Salty Matters

The Blog is written by me, John Warren. Once every three or four weeks or so I will post an article or two on an evaporite topic that has piqued my interest. On the Saltwork Publications webpage (under "the Works") there is a growing library of pdfs and epubs based on these blogs. These articles on the website have much higher resolution extractable graphics in than in the blog. There is also a link to this set of pdfs and epubs on the home page (

Red Sea metals: what is the role of salt in metal enrichment?

John Warren - Friday, April 29, 2016



Work over the past four decades has shown many sediment-hosted stratiform copper deposits are closely allied with evaporite occurrences or indicators of former evaporites, as are some SedEx (Sedimentary Exhalative) and MVT (Mississippi Valley Type) deposits (Warren, 2016). Some ore deposits, especially those that have evolved beyond greenschist facies, can retain the actual salts responsible for the association, primarily anhydrite relics, in proximity to the ore. Such deposits include the Zambian and Redstone copper belts, Creta, Boleo, Corocoro, Dzhezkazgan, Kupferschiefer (Lubin and Mansfeld regions), Largentière and the Mt Isa copper association. All these accumulations of base metals are associated with the formation of a burial-diagenetic hypersaline redox/mixing front, where either copper or Pb-Zn sulphides tended to accumulate. Mechanisms that concentrate and precipitate base metal ores in this evaporite, typically halokinetic, milieu are the topic of upcoming blogs. Then there are deposits that are the result from hot brine fluids, tied to dissolving evaporites and igneous activity, mixing and cooling with seawater, so precipitating a variety of hydrothermal salts, sometimes in including economic levels of copper, lead and zinc (Warren, 2016)

In this article, I focus on one such hypersaline-brine deposit, the cupriferous hydrothermal laminites of the Atlantis II Deep in the Red Sea and look at the role of evaporites in the enrichment of metals in this deposit. It is a modern example of a metalliferous laminite forming in a brine lake sump on the deep seafloor where the brine lake and the stabilisation of the precipitation interface is a result of the dissolution of adjacent halokinetic salt masses. Most economic geologists classify the metalliferous Red Sea deeps as SedEx deposits, but the low levels of lead and high levels of copper, along with its stratigraphic position atop seafloor basalts, place it outside the usual Pb-Zn dominant system that typifies ancient SedEx deposits. Some economic geologists use the Red Sea deeps as analogues for volcanic massive sulphides, and some argue it even illustrates aspects of some stratiform Cu accumulations. Many such economic geology studies have the propensity to ignore the elephant in the room; that is the Red Sea deeps are the result of brine focusing by a large Tertiary-age halokinetically-plumbed seafloor brine association. This helps explain the large volume of metals compared to Cyprus-style and mid-ocean ridge volcanic massive sulphides (Warren 2016, Chapters 15 and16).

In my mind what is most important about the brine lakes on the deep seafloor of the Red Sea is the fact that they exist with such large lateral extents only because of dissolution of the hosting halokinetic slope and rise salt mass. Seismic surveys conducted in the past decade in the Red Sea show extensive salt flows (submarine salt glaciers) along the whole of the Red Sea Rift (at least from 19–23°N; Augustin et al., 2014; Feldens and Mitchell, 2015)). In places, these salt sheets flow into and completely blanket the axial region of the rift. Where not covered by namakiers, the seafloor comprises volcanic terrain characteristic of a mid-ocean spreading axis. In the salt-covered areas, evidence from bathymetry, volume-balance of the salt flows, and geophysical data all seems to support the conclusion that the sub-salt basement is mostly basaltic in nature and represents oceanic crust (Augustin et al., 2014).


The Rift

The Red Sea, located between Egypt and Saudi Arabia, represents a young active rift system that from north to south transitions from continental to oceanic rift (Rasul and Stewart, 2015). It is one of the youngest marine zones on Earth, propelled by an area of relatively slow seafloor spreading (≈1.6 cm/year). Together with the Gulf of Aqaba-Dead Sea transform fault, it forms the western boundary of the Arabian plate, which is moving in a north-easterly direction (Figure 1; Stern and Johnson, 2010). The plate is bounded by the Bitlis Suture and the Zagros fold belt and subduction zone to the north and north-east, and the Gulf of Aden spreading center and Owen Fracture Zone to the south and southeast. The Red Sea first formed about 25 Ma ago in response to crustal extension related to the interface movements of the African Plate, the Sinai Plate, and the Arabian Plate (Schardt, 2016). The present site of Red Sea rifting is controlled, or largely overprinting, on pre-existing structures in the crust, such as the Central African Fault Zone. In the area between 15° and 20° along the rift axis, active seafloor spreading is prominent and is characterized by the formation of oceanic crust with Mid-Ocean Ridge Basalt (MORB) composition for the last 3 Ma (Rasul and Stewart, 2015). In contrast, the northern portion of the Red Sea sits in a magmatic continental rift in which a mid-ocean ridge spreading centre is just beginning to form. That is, the split in the crust that is the Red Sea is unzipping from south to north (Figure 1).

