Muriel Erambert

Processing of crust in the root of the Caledonian continental collision zone: the role of eclogitization

Abstract Segments of the root to the Caledonian collision zone are exposed along the west coast of Norway, allowing the study of processes and petrophysical conditions at the deepest crustal levels. P-T conditions prevailing in these crustal root zones correspond to eclogite facies. Eclogitization is associated with marked changes in petrophysical properties, notably a density (10–15%) and Vp increase which may give a mantle signature to an eclogitized crust. The observed ductility enhancement may be caused by transformation plasticity, reduction in grain size, formation of new mineral assemblages and presence of fluid. In the Western Gneiss Region and the Bergen Arcs, eclogites form along shear zones, interpreted as fluid pathways with sharp boundaries to their metastable protolith. This makes eclogite shear zones potential deep crustal reflectors. Dry crust subducted into roots of continental collision zones will undergo eclogitization only if hydrous fluids are available. Timing of metamorphic reactions does not depend on crossing of equilibrium boundaries in the P-T space, but rather on the introduction of fluids into the system. Eclogites are not only passive recorders of P-T-t paths illustrating very deep subduction of crust (> 100 km); rapid changes in petrophysical properties make fluid-induced eclogitization a dynamic process that influences geodynamics. Stresses set up by the volume changes or by tectonic forces are released by ductile deformation along shear zones and by seismic faulting as evidenced by syn-eclogitic pseudotachylytes in the metastable protolith. Deformation, including seismicity, enhances fluid infiltration and further eclogitization. Dense and rheologically weak eclogites may delaminate and sink into the mantle. In the absence of fluids, crustal doubling may occur without metamorphic re-equilibration. The evolution of a collision zone depends on the fluid regime which should be considered in geodynamic modelling in addition to P and T.

The effect of fluid and deformation on zoning and inclusion patterns in poly-metamorphic garnets

Within the Bergen Arcs of W Norway, Caledonian eclogite facies assemblages (T≥650°C, P≥15 kbar) have formed from Grenvillian granulites (T= 800–900°C, P≥10 kbar) along shear zones and fluid pathways. Garnets in the granulites (grtI: Pyr56–40 Alm45–25Gro19–14) are unzoned or display a weak (ca. 1 wt% FeO over 1000μm) zoning. The eclogite facies rocks contain garnets inherited from their granulite facies protoliths. These relict garnets have certain areas with compositions identical to the garnets in their nearby granulite, but can be crosscut by bands of a more Almrich composition (grtII: Pyr31–41Alm40–47Gro17–21) formed during the eclogite facies event. These bands, orientated preferentially parallel or perpendicular to the eclogite foliation, may contain mineral filled veins or trails of eclogite-facies minerals (omphacite, amphibole, white mica, kyanite, quartz and dolomite). Steep compositional gradients (up to 9 wt% FeO over 40 μm) separate the two generations of garnets, indicating limited volume diffusion. The bands are interpreted as fluid rich channels where element mobility must have been infinitely greater than it was for the temperature controlled volume diffusion at mineral interfaces in the granulites. The re-equilibration of granulite facies garnets during the eclogite facies event must, therefore, be a function of fracture density (deformation) and fluid availability. The results cast doubts on modern petrological and geochronological methods that assume pure temperature controlled chemical re-equilibration of garnets.

Garnets recording deep crustal earthquakes

Syn-eclogite facies pseudotachylytes of the Bergen Arcs, western Norway, were formed during co-seismic faulting at a depth of 60 km during Caledonian continent collision, concurrent with fluid-induced partial eclogitization of Precambrian anorthositic granulites. The pseudotachylytes contain the high-pressure assemblage omphacite, garnet and kyanite and formed in the metastable granulites at the border of eclogitized areas. Rock volumes corresponding to outcrops of 100 m2 in area were shattered by pseudotachylytes associated with ultramylonites, cataclasites and shear zones. The granulite facies garnets, unzoned and inclusion-free away from the pseudotachylyte veins, display planar and curved fractures, brecciation and melting with decreasing distance from the veins. Sieve-textured garnet developed by the disintegration of granulite facies garnet, crystallization of eclogite facies mineral inclusions and welding together of garnet fragments. Fractured granulite facies garnets trend to more almandine-rich compositions, with variable grossular contents. In the pseudotachylyte veins, a second generation of minute (10–200 μm) euhedral or dendritic garnets has a higher FeMg ratio than the granulite facies garnet and a wide range in grossular content (Grs11Grs89), suggesting rapid growth and cooling. Fracturing of garnet and its preferential disappearance in the pseudotachylyte show that garnet has a relatively low fracture toughness. Seismic faulting caused enhanced reaction rates during ongoing fluid-induced eclogitization of the deep crust by increasing the surface to volume ratio of garnet and its dislocation density. The sudden volume reduction associated with eclogitization of metastable crust may have nucleated further faulting and pseudotachylyte formation.

