite precipitation, followed by a high-temper- por phase (V), and the solid state, e.g., halite with (H).
ature reaction period.” Furthermore, he concludes that: “A brine pro- As simulated numerically by Coumou et al. (2009) (Fig. 2), saline

OO
duced by condensation from seawater-derived fluid at 450 °C would water may experience a series of phase separation events during its
have a salinity five times greater than that of seawater. Brines produced movement downwards through the fractured rock. Consequently, the
by exsolution of an aqueous phase from a magma at temperatures in ex- brine is subject to phase separation, with the partitioning of low-salin-
cess of 700 °C have salinities of more than 10 times that of seawater.” ity vapors and dense (high salinity) brines. Physical movement of the
(Butterfield, 2000). brine into domains of slightly higher pressures and temperatures leads
This realization, lead Lecumberri-Sanchez et al. (2015) to conclude to the next stage, resulting in increased brine density. Retaining the
their world-wide study of rising upper crustal plutons with the state- brine in the fracture system depends on overcoming the lifting capacity
of the escaping vapors. Hence, pressure (hydraulic head) plays a role in

PR
ment: “Salt precipitation appears to be a ubiquitous feature in hydrous
regulating vapor volume and, therefore, hydrodynamic lifting capacity.
volcanic and geothermal environments, invoking dynamic behavior due
Coumou et al. (2009) state that both simulations and evidence from na-
to rapid multiphase reactions involving liquids, vapor, and solids.”
ture suggests that brines with higher salinities than ∼12% will tend to
The implications of these observations are that the interaction be-
sink in and not be lifted out of the system, provided that phase sepa-
tween seawater and hot igneous bodies at sufficiently large water depth
ration takes place deep enough, e.g. below 3,5 km in their simulations.
(>300 bar pressure) unequivocally produces concentrated brines and
The salts and brines formed at such conditions may, however, be ‘mined
local precipitation of halite and other salts in pores and fractures
(Hovland et al., 2006a, 2016, b, c; Gruen et al., 2014; Scott et al., 2017; out’ by later influx of colder waters, or by mechanical (tectonic) forces

D
Scribano et al., 2017). It is this fundamental revelation that forms the acting on the rocks. Such waters might be the result of condensing va-
basis for the current hydrothermal salt theory, where the fact that sea- pors from deeper phase separation processes.
water normally phase separates somewhere within every conceivable To explain the phase behavior (Fig. 1), three simplified examples are
included below. Only Example 1 with lines in blue, and points A, B, and
TE
type of hydrothermal circulation cell, due to high temperatures, often
combined with high pressures, plays the central role. C,- is illustrated in Fig. 1.

2.4.6. Example 1
2.4.5. Application of thermodynamics In Fig. 1, a brine of 10% NaCl concentration encounters a pressure of
Hydrothermal systems occur in highly fractured, permeable and 350 bars and a temperature of 450 °C (point A). Here, the brine will sep-
EC

porous rocks that typically occur above strong heat-sources, such as arate into a vapor containing traces of salt (B) and a denser brine having
hot oceanic crust, rising igneous bodies (‘magma chambers’ or plutons). 40% of salt (C). In a rift system, given the right hydro-dynamical condi-
Water trapped in the fissures, fractures, and cracks will be forced by the tions, this brine may now migrate downwards (due to its higher density)
steep temperature gradient to convect, often vigorously. Such locations and encounter even more extreme conditions leading to another stage in
on the deep ocean floor are also described as: ‘zones of high heat-flow’. the phase separation. This may eventually lead to solid salt precipitation
If brines are present in hot fractured fault zones, thermo- and and accumulation. In real systems, the fluid flow is tortuous and convo-
fluid-dynamics will control the brine behavior. As it most often leads to luted, depending on local conditions within the rock fracture network.
RR

dynamic fluid flow (forced convection), interaction with side wall rocks, The thermodynamically (PT)-determined alterations and phase separa-
and precipitation of salts, will occur, as documented by numerical mod- tion stages in real systems are therefore gradual and infinitive.
elling (Hovland et al., 2006a; Driesner and Heinrich, 2007; Coumou et If conditions take the fluids outside, or above the critical conditions
al., 2009; Gruen et al., 2014; Lecumberri-Sanchez et al., 2015; Scott et for any salinity, represented by the “critical curve” in Fig. 1, only fresh
al., 2017). supercritical vapor and solid salt may exist.
Scott et al. (2017) studied the exploitation of geothermal energy on
CO

Iceland, and state: “Numerical simulation of subaerial, magma-driven, 2.4.7. Example 2


saline hydrothermal systems reveals that fluid phase separation near If subsurface brines are exposed to a pressure drop due to, for ex-
the intrusion is a first order control on the dynamics and efficiency of ample rifting, the conditions may move below the V + L + H surface of
heat and mass transfer. Above shallow intrusions emplaced at <2.5 km the phase diagram. This will lead to boiling and instantaneous solid salt
depth, phase separation through boiling of saline liquid leads to ac- precipitation.
cumulation of low-mobility hypersaline brines and halite precipitation,
thereby reducing the efficiency of heat and mass transfer. Above deeper 2.4.8. Example 3
UN

intrusions (>4 km), where fluid pressure is > 30 Mpa, phase separation In a subduction setting, the slab with its contained seawater is ex-
occurs by condensation of hypersaline brine from a saline intermedi- posed to gradually higher pressures while maintaining a relatively low
ate-density fluid. The fraction of brine remains small, and advective, va- (inside) temperature compared to its surroundings (thermal inertia).
por-dominated mass and heat fluxes are maximized. We thus hypothe- However, - since the fluids will experience compaction of the slab, they
size that, in contrast to pure water systems, for which shallow intrusions may also be squeezed out and meet the conditions in the warmer hang-
make better targets for supercritical resource exploitation, the optimal ing wall above the slab. This may move brines with certain salinities
targets in saline systems are located above deeper intrusions.” This is into warmer regions where they separate into vapor and even denser
fully in line with the hydrothermal salt theory applied here. brines.
To understand the transition and transformation of NaCl-solutions Coumou et al. (2009) provide ample evidence for deep salt forma-
in high-temperature, high-pressure hydrothermal systems, it is neces tion in the Earth's crust/mantle during stages of the Wilson cycles where
saline waters meets a heat source. See Fig. 2 below.

7
M. Hovland et al. Marine and Petroleum Geology xxx (2017) xxx-xxx

2.4.9. Fresh water formation (vapor condensation)


It should come as no surprise that when brines or seawater are in-
creasingly concentrated in salinity, some fresh water must leave the sys-
tem and may interact with adjacent salts or rocks on its way. Even if the
water leaves the hydrothermal system as vapor with little or no salt-car-

F
rying capacity, it eventually cools and condenses to pure (distilled) wa-
ter and may start dissolving (‘picking up’) salts again. Wherever de-
posited salts are exposed to this flow of fresh water, they may be re-dis-

OO
solved along the flow path of this ‘reflux’ water. This may lead to the
formation of brines that move on and even deposit salts further up in
the stratigraphy. In the hydrothermal salt model, the flow of fresh wa-
ter is not located arbitrarily in relation to the circulation cell, and very
often, it is located immediately above the centre of the hottest portions.
Numerical simulations also confirm this (e.g., Gruen et al., 2014).
Depending on the relative amount of water and salt, lean or rich

PR
brines will be the result. At the seabed, salt ‘glaciers’, as seen in the
Red Sea, may seemingly end their movement towards deeper parts of
the rift basin. This is also where this fresh water rises upwards and
emanates. During oceanic rifting and subduction, the only water in-
jected into the hydrothermal system, is seawater. This implies that any
fresh-water flows observed, must be the result of loss of salt from the
seawater. Even if the water has been involved in hydroxylation of min-
erals, it still came from seawater that lost its salt somewhere.

D
There are, however, exemptions to this ‘rule’ in shallow, meteoric Fig. 3. Diagram modified from Lewis and Holness (1996) showing the equilibrium halite/
systems involving ground water and in early stage continental rifting water dihedral angles (θ) as a function of P and T and the corresponding distribution of
pore water. In the white area (upper right), θ is less than 60°, resulting in the formation
where water input is often meteoric. The output, however, is rarely
of continuous liquid films in the halite-water system. The light grey area (middle), defines
fresh, as it usually leaches and transports different solutes upon return-
TE
the P-T region in which the permeability of halite may be significant. The dark grey area
ing to the surface. Rift lakes of East Africa provide evidence for this. (lower), θ >60°, represents the region with (‘normal’) impermeable halite. The dashed
The returning meteoric water has picked up salts from deeper down, line indicates a geothermal P-T gradient. Drawings of halite crystals illustrating continu-
and, therefore, several saline (alkaline) lakes exist along the rift. Each ous liquid films (upper) and disconnected liquid (lower) are also shown.
lake tells a story of what substances the water encountered in the deep
(Warren, 2006). this mechanism allows refining of salt, by the fact that the various salt
types are precipitating according to their solubility at each specific tem-
EC

perature interval.
2.4.10. Conditioned permeability of halite
Another effect of the water saturated inter-connected pore system of
When hot saturated brines migrate upwards from the deep, they will
salt is the reduction of the effective stresses in the salt body, i.e., the to-
be cooled and salts may precipitate from the solution. This crystalliza-
tal stress on the salt crystals from the overburden pressure is now partly
tion of salt will exert a pressure on the overburden and is suspected
supported by the pore liquids.
to be an important displacement mechanism for deep salt and for the
This may, in fact, well cause a dramatic change in seismic imped-
RR

formation of salt domes. Salt formations are traditionally treated as be-


ance. Thus, when the seismic waves pass from the upper ‘brittle salt’
ing nearly impermeable for brines and hydrocarbons. However, accord-
ing to experimental work first carried out by Lewis and Holness (1996), with higher sonic velocities to a more ductile and permeable salt deeper
halite buried at depths greater than ∼3 km and elevated temperatures down with lower sonic velocities, the impedance shift will be clearly
seen on the seismic records. The so-called decollement seen on many in-
becomes permeable due to alteration of the dihedral angles of the halite
terpreted seismic images of deep-lying salt deposits (‘evaporites’) may
crystals.
When water is present in the salt, which is always the case in nat- actually represent the ‘Holness Zone’, which causes a seismic response
CO

ural salt deposits, the dihedral angle of the salt crystals is determinant to the distribution of pore water.
for the connectivity of this water (Holness and Lewis, 1997). When the
dihedral angle is less than 60° the water is spreading out on the crys- 2.4.11. Refining of salts
tal surfaces as continuous liquid films and the salt becomes permeable The various solubilities of sea salts, which also change significantly
(Fig. 3). Vice-versa, when the dihedral angle is larger than 60° the wa- at higher temperatures and pressures, may result in sorting of the salt
types in the hydrothermal circulation of sea water. A water-deprived
ter film becomes disconnected and the salt is impermeable: “Our results
salt accumulation in a hydrothermal system will, therefore, lose the
explain recent observations of major salt-fluid interactions at depth,
UN

more soluble salts, e.g. magnesium and calcium chlorides, in contact


and suggest that deep-rooted salt diapirs may act as conduits for basi-
with any available water. Thus, refined halite may precipitate and ac-
nal formation water” (Holness and Lewis, 1997). We are referring to
cumulate, e.g. i) inside salt stocks (domes/diapirs); ii) by venting into
this PT-dependent permeable zone as the Holness Zone (Hovland et al.,
ponds of concentrated calcium and magnesium chloride brines on the
2015). In the hydrothermal model for salt formation, one of the con-
sea floor; or iii) by venting up onto the surface where it evaporates in
sequences of the Holness Zone is that any upward-migrating brine in
a hot and dry climate, as reported from Dallol (Talbot, 2007). In this
this zone will transfer the pressure in the liquid further upwards and
way salt accumulations found as salt stocks or other salt formations have
may cause hydraulic fracturing of the above lying impermeable rock
compositions significantly different from sea salt.
salt, as also pointed out by Schoenherr et al. (2007a, b), Ghanbarzadeh
Another plausible scenario is if an accumulation of different salts
et al. (2015), Warren (2016, 2017). Hot brines at depth can flow
is reheated in the subsurface, e.g. by volcanism or sill intrusions, then
through ‘solid halite’ and displace salt from the high temperature re-
the salts containing crystal water have a tendency to melt, forming hot,
gion to a cooler region where dissolved salts precipitate. Furthermore,