The Salt

The rift basement is covered a thick sequence of middle Miocene evaporites that precipitated in the earlier hydrographically isolated stage of rifting (Badenian – Middle Miocene). The maximum thickness of rift-fill sediments, including halokinetic salt, is around 8,000 m in the Morgan basin in the southern Red Sea (Farhoud, 2009; Ehrhardt et al., 2005). Girdler and Southren (1987) conclude that Miocene evaporites first accumulated on Red Sea transitional crust but must have later flowed downdip to now cover parts of the axial zone (basaltic) of the Plio-Pleistocene oceanic crust. At latitudes of 20° to 23° N, transform fracture zones provide focused passage-ways for salt flow. They also enable the involvement of dissolving salt in axial hydrothermal circulation, so producing pools of dense hot brines and the topographic isolation of spreading segments into evaporite-enclosed deeps (Feldens and Mitchell, 2015). So today, flow-like features cored by Miocene evaporites are situated along the axis of the Red Sea atop younger magnetic seafloor spreading anomalies. However, not all brine seeps occur in or near the deep axis of the Red Sea on the downdip edge of flowing Miocene salt, some occur in much shallower suprasalt positions nearer the coastal margins of the Red Sea, in waters just down dip of actively-growing well-lit coral reefs (Batang et al., 2012).

Six salt flows, most showing rounded fronts in plan-view, with heights of several hundred meters and widths between 3 and 10 km, are seen in high-resolution bathymetry and DSDP core material around Thetis Deep and Atlantis II Deep, and between Atlantis II Deep and Port Sudan Deep (Figure 2; Feldens and Mitchell, 2015; Mitchell et al., 2010). Relief on the underlying volcanic basement surface likely controls the positions of individual salt flow lobes. On the flow surfaces, along-slope and downslope ridge and trough morphologies have developed parallel to the local seafloor gradient, presumably due to the extension of the hemiplegic sediment cover or strike-slip movement within the evaporites.

Some sites with irregular seafloor topography are observed close to the flow fronts, interpreted to be the result of dissolution of Miocene evaporites, which contributes to the formation of brine lakes in several of the endorheic deeps (Feldens and Mitchell, 2015). Based on the vertical relief of the flow lobes, deformation is still taking place in the upper part of the evaporite sequence. Considering the salt flow that creates the Atlantis II Deep in more detail, strain rates due to dislocation creep and pressure solution creep are estimated to be 10−14 sec-1 and 10−10 sec-1, respectively, using given assumptions of grain size and deforming layer thickness (Feldens and Mitchell, 2015). The latter strain rate is comparable to strain rates observed for onshore salt flows in Iran and signifies flow speeds of several mm/year for some offshore salt flows. Thus, salt flow movements can potentially keep up with Arabia–Nubia tectonic half-spreading rates across large parts of the Red Sea (Figure 1)

The Deeps

Beneath waters more than a kilometre deep, along the deep rift axis, there are 26 brine pools and deeps, some of which are underlain by metalliferous sediments (Figure 3; Blanc and Anschutz 1995, Blum and Puchelt, 1991). Because of varying size, age, and formation history between the various deeps, Ehrhardt and Hübscher (2015) discriminate between central and northern Red Sea deeps. The larger central Red Sea deeps are located in the axial trough and are separated by inter-trough zones. They are floored by young basaltic crust and exhibit magnetic anomalies not older than 1.7 Ma. The northern Red Sea deeps are smaller and form only isolated deeps within the axial depression. Some of them are accompanied by volcanic activity. Many of the central Red Sea deeps contain bottom-water brines and metalliferous sediments, pointing to hydrothermal circulation of seawater (Schmidt et al., 2015). The largest and most prominent deep is the Atlantis II Deep, located in the central part of the Red Sea in the vicinity of other large deeps such as the Chain Deep and Discovery Deep. Other prominent deeps are the Tethys and Nereus Deeps further north, but still in the central part of the Red Sea.

Historically, the various deeps along the Red Sea rift axis are deemed to be initial seafloor spreading cells that will accrete sometime in the future into a continuous spreading axis. Northern Red Sea deeps are isolated structures often associated with single volcanic edifices in comparison to the further-developed and larger central Red Sea deeps where small spreading ridges are locally active (Ehrhardt and Hübscher, 2015). But not all deeps are related to initial seafloor spreading cells, and there are two types of ocean deeps: (a) volcanic and tectonically impacted deeps that opened by a lateral tear of the Miocene evaporites (salt) and Plio-Quaternary overburden; (b) non-volcanic deeps built by subsidence of Plio-Quaternary sediments due to evaporite subrosion (dissolution) processes. Type b) deeps develop as evaporite collapse structures (Figure 4: Ehrhardt and Hübscher, 2015). In contrast, the type (a) volcanic deeps can be correlated with their positions in NW–SE-oriented segments of the Red Sea, which are daylighted volcanic segments. The N–S segments, between these volcanically active NW–SE segments, is called a “non-volcanic segment” as no volcanic activity is known, in agreement with the magnetic data that shows no major anomalies. Accordingly, the deeps in the "nonvolcanic segments" are evaporite collapse-related structures creating discontinuities and brine breakout zones in and atop the salt sheets without the need for a seafloor spreading cell.

Such evaporite collapse-type ocean deeps are not limited to the non-volcanic segments, as subrosion processes driven by upwells in hydrothermal circulation are possible at any part of the axial depression, especially along fault damage zones. The combined interpretation of bathymetry and seismic reflection profiles gives further insight into the nature of lateral salt gliding in the Red Sea. Salt rises are typically present where the salt flows above basement faults. The internal reflection characteristic of the salt changes laterally from reflection-free to stratified, which suggests significant salt deformation during the salt deposition. Acoustically-transparent halite accumulated locally and evolving rim synclines were filled by stratified evaporite-related facies. (Figure 5)

Both types of deeps, as defined by Ehrhardt and Hübscher (2015), are surrounded by thick halokinetic masses of Miocene salt with brine chemistry in the bottom brine layer that signposts ongoing halite subrosion and dissolution. Red Sea deeps were discovered in the 1960s at a time when lateral translation of salt (gliding and spreading) and the formation of density stratification that define deepsea hypersaline anoxic lakes (DHALS) were not known (Warren, 2016). Today, with our knowledge of seeps and hypersaline seafloor depressions in halokinetic terranes on the slope and rise in the Gulf of Mexico and accretionary ridges in the parts of the Mediterranean Sea, we now know that the brine-filled deeps on the floor of the Red Sea are just another example of DHALs. What is most interesting in the chemical make-up Red Sea DHALS are the elevated levels of iron, copper and lead that occur in some deeps, especially the deepest and one of the most hypersaline set of linked depressions known as the Atlantis II deep (Figure 6).