Cl-scapolite, Cl-amphibole, and plagioclase equilibria in ductile shear zones at Nusfjord, Lofoten, Norway: implications for fluid compositional evolution during fluid-mineral interaction in the deep crust

Abstract Cl-rich scapolite and amphibole formed during ductile shear deformation associated with the infiltration of an externally derived Cl-bearing fluid in a gabbroanorthosite of the Flakstadoy Basic Complex, Lofoten, Norway. Amphibole and scapolite formed along the contacts between incompletely altered igneous mafic minerals (orthopyroxene, clinopyroxene, biotite, and ilmenite) and plagioclase and internally in the grains of primary igneous plagioclase. The secondary minerals show large compositional variations on thin-section scale. The Cl content of scapolite varies between 0.3 and 0.95 apfu (atoms per formula unit), whereas amphibole shows Cl concentrations from 0 to 1.5 apfu. The primary igneous plagioclase (An50-An60) underwent extensive recrystallization and alteration during the fluid–rock interactions. Secondary plagioclase shows compositions in the range from An20 to An55. In general, plagioclase did not equilibrate with the fluid phase during alteration, due to the sluggishness of the cation exchange reactions between plagioclase and fluid. Occasionally, however, equilibrium among plagioclase, scapolite, and the fluid phase was attained. The compositional variations of amphibole, scapolite, and plagioclase that equilibrated with the fluid are principally related to variations in the fluid activity ratios aCl−/a(CO3)2− and aCl−/aOH−. The large compositional variations of the minerals on thin-section scale thus indicate steep gradients in the fluid activity ratios aCl−/a(CO3)2− and aCl−/aOH−. The activity gradients of the fluid phase developed as a result of the preferential extraction of water from the grain-boundary fluid during the formation of hydrous silicates. Scapolite and the most Cl-enriched amphibole formed in equilibrium with an evolved fluid phase, enriched in Cl and CO2.

Interaction between fluid flow, fracturing and mineral growth during eclogitization, an example from the Sunnfjord area, Western Gneiss Region, Norway

In Sunnfjord, Western Gneiss Region of south Norway, a Proterozoic layered gabbro complex displays various degrees of transformation to eclogite. In the unreacted parts, layering in the gabbro is defined by modal variations of plagioclase, olivine, pyroxenes and minor Fe–Ti oxide and spinel. Coronitic and foliated eclogite formed from the gabbro by hydration reactions at T=510–620°C, causing a volume decrease of c. 13%. In the coronitic eclogite, the mafic magmatic phases are replaced by aggregates of omphacite, barroisite, tremolite, talc and rutile, whereas the plagioclase domains are pseudomorphed by omphacite, barroisite, clinozoisite, kyanite, paragonite and garnet. The felsic and mafic domains are separated by a garnet rim up to 5-mm thick. Garnet was also formed along dilational veins connecting and radiating out of coronas, including the same eclogite facies minerals. In addition, microfractures filled by amphibole and omphacite cut through the corona and vein garnet, oriented perpendicular to the garnet layering. The transformation of dry gabbro to eclogite with hydrous minerals requires supply of water. The timing of metamorphic reactions is therefore dependent on the timing of fluid introduction. The inclusion pattern of garnet in the Holt–Tyssedalsvatnet metagabbro complex indicates that transformation started under eclogite facies conditions. Brittle deformation, in form of fractures allowing infiltration of fluids and mobilisation of elements, is shown to be the most important process initiating transformation. Brittle deformation is thereby active in deep crustal levels corresponding to eclogite facies conditions. Fracturing is interpreted as having been caused by a combination of high fluid pressure, volume changes during mineral transformations and external stresses. Ductile deformation started after the initial metamorphic transformation. Garnet chemistry and zoning pattern are controlled by the chemistry of the growth place, the fluid influx and element supply. Abrupt chemical variations across corona garnet may have been formed by an interaction between rapid growth and transport along microfractures between plagioclase and mafic domain. The metamorphic transition proceeded rapidly. The transformation of the deep crust to eclogite can be viewed as a dynamic process where the volume changes result in fracturing, and these fractures funnelling fluid and element enhancing mineral reaction. The densification causes further fracturing, indicating that the process is self-driven provided that fluid is available.