8
M. Hovland et al. Marine and Petroleum Geology xxx (2017) xxx-xxx

dense brines that may migrate out, driven by tectonic forces or lifted by covery (Corselli et al., 1996; Winckler et al., 2001; Wallmann et al.,
volcanic gases. An example of such a hydrated salt is bischofite, which 2002).
melts at temperatures above 120 °⁠ C (Schofield et al., 2014). Upon cool- These ponds are located on top of an oceanic subduction zone (the
ing, these brines will again crystallize to form solid salts. Such salts will Mediterranean Ridge), and the brines are venting out of these zones
accumulate stratigraphically above the halite and anhydrite where they (Figs. 4 and 5). It is most likely that these brines were originally part
initially were deposited. It should be noted, however, that if volcanic of the salty pore water going down with the oceanic slab and therefore

F
gases like CO2⁠ or SO2⁠ are exposed to the brine, new salt types may pre- likely the result of refining processes in the subduction zone, leaving be-
cipitate; e.g. carbonates and sulfates. hind major parts of the sea salt, mainly halite, below ground.
2.4.11.1. Refining examples It is very difficult to directly observe refin-

OO
ing processes in the subsurface, but the results of refining can be ob- 2.4.12. Preservation of salts
served some places on the surface, for instance in the Mediterranean The previous explanation given for the formation of high-magnesium
(Cita et al., 1988; Woodside et al., 2001). The Mediterranean Sea is salts is an example of solubility-based refining of salts, but also pro-
unique in that it overlies a zone of mantle convection and has under- vides an example of a mechanism for the preservation of salt. Obser-
gone a long period of tectonic deformation which includes slab roll- vation of seafloor hypersaline ponds (or ‘lakes’) in depressions on the
back and the collision of two large, slowly moving plates (Faccenna et Mediterranean seafloor, and elsewhere, shows that lack of mixing en-
al., 2014). Several brine ponds are located on the seabed, which con- ergy in combination with gravitational stability, prevents mixing of the
tain concentrated brines of high-solubility salts (magnesium and cal- two. Over time, supersaturation as a result of cooling may occur in the

PR
cium chlorides), e.g. the brine ponds Bannock, Urania, Nadir, and Dis brine lakes, leading to precipitation of solid salts that are protected by
already saturated brines above.

D
TE
EC
RR

Fig. 4. Image modified from Huguen et al. (2006). It clearly shows the topography of the Mediterranean Ridge accretionary wedge and its location in relation to the deep hypersaline
brine ponds: Bannock (B), Atlantis (A), Urania (U), and Discovery (D), which reach nearly 4000 m water depth. The brines are magnesium rich, and halite depleted, signifying refined
salts, due to hydrothermal processes, see text.
CO
UN

Fig. 5. Geological profile across the Mediterranean Rige, modified from Westbrook and Reston (2002). The red arrow indicates rising warm hydrothermal hypersaline brines feeding into
the anomalous seafloor brine ponds. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

9
M. Hovland et al. Marine and Petroleum Geology xxx (2017) xxx-xxx

In more demanding environments, dry climate may protect the salts


from dissolving, even if the salts were not the result of this same climate
(but of hydrothermal activity). In the ocean, salts may also be protected
by a cover of anhydrite between seawater and other salts. This anhy-
drite might be the result of brine outflow carrying particulate anhydrite
with it, or anhydrite remaining from dissolution of anhydrite stained

F
salts on top of salt structures. Normal hemipelagic sediments may also
cover a salt structure and thereby protect it from the actions of the sea-
water. Furthermore, some salt structures are located next to leakage of

OO
hydrocarbons into the sea. The biologic community may then interact
and provide the CO2⁠ necessary to produce a protective cover of authi-
genic carbonate (Judd and Hovland, 2007).

2.4.13. Formation of leaky salt stocks (domes)


As explained in Chapter 2.4.6, saline brines will occupy the contin-
uous pore space of the permeable salts within the ‘Holness Zone’. In
this zone the salt will be more plastic than the salt above (due to less

PR
grain-to-grain contacts) and will be pressurized by the lithostatic con-
fining pressure from the surrounding rocks. This pressure will also be
transferred to the pore water, leading to higher pore pressure in the salt
body, causing also the ‘Holness zone’ to move upwards. This increased
Fig. 6. The southern Dead Sea located 400 m below sea level. SD = the Sedom diapir;
internal pressure is causing expansion of the salt body and may also LD = the Lisan diapir. The fault-lines are: JF = Jerico Fault; SF = Sedom Fault; BF = Boqeq
lead to salt extrusion/intrusion, as often observed in nature (Schoenherr fault, and AF = Arava Fault. Based on Gradmann et al. (2005). (Modified from GeoMa-
et al., 2007a; Hovland et al., 2015; Warren, 2017). Eventually, this in- pApp, e.g., www.Geomapapp.org; Ryan et al., 2009).

D
creased pore pressure may reach the solid salt above the ‘Holness zone’
with sufficient energy to cause hydraulic fracturing and allow periodic low velocity zone extending to a depth of 13–18 km under the basin,
leakage of brine, as stated previously (Table 1) and discussed in more
while the lower crust and Moho was not perturbed. These observa-
detail in Part 2.
TE
tions are supposed to be due to strain softening in the middle crust, in-
If the hydraulic fracturing happens at great depth, the fractures tend
voked to explain the isostatic compensation and the rapid subsidence of
to be vertical, while fracturing at shallow depths is likely to cause
the basin during the Pleistocene (Al-Zoubi and ten Brink, 2002). They
semi-lateral fractures, creating ‘mushroom like’ salt bodies towards the
also produced a conceptual model of the crustal deformation during the
surface (or sea floor). opening of the basin, illustrating the subsidence of the crustal blocks
Hot brines coming from the base of the salt stock (dome) will cool along listric faults and the sedimentary filling (Fig. 7).
EC

and reach over-saturation during the ascent and salts will precipitate ac- The Lisan salt diapir (in the east) and the Sedom diapir (in the west)
cording to their specific solubilities at the different temperature/pres- are the largest salt structures in the DSB (Figs. 6 and 8), up to 7 km deep,
sure conditions. However, this precipitation of salt will not necessarily
and occur as buried structures in the sediments. According to Al-Zoubi
block the flow completely; instead the entire salt body is likely to ex-
and ten Brink (2001) the formation of these structures may have been
pand, and new flow paths are formed. Some salts with higher solubil-
as short as 0.2 million years. They also report the composition of the Se-
ities, e.g. calcium chloride, may dissolve in the migrating brine, to be
dom diapir to be ∼80% halite, the remaining being anhydrite/gypsum,
transported further up; i.e. a refining of the salt body is taking place.
RR

marls, chalk, dolomite and shale.


The entire process works in concert with the driving pressure that feeds
2.4.14.1. The Dead Sea: accumulating salts in a non-marine
brine from below, into the salt accumulation (e.g., Holness and Lewis,
environment As pointed out by Hardie (1990) the bulk of the solutes in
1997).
the lake brines are added by the upwelling of saline springs around and
up through the lake bottom: “Thus, the Dead Sea stands as a clear-cut
2.4.14. The Dead Sea diapirs
case of a non-marine evaporite depositional site where the evaporating
The Dead Sea is a well-known site for ongoing precipitation and ac-
brines are CaCl2⁠ -rich and SO4⁠ -poor because the inflow-waters are
CO

cumulation of salts in a non-marine environment. This saline lake is sit-


CaCl2⁠ -rich and SO4⁠ -poor” (Hardie, 1990). These features are illustrated
uated in the large fault system referred to as the ‘Dead Sea – Jordan
in Fig. 8, with the sedimentary stratigraphy and the two main salt
Transform rift’; a pull-apart basin, where two overlapping faults (or a
domes, Sedom and Lisan. The structural features are also confirmed by
fault bend) creates an area of crustal extension, causing large scale sub-
Weinberger et al. (2006).
sidence of the basin, thus, providing accommodation space for the depo-
Talbot et al. (1996) made a study of the more than 600 m thick salt
sition of thick sequences of sediments and salts. The current elevation of
beds in the basin and focussed on the large, bulbous halite and carnallite
the lake surface is ∼400 m below mean sea level. A map of the southern
bodies on the bottom. These types of anomalous salt bodies had never
UN

part of the Dead Sea shows the location of the main tectonic and salt
been described before and they suggested a formation caused by the
features (Fig. 6).
slow seepage of groundwater brines through cracks, inferring conduits
Rabinowitz and Mart (1999) were able to reveal the occurrence of
through the Dead Sea lake bottom. They surmised that the band of dens-
shallow mantle masses under the southern Dead Sea using seismic to-
est bulbous salt growths (‘reefs’) occurred in the area between the Lisan
mography, and discloses that: “… the Moho under the Dead Sea is found
and Sedom diapirs (Fig. 8) and were most probably fed by “groundwa-
at a depth of (only) 22 km, and the seismic velocity in the upper lithos-
ter rising from sources deeper than 50 m, …” “… along the Boqeq fault,
pheric mantle is anomalously low”. The data actually suggest that hot
one of the en-echelon transverse fault zones that account for the rhom-
magmatic diapirs ascend along the boundary faults of the rift into the
bic shape of the southern basin”. (Talbot et al., 1996).
intermediate crust, giving a heat-flow in the basin of 40–50 mW/m2⁠ , ac-
This suggestion was further strengthened by the finding of arte-
cording to ten Brink et al. (2006).
sian pressurized groundwater brines that caused precipitation of white
The Dead Sea Basin (DSB) structures were further investigated by
halite by natural springs along the “… brown floor of the Lisan Straits.”
Al-Zoubi and ten Brink (2002), and their seismic surveys also showed a

10
M. Hovland et al. Marine and Petroleum Geology xxx (2017) xxx-xxx

F
OO
PR
Fig. 7. Conceptual model of the crustal deformation associated with the opening of the Dead Sea Basin (DSB), indicating a shear zone with fluid flow and earthquake epicentres at the

D
base of the upper crust. (From Al-Zoubi and ten Brink, 2002).

with fluid flow at the base of the upper crust, at 13–18 km depth. It is
suggested that the hydrothermal brines are coming from this zone, as it
TE
may explain the low temperature and high salinity of the venting brines
(29 °⁠ C), which is an effect of cooling along long migration conduits dur-
ing the ascent. The temperature in this deep shear zone is not known,
but global average vertical temperature gradients are 20–30 °⁠ C per km
(Levitte and Greitzer, 2000), indicating temperatures above 300 °⁠ C at
15 km depth.
EC

This concept of a deep hydrothermal source is supported by the find-


ings of Eckstein and Simmonsi (1977): “Deep faults associated with the
Jordan-Dead Sea Rift system act as conduits for hot waters ascending
from deep confined aquifers …” “Most of the ascending thermal waters
are absorbed by shallow aquifers with lower hydraulic potential. Such
regions are characterized by anomalously high heat-flow; several values
RR

exceed 2 and one value is 11 μcal/cm2⁠ s”.