Brine Chemistry in Red Sea DHALS

Most Red Sea deeps contain waters with somewhat elevated salinities, compared to normal seawater. Bulk chemistry of major ions in bottom brines from the various Red Sea DHALS are covariant and are derived by dissolution of the adjacent and underlying Miocene halite (Figure 7; replotted from Schmidt et al., 2015).

Mineralization in Red Sea DHALS

Economically, the most important brine pool is the Atlantis II Deep; other smaller deeps, with variable development of metalliferous muds and brine sumps, include; Commission Plain, Hatiba, Thetis, Nereus, Vema, Gypsum, Kebrit and Shaban Deeps (Figure 3; Chapter 15, Warren 2016). Laminites of the Atlantis II Deep are highly metalliferous, while the Kebrit and Shaban deeps are of metalliferous interest in that fragments of massive sulphide from hydrothermal chimney sulphides were recovered in bottom grab samples (Blum and Puchelt, 1991). All Red Sea DHALS are located in sumps along the spreading axis, in the region of the median valley. Most of these axial troughs and deeps are also located where transverse faults, inferred from bathymetric data, seismic, or from continuation of continental fracture lines, cross the median rift valley in regions that are also characterised by halokinetic Miocene salt. Not all Red Sea deeps are DHALS and not all Red Sea DHALS overlie metalliferous laminites.

The variably metalliferous seafloor deeps or deepsea hypersaline anoxic lakes (DHALs) in the deep water axial rift of the Red Sea define the metalliferous end of a spectrum of worldwide DHALs formed in response to sub-seafloor dissolution of shallowly-buried halokinetic salt masses. What makes the Red sea deeps unique is that they can host substantial amounts of metal sulphides, and, as Pierre et al. (2010) show, a Red Sea deep without the seafloor brine lake, is not significantly mineralised.

In my opinion, it is the intersection of the DHAL setting with an active to incipient midocean ridge (ultimate metal source), and a lack of sedimentation in the DHAL, other than hydrothermal precipitates (including widespread hydrothermal anhydrite), that explains the size and extent of the Atlantis II deposit. Its salt-dissolution-related brine hydrology, with a lack of detrital input, changes the typical mid-ocean massive-sulphide ridge deposit (with volumes usually around 300,000 and up to 3 million tonnes; Hannington et al., 2011) into a more stable brine-stratified bottom hydrology, which can fix metals over longer time and stability frames, so that the known sulphide accumulation in the Atlantis II Deep today has a metal reserve that exceeds 90 million tonnes.

The Red Sea DHAL evaporite-metal-volcanic association underlines why vanished evaporites are significant in the formation of giant and supergiant base metal deposits. Most thick subsurface evaporites in any tectonically-active metalliferous basin tend to flow and ultimately dissolve. Through their ongoing flow, dissolution and alteration, chloride- and sulphate-rich evaporites can create stable brine-interface conditions suitable for metal enrichment and entrapment. This takes place in subsurface settings ranging from the burial diagenetic through to the metamorphic and into igneous realms. An overview of a selection of the large-scale ore deposits associated with hypersaline brines tied to dissolving/altered and "vanished" salt masses, plotted on a topographic and salt basin base, shows that the majority of evaporite-associated ore deposits lie outside areas occupied by actual evaporite salts (Figure 8; see Warren Chapters 15 and 16 for detail). Rather, they tend to be located at or near the edges of a salt basin or in areas where most or all of the actual salts have long gone (typically via subsurface dissolution or metamorphic transformation). This widespread metal-evaporite association, and the enhancement in deposit size it creates, is not necessarily recognised as significant by geologists not familiar with the importance of "the salt that was." So evaporites, which across the Phanerozoic constitute less than 2% of the world's sediments, are intimately tied to (Warren, 2016):


  • All supergiant sediment-hosted copper deposits (halokinetic brine focus)
  • More than 50% of world’s giant SedEx deposits (halokinetic brine focus)
  • More than 80% of the giant MVT deposits (sulphate-fixer & brine)
  • The world's largest Phanerozoic Ni deposit
  • Many of the larger IOCG deposits (meta-evaporite, brine and hydrothermal)


Augustin, N., C. W. Devey, F. M. van der Zwan, P. Feldens, M. Tominaga, R. A. Bantan, and T. Kwasnitschka, 2014, The rifting to spreading transition in the Red Sea: Earth and Planetary Science Letters, v. 395, p. 217-230.

Batang, Z. B., E. Papathanassiou, A. Al-Suwailem, C. Smith, M. Salomidi, G. Petihakis, N. M. Alikunhi, L. Smith, F. Mallon, T. Yapici, and N. Fayad, 2012, First discovery of a cold seep on the continental margin of the central Red Sea: Journal of Marine Systems, v. 94, p. 247-253.

Blanc, G., and P. Anschutz, 1995, New stratification in the hydrothermal brine system of the Atlantis II Deep, Red Sea: Geology, v. 23, p. 543-546.