Petrology of nepheline syenite pegmatites in the Oslo Rift, Norway: Zr and Ti mineral assemblages in miaskitic and agpaitic pegmatites in the Larvik Plutonic Complex

Abstract Agpaitic nepheline syenites have complex, Na-Ca-Zr-Ti minerals as the main hosts for zirconium and titanium, rather than zircon and titanite, which are characteristic for miaskitic rocks. The transition from a miaskitic to an agpaitic crystallization regime in silica-undersaturated magma has traditionally been related to increasing peralkalinity of the magma, but halogen and water contents are also important parameters. The Larvik Plutonic Complex (LPC) in the Permian Oslo Rift, Norway consists of intrusions of hypersolvus monzonite (larvikite), nepheline monzonite (lardalite) and nepheline syenite. Pegmatites ranging in composition from miaskitic syenite with or without nepheline to mildly agpaitic nepheline syenite are the latest products of magmatic differentiation in the complex. The pegmatites can be grouped in (at least) four distinct suites from their magmatic Ti and Zr silicate mineral assemblages. Semiquantitative petrogenetic grids for pegmatites in log aNa2SiO5 - log aH2O - log aHF space can be constructed using information on the composition and distribution of minerals in the pegmatites, including the Zr-rich minerals zircon, parakeldyshite, eudialyte, låvenite, wöhlerite, rosenbuschite, hiortdahlite and catapleiite, and the Ti-dominated minerals aenigmatite, zirconolite (polymignite), astrophyllite, lorenzenite, titanite, mosandrite and rinkite. The chemographic analysis indicates that although increasing peralkalinity of the residual magma (given by the activity of the Na2Si2O5 or Nds component) is an important driving force for the miaskitic to agpaitic transition, water, fluoride (HF) and chloride (HCl) activity controls the actual mineral assemblages forming during crystallization of the residual magmas. The most distinctive mineral in the miaskitic pegmatites is zirconolite. At low fluoride activity, parakeldyshite, lorenzenite and wöhlerite are stable in mildly agpaitic systems. High fluorine (or HF) activity favours minerals such as låvenite, hiortdahlite,rosenbuschite and rinkite, and elevated water activity mosandrite and catapleiite. Astrophyllite and aenigmatite are stable over large ranges of Nds activity, at intermediate and low water activities, respectively.

Heulandite-Ba, a new zeolite species from Norway

Heulandite-Ba, ideally (Ba,Ca,Sr,K,Na) 5 Al 9 Si 27 O 72 ·22H 2 O, is a new zeolite species in the heulandite series, occurring as an accessory mineral in hydrothermal veins of the Kongsberg silver deposit type at the Northern RavnAs prospect, southern Vinoren, 14 km NNW of Kongsberg town, Kongsberg ore district, Flesberg community, Buskerud county, Norway. The mineral has also been found at the Bratteskjerpet mine, Saggrenda near Kongsberg, and in hydrothermal veins in quartzite at Sjoa in Sel community, Oppland county. Heulandite-Ba occurs as well developed, thick tabular, trapezoidal crystals up to 4 mm across, showing the forms {100}, {010}, {001}, 111} and {201}. The mineral is colourless to white, rarely very pale yellowish white or pale beige, with a white streak; transparent to translucent, with a vitreous lustre, pearly on {010}. The mineral has a perfect {010} cleavage; subconchoidal to uneven fracture. It is non-fluorescent in long- or short-wave ultraviolet light. The Mohs9 hardness is 3½; D meas = 2.35(1) and D calc = 2.350 g/cm 3 . Heulandite-Ba is biaxial positive with n α = 1.5056(5), n β = 1.5064(5) and n γ = 1.5150(5); Δ = 0.0094, n (mean) = 1.5090. 2V γ (calc) = 34.1°, 2V γ (meas) = 38(1)°; distinct dispersion, r > v; α Λ c varying from ≅ 39° to ≅ 51° in obtuse angle β, γ = b. An average of 14 electron microprobe analyses on heulandite-Ba from the Northern RavnAs prospect, Kongsberg, gave SiO 2 54.26, Al 2 O 3 15.27, MgO 2 O 0.34, K 2 O 0.58, H 2 O 13.1 (from TGA), total 99.99, corresponding to (Ba 2.49 Ca 1.41 Sr 0.30 K 0.37 Na 0.33 ) Σ4.90 Al 8.96 Si 27.00 O 72.00 ·21.75H 2 Oon the basis of 72 framework oxygen atoms. Chemical zoning is frequent, with transitions to heulandite-Ca and heulandite-Sr. Heulandite-Ba is monoclinic, C 2/ m , with a = 17.738(3), b = 17.856(2), c = 7.419(1) A, β = 116.55(2)°, V = 2102.0(7) A 3 , Z = 1. The strongest five X-ray diffraction lines of the powder pattern [ d in A( I )( hkl )] are: 2.973(100)(151), 3.978(97)(131), 7.941(66)(200), 4.650(66)(-131), 2.807(65)(-621). The crystal structure refinements (R = 3.5%) of heulandite-Ba were done in space groups C 2/ m, Cm, C 2, and C 1, but refinements in space groups with lower symmetry than C 2/ m did not improve the structural model.