However, the extensive hydrothermal system which is filling the
Fig. 8. The southern Dead Sea Basin: Illustration of the stratigraphy and the salt domes
Lisan and Sedom. See text for further information (Based on Gradmann et al., 2005).
basin with salts must also be sourced from large accumulations of ‘hid-
den salt’ deep in the subsurface; probably below the upper crust. Such
(Talbot et al., 1996). Here, groundwater with a constant temperature of occurrences of ‘hidden salt’ below the upper crust, interacting with an
29 °C rises to about the level of the Jordanian salinity ponds on the east- extensive hydrothermal system must have affected the entire system sig-
CO

ern flank of the straits. nificantly, both with respect to the tectonic processes and also the seis-
The most spectacular and obvious seepage-related salt structures are mic properties; i.e. lower seismic velocities. This is actually indicated
the polygonal networks of salt walls on the seafloor they document: on the model of crustal deformation associated with the opening of the
“Thus, patches of distinct grey halite associated with wafts (smell) of hy- Dead Sea Basin, given by Al-Zoubi and ten Brink (2002) (Fig. 7). Here,
drogen sulphide probably signal venting of foetid groundwater in small a shear zone at the base of the upper crust with fluid flow is indicated at
groups of reef compartments hundreds of metres apart. More spectac- around 14 km depth, and they conclude: “The lower lithostatic pressure
ularly, a particular screen of halite, a few decimetres thick and 1 m (rock overburden) at mid-crust levels under the basin relative to the sur-
UN

high, separates brines that crystallize halite for kilometres on one side rounding crust and the fracturing of rock under the basin due to earth-
from brines that crystallize carnallite for kilometres on the other side.” quakes at that depth (13–18 km) promote fluid flow from the surround-
(Talbot et al., 1996). These features demonstrate that the seepage of ing crust into the middle crust below the basin. The presence of fluids
brines on the seafloor is derived from different compartments in the sub- within the fractured crust helps create weak shear zones along which
surface, where refining processes have isolated various types of salts. the opening is concentrated.”
2.4.14.2. Where is the source of the brines? The pattern of the venting The large scale hydrothermal system driven by the heat of shallow
brines into the Dead Sea strongly indicates that the sources are of mantle masses beneath the basin and the occurrences of large scale ac-
deep-seated hydrothermal origin, where the circulation is driven by the cumulations of ‘hidden salt’ below the upper crust can also explain the
magmatic activity under the basin. The 3D model of the fault zone various types of salts depositing on the seafloor: halite and carnallite.
(Fig. 7) from Al-Zoubi and ten Brink (2002) actually indicates a shear It is cautiously suggested that such accumulations of ‘hidden salts’ may
zone be the result of subduction in a previous Wilson cycle. Furthermore, it

11
M. Hovland et al. Marine and Petroleum Geology xxx (2017) xxx-xxx

is suggested that such salt accumulations will form zones of rock-me- plate is being pushed on top of the other. Eventually, the subducting
chanical weakness in the underground, thus, promoting new rifting to slab tears off and the mountains stop growing.
preferably propagate through such zones. These items will be discussed Stage G –brings us back to a stable craton, where erosion eventually
further in the following chapter (Chapter 3). grinds down the mountains and the stage is prepared for a new rifting
episode.
3. Relationships between salt formation and Wilson cycles

F
3.2.1. Salt-forming stages in the Wilson cycles
3.1. Introduction The stages that provide the opportunity for water or seawater to con-
tact hot mantle, are the most important stages for salt formation. The

OO
Following Alfred Wegener's assertion that the Earth's continents initial continental rifting (Stage B) is usually involving influx of mete-
had fitted together, and that they later had split up and drifted apart oric water into deep faults within the rift system. This may easily mobi-
(Wegener, 1915, 1922), geophysicists and geologists took another 40 lize soluble substances in the mantle/deep crust and bring them to the
years to resume this model. It was the unequivocal inference from mag- surface via hydrothermal processes. Water may leach out elements con-
netometer transects across mid ocean-ridges with the vessel Eltanin in stituting salts when interacting with pristine mantle and other rocks, or
1966 that turned the table and provided the ‘smoking gun’ that Wegen- bring out salts from previous stages of the Wilson cycle. Burke et al.
er's model was, indeed, generally correct (Pitman and Talwani, 1972). (2003) have used the presence of obducted carbonatites to show that
The magnetic anomalies on both sides of the mid-ocean ridges came as suture zones in the vicinity of the East African Rift represent older sub-

PR
proof that the ocean bottom was young and continuously was renewed duction zones that are now located at the center of rifting activities. The
along the ridges (Talwani and Eldholm, 1977). implication of their work is that old salts accumulated during much ear-
During the same year (1966), John Tuso Wilson also published his lier stages, may be mobilized and re-appear on the surface, from deep,
benchmark paper: “Did the Atlantic close and then reopen?” Here, he hidden sources.
presented the question: “How can the plate-tectonic history of the Earth When seawater starts filling the rift during Stage C, the same
be deciphered for times earlier than the oldest in situ Ocean floor?” processes as above will continue and, in addition, the seawater brings
(Burke, 2016). His answer was “In terms of rocks and structures char- in more salts that may form heavy brines and precipitate solid salts in

D
acterizing stages in the life cycle of the ocean basins” (Wilson, 1968). the hydrothermal processes. An example of halite production from Lake
Magadi in the East African Rift clearly represents evidence of halite be-
Soon this definition was termed “The Wilson cycle” (Dewey and Burke,
ing accumulated despite occurring far away from current seawater.
1974; Burke, 2016) and today, the concept of Wilson cycles has become
During Stage D, the cooling oceanic crust is continually being ex-
TE
a fundamental term in plate tectonics, despite still being debated.
posed to seawater down to considerable depth due to contraction and
Questions about the start of the Wilson cycles were addressed by
faulting. Oceanic crust is known to have porosities in the range 20–25%
Piper (2013) who suggests a concept of Lid tectonics before Plate tec-
(Becker and Fisher, 2000). While still remaining hot, ‘off-axis’ hy-
tonics. This suggestion is based on the fact that mobile tectonic plates
are recognized by their paleomagnetic patterns of polar wander paths, drothermal processes continue at depth within the oceanic crust, thus,
while stagnant tectonic lids are not. These paleomagnetic patterns can making it ready for the next stage, Stage F. The oceanic crust is now
loaded with seawater, brines, and even some solid salts (Butterfield,
EC

be mapped, and according to Piper (2013) the change from ‘Lid tecton-
2000).
ics’ to ‘Mobile plate tectonics’ was transitional, and the Wilson cycle style
During Stage F, the oceanic crust and its content of components from
plate tectonics developed mainly in the Proterozoic eon that began at
former seawater is subducted. It is heated and pressurized, thereby al-
2.5 Ga.
lowing a new set of chemical and physical reactions to take place.
A subset of the Wilson cycles is commonly referred to as the “Su-
3.2. What are Wilson cycles?
percontinent cycle”. During the history of Earth, several supercontinents
RR

The stages in the Wilson cycles in their simplest form are, generally, have assembled large landmasses, for thereafter to rift and drift apart
as follows (Wilson, 1968; Torsvik et al., 2010; Burke, 2011, 2016): Stage again. From complete assembly of landmasses to rift, it has taken on av-
A – a stable craton of continental crust exists. Due to its lower specific erage ∼250 Ma.
gravity, it floats on top of the mantle. These events are of great importance to the total energy input to
Stage B – the craton experiences uplift from a rising hot-spot and the formation of solid salts from dilute brines such as seawater, and
there is evidence that supercontinents are an integral part of Wilson
starts to rift. A system of branched rifts occurs, thereby creating a
CO

cycles, involving rifting, sea floor spreading, and subduction, and that
4–5 km relief between the edges and the bottom of the rift. Eventually,
these processes have been going on for at least 2.7 Ga (Piper, 2013;
hot, mafic or ultramafic magma flows in through its bottom.
Buiter, 2016). This provides ample evidence for the existence of numer-
Stage C – an early divergent margin evolves. The dense magma
ous stages of Wilson cycles and vast areas of Earth that potentially have
starts forming oceanic crust as it cools. Its density prevents it from ris-
been involved in salt formation.
ing above sea level. As extension continues, the margins become more
distant from the rift. The oceanic crust cools with time and becomes
denser, sinking deeper into the mantle, thus forming the deep oceans. 3.3. Subduction and salt
UN

Stage D – a fully divergent margin develops. By now, the first


oceanic crust is too dense to be supported by the underlying mantle. It 3.3.1. Characteristics of the oceanic crust
starts to sink in, and thereby, starts pulling the rest of the oceanic plate The oceanic crust has an average thickness of ∼8 km and a regional
with it. As it subducts, it brings water with it that feeds volcanism at the porosity of up to 25%, a porosity, which is considered mainly to be
plate boundary. filled with seawater (Becker and Davis, 1998). Whereas it has proved
Stage E – a volcanic margin starts building mountains, either as an relatively easy to find ocean floor vent sites for hydrothermal flow,
island arc (if the subduction is oceanic), or as a mountain arc if the sub- it is much more difficult to document where seawater enters into the
duction is of Cordilleran type. oceanic crust, the ‘recharge’ zones. However, one such site was found
Stage F – eventually, the subducting plate has brought the two con- by chance, during the drilling of DSDP Hole 395A, at Site 395 in the
tinents on either side of the oceanic crust into contact with each other, Atlantic Ocean. The hole is located at an isolated sediment pond, the
and collisional mountain building starts. Continental crust from one

12
M. Hovland et al. Marine and Petroleum Geology xxx (2017) xxx-xxx

‘North Pond’, measuring 8 by 15 km. The hole was drilled ∼70 km west
of the Mid Atlantic Ridge (MAR) crest, at 4440 m water depth. After 3.3.2. Hidden- and sometimes visible salts in subduction zones
penetrating 92 m into the sediments and over 500 m into the oceanic A study by Scribano et al. (2017) states that serpentinites may act as
crust, mainly consisting of basalts, to their surprise, the researchers reservoirs for dense brines and salts due to the mineralogical processes
found drawdown of water instead of the expected over-pressure (Becker involved in serpentinization of hot peridotite in contact with seawater.
and Davis, 1998). This salt may later be driven out to become visible salt deposits.