Blum, N., and H. Puchelt, 1991, Sedimentary-hosted polymetallic massive sulphide deposits of the Kebrit and Shaban Deeps, Red Sea.: Mineralium Deposita, v. 26, p. 217-227.

Ehrhardt, A., and C. Hübscher, 2015, The Northern Red Sea in Transition from Rifting to Drifting-Lessons Learned from Ocean Deeps, in N. M. A. Rasul, and I. C. F. Stewart, eds., The Red Sea: Berlin Heidelberg, Springer p. 99-121.

Ehrhardt, A., C. Hübscher, and D. Gajewski, 2005, Conrad Deep, Northern Red Sea: Development of an early stage ocean deep within the axial depression: Tectonophysics, v. 411, p. 19-40.

Farhoud, K., 2009, Accommodation zones and tectono-stratigraphy of the Gulf of Suez, Egypt: a contribution from aeromagnetic analysis: GeoArabia, v. 14, p. 139-162.

Feldens, P., and N. C. Mitchell, 2015, Salt Flows in the Central Red Sea, in N. M. A. Rasul, and I. C. F. Stewart, eds., The Red Sea: Springer Earth System Sciences: Berlin Heidelberg, Springer p. 205-218.

Girdler, R. W., and T. C. Southren, 1987, Structure and evolution of the northern Red Sea: Nature, v. 330, p. 716-721.

Hannington, M., J. Jamieson, T. Monecke, S. Petersen, and S. Beaulieu, 2011, The abundance of seafloor massive sulfide deposits: Geology, v. 39, p. 1155-1158.

Pierret, M. C., N. Clauer, D. Bosch, and G. Blanc, 2010, Formation of Thetis Deep metal-rich sediments in the absence of brines, Red Sea: Journal of Geochemical Exploration, v. 104, p. 12-26.

Rasul, N. M. A., and I. C. F. Stewart, 2015, The Red Sea: Springer Earth System Sciences, Springer, 638 p.

Rowan, M. G., 2014, Passive-margin salt basins: hyperextension, evaporite deposition, and salt tectonics: Basin Research, v. 26, p. 154-182.

Schardt, C., 2016, Hydrothermal fluid migration and brine pool formation in the Red Sea: the Atlantis II Deep: Mineralium Deposita, v. 51, p. 89-111.

Schmidt, M., R. Al-Farawati, and R. Botz, 2015, Geochemical Classification of Brine-Filled Red Sea Deeps, in N. M. A. Rasul, and I. C. F. Stewart, eds., The Red Sea: Berlin Heidelberg, Springer-Verlag, p. 219-233.

Stern, R. J., and P. R. Johnson, 2010, Continental lithosphere of the Arabian Plate: a geologic, petrologic, and geophysical synthesis: Earth Science Reviews, v. 101, p. 29-67.

Warren, J. K., 2016, Evaporites: A compendium (ISBN 978-3-319-13511-3) Released Feb. 2016: Berlin, Springer, 1854 p.







Lapis Lazuli: A metamorphosed evaporite

John Warren - Friday, November 13, 2015


Precious stones and [1]gems are rare by definition; hence need exceptional geologic conditions to give rise to gem-quality materials. A nexus across most natural gem-forming environments is the requirement for hydrous typically saline to hypersaline solutions, apt to precipitate euhedral crystals in a void or a pressure shadow, from fluids that contain elevated and unusual levels of particular constituents, including chromophores; hence pegmatites, volcanics and meta-evaporites are commonplace hosts for natural gemstones. Fluids promoting the growth of gem-quality crystals typically include the availability of uncommon major constituents, along with the presence of adequate chromophores [2] , as well limited concentrations of undesirable elements. Another need is open fluid space in an environment conducive to growing crystals of sufficient size and transparency. This general statement of requirements to form a precious stone also encapsulates why some gems have meta-evaporitic associations.

We know that depositional units of evaporite salts typically disappear or transform into other mineral phases by the early greenschist phase (Warren, 2016). As this happens the dissolution/transformation releases a pulse of hot basinal chloride waters that can carry gold and base metals (topics for another blog) as well as leaching and carrying elements such as beryllium, chromium and vanadium (chromophores) from adjacent organic-rich shales. Trace elements also tend to be enriched in the more evolved depositional brines that precipitate in later minerals in a primary evaporite precipitative series. Later, at the same time as halite dissolves or transforms on entry into the metamorphic realm, anhydrite layers and masses typically remain into the more evolved portions of the metamorphic realm (amphibolite-granulite facies). The volume loss associated with the dissolution/transformation of meta-evaporites facilitates the formation of open fluid space (sometimes pressurized) in veins and fractures so favoring sites that then allow the free growth of precious stones and euhedral ruby, tourmaline and emerald gemstones.

I am not saying all precious stones and gems are associated with evaporites, many natural gemstone settings are not, but the lapis lazuli of Afghanistan, the melon-sized perfect rubies of Myanmar, the prolific emerald fields in Columbia, and the tsavorite deposits of east Africa likely are (Garnier et al., 2008; Giuliani et al 2005; Feneyrol et al., 2013). In this article I will focus on lapis lazuli, for a discussion of other semi-precious stones and gems that are meta-evaporites see Chapter 14 in Warren (2016).


