June 2024 LIP of the Month

Constraints on the source of Siberian Trap magmas from Mo isotope evidence

Aleksandr E. Marfin^a*, Michael Bizimis^a, Peter C. Lightfoot^b, Gene Yogodzinski^a, Alexei Ivanov^c, Matthew Brzozowski^d, Anton Latyshev^e,f, Tatiyana Radomskaya^g

^a School of Earth, Ocean, and Environment, University of South Carolina, Columbia, SC 29208, USA

^b Department of Earth Sciences, University of Western Ontario, ONT, N6A 5B7, Canada

^c Institute of the Earth’s Crust SB RAS, Irkutsk, 664033, Russia

^d Key Laboratory of Western China's Mineral Resources and Geological Engineering, Ministry of Education, School of Earth Science and Resources, Chang'an University, Xi'an 710054, China

^e Schmidt Institute of the Physics of the Earth, Moscow 123242, Russia

^f Lomonosov Moscow State University, Moscow, 119991, Russia

^g Vinogradov Institute of Geochemistry SB RAS, Irkutsk, 664033, Russia

Extracted and modified from:

Abstract

The Siberian Trap Large Igneous Province (ST-LIP) covers a large region in Siberia (Russia); it is the most voluminous continental flood magmatic province in Phanerozoic. Despite the long history of the study of ST-LIP magmatism, it still needs to be fully understood what role the asthenospheric mantle, lithospheric mantle, and continental crust play in magma genesis. Understanding different impacts from different sources is essential because the ST-LIP event links with the end-Permian mass extinction (Burgess et al., 2017) and hosts large Cu-Ni-PGE magmatic sulfide deposits (Norilsk area) (Naldrett, 1992). The complete sequence of the trap basalts is preserved in the Norilsk region and has a total thickness of about 4 km (Krivolutskaya and Sobolev, 2016). All basalts can be grouped into three parts: lower, middle, and upper parts (Fedorenko et al., 1996). We present a new dataset of whole rock radiogenic Sr-Nd-Hf and stable Mo isotopes from surface and drill hole samples. Our new Sr-Nd isotopic data is consistent with previous work (Lightfoot et al., 1990, 1993) and generally plots within the MORB-OIB field except for one volcanic formation which is contaminated by upper continental crust. Our new Hf radiogenic data is consistent with the Nd isotopic data and plots within the terrestrial array with an apparent slope nearly identical to the mantle array. The Mo isotopic data (reported as δ98Mo ratio relative to NIST SRM 3134) vary from -0.62 ‰to 0.07 ‰ for the lower and middle parts and from –0.25 ‰ to –0.03 ‰ for the upper part of the volcanic sequence. Together with trace elements and radiogenic isotope systematic, we attribute variation in the earliest type of magma to contribution from a dehydrated eclogitic component. Thus, during ST-LIP event magmatism, some magmas were derived from sources that were not plume, emphasizing the importance of non-asthenospheric sources even for voluminous LIP magmatism.

Geology

The ST-LIP covers an area of about 7×10⁶ km², with an approximate volume of 4 × 10⁶ km³ (Ivanov, 2007; Masaitis, 1983), but it was shown that about 2 km of the eastern lobe of the ST-LIP has been eroded; thus the estimated volume should be increased (Bagdasaryan et al., 2022) (Figure 1A,B). Such a gigantic volume of magma was emplacement in a short period: no more than two million years (Burgess and Bowring, 2015; Ivanov et al., 2021). We focused our research on the Norilsk area because it has the highest thickness of basalts. The basalts can be divided into 11 formations (Figure 1C).

These formations are grouped into lower, middle, and upper parts. All rocks are classified as trachybasalts-basaltic trachy andesites, subalkaline basalts, and basalts. The Gudchikhinsky formation is characterized by a high concentration of MgO (10-20%).

The lower (earlier) basalts (Ivakinsky (Iv), Syverminsky (Sv), Gudchikhinsky (Gd)) are characterized by positive and negative Pb anomalies, and depletion of Nb and Ta compared to Th and U (Figure 2A,B). Rare earth element patterns are less enriched relative to continental crust estimates. The middle part (Khakanchansky (Kk), Tuklonsky (Tk), and Nadezhdinsky (Nd) basalts show Ta-Nb depletion and positive Pb anomaly (Figure 2C,D). The upper (younger) basalts (Morongovsky (Mr), Mokulaevsky (Mk), Kharaelakhsky (Kh), Kumginsky (Km), Samoedsky (Sm)) represent the most voluminous part and is characterized by a flat REE distribution pattern and slightly enriched incompatible elements. Wee estimate the composition of the Ergalakh sill complex, which is an intrusive analog of the Ivakinsky formation, and the mineralized Norilsk 1 intrusion, whose formation age overlaps the Morongovsky-Mokulaevsky boundary and it can be considered comagmatic to this lava.