F
During a re-visit to this hole, they measured that seawater was dis- Similarly, a study on the fate of salts in peridotites and serpentinites
appearing into the open hole at a rate of ∼1000 L per hour. The hole exposed to seawater was undertaken by Sharp and Barnes (2004). Some
was then fitted with a CORK (Circulation Obviation Retrofit Kit, e.g., an of their conclusions are:

OO
instrumented borehole seal) and monitored: “In the 21 years since the
(1) Serpentinite may be the major conduit for surficial chlorine transfer
initial drilling, Hole 395A has been re-entered 4 times: first during Leg
to mantle depths and an important part of the chlorine cycle. The
78B, and most recently, during Leg 174B. Each time, repeat temperature
Cl flux in serpentinites into the mantle is larger than all other previ-
logs, fluid samples, and flowmeter logs clearly demonstrated that ocean
ously identified fluxes.
bottom water was flowing down the hole at a consistent rate of about
(2) Cl is sited in both water-soluble and –insoluble components.
1000 L/hr” (Becker and Davis, 1998). Results from similar settings, i.e.,
(3) A high-salinity fluid “plume” is evolving during serpentinite dehy-
the flank of the Endeavour Ridge (ODP Leg 168), strongly supports the

PR
notion that there are huge fluxes of low-temperature seawater flowing dration of a subducting slab. The high-salinity fluid affects meta-
through very transmissive upper basement sedimented young oceanic morphic reactions and melting properties above the slab and can
crust, regardless of whether the sediment cover is continuous or patchy cause intense metasomatism.
and regardless of spreading rate.
Oceanic/continental subduction occurs where parts of the oceanic Deeper processes within subduction zones involving granites have
lithosphere descends into the mantle beneath an overriding continental been investigated by Srikantappa et al. (1992). They observe fluid in-
plate. During subduction, the oceanic plate becomes dewatered as its clusions with halite contents of up to 50–60% held in rocks having been
hydrous fluids are released and permeate the overlying mantle wedge exposed to 500–550 °C. Based on appearance and orientation of the in-

D
(Fig. 9). Similar dewatering also occurs for intra-oceanic subductions, clusions, the authors suggest that their content is of magmatic origin,
where oceanic lithosphere descends below another oceanic lithosphere. and not meteoric.
Even if the subducting slabs release some of their pore water con- Studies of Eclogites from even deeper portions of subduction zones
have been performed by numerous authors, including i) Svensen et al.
TE
tent early in the subduction process, large masses of water and salts are
brought down to great depth and may undergo hydrothermal conver- (2001), and their study of the eclogites of western Norway, where they
sion. To illustrate the potential for salt formation in just one subduction observed salinities in eclogite inclusions of 15–25% salt content; ii)
zone, the Andean subduction can provide some hints. Faccenna et al. Xiao et al. (1998) who studied eclogites in eastern China and found
(2017) have calculated the length of subducted lithosphere in the An- high salinity inclusions in rocks having been exposed to more than
dean subduction zone to be in the range of 2200 km to 7800 km, over 700 °C; and iii) Alexeev et al. (2017) who report from a study on dis-
EC

a period of 25–80 Ma. This subduction zone stretches along the entire posal of highly saline brines flowing into a diamond mine in an eclog-
South America over a width of more than 15,000 km. With a suspected ite hosted kimberlite, situated within the Siberian platform. Inflow of
porosity of 5–10% (from initially 20–25%), filled with saline porewa- salt brines into the mine exceeded 75 m3⁠ per day of brines with solid
ters within the subducted, multiple kilometres thick slabs of the Pacific contents of more than 350 g/litre. These examples leave no doubt that
Ocean crust, the amount of potentially formed salt will be in the size of
millions of cubic kilometres.
RR
CO
UN

Fig. 9. Left, Zellmer et al.’s figure, showing subduction-related arc volcanism that is triggered by fluids and minerals ‘squeezed’ out of the down-going oceanic lithosphere under the sub-
duction wedge of the overlying continental plate. ‘AW’ = Accretionary wedge; ‘OC’ = Oceanic crust; ‘CC’ = Continental crust. Right, diagram from Manning (2004), which schematically
shows the path of a slab-derived fluid, with mantle H2⁠ O contents (wt%). The fluid migrates into the mantle wedge (solid orange arrows), where it is absorbed through formation of hydrous
minerals. Downward flow of solid mantle (dashed arrows) causes dehydration. After multiple hydration/dehydration steps, the fluid enters a region where it is stable with anhydrous
minerals, which allows greater travel distances. Both images illustrate important aspects of the inner ‘anatomy’ of subductions. These processes occur within the green rectangular zone
indicated on the lefhand figure (Modified from Zellmer et al., 2015; Manning, 2004). (For interpretation of the references to color in this figure legend, the reader is referred to the Web
version of this article.)

13
M. Hovland et al. Marine and Petroleum Geology xxx (2017) xxx-xxx

salts are transported to great depths in subduction zones, from where gle volcano is equivalent to 500,000 km3⁠ . The volcano is located south
they may reappear during later stages of the Wilson cycle. of Lake Uyuni.
It might be stated that the fraction of salt within inclusions is rela- This is what Laumonier et al. (2017) state: “Similar high conductiv-
tively small compared to the volume of the solid rocks and hence that ity features are observed beneath the Cascades volcanic arc and Taupo
they may only represent a minor source of salt. This view does not cap- Volcanic Zone. This suggests that large amounts of water in super-hy-
ture the full extent of the processes behind the formation of these rocks drous andesitic magmas could be a common feature of active continen-

F
and their inclusions. There is ample evidence that the trapped inclusions tal arcs and may illustrate a key step in the structure and growth of the
represent just a minute part of the salts and fluids that were present dur- continental crust.” The investigation by Laumonier et al. rests on the in-
ing crystallization of the rocks. The relatively low crystallizing tempera- terpretation of water as the cause for the high conductivity. The authors

OO
tures tell us that these rocks were not formed in dry melts and that much performed high pressure, high temperature experiments with relevant
more water was present before hardening. This is being recognized in rocks and water to establish the water content needed to produce the
several studies, for example Alderton and Harmon (1991), O'Reilly et al. observed conductivity. It may also cautiously be suggested that water at
(1997) and Harlov et al. (2012). these depths is most likely saturated with salts, e.g., it is likely that hid-
They all observe how different fluids, including brines, are trapped den salts or brines play a yet undetected but important role here.
during the crystallizing of the rocks, thereby proving presence of fluxes During later tectonic events, e.g., rifting (like in the East African
of brines within the subducting system. When quartz is crystallizing Rift), the hidden salts may become activated and mixed with newly
from solution, it produces water that tends to dilute the brines pre-

PR
formed salts from rift-related hydrothermal processes (e.g., Lake Mag-
sent. It is therefore likely that many inclusions in granites and eclogites adi), although, however, it is not straightforward to distinguish between
have captured slightly diluted brines, relative to those present at the on- the two generations of salt. Generally, however, the most soluble salt
set of crystallization. Regarding the question of melting versus solution types, e.g., magnesium chloride, disappear out of the system.
of minerals in the subduction zones, Manning (2004) states that: “Un-
der appropriate conditions, silicate solubilities may become so high that 3.4. Rifting and salt
there is complete miscibility between hydrous melts and dilute aqueous
solutions.” 3.4.1. Introduction

D
Laumonier et al. (2017) confirm that the quantity of brines in rocks In the Wilson cycle salt model, rifting represents the ‘second phase’,
within the deeps of subduction zones is in the order of 8–10% relative following that of subduction. Rifting occurs where parts of the continen-
to rock volume. They have located a vast volume of water (that origi- tal or oceanic crust is stressed to such an extent that it is pulled apart,
nally subducted as seawater) beneath one volcano - the Uturuncu vol- and causes hot mantle to be drawn upwards, so that it partially melts
TE
cano on the Bolivian Altiplano. The volume of fluids beneath this sin (Fig. 10) (Asimow, 2017).
EC
RR
CO
UN

Fig. 10. Conceptual drawing showing the geology of mid ocean rifting, modified from Asimow (2017). The age (millions of years) of the oceanic crust is indicated on the horizontal
axis. “The solid-state flow induced by plate spreading brings hot material upward. Upwelling mantle first crosses the hydrous solidus, where melting begins as a result of trace water in
the mantle, and subsequently the dry solidus, where melting accelerates. The thickness and composition of the crust, geophysical probes, and experiments locating the solidus constrain
estimates of mantle temperature. Sarafian et al. infer that the mantle is 60 K hotter than previously thought.” (Asimow, 2017).

14
M. Hovland et al. Marine and Petroleum Geology xxx (2017) xxx-xxx

This chapter reviews how brines and solid salts may become integral surface must contain a higher portion of materials from the slab grave-
in the rifting process in both continental (non-marine) and oceanic rift- yard itself. This would imply a certain content of water and salts.
ing. As a hot-spot develops into a spreading oceanic crust, the initial
surge of volatile materials may become less dominant and it releases
3.4.2. Hot-spots and mantle plumes: precursors for rifting magma with a typical basaltic content of chlorides. The difference in
According to Bosworth (2015) and references therein, the Red Sea salinity between hot-spots and oceanic basalts has been documented by

F
rift occurred between two hot-spots, or hot mantle plumes, the ‘Cairo Stroncik and Haase (2004). They state: “The majority of hotspot lavas
Mini-Plume’ and the ‘Afar Plume’. What do such plumes represent in have higher Cl/K ratios than depleted mid-ocean ridge basalts, con-
the current context? sistent with the presence of recycled crustal components in the man-

OO
Hot-spots arise as a consequence of physical instabilities in mantle tle-plume sources of hotspots.”
materials. During subduction, the first instability to appear is the com- Kovalenko et al. (2006) studied the zonal differences within three
pression of the subducting slab, leading to the escape of some of the mantle hot-spots. They report on the source of the plumes as well as the
saline fluids. During oceanic subduction, the fluids have a relatively distribution of water, chlorine and other elements within it. They con-
easy access to the surface - or seabed, where they escape to the water clude: “The plume mantle was formed mainly by the mixing of three
column. This especially occurs along fault planes leading down to the sources: ultradepleted mantle, moderately enriched relatively dry man-
subduction zone. tle, and moderately enriched H2⁠ O-rich mantle. In addition to the three

PR
As the slab decends the saline brines are causing basaltic volcanism main components of the plume mantle, there are probably minor com-
and the build-up of island arcs, by penetrating the hanging wall (Fig. 9). ponents enriched in chlorine and depleted in fluorine. It is supposed that
During Cordilleran subduction, the situation is similar, but the porewa- all these components are entrained into the plume mantle through the
ter contains some extra elements due to slightly different chemistry in mantle recycling of components of the oceanic and continental crust.
the system. Some saline fluids also escape the subduction zone by flow The established relationships are in agreement with the zonal model of
to the surface via hydrothermal activity and conductive faults. Brines a mantle plume, which includes a hot central part poor in H2⁠ O, Cl, and
escaping the subduction zone in this situation have to penetrate much S; an outer part enriched in volatile and nonvolatile incompatible ele-
thicker crust in order to emanate on the surface. Salty brines emanating ments; and enclosing mantle material interacting with the plume.” This