Lapis lazuli

Lapis Lazuli is the metamorphic remnant of a sodic-rich, quartz-absent, evaporite mineral assemblage. It is composed of an accumulation of minerals not a single mineral (unlike rubies and emeralds); it is mostly lazurite (Na, Ca)8(AlSiO4)6(S, SO4,Cl)1-2), typically at levels of 30-40%. Lapis gemstone also contains calcite (white veins), sodalite (blue), and pyrite (gold flecks of color). Dependent on metamorphic history and protolith chemistry, other common minerals in lapis include; augite, diopside, enstatite, mica, haüyanite, hornblende, and nosean. Some specimens also contain trace amounts of the sulphur-rich mineral lollingite (var. geyerite). Lazurite is a member of the sodalite group of feldspathoid minerals (Table 1). Feldspathoids have chemistries that are close to those of the alkali feldspars, but are poor in silica. If free quartz were present at the time of formation it would have reacted with any feldspathoid precursor to form feldspar not lazurite. Natural lazurite contains both sulphide and sulphate sulphur, in addition to calcium and sodium, and so is sometimes classified as a sulphide-bearing haüyne (Figure 1). Sulphur gives lazurite its characteristically intense blue color, which comes from three polysulphide units made up of three sulphur atoms having a single negative charge. The S3- ion in the sulphur has a total of 19 electrons in molecular orbitals and a transition among these orbitals produces a strong absorption band at 600 nm, giving a strong blue color, with yellow overtones. The intensity of the gem’s blue is increased with increasing sulphur and calcium content, while a green color is the result of insufficient sulphur (O’Donoghue, 2006, p. 329).


Other members of the sodalite group include sodalite and nosean (Table 1). Sodalite is the most sodium-rich member of the sodalite group and differs from the other minerals of the group in that its lattice retains chlorine. Interestingly, sodalite can be created in the laboratory by heating muscovite or kaolinite in the presence of NaCl at temperatures of 500°C or more. In the literature, the commonly accepted origin of lazurite is through contact metamorphism and metasomatism of dolomitic limestone. Such a metasedimentary system also requires a source of sodium, chlorine and sulphur; the obvious source is interbedded evaporites in the protolith, as is seen in plots of its molecular constituents (Aleksandrov and Senin, 2006).

Lapis lazuli from the Precambrian of Baffin Island, Canada (Figure 1), and from Edwards, New York, are meta-evaporites with evaporite remnants (anhydrites) remaining in the same series, as are the lapis lazuli deposits at Sar-e-Sang in the Kokcha valley, Afghanistan and the lapis deposits in Liadjuar-Dara region (“River of Lazurite”) at an altitude of 5000 m in the Pamir Mountains, Tajikistan (Webster, 1975). Throughout history its bright blue color has made lapis, mostly from Sar-e-Sang, a valued gem commodity. First mined 6000 years ago, the Sar-e-Sang lapis was transported to Egypt and present day Iraq and later to Europe where it was used in jewelry and for ornamental stone[3]. Europeans even ground down the rock into an expensive powdered pigment for paints called “ultramarine”.

Lapis deposits in Lake Harbor on Baffin Island and in the Edwards Mine, New York, were produced by high-grade metamorphism of a sulphate-halite-marble protolith (Hogarth and Griffin, 1978). The anhydrites preserved near Balmat are remnants of this sequence. On Baffin Island the two main lapis lazuli lenses, some 1.6 km apart, lie at the structural top of two sequences of dolomitic marble, the thicker lens being approximately 150 m across (Figure 1b). The elongation of both lenses parallels the local layering and foliation and shows a well-developed layering parallel to the regional foliation, giving additional evidence of its sedimentary protolith to the deposits. The Main and Northern bodies constitute diopside–lazurite rocks of variable gem quality and are localized in marbles among biotite gneisses. The Main (Southern) occurrence is as long as 170 m and 6 m thick. In these deposits, sheets of high-quality lazurite (up to 1 m thick) contain variable amounts of relict diopside and plagioclase, as well as newly formed haüyne, nepheline, or phlogopite. The quantitative proportions of these minerals define the color of the rock, which changes to a more intense blue with heating. The Northern occurrence (25×36 m in size) is less rich than the Main occurrence and consists of small (no more than 1 m) lenses showing disseminated lazurite, which imparts a bluish green color to the polished surface of the rock. Chlorine and sulphur in the various lazurites, accessory pyrite, and pyrrhotite were derived from metamorphosed gypsum-, anhydrite-, and evaporitic-carbonate protoliths (Hogarth and Griffin, 1978).

In the Lake Baikal lazurite occurrences, there is once again a strong association between marble of the Perval’na Group and lazurite occurrence (Figure 2a). For example, the Slyudyanka deposit is hosted in diopside skarns and spinel–forsterite calciphyres, developed from metamorphically-evolved evaporitic dolomites (Aleksandrov and Senin, 2006). The Slyudyanka deposit shows clearly pronounced metasomatic zoning, which was associated with the prograde magnesian skarn stage and was overprinted by retrograde postmagmatic assemblages, that formed together with lazurite-bearing rocks under the influence of saline alkaline S–Cl-bearing hydrothermal solutions. These solutions also caused microclinization of blocks of leucocratic granite with the formation of lazurite in the some of the inner skarn zones. Potassium solutions caused phlogopitization of the host rocks.