Solid solution between potassic-obertiite and potassic-fluoro-magnesio-arfvedsonite in a silica-rich lamproite from northeastern Mozambique

A fine-grained dike rock was collected during regional geologic mapping in the Xixano region, northeastern Mozambique. Observation in thin section and SEM and EMP analyses showed a fine-grained igneous texture dominated by optically unusual zoned amphibole and Fe 3+ -bearing low sanidine, with Sr-bearing fluorapatite, rutile, Sr-bearing barite, a silica mineral, hematite and zircon. The mineralogy and major element data indicate an unusual high-silica lamproite. The amphibole is concentrically zoned and pleochroic from pale brown to pale greenish blue, with common abnormal blue and brown interference colours, indicative of high dispersion commonly associated with Fe 3+ . Results of 29 EMP analyses, including two traverses across zoning, were formulated in different modes. The most satisfactory formulae were obtained assuming high Fe 3+ /(Fe 2+ + Fe 3+ ) ratios and significant oxo component (O 2− at the O(3) site substituting for (OH, F)). The most Ti-rich core composition approximates: K Na 2 (Mg 3 Fe 2+ 0.5 Fe 3+ 1 Ti 0.5 ) Si 8 O 22 (F 1 O 1 ), and can be classified as 1:1 solid solution between potassic-obertiite and potassic-fluoro-magnesio-arfvedsonite. Zoning toward the rims increases the potassic-fluoro-magnesio-arfvedsonite component up to K Na 2 (Mg 3.5 Fe 3+ 1.5 ) Si 8 O 22 (F 1 (OH) 0.5 O 0.5 ).

Solid solution between potassic alkali amphiboles from the silica-rich Kvaløya lamproite, West Troms Basement Complex, northern Norway

Alkali amphibole of rare compositions occurs as a rock-forming mineral in a high-Si phlogopite lamproite from Kvaloya, northern Norway. The amphibole typically occurs as small grains forming irregular and rosette-shaped aggregates in a matrix dominated by Fe-rich K-feldspar and quartz. Amphibole shows compositions ranging between the three limiting compositions: A : A,B ( K 1.01 Na 1.99 ) C ( Na 0.26 Mg 1.58 Mn 0.03 Fe 2 + 0.91 Fe 3 + 1.6 Ti 0.47 □ 0.13 ) T Si 8 O 22 W [ F 0.97 O 1.03 ] B : A,B ( KNa 2 ) C ( Na 0.04 Mg 1.04 Mn 0.22 Fe 2 + 0.65 Fe 3 + 2.07 Ti 0.25 □ 0.7 ) T Si 8 O 22 W [ F 0.68 Cl 0.01 O 0.13 ( OH ) 1.18 ] C : A,B ( K 0.9 Na 2.1 ) C ( Na 0.04 Mg 3.54 Mn 0.02 Fe 2 + 0.28 Fe 3 + 1.04 Ti 0.03 □ 0.02 ) T Si 8 O 22 W [ F 1.34 O 0.06 ( OH ) 0.6 ] Composition C shows significant content of fluoro-potassic-magnesio-arfvedsonite, while composition A is a Fe 2+ , Fe 3+ and C Na rich variety of potassic-obertiite. Composition B is characterized by an exceptional high value of C □. It is emphasized that the presence of C □ and C Na in amphibole needs to be confirmed by other methods. The relationship between W O 2− , C □,Ti 4+ and Fe 3+ of amphibole can be expressed by the following exchange operators, choosing potassic-magnesio-arfvedsonite [KNa 2 (Mg 4 Fe 3+ )Si 8 O 22 (OH) 2 ] as the additive component: Ti 4 + Mg 2 + − 1 H + − 2 Fe 3 + Mg 2 + − 1 H + − 1 Ti 4 + □ Mg 2 + − 2 Fe 3 + 2 □ Mg 2 + − 3 The two first exchange operators result in deprotonation of OH, while the two others result in the formation of vacancies on the C sites. The presence of amphibole both in the lamproite and in the adjacent fenitized granite suggests that the mineral formed during reactions between rock and fluids derived from the volatile-rich lamproite magma. Possibly, amphibole core (composition A and B), formed in equilibrium with the fluid phase during crystallization of the melt, while amphibole rim (composition C) formed during subsequent mineral-fluid reactions. Presence of hematite in the lamproite matrix in addition to oxo-amphibole indicates that the rock formed during highly oxidizing conditions.