D
onto surfaces with arid climate may lead to accumulation of salt, while description of mantle plumes is in line with our model, where water
those venting out on locations with wet climate will be washed back to contained in the outer parts of the plume most likely originates from
the sea. the subducted oceanic plate and that chlorine and sulfur are remnants
Hot brines emanating on the bottom of deep oceans tend to be bet- of marine salts associated with pore water in the oceanic slab.
TE
ter protected against dissolution than brines ending up on land. This is Both of these two studies on chloride contents and origin of hot-spot
because brines that are venting on the sea floor will cool and precipitate magmas are basing their results on studies of inclusions trapped in vol-
solid salts that may later be protected permanently by sediments or ma- canic glass as the magma crystallized. It is well known that water and
terial of biogenic origin such as carbonates. basalts exist in full miscibility at elevated pressures and temperatures,
During the next step of development in the Cordilleran subduction, as the water lowers the melting temperature of the mixture. However,
during ascent the magma is gradually depressurized and ultimately a
EC

the brines interact with the hanging wall to produce the physical in-
stability that leads to granitic plutons rising to the surface. The plutons higher percentage of water escapes and may transport salts with it as
also act as a transport mechanism for saline brines. If the subduction has the lava solidifies. Thus, the amount of water and salts found trapped
been flat for extended periods of time, the original hanging wall mantle in inclusions probably represents the later stage amount and concentra-
may be pushed away, and a different form of magmatism is seen. But tions, rather than the initial stage.
the water and accompanying salts are still important components of this Because hot spot plumes from deep in the mantle (e.g. 3000 km) in-
inventory. clude subducted oceanic crust, also sea salts are found in these lavas.
RR

During the last stages in the life of a subducting slab it sinks into Sobolev et al. (2011) report the analysis of strontium isotopes
the so-called ‘slab graveyard’ at the lower boundaries of the outer man- (8⁠ 7Sr/8⁠ 6Sr) in lava from Mauna Loa, Hawaii, which matches the com-
tle (Fig. 11). Here, the relatively cool remains of the slab start heat- position of 0.3–0.5 Ga old seawater. Seawater older than 0.5 Ga is sig-
ing up again, and with its higher content of remaining hydroxyl-bear- nificantly less radiogenic. Direct contact with modern seawater is not
ing minerals may become buoyant and form a volcanic hot-spot that consistent with the low values of B and Cl. Thus, they conclude that
rises to the surface. During ascent, it will interact with the surround- seawater must have ‘contaminated’ the Mauna Loa source rock prior to
CO

ing mantle and become slightly mixed with it. This does not preclude subduction (Sobolev et al., 2011). These results indicate that sea salts
the fact that the first parts of the hot-spot magma arriving at the
UN

Fig. 11. Cartoon modified from Svensen et al. (2017), showing subducted lithospheric slabs that may interact with the margins of thermochemical provinces in the deep mantle, including
‘TUZO’ (in red), located beneath much of Africa and India. This may trigger the formation of plumes, as discussed in the text. (For interpretation of the references to color in this figure
legend, the reader is referred to the Web version of this article.)

15
M. Hovland et al. Marine and Petroleum Geology xxx (2017) xxx-xxx

used 0.5 Ga to re-appear at surface; i.e., much shorter time than earlier
expected. Based on their study, Sobolev et al. (2011) also calculated the 3.4.4. Numerical modeling of the South Atlantic rift
average circulation rate of crust through the mantle to be in the order In a recent numerical modeling of the ‘rift-to-drift phase’ of the
of 1–2 cm/year. South Atlantic salt basin, Norton et al. (2016) mapped the limits of
The differences in salinity of hot-spot volcanism, therefore, may be oceanic crust and salt (Fig. 12). In accordance with the model proposed
explained by the origin of their magma. Hot-spots that arise as a result by Torsvik et al. (2009), they suggest that salt was deposited when the

F
of subducted materials being recycled should have higher salinity than seafloor spreading began, that is, when the South Atlantic opened. In
others. However, this salinity may not last forever as the more volatile addition, Norton et al. (2016), suggest that precipitated salt flowed over
components must be expected to arrive early and then diminish in con- and covered the spreading ridge axis, thus, sealing off the ‘extrusive

OO
centration over time. Initial hot-spot volcanism of recycled mantle ma- components’ of the oceanic crust, resulting in the formation of intrusive
terials should therefore be expected to produce salts with the magma. oceanic crust: “Seafloor spreading eventually broke through the thin-
This salt comes in addition to other salts produced by hydrothermal ac- ning salt, forming breakthrough volcanoes preserved today as basement
tivity from heat and rock-water interaction. ramps at the distal limits. These ramps formed diachronously, so the dis-
The composition of evaporites and various rock types were investi- tal salt limits are not isochrones, explaining the poor fit of these features
gated by Kendrick et al. (2017) and compared to seawater. Some of the in plate reconstructions.” (Norton et al., 2016). Their results are actu-
data are shown in Table 3 together with calculated ratios between the ally much in line with recent descriptions from the Red Sea by Augustin
three elements Cl, Br, and F. Interestingly, these data show very large

PR
et al. (2016), where giant salt flows seal off large tracts of the actively
differences between the ratios of these elements for the various sources. rifting central axis of the Red Sea (to be discussed in Part 2).
The Cl/Br ratios for evaporites are one order of magnitude higher than Norton et al. (2016) actually suggest that the salt formation, in sce-
for seawater, while the corresponding Cl/F ratios for evaporites are nario C, of Fig. 12, is more-or-less ‘instantaneous’: “Estimates of du-
nearly four hundred times higher than for seawater. The Cl/Br ratios
ration of salt deposition are poorly constrained, ranging from 600 k.y.
for ‘primitive mantle’ are comparable to those of seawater while the
to 5 m.y. (Davison et al., 2012). Given Aptian plate spreading rates of
Cl/F ratio of ‘primitive mantle’ is two orders of magnitude lower than 50–60 km/m.y. (Heine et al., 2013), longer intervals of salt deposition
seawater. These elemental ratios demonstrate that evaporites have gone would mean that the conjugate salt basins separated before the end of

D
through pervasive refining processes and interactions with the side-wall salt deposition; this is implausible. We thus prefer a short period of
rocks in the hydrothermal system and are no longer similar to seawater salt deposition. We also assume that isostatic loading, occurring as it
composition. does over thousands of years as opposed to hundreds of thousands of
It is important to be aware that even if concentrated brines are
TE
years for salt deposition (van den Belt and De Boer, 2007), was instan-
formed from seawater circulating in ascending mantle plumes, special taneous”. However, to us, this seems very improbable, especially when
geological conditions are required for this salt to precipitate as solid
we know that isostatic forces will not be powerful enough to provide the
salt and to be preserved. Such relationships cannot be found every-
needed subsidence.
where. Along the MAR, for example, there is no concentration of salt,
Thus, a short period of salt deposition is likely, combined with in-
because the seawater-rock interactions are near the seabed and the salt
stantaneous isostatic compensation. According to van den Belt and De
is washed away. This will often be the case if the serpentinization takes
EC

Boer (2007) this scenario seems improbable, because the weight ex-
place near the seabed.
certed by the low density salt is not sufficient to isostatically push down
the heavy rocks beneath the basement floor. However, according to our
model the salt filling the basin originates from ‘hidden salt’ beneath the
3.4.3. The South Atlantic rift basins
basin-floor, and the removal of this salt up into the basin will give an
In their analysis of the opening of the South Atlantic Ocean and the
instantaneous subsidence of the basin floor due to mass balance condi-
dissection of an Aptian salt basin, Torsvik et al. (2009), argue for a con-
RR

tions. This process is referred to as subrosion, e.g., below-ground salt dis-


ventional evaporite formation: “ … that the Aptian salt accumulations
solution (Ehrhardt and Hübscher, 2015). Furthermore, the circulation of
belonged to a single pre-breakup (syn-rift) basin confined to continen-
seawater during the transportation phase added more salt to the salt al-
tal and/or subaerial basaltic substrates.” However, water depths of up
ready accumulated beneath the basin, as is also seen in the Red Sea, to
to 500 m in this single syn-rift basin are not excluded: “It is widely ac- be described and discussed in more detail in Part 2.
cepted that the salt was deposited in a shallow-water environment, and Thus, it is suggested that the salt basin was filled by salt deposits
palaeontological evidence from the Aptian to earliest Albian dolomite during periods of high heat-flow, e.g., during SDR-formation, and that
CO

and sapropel sequence of the Angolan margin, deposited just above the salt was brought into the basin by hydrothermal saturated brines from
Aptian salt, suggests no more than 500 m paleo-water depth (Marton et below.
al., 2000)”.
In order to explain evaporation in such a deep basin, Torsvik et al. 3.4.5. Indications of the same process other places
(2009) refer to work by Nunn and Harris (2007): “These authors pro- It is inferred that a similar process is currently taking place in the
posed a model of subsurface seepage of seawater across a barrier to ex- northern Red Sea and Afar regions (see Part 2) and is thus, demonstrat-
plain this apparent dilemma, which also would constitute a possible ad- ing the close relationship between hydrothermal salt production and
UN

ditional mechanism for the syn-rift deposition of the great evaporite se- volcanism, as well as brine migration and refining of salts, whether they
quence.” In our opinion, however, the potential effects of hydrothermal are deposited on the surface in a dry climate or on the seafloor by pre-
processes associated with high heat-flow are more plausible mechanisms cipitation in high-density brine pools. However, the most interesting as-
for explaining the salt accumulation in the basin. pect of the conclusions of Norton et al. (2016) is their transition from

Table 3
Contents of Cl, Br and F in seawater, evaporites and ‘primitive mantle’. Ratios between the three elements are also shown. (From Kendrick et al., 2017).

Cl (ppm) Br (ppm) F (ppm) Cl/Br Cl/F

Seawater 19,300 ± 970 66 ± 3.3 130 ± 0.07 292 148


Evaporites 550,000 ± 50,000 150 ± 100 1000 ± 300 3667 55,000
Primitive mantle 5±2 0.076 ± 0.025 17 ± 6 342 1.5

16
M. Hovland et al. Marine and Petroleum Geology xxx (2017) xxx-xxx

F
OO
PR
D
TE
EC

Fig. 12. Sketch, modified from Norton et al. (2016), represents a modern view, based on numerical geo-modeling, on how and when salt was deposited in the Central Atlantic, Campos
and Kwanza basins. ‘RA’ is the depth of the Rift Axis below sea level. The figure text of Norton et al. (2016) is as follows:”The case of seaward-dipping reflector (SDR) margins like part
of the Campos Basin is illustrated. A: Rift basin is subaerial and transitioning to oceanic crust formation. Rift axis is ∼1 km below sea level. B: Same time as in A, but now seawater has
RR

flooded the basin. Rift axis is ∼1500 m below sea level. C: Salt has filled the basin to sea level; rift axis is ∼3500 m below sea level. D: 3–4 Ma after C, with ∼100 km total motion since C.
Intrusive oceanic crust is being created under salt. E: Shortly after D, with volcanism breaking through the salt, forming the breakthrough volcanoes. F: A few million years after E, with
normal oceanic spreading”.