Likewise, scapolite and magnesian whiteschists are typically saline mineral phases in the classic deposits of the Sar-e-Sang District (Figure 2b; Faryad, 2002). There, the lapis is composed of a combination of lazurite, diopside, calcite and pyrite and occurs in beds and lenses up to 4 meters thick within a scapolitic magnesian-marble skarn near the center of the Hindu Kush granitic massif. It is typically interlayered with, or forms veins and lenses within a gneissic and pegmatitic host. Lens-shaped lodes are typically hosted in orthoclase–microcline–perthite hornfels containing albite and quartz (Figure 2b). Lazurite bodies at the Sar-e-Sang deposit are associated with diopside metasomatites bearing nepheline, pale blue haüyne, and blue lazurite, and some lazurite-rich zones can contain up to 40-90 vol% lazurite. The rocks also contain diopside, haüyne, afghanite, and nepheline, as well as disseminated pyrite replaced by pyrrhotite. Pockets of near pure lapis lazuli can be up to 40m across and occasionally up to a meter.

Lapis lazuli in the North Italian Mountains of Colorado occurs in impure marbles in a meta-evaporitic skarn near the contact with the Eocene-age quartz monzonite and quartz diorites of the Italian Mountain stocks (Hogarth and Griffin, 1980; Mauger, 2007). There, near vertical Pennsylvanian black shales and carbonates along the west margin of the intrusive have been converted to phlogopite-diopside-andalusite hornfels and scapolite-diopside skarns with minor analcime. Compared to Sar-e-Sang, lapis in this skarn deposit is of inferior quality. It forms as deep blue lazurite granules in fine-grained forsterite-Ti phlogopite-calcite skarns and calcite marbles with diopside, Ti-phlogopite and pyrite. The hosting sediments (Mississippian limestones and Devonian sandstones) define along the NE margin of the pluton, while the NaCl came from dissolution of once nearby halite or dissolution-derived saline surface waters and shallow groundwaters moving south from the Eagle Basin.

High quality lapis is also mined from a limestone-granite skarn contact in the Chilean Andes (3500 m elevation) in the headwaters of the Cazadero and Vias River, Ovalle, Coquimbo, Chile. The lapis there is good quality, although somewhat paler than Sar-e-Sang and, like the Baikal lapis deposits of Russia, is associated with wollastonite not diopside, making it a less attractive gem. The Chilean lapis occurs in an association of phlogopite, sodalite, calcite and pyrite (Coenraads and Canut de Bon, 2000).

Meta-evaporites in the Sar-e-Sang region of Afghanistan exhibit mosaic equilibria across small volumes (in the cm3 range) within a talc-kyanite schist (whiteschist) host. The microscale mineral variations are characterized by variations in mineral assemblages conventionally attributed to vastly different pressure/temperature conditions during regional metamorphism.

On the basis of petrographic and microprobe studies, these assemblages are attributed to three consecutive stages of metamorphism of a chemically exceptional rock with a composition that falls largely into the model system MgO-Al2O3-SiO2-H2O (Figure 3; Schreyer and Abraham, 1976). Stage 1, typified by Mg chlorite-quartz -talc and some paragonite, was followed during stage 2 by talc-kyanite, Mg [4]gedrite-quartz and the growth of large dravites (magnesian tourmalines). Microprobe analyses of the phases, gedrite and talc, indicate variable degrees of sodium incorporation into these phases according to the substitution NaAl—>Si. In stage 3, pure Mg cordierite formed with or without corundum and/or talc, and the kyanite was partly converted into sillimanite. Pressure and temperature during this final stage of metamorphism was near 5-6 kb and 640°C.

Schreyer and Abraham (1976) concluded that chemical variations in the metamorphic fluids were generated by progressive metamorphism and mobilization of an evaporite deposit. Relict anhydrite and gypsum(rehydrated anhydrite) still occur in the Sar-e-Sang area. Whiteschists and the associated lapis lazuli deposits of the region are part of a highly metamorphosed evaporitic succession. Salts have largely vanished due to ongoing melting and volatilizations. The preservation of the three stage succession of mineral assemblages, across such small scales and yet related to each other through isochemical reactions, means that the main factors governing the metamorphic history of this whiteschist were compositional changes of the coexisting fluids with time. Under this scenario any pressure-temperature variations were subordinate and the chemistry of the fluids evolved as the evaporites underwent metasomatic alteration.

The sedimentary pelitic layers of this precursor evaporitic sequence first underwent a period of metamorphism in which fluid pressures approached lithostatic (stage 1). Subsequently at higher metamorphic grades, with the beginning of mobilization of the salts, the metamorphic fluids became increasingly enriched in ions such as Na+, Mg2+, Cl-, SO42-, BO33-, etc., so that water fugacity dropped considerably. This period is represented by stage 2 of the whiteschist metamorphism and was characterized by strong metasomatism that led, for example, to the growth of dravite and the amphibolite, gedrite. The physical and chemical character of stage 3 is less clearly defined. Kyanite/sillimanite inversion requires an increase in temperature or a decrease in pressure, or both; but changes in the composition of a coexisting gas phase may have played an additional role in the formation of cordierite.

Unlike classic metamorphic associations, the meta-evaporite-derived assemblage in Afghanistan may in a single thin section entrain mineral assemblages that conventionally would be assigned to the greenschist facies, the hornfels facies, and to a high pressure (amphibolite) regime. The assemblages are in effect mosaic equilibria that reflect changes in fluid composition generated from a metamorphosing evaporite pile over time and only to a lesser degree by regional evolution of total temperature and pressure. Once again, evaporites generate unusual responses compared to the general responses of metasediments.