Stage B to Stage C (Fig. 12), where enormous salt deposits are formed tonic activity preceding salt deposition. Over the last decade it has be-
“instantaneously” (Norton et al., 2016). This is indeed parallel to what come a well established fact that heavy brines form and solid salts pre-
seems to have occurred also in the Mediterranean Sea (e.g., during or cipitate within hydrothermal systems, due to the thermodynamic and
CO

prior to the Messinian Salinity Crisis) and in the Red Sea a few million physico-chemical properties of seawater at high temperatures and pres-
years earlier. sures. It is suggested that volcanism and hydrothermal activity events
But even though the bulk of the salt was deposited in the late Ter- associated with different stages of the Wilson cycle are the driving forces
tiary era, salt formation continues. In the Red Sea basin, precipitation behind salt accumulations in giant basins.
of salt occurs in the subsurface, and slurries of solid salt and satu- The Wilson cycle is an expression of the geodynamics of Earth,
rated brines rise up and are observed venting out along the sides of and based on the conceptual elements highlighted in this work, it is
the basin, forming “namakiers” of salt that move down the slopes to- suggested that the formation of hydrothermal salt is associated with
UN

wards the embryonic spreading centers of the Red Sea. This is in spite deeply circulating seawater within the hydrothermal systems in the
of the fact that the basin water has salinities just above that of seawa- Earth's crust and upper mantle, driven by the internal heat of the Earth.
ter (35,000–37,000 ppm). A completely different scenario is seen in the These seawater-based processes also include serpentinization of mantle
Dead Sea, where the salinity in the water exceeds 300,000 ppm, and masses and water-rock interactions by the circulating hot seawater in
halite and carnallite is observed to precipitate in the margins and at the the oceanic crust. The immense recycling of crustal materials, includ-
seabed (Talbot et al., 1996). ing the subduction of the porous oceanic crust charged with brines and
probably also solid salt, brings water and salt deep down into the man-
4. Concluding remarks tle.
The tectonic activity, including volcanism and associated hydrother-
The formation of large marine salt deposits is mainly registered in mal processes bring water and salts upwards towards the surface where
ancient basins, especially in areas with a geological history of high tec huge amounts of salts may precipitate; e.g. in the high An

17
M. Hovland et al. Marine and Petroleum Geology xxx (2017) xxx-xxx

des Mountains. It is also acknowledged that such brines and salts may be Christeleit, E.C., Brandon, M.T., Zhuang, G., 2015. Evidence for deep-water deposition of
abyssal Mediterranean evaporites during the Messinian salinity crisis. Earth Planet Sci.
stored sub-surface for long periods of time, e.g. extending from one Wil- Lett. 427, 226–235.
son cycle to another. Thus, it is cautiously suggested that large amounts Cita, M.B., Fornaciari, M., Camerlenghi, A., Corselli, C., Erba, E., McCoy, F.W., Vezzoli,
of salts remain hidden inside the subduction zone including the volcanic L., 1988. Anoxic basins of the eastern meditarranean: new evidence from the bacino
Bannock area. Soc. Geol. Ital. 36, 131–144.
and sedimentary rocks above the subducted slab. During later rifting
Corselli, C., Basso, D., Lange, G., Thomson, J., 1996. Mediterranean Ridge accretionary
and oceanization processes, the subducted rocks, including their con- complex yields rich surprises. Eos Trans. Am. Geophy. Union 77, 227, 227.

F
tents of hidden salts, may again be mobilized and deposited at the sur- Coumou, D., Driesner, T., Weis, P., Heinrich, C., 2009. Phase separation, brine formation,
and salinity variation at Black Smoker hydrothermal systems. J. Geophys. Res. Solid
face.
Earth 114.
Davison, I., Anderson, L., Nuttall, P., 2012. Salt Deposition, Loading and Gravity Drainage

OO
Uncited references in the Campos and Santos Salt Basins. Geological Society, London, 159–174, Special
Publications 363.
Dewey, J.F., Burke, K., 1974. Hot spots and continental break-up: implications for colli-
Bird, 2003; Burke, 1975; Burke et al., 1977; Ehrhardt et al., 2005; sional orogeny. Geology 2, 57–60.
Engvik et al., 2011; Garfunkel et al., 2014; Gutscher et al., 2000; Dick, H.J., Lin, J., Schouten, H., 2003. An ultraslow-spreading class of ocean ridge. Nature
Hannington et al., 2001; Pirajno, 2009; Searle and Ross, 1975; Wilson, 426, 405.
Driesner, T., Heinrich, C.A., 2007. The system H 2 O–NaCl. Part I: correlation formulae for
1966. phase relations in temperature–pressure–composition space from 0 to 1000° C, 0 to
5000bar, and 0 to 1 X NaCl. Geochem. Cosmochim. Acta 71, 4880–4901.
Acknowledgements

PR
Eckstein, Y., Simmonsi, G., 1977. Measurement and interpretation of terrestrial heat
flow in Israel. Geothermics 6 (3), 117–142. https://doi.org/10.1016/
0375-6505(77)90023-2.
We would like to thank editors Massimo Zecchin and Timothy Ehrhardt, A., Hübscher, C., 2015. The northern Red Sea in transition from rifting to drift-
Horsecroft for suggesting publishing this review article. Also we thank ing – lessons learned from ocean deeps. In: Rasul, N.M.A., Stewart, I.C.F. (Eds.), The
V. Scribano and W. Mohriak for their thorough and constructive re- Red Sea: the Formation, Morphology, Oceanography and Environment of a Young
Ocean Basin. Springer Earth System Sciences, Berlin Heidelberg, pp. 99–121.
views.
Ehrhardt, A., Hübscher, C., Gajewski, D., 2005. Conrad Deep, northern Red Sea: develop-
ment of an early stage ocean deep within the axial depression. Tectonophysics 411,
References 19–40.

D
Engvik, A., Mezger, K., Wortelkamp, S., Bast, R., Corfu, F., Korneliussen, A., Ihlen, P., Bin-
Alderton, D., Harmon, R., 1991. Fluid inclusion and stable isotope evidence for the origin gen, B., Austrheim, H., 2011. Metasomatism of gabbro–mineral replacement and el-
of mineralizing fluids in south-west England. Mineral. Mag. 55, 605–611. ement mobilization during the Sveconorwegian metamorphic event. J. Metamorph.
Alexeev, S.V., Alexeeva, L.P., Shvartsev, S.L., Sidkina, E.S., 2017. Specifics of the late Geol. 29, 399–423.
Cenozoic geochemical evolution of chloride calcium brines in the Olenek cryoartesian Evans, B.W., 2004. The serpentinite multisystem revisited: chrysotile is metastable. Int.
TE
basin. Geochem. Int. 55 (5), 442–456. https://doi.org/10.1134/S0016702917050020. Geol. Rev. 46, 479–506.
Alling, H.L., 1928. The Geology and Origin of the Silurian Salt of New York State, vol. 275, Faccenna, C., Oncken, O., Holt, A.F., Becker, T.W., 2017. Initiation of the Andean orogeny
New York State Museum And Science Service, Bulletin. by lower mantle subduction. Earth Planet Sci. Lett. 463, https://doi.org/10.1016/j.
Al-Zoubi, A., ten Brink, U.S., 2001. Salt diapirs in the Dead Sea basin and their relation- epsl.2017.01.041.
ship to quaternary extensional tectonics. Mar. Petrol. Geol. 18, 779, 292. Faccenna, C., Becker, T.W., Auer, L., Billi, A., Boschi, L., Brun, J.P., Capitanio, F.A., Fu-
Al-Zoubi, A., ten Brink, U., 2002. Lower crustal flow and the role of shear in basin subsi- niciello, F., Horvàth, F., Jolivet, L., Piromallo, C., Royden, L., Rossetti, F., Serpelloni,
dence: an example from the Dead Sea basin. Earth Planet Sci. Lett. 199, 67–79. E., 2014. Mantle dynamics in the mediterranean. Rev. Geophys. 52, 283–332. https:
//doi.org/10.1002/2013RG000444.
EC

Asimow, P.D., 2017. A measure of mantle melting. Science 355, 908–909.


Augustin, N., van der Zwan, F.M., Devey, C.W., Ligi, M., Kwasnitschka, T., Feldens, P., Früh-Green, G.L., Connolly, J.A.D., Plas, A., Kelley, D.S., Groberty, B., 2004. Serpentiniza-
Bantan, R.A., Basaham, A.S., 2016. Geomorphology of the central red sea rift: de- tion of oceanic peridotites: implications for geochemical cycles and biological activity.
termining spreading processes. J. Geomorphol. https://doi.org/10.1016/j.geomorph. In: Wilcock, W.S.D., Delong, E.F., Kelley, D.S., Groberty, B. (Eds.), The Sub Seafloor
2016.08.028. Biosphere at Mid-ocean Ridges. https://doi.org/10.1029/144GM08.
Becker, K., Davis, E.E., 1998. Leg 174B revisits hole 395A: logging and long-term monitor- Garfunkel, Z., Ben-Avraham, Z., Kagan, E., 2014. Dead Sea Transform Fault System: Re-
ing of off-axis hydrothermal processes in young oceanic crust. JOIDES J. 24, 1–13. views. Springer: Modern approaches in solid earth sciences, v.6.
Becker, K., Fisher, A.T., 2000. Permeability of upper oceanic basement on the eastern flank Ghanbarzadeh, S., Hesse, M.A., Prodanović, M., Gardner, J.E., 2015. Deformation-assisted
of the Juan de Fuca Ridge determined with drill-string packer experiments. J. Geo- fluid percolation in rock salt. Science 350, 1069–1072.
RR

phys. Res. Solid Earth 105, 897–912. Gradmann, S., Hübscher, C., Ben-Avraham, Z., Gajewski, D., Netzeband, G., 2005. Salt tec-
Bellissent-Funel, M.-C., 2001. Structure of supercritical water. J. Mol. Liq. 90, 313–322. tonics off northern Israel. Mar. Petrol. Geol. 22 (5), 597–611.
Bird, P., 2003. An updated digital model of plate boundaries. Geochem. Geophys.Geosys. Graedel, T.E., Keene, W., 1996. The budget and cycle of Earth's natural chlorine. Pure
4, 1027. Appl. Chem. 68, 1689–1697.
Bischoff, J.L., Rosenbauer, R.J., 1989. Salinity variations in submarine hydrothermal sys- Gruen, G., Weis, P., Driesner, T., Heinrich, C.A., de Ronde, C.E., 2014. Hydrodynamic
tems by layered double-diffusive convection. J. Geol. 97, 613–623. modeling of magmatic–hydrothermal activity at submarine arc volcanoes, with impli-
Bosworth, W., 2015. Geological evolution of the Red Sea: historical background, review, cations for ore formation. Earth Planet Sci. Lett. 404, 307–318.
and synthesis. In: Rasul, N.M.A., Stewart, I.C.F. (Eds.), The Red Sea: the Forma- Gutscher, M.A., Spakman, W., Bijwaard, H., Engdahl, E.R., 2000. Geodynamics of flat sub-
CO

tion, Morphology, Oceanography and Environment of a Young Ocean Basin. Springer, duction: seismicity and tomographic constraints from the Andean margin. Tectonics
Berlin Heidelberg, pp. 45–78. 19, 814–833.
Braitsch, O., 1971. Salt Deposits - Their Origin and Composition. Vol. 4 of ‘Minerals, Rocks Hannington, M., Herzig, P., Stoffers, P., Scholten, J., Botz, R., Garbe-Schönberg, D., Jonas-
and Inorganic Materials’, 297 pp Springer-Verlag, Berlin, Heidelberg, New York. son, I., Roest, W., 2001. First observations of high-temperature submarine hydrother-
Buiter, S., 2016. Geodynamic models highlighting the roles of inheritance during Wilson mal vents and massive anhydrite deposits off the north coast of Iceland. Mar. Geol.
Cycle. In: Abstract. Proceedings of the Arthur Holmes Meeting. Geology Society of 177, 199–220.
London, London, p. 26, May 23-25, 2016. Hardie, L.A., 1990. The roles of rifting and hydrothermal CaCl2⁠ brines in the origin of
Burke, K., 1975. Atlantic evaporites formed by evaporation of water spilled from Pacific, potash evaporites: a hypothesis. Am. J. Sci. 290, 43–106.
Tethyan, and Southern oceans. Geology 3, 613–616. Harlov, D.E., Van Den Kerkhof, A., Johansson, L., 2012. The Varberg–Torpa charnock-
UN

Burke, K., 2011. Plate tectonics, the Wilson cycle, and mantle plumes: geodynamics ite–granite association, SW Sweden: mineralogy, petrology, and fluid inclusion chem-
from the top. Annu. Rev. Earth Planet Sci. 39, 1–29. https://doi.org/10.1146/ istry. J. Petrol. 54, 3–40.
annurev-earth-040809-152521. Hayashi, H., Hakuta, Y., 2010. Hydrothermal synthesis of metal oxide nanoparticles in su-
Burke, K., Ashwal, L.D., Webb, S.J., 2003. New way to map old sutures using deformed percritical water. Materials 3, 3794–3817.
alkaline rocks and carbonatites. Geology 31, 391–394. Heine, C., Zoethout, J., Müller, R.D., 2013. Kinematics of The south atlantic rift. Solid
Burke, K., Dewey, J., Kidd, W., 1977. World distribution of sutures—the sites of former Earth 4, 215–253.
oceans. Tectonophysics 40, 69–99. Holness, M.B., Lewis, S., 1997. The structure of the halite-brine interface inferred from
Burke, K., 2016. The Wilson Cycle of opening and closing of the ocean basins. In: Abstract. pressure and temperature variations of equilibrium dihedral angles in the
Proceedings of the Arthur Holmes Meeting. Geology Society of London, London, p. 8, halite-H2O-CO2 system. Geochem. Cosmochim. Acta 61, 795–804.
May 23-25, 2016. Hovland, M., Fichler, C., Rueslåtten, H., Johnsen, H.K., 2006a. Deep rooted piercement
Butterfield, D.A., 2000. Deep ocean hydrothermal vents. In: Sigurdsson, H., Houghton, structures in deep sedimentary basins—manifestations of supercritical water genera-
B.F., McNutt, S.R., Rymer, H., Stix, J. (Eds.), Encyclopedia of Volcanoes. Academic
tion at depth. J. Geochem. Explor. 89, 157–160.
Press, San Diego, pp. 857–875.
Hovland, M., Kuznetsova, T., Rueslåtten, H., Kvamme, B., Johnsen, H.K., Fladmark, G.E.,
Charnock, H., 1964. Anomalous bottom water in the red sea. Nature 203, 591, 591.
Hebach, A., 2006b. Subsurface precipitation of salts in supercritical seawater. Basin
Res. 18, 221–230.

18
M. Hovland et al. Marine and Petroleum Geology xxx (2017) xxx-xxx

Hovland, M., Rueslåtten, H.G., Johnsen, H.K., Kvamme, B., Kuznetsova, T., 2006c. Salt for- O'Reilly, C., Jenkin, G., Feely, M., Alderton, D., Fallick, A., 1997. A fluid inclusion and
mation associated with subsurface boiling and supercritical water. Mar. Petrol. Geol. stable isotope study of 200 Ma of fluid evolution in the Galway Granite, Connemara,
23, 855–869. Ireland. Contrib. Mineral. Petrol. 129, 120–142.
Hovland, M., Rueslåtten, H., Johnsen, H.K., 2014. Buried hydrothermal systems: the po- Pan, D., Spanu, L., Harrison, B., Sverjensky, D.A., Galli, G., 2013. Dielectric properties of
tential role of supercritical water,“ScriW”, in various geological processes and occur- water under extreme conditions and transport of carbonates in the deep Earth. Proc.
rences in the sub-surface. Am. J. Anal. Chem. 5, 128–139. Natl. Acad. Sci. Unit. States Am. 110, 6646–6650.
Hovland, M., Rueslåtten, H., Johnsen, H.K., 2015. Red Sea salt formations - a result of hy- Pearce, J., 2002. The oceanic lithosphere, achievements and opportunities of scientific
drothermal processes. In: Rasul, N.M.A., Stewart, I.C.F. (Eds.), The Red Sea. Springer ocean drilling. JOIDES 28, 61–66.

F
Earth System Sciences, Berlin, Heidelberg, pp. 187–203. Piper, J.D.A., 2013. A planetary perspective on earth evolution: lid tectonics before Plate
Hovland, M., Rueslåtten, H., Johnsen, H.K., Manuella, F., 2016. Possible role of salt accu- tectonics. Tectonophysics 589, 44–56.
Pirajno, F., 2009. Hydrothermal Processes and Mineral Systems. Geological Survey of
mualtions in Wilson cycles. In: (abstract) Proceedings of the Arthur Holmes Meeting,
Western Australia Springer.

OO
2016. Geol. Soc., London, May 23-25.
Pitman, W.C., Talwani, M., 1972. Sea-floor spreading in the north atlantic. Geol. Soc. Am.
Hsü, K.J., Cita, M.B., Ryan, W.B.F., 1973a. The Origin of the Mediterranean Evaporite. Ini-
Bull. 83, 619–646.
tial Reports of the Deep Sea Drilling Project. Government Printing Office, 1203–1231,
Rabinowitz, N., Mart, Y., 1999. Seismic tomography of the Dead Sea region: thinned crust,
V. 13.
anomalous velocities and possible magmatic diapirism. Geol. Soc. Lond. Spec. Publ.
Hsü, K., Ryan, W., Cita, M., 1973b. Late Miocene desiccation of the mediterranean. Nature
174, 79–92.
242, 240–244. Ryan, W.B.F., Carbotte, S.M., Coplan, J.O., O'Hara, S., Melkonian, A., Arko, R., Weissel,
Hsü, K.J., Montadert, L., Bernoulli, D., Cita, M.B., Erickson, A., Garrison, R.E., Kidd, R.B., R.A., Ferrini, V., Goodwillie, A., Nitsche, F., Bonczkowski, J., Zemsky, R., 2009. Global
Mèlierés, F., Müller, C., Wright, R., 1977. History of the Mediterranean salinity crisis. Multi-resolution topography synthesis. Geochem. Geophys. Geosyst. 10, Q03014.
Nature 267, 399–403. https://doi.org/10.1029/2008GC002332.
Huguen, C., Chamot-Rooke, N., Loubrieu, B., Mascle, J., 2006. Morphology of a pre-colli- Ryan, W.B.F., Hsü, K.J., et al., 1973. Initial Report Deep Sea Drilling Project, vol. 13, U.S.

PR
sional, salt-bearing, accretionary complex: the Mediterranean Ridge (eastern mediter- Government Printing Office, Washington.
ranean). Mar. Geophys. Res. 27, 61–75. Schofield, N., Alsop, I., Warren, J., Underhill, J., Lehné, R., Beer, W., Lukas, V., 2014. Mo-
Ildefonse, B., Blackman, D., John, B., Ohara, Y., Miller, D., MacLeod, C.J., 2007. Oceanic bilizing salt: magma-salt interactions. Geology https://doi.org/10.1130/G35406.
core complexes and crustal accretion at slow-spreading ridges. Geology 35, 623–626. Schreiber, B.C., Lugli, S., Babel, M. (Eds.), 2007. Evaporites through Space and Time, vol.
Judd, A.G., Hovland, M., 2007. Submarine Fluid Flow, the Impact on Geology, Biology, 285, Geological Society of London, London, pp. 1–13, Special Publication.
and the Marine Environment. Cambridge University Press, 475 pp. Schoenherr, J., Littke, R., Urai, J.L., Kukla, P.A., Rawahi, Z., 2007a. Polyphase thermal
Kendrick, M., Hémond, C., Kamenetsky, V., Danyushevsky, L., Devey, C.W., Rodemann, T., evolution in the infra-Cambrian Ara Group (South Oman Salt Basin) as deduced by
Jackson, M., Perfit, M., 2017. Seawater cycled throughout Earth/'s mantle in partially maturity of solid reservoir bitumen. Org. Geochem. 38, 1293–1318.
serpentinized lithosphere. Nat. Geosci. 10, 222–228. Schoenherr, J., Urai, J.L., Kukla, P.A., Littke, R., Schléder, Z., Larroque, J.-M., Newall,
Kovalenko, V.I., Naumov, V.B., Girnis, A.V., Dorofeeva, V.A., Yarmolyuk, V.V., 2006. Es-

D
M.,J., Al-Abry, N., Al-Siyabi, H., Rawahi, Z., 2007b. Limits to the sealing capacity of
timation of the average contents of H2O, Cl, F, and S in the depleted mantle on the rock salt: a case study of the infra-Cambrian Ara salt from the south Oman salt basin.
basis of the compositions of melt inclusions and quenched glasses of mid-ocean ridge Am. Assoc. Petrol. Geol. Bull. 91 (11), 1541–1557.
basalts. Geochem. Int. 44, 209–231. Scott, S., Driesner, T., Weis, P., 2017. Boiling and condensation of saline geothermal fluids
Laumonier, M., Gaillard, F., Muir, D., Blundy, J., Unsworth, M., 2017. Giant magmatic above magmatic intrusions. Geophys. Res. Lett. 44, 1696–1705.
water reservoirs at mid-crustal depth inferred from electrical conductivity and the
TE
Scribano, V., Carbone, S., Manuella, F.C., Hovland, M., Rueslåtten, H., Johnsen, H.-K.,
growth of the continental crust. Earth Planet Sci. Lett. 457, 173–180.
2017. Origin of salt giants in abyssal serpentinite systems. Int. J. Earth Sci. 1–14.
Lecumberri-Sanchez, P., Steele-MacInnis, M., Weis, P., Driesner, T., Bodnar, R.J., 2015.
Scribano, V., Viccaro, M., 2014. Sicily. Nicolosi (Catania) 29 Miscellanea INGV ISSN
Salt precipitation in magmatic-hydrothermal systems associated with upper crustal
2039-6651 En-route Formation of Highly Silica-undersaturated Melts through Interac-
plutons. Geology 43, 1063–1066.
tion between Ascending Basalt and Serpentinite-related Saline Brines: Inference from
Levitte, D., Greitzer, Y., 2000. Geothermal update report from Israel 1999. In: Proc. World
Hyblean Cenozoic Nephelinites, vol. 25, 107.
Geothermal Congress.
Searle, R.C., Ross, D.A., 1975. A geophysical study of the Red Sea axial trough between
Lewis, S., Holness, M., 1996. Equilibrium halite-H2O dihedral angles: high rock-salt per-
20.5 and 22 N. Geophys. J. Int. 43, 555–572.
EC

meability in the shallow crust?. Geology 24, 431–434.