In a refinement paper discussing the likely relationships between evaporites and whiteschists, Franz et al., 2013 note that whiteschist mineral assemblages are stable up to pressures of more than 4 GPa, but may already form at pressures of 0.5 GPa. Their formation largely depends on the composition of the protolith and requires elevated contents of Al and Mg as well as low Fe, Ca, and Na contents, as otherwise chloritoid, amphibole, feldspar, or omphacite are formed instead of kyanite or talc. They go on to note that the stability field of a whiteschist mineral assemblage strongly depends on XCO2 and fO2: at low values of XCO2, CO2 binds Mg to carbonates strongly reducing the whiteschist stability field, whereas high fO2 enlarges the stability field and stabilizes yoderite [Mg(Al,Fe3+)3(SiO4)2O(OH)].

The scarcity of whiteschist is not necessarily due to unusual P–T conditions, but to the restricted range of suitable protolith compositions and the spatial distribution of these protoliths: (1) continental sedimentary rocks and (2) hydrothermally and metasomatically altered felsic to mafic rocks. They argue continental sedimentary rocks that may produce whiteschist mineral assemblages typically have been deposited under arid climatic conditions in closed evaporite basins and may be restricted to relatively low latitudes. These rocks typically contain large amounts of palygorskite and sepiolite. Franz et al., (2013) conclude whiteschist assemblages typically are only found in settings of continental collision or where continental lacustrine fragments were involved in subduction.

In my opinion, the mosaic signature of the precursor mineral phases in the typical Sar-e-San lapis lazuli is a metamorphically-evolved response to the combination of precursor permeability and stability contrasts typical of variably-cemented halite mosaic sediments in what were likely haloturbated and variably cemented saline continental lacustrine precursors.


Aleksandrov, S., and V. Senin, 2006, Genesis and composition of lazurite in magnesian skarns: Geochemistry International, v. 44, p. 976-988.

Coenraads, R., and C. C. de Bon, 2000, Lapis Lazuli from the Coquimbo Region, Chile: Gems & Gemology, v. 36, p. 28-41.

Faryad, S. W., 2002, Metamorphic Conditions and Fluid Compositions of Scapolite-Bearing Rocks from the Lapis Lazuli Deposit at Sare Sang, Afghanistan: Journal of Petrology, v. 43, p. 725-747.

Feneyrol, J., G. Giuliani, D. Ohnenstetter, A. E. Fallick, J. E. Martelat, P. Monié, J. Dubessy, C. Rollion-Bard, E. Le Goff, E. Malisa, A. F. M. Rakotondrazafy, V. Pardieu, T. Kahn, D. Ichang'i, E. Venance, N. R. Voarintsoa, M. M. Ranatsenho, C. Simonet, E. Omito, C. Nyamai, and M. Saul, 2013, New aspects and perspectives on tsavorite deposits: Ore Geology Reviews, v. 53, p. 1-25.

Franz, L., R. L. Romer, and C. Capitani, 2013, Protoliths and phase petrology of whiteschists: Contributions to Mineralogy and Petrology, v. 166, p. 255-274.

Garnier, V., G. Giuliani, D. Ohnenstetter, A. E. Fallick, J. Dubessy, D. Banks, H. Q. Vinh, T. Lhomme, H. Maluski, A. Pecher, K. A. Bakhsh, P. Van Long, P. T. Trinh, and D. Schwarz, 2008, Marble-hosted ruby deposits from Central and Southeast Asia: Towards a new genetic model: Ore Geology Reviews, v. 34, p. 169-191.

Giuliani, G., A. E. Fallick, V. Garnier, C. France-Lanord, D. Ohnenstetter, and D. Schwarz, 2005, Oxygen isotope composition as a tracer for the origins of rubies and sapphires: Geology, v. 33, p. 249-252.

Hogarth, D. D., and W. L. Griffin, 1978, Lapis lazuli from Baffin Island; a Precambrian meta-evaporite: Lithos, v. 11, p. 37-60.

Mauger, R. L., 2007, Contact metamorphism-metasomatism associated with the latest Eocene northern Italian Mountain granite intrusion, Gunnison County, Colorado: Abstracts with Programs - Geological Society of America, v. 39, p. 394.

O'Donoghue, M., 2006, Gems; Their Sources, Descriptions and Identification (6th Edition): Amsterdam, Elsevier, 873 p.

Schreyer, W., and K. Abraham, 1976, Three-stage metamorphic history of a whiteschist from Sar e Sang, Afghanistan, as part of a former evaporite deposit: Contributions to Mineralogy & Petrology, v. 59, p. 111-130.

Von Rosen, L., 1990, Lapis lazuli in archaelogical contexts, in P. Aströms, ed., Studies in Mediterranean Archaeology and Literature, v. 93, Partille, Sweden.

Warren, J. K., 2016, Evaporites: A compendium (ISBN 978-3-319-13511-3) Released Dec. 2015: Berlin, Springer, 1600 p.

Webster, R., 1975, Gems, their sources, descriptions and identification: London, Newnes, Butterworths.

Wood, J., 1841, John Wood. A Personal Narrative of a Journey to the Source of the River Oxus by the Route of the Indus, Kabul, and Badakhshan, Performed under the Sanction of the Supreme Government of India, in the Years 1836, 1837, and 1838. Avaialble as Elibron Classics, 2001, 458 pages. Replica of 1841 edition by John Murray, London.

[1] The Ancient Greeks, distinguished between precious and semi-precious stones; similar distinctions were made in other ancient cultures. In modern usage, the precious stones are diamond, ruby, sapphire and emerald, with all other gemstones, including lapis lazuli, being semi-precious.

[2] Chromophore; is the part of a gem lattice responsible for its color. A gem’s color arises when a molecule absorbs certain wavelengths of visible light and transmits or reflects others. It is a structural feature in the lattice indicative of the presence of a gem-specific electron configuration of the ions in its crystal lattice; such as transition metal ions (Cr, V or Fe) occupying several different coordination sites. For example, ferrous iron (Fe2+) or ferric iron (Fe3+), the ferrous ion in peridot causes the green color and ferric ion causes the yellow color in chrysoberyl. This color effect has important uses in heat-treatment gemstones such as blue color in heat-treated sapphire.

[3] Lapis is the Latin word for “stone” and lazuli is the genitive form of the Medieval Latin lazulum meaning blue, which was taken originally from the Persian lāžaward, the name of a place where lapis lazuli was mined. Taken as a whole, lapis lazuli originally meant “stone of Lāzhward.” With time, the name of the place came to be associated with the stone mined there and, eventually, with its bright blue color.

Lapis lazuli’s use as jewelery can be traced back to the 5th millennium B.C.E. with the discovery of beads at a cemetery outside the temple walls of Eridu (Sumer) in southern Babylonia (Von Rosen, 1990) and was used as glyptic from then until now in the manufacture of jewels, amulets, seals and inlays. To the ancient Egyptians, it was considered a gem representing the skies or heaven, thus was thought to denote light, truth and wisdom. It was thus often shaped into eye-shaped gems and was worn by judges in ancient Egypt. A lapis amulet graced the brow of Ra. Lapis is noted in Revelations in Christian mythology as a stone in the Breastplate of Aaron. In China, lapis was worn during the Manchu dynasty for services in the Temple of Heaven. The Romans and Greeks used it as a cure for fever and melancholy. It also glazed the bricks that formed the spectacularly blue “Gate of Kings” or Ishtar entryway to the ancient city of Babylon (≈1800 BC).

The Sar-e-Sang region has supplied much of the gem quality lapis to the world. One of the first European explorers to the region (Wood, 1841) described mining methods in use at that time. Camel-thorn and tamarisk twigs were collected from the valley below and carried up the steep path to the mine. When sufficient fuel had been collected, it was piled against the rock face and a fire was lit. When the rock was hot, cold water, which also had to be carried up the steep 350 m ascent from the valley floor, was thrown onto it. The rock cracked and split, enabling further work to be done with the primitive tools available (pick, hammer and chisel) in order to extract the lapis lazuli from its marble host rock.

[4] Gedrite is a silicate mineral of the amphibole group with formula: (Mg;Fe2+)2[(Mg;Fe2+)3Al2](Si6Al2)O22(OH)2

Recent Posts


NaSO4 salts vadose zone kainitite Lake Peigneur Ripon subsidence basin venice lapis lazuli dihedral angle hydrothermal potash Kara bogaz gol intersalt evaporite karst auto-suture anthropogenically enhanced salt dissolution Kalush Potash knistersalz Dead Sea caves lot's wife Crescent potash alkaline lake sodium silicate Evaporite-source rock association gassy salt High Magadi beds antarcticite trona Messinian Dead Sea karst collapse sinjarite crocodile skin chert H2S Platform evaporite Archean Warrawoona Group salt tectonics Badenian Pangaea lunette namakier McArthur River Pb-Zn epsomite Stebnik Potash methane oil gusher gas in salt palygorskite rockburst bischofite Mega-monsoon Realmonte potash gypsum dune perchlorate Zaragoza magadiite Jefferson Island salt mine seal capacity salt periphery salt trade well blowout saline clay Precambrian evaporites anomalous salt zones waste storage in salt cavity Sulphate of potash water on Mars nuclear waste storage mummifiction salt seal mass die-back silica solubility SOP Ure Terrace deep meteoric potash tachyhydrite Sumo lithium battery African rift valley lakes dissolution collapse doline nitrogen MVT deposit capillary zone flowing salt potash ore meta-evaporite salt ablation breccia Paleoproterozoic Oxygenation Event North Pole Proterozoic climate control on salt jadarite cryogenic salt eolian transport Great Salt Lake halite Catalayud evaporite-metal association allo-suture MOP solikamsk 2 Magdalen's Road doline recurring slope lines (RSL) vanished evaporite blowout potash ore price brine evolution stevensite evaporite base metal hectorite potash chert bedded potash basinwide evaporite mirabilite dark salt Neoproterozoic Oxygenation Event natural geohazard Corocoro copper zeolite Lake Magadi hydrogen collapse doline Koppen climate carnallitite Weeks Island salt mine cauliflower chert marine brine CaCl2 brine Belle Isle salt mine evaporite dissolution Stebnyk potash Lomagundi Event salt karst authigenic silica lazurite salt suture evaporite-hydrocarbon association Pilbara SedEx silicified anhydrite nodules Ethiopia halite-hosted cave DHAL Mesoproterozoic Ingebright Lake geohazard source rock mine stability lithium carbonate black salt Salar de Atacama CO2 Karabogazgol intrasalt gem circum-Atlantic Salt Basins hydrological indicator extrasalt Dallol saltpan nacholite salt leakage, dihedral angle, halite, halokinesis, salt flow, organic matter Muriate of potash hydrohalite HYC Pb-Zn supercontinent Five Island salt dome trend Turkmenistan Hyperarid Clayton Valley playa: ancient climate Zabuye Lake Deep seafloor hypersaline anoxic lake sepiolite Neoproterozoic astrakanite Danakhil Depression, Afar snake-skin chert sedimentary copper gas outburst freefight lake water in modern-day Mars Quaternary climate halotolerant Hadley cell: Musley potash lithium brine Atlantis II Deep sinkhole salt mine Hell Kettle halokinetic hydrothermal karst Deep Koeppen Climate Red Sea