Selley, R.C., 2005. Mineralogy and classification. In: Selley, R.C., Cocks, L.R.M., Plimer,
Lide, D.,R., 1990. CRC Handbook of Chemistry and Physics, 71th Edition CRC Press, Boca
I.R. (Eds.), Encyclopedia of Geology. Elsevier, Amsterdam, pp. 27–37.
Raton.
Sharp, Z., Barnes, J., 2004. Water-soluble chlorides in massive seafloor serpentinites: a
Liebscher, A., Heinrich, C.,A. (Eds.), 2007. Fluid-fluid Interactions. Reviews in Mineralogy
source of chloride in subduction zones. Earth Planet Sci. Lett. 226, 243–254.
and Geochemistry, vol. 65.
Snow, J.E., Edmonds, H.N., 2007. Ultraslow-spreading ridges rapid paradigm changes.
Liou, J.G., Tsujimori, T., Yang, J., Zhang, R., Ernst, W., 2014. Recycling of crustal materi-
Oceanography 20, 90–101.
als through study of ultrahigh-pressure minerals in collisional orogens, ophiolites, and
Sobolev, A., Hofmann, A.W., Jochum, K.P., Kuzmin, D.V., Stoll, B., 2011. The age of sub-
mantle xenoliths: a review. J. Asian Earth Sci. 96, 386–420.
ducted components in the source of Hawaiian plume. Nature 476.
Lonsdale, P., 1977. Clustering of suspension-feeding macrobenthos near abyssal hy-
RR

Sozansky, V., 1973. Origin of salt deposits in deep-water basins of Atlantic Ocean. AAPG
drothermal vents at oceanic spreading centers. Deep Sea Res. 24,
Bull. 57, 589–590.
857IN3859–3858IN4863.
Srikantappa, C., Narasimha, K.P., Basavarajappa, H., 1992. Highly saline fluid inclusions
Lugli, S., Manzi, V., Roveri, M., Schreiber, B.C., 2015. The deep record of the Messinian
in Chamundi granite, South India. Curr. Sci. 307–309.
salinity crisis: evidence of a non-desiccated Mediterranean Sea. Palaeogeogr. Palaeo-
Stroncik, N.A., Haase, K.M., 2004. Chlorine in oceanic intraplate basalts: constraints on
climatol. Palaeoecol. 433, 201–218.
mantle sources and recycling processes. Geology 32, 945–948.
Malvoisin, B., Brantut, N., Kaczmarek, M.A., 2017. Control of serpentinisation rate by re-
Svensen, H., Jamtveit, B., Banks, D.A., Austrheim, H., 2001. Halogen contents of eclog-
action-induced cracking. Earth Planet Sci. Lett. 476, 143–152.
ite facies fluid inclusions and minerals: caledonides, western Norway. J. Metamorph.
Martin, B., Fyfe, W.S., 1970. Some experimental and theoretical observations on the kinet-
Geol. 19, 165–178.
CO

ics of hydration reactions with particular reference to serpentinization. Chem. Geol. 6,


Svensen, H.H., Torsvik, T.H., Callegaro, S., Augland, L., Heimdal, T.H., Jerram, D.A.,
185–202.
Planke, S., Pereira, E., 2017. Gondwana large igneous provinces: plate reconstruc-
Manning, C.E., 2004. The chemistry of subduction-zone fluids. Earth Planet Sci. Lett. 223,
tions, volcanic basins and sill volumes. In: In: Sensarma, S., Storey, B.C. (Eds.), Large
1–16.
Igneous Provinces from Gondwana and Adjacent Regions, vol. 463, Geology Society
Manuella, F.C., Ottolini, L., Carbone, S., Scavo, L., 2016. Metasomatizing effects of serpen-
of London, Special Publication.
tinization-related hydrothermal fluids in abyssal peridotites: new contributions from
Talbot, C.J., 2007. Hydrothermal salt—but how much?. Mar. Petrol. Geol. 25, 191–202.
Hyblean peridotite xenoliths (southeastern Sicily). Lithos 264, 405–421.
Talbot, C.J., Stanley, W., Soub, R., Al-Sadoun, N., 1996. Epitaxial salt reefs and mush-
Marton, L.G., Tari, G.C., Lehmann, C.T., 2000. Evolution of the Angolan passive margin,
rooms in the southern Dead Sea. Sedimentology 43, 1025–1047.
West Africa, with emphasis on post-salt structural styles. In: In: Mohriak, W., Talwani,
Talwani, M., Eldholm, O., 1977. Evolution of the Norwegian-Greenland sea. Geol. Soc.
UN

M. (Eds.), Atlantic Rifts and Continental Margins, vol. 115, American Geophysical
Am. Bull. 88, 969–999.
Union Geophysical Monograph, pp. 129–149.
Taylor, S.R., McLennan, M., 1985. The Continental Crust: its Composition and Evolution.
Mével, C., 2003. Serpentinization of abyssal peridotites at mid-ocean ridges. Compt. Ren-
Blackwell Scientific Publications, UnitedStates, 1–328.
dus Geosci. 335, 825–852.
Tester, J.W., Holgate, H.R., Armellini, F.J., Webley, P.A., Killilea, W.R., Hong, G.T.,
Miranda, E.A., Dilek, Y., 2010. Oceanic core complex development in modern and ancient
Barner, H.E., 1993. Supercritical water oxidation technology. In: In: Tedder, D.W.,
oceanic lithosphere: gabbro-localized versus peridotite-localized detachment models.
Pohland, F.G. (Eds.), Emerging Technologies in Hazardous Waste Management III, vol.
J. Geol. 118, 95–109.
518, American Chemical Society, pp. 35–76.
Mohriak, W.U., Szatmari, P., Anjos, S., 2012. Salt: geology and tectonics of selected Brazil-
ten Brink, U.S., Al-Zoubi, A.S., Flores, C.H., Rotstein, Y., Qabbani, I., Harder, S.H., Keller,
ian basins in their global context. Geol. Soc. Lond. Special Publ. 363, 131–158.
G.R., 2006. Seismic imaging of deep low-velocity zone beneath the Dead Sea basin
Momenzadeh, M., 1990. Saline deposits and alkaline magmatism: a genetic model. J.
and transform fault: implications for strain localization and crustal rigidity. Geophys.
Petrol. Geol. 13, 341–356.
Res. Lett. 33, L24314. https://doi.org/10.1029/2006GL027890.
Norton, I.O., Carruthers, D.T., Hudec, M.R., 2016. Rift to drift transition in the South At-
Torsvik, T.H., Burke, K., Steinberger, B., Webb, S.J., Ashwal, L.D., 2010. Diamonds sam-
lantic salt basins: a new flavor of oceanic crust. Geology 44, 55–58.
pled by plumes from the core-mantle boundary. Nature 466, 352–355.
Nunn, J.A., Harris, N.B., 2007. Subsurface seepage of seawater across a barrier: a source
Torsvik, T.H., Rousse, S., Labails, C., Smethurst, M.A., 2009. A new scheme for the open-
of water and salt to peripheral salt basins. Geol. Soc. Am. Bull. 119, 1201–1217.
ing of the South Atlantic Ocean and the dissection of an Aptian salt basin. Geophys. J.
Int. 177, 1315–1333.

19
M. Hovland et al. Marine and Petroleum Geology xxx (2017) xxx-xxx

van den Belt, F.J.G., De Boer, P.L., 2007. A shallow-basin model for saline giants based Weinberger, R., Begin, Z.B., Waldmann, N., Gardosh, M., Baer, G., Frumkin, A., Wdowin-
on isostasy-driven subsidence. In: Paola, C., Nichols, G., Williams, E. (Eds.), Sedimen- ski, S., 2006. Quaternary rise of the sedom diapir, Dead Sea Basin. In: Enzel, Y.,
tary Processes, Environments and Basins: a Tribute to Peter Friend. Wiley Blackwell, Agnon, A., Stein, M. (Eds.), New Frontiers in Dead Sea Paleoenvironmental Re-
Malden, pp. 241–252, International Association of Sedimentology, Special Publica- search. pp. 33–51, Geol. Soc. America Special Paper 401 https://doi.org/10.1130/
tion. 2006.2401(03).
Vogel, M., Früh-Green, G.L., Boschi, C., Schwarzenbach, E.M., 2014. Serpentinization and Westbrook, G.K., Reston, T.J., 2002. The accretionary complex of the Mediterranean
fluid-rock interaction in Jurassic mafic and ultramafic sea-floor: constraints from Lig- Ridge: tectonics, fluid flow and the formation of brine lakes–an introduction to the
urian ophiolite sequences. In: EGU General Assembly Conference Abstracts, vol. 16. special issue of Marine Geology. Mar. Geol. 186, 1–8.

F
Wallace, P., Anderson Jr., A.T., 2000. Volatiles in magmas. In: Sigurdsson, H., Haughton, Wilson, J.T., 1966. Did the Atlantic close and then re-open?. Nature 211, 676–681.
B., McNutt, S.R., Rymer, H., Stix, J. (Eds.), The Encyclopedia of Volcanoes. Academic Wilson, J.T., 1968. Static or mobile earth: the current scientific revolution. Proc. Am. Phil.
Press, San Diego, pp. 149–170. Soc. 112, 309–320.
Wallmann, K., Aghib, F., Castradori, D., Cita, M., Suess, E., Greinert, J., Rickert, D., 2002. Winckler, G., Aeschbach-Hertig, W., Kipfer, R., Botz, R., Rübel, A.P., Bayer, R., Stoffers, P.,

OO
Sedimentation and formation of secondary minerals in the hypersaline Discovery 2001. Constraints on origin and evolution of Red Sea brines from helium and argon
Basin, eastern Mediterranean. Mar. Geol. 186, 9–28. isotopes. Earth Planet Sci. Lett. 184, 671–683.
Warren, J.K., 1999. Evaporites: Their Evolution and Economics. Blackwell Science, Ox- Woodside, J.M., Shipboard Scientists of the MEDINAUT/MEDINETH Projects, 2001. Nau-
ford. tile observations of Eastern Mediterranean mud volcanoes and gas seeps – results from
Warren, J.K., 2006. Evaporites: Sediments, Resources and Hydrocarbons. Springer, the MEDINAUT and MEDINETH projects. In: Akhamanov, G., Suzyumov, A. (Eds.),
Würzburg. Geological Processes on Deep-water European Margins. UNESCO, Paris, pp. 39–40,
Warren, J.K., 2010. Evaporites through time: tectonic, climatic and eustatic controls in International Oceanographic Commission Workshop Report No 175 on the Interna-
marine and nonmarine deposits. Earth Sci. Rev. 98, 217–268. tional Conference and ninth post-cruise meeting of the Training Through Research
Warren, J.K., 2016. Evaporites: a Compendium (USBN 978-3-319-13511-3). Springer, Programme, Moscow-Mozhenka, Russia, 28 January – 2 February 2001. Intergovern-

PR
Berlin, Heidelberg. mental Oceanographic Commission Workshop Report, 175.
Warren, J.K., 2017. Salt usually seals, but sometimes leaks: implications for mine and cav- Xiao, Y.-L., Hoefs, J., van den Kerkhof, A.M., Zheng, Y.-F., 1998. Fluid Inclusions in Ultra
ern stabilities in the short and long term. Earth Sci. Rev. 165, 302–341. High-pressure Eclogites from the Dabie Shan, Eastern China. 1667–1668, Godschmidt
Wegener, A., 1915. Die Entstehung der Kontinente und Ozeane, first ed. Friedrich Vieweg conference Toulouse.
& Sohn, Braunschweig. Zellmer, G.F., Edmonds, M., Straub, S.M., 2015. The role of volatiles in the genesis, evolu-
Wegener, A., 1922. Die Einstehung der Kontinente und Oceane, third ed., Methuen, Lon- tion and eruption of arc magmas. Geol. Soc. Lond. Special Publ. 410.
don.

D
TE
EC
RR
CO
UN

20

S-ar putea să vă placă și