Radiogenic and Stable Isotopes

Figure 3 shows the variation of radiogenic Sr-Nd-Hf and stable Mo isotopes and the variation in concentration of Mo, Cu, and Ni. The Ivakinsky and Syverminsky formations have slightly radiogenic Sr and unradiogenic Nd. The Gudchikhinsky formation has similar Sr signatures but more radiogenic Nd (Figure 3A,B). The most variable part is the Nadezdinsky formation (green square) which shows a crustal signature and Cu-Ni depletion (Figure 3A,B,F). It was interpreted as extracting chalcophile metals and platinum group elements from Nadezhdinsky type of magma due assimilation processes. The most voluminous part (blue dots) has slightly radiogenic Sr and unradiogenic Nd. The Hf isotopic data strongly correlate with the Nd isotopic data (Figure 3C): The Ivakinsky and Syverminsky formation and the Ergalakh sill complex are characterized by unradiogenic Hf. The Gudchikhinsky formation has radiogenic Hf, similar to the upper part of the lavas (Figure 3C). The most unradiogenic Hf is in the middle part (Nadezhdensky formation), which confirms the crustal contamination process.

The δ⁹⁸Mo varies more significantly in the lower part, and, in general, is characterized by light signatures from –0.62 ‰ to 0.07 ‰, which is outside the MORB type mantle range of –0.24 ‰ to 0.08 ‰. The δ⁹⁸Mo in upper part vary from –0.25 ‰ to –0.03 ‰ and, in general, is characterized by a MORB-like signature (Figure 3D).

Mixing models

The variability of δ⁹⁸Mo isotopes cannot be explained by weathering, fractional crystallization, or sulfide saturation (in this setting, Mo behaves like a lithophile element). Also, contamination by anhydrite and Paleozoic sediments, like black shales or coal, requires unrealistic amounts of addition (>30%) to a parental Tuklonsky type of magma to explain the δ⁹⁸Mo range of the basalts (Figure 4).

We can explain the variability of δ⁹⁸Mo in the Nadezhdynsky formation by contamination of the Upper continental crust. Approximately 20-30% contamination of the upper crust is needed to explain the range of radiogenic and stable isotopes in Nadezdinsky basalts. The most challenging part is explaining the light isotopic composition of the Gudchikhinsky formation is significantly lighter than <0.2‰.

Origin of the δ⁹⁸Mo variability of the ST-LIP magma sources

Based on high Ni content and Fe/Mn ratio in olivine it was suggested that ST-LIP magma sources contain pyroxene-rich components (Sobolev et al., 2011). Pyroxene-rich lithologies start to melt in a greater depth and will melt to a larger degree than peridotite at a given amount of decompression (Bizimis and Peslier, 2015 and references therein). Existing δ⁹⁸Mo data on eclogite composition shows that they are characterized by light values (average -0.51 ‰, Ahmad et al., 2021; Chen et al., 2019), which is explained by the removal of isotopically heavy Mo by fluid during slab dehydration. Given that Earth’s upper mantle (δ⁹⁸Mo = -0.21 ± 0.06 ‰; Yang et al., 2017) and MORB (δ⁹⁸Mo -0.24 ‰ to 0.08 ‰; Bezard et al., 2016) have small variability of Mo, and which is heavier than for eclogites, then the light-δ⁹⁸Mo in the lower formations (including the Ergalakh sill complex) can reflect an eclogitic (recycled oceanic crust) component in the source (Figure 6). Some modern OIB basalts are explained by the mixing between peridotite- and eclogite-derived melts (Gaschnig et al., 2021; Willbold and Elliott, 2023). Also, some komatiites have light-δ⁹⁸Mo, which means that isotopically the light-δ⁹⁸Mo component may have existed in the mantle since the Archean. Thus, we have attributed the light-Mo in the lower formations to recycled eclogites in the source.

It is difficult to determine where the eclogites are located. They may be part of the subcontinental lithospheric mantle or part of the upwelling plume. Based on the eclogite’s xenoliths in the Siberian kimberlite pipes, we suggested that eclogites were part of the subcontinental lithospheric mantle (SCLM) and started to melt when the plume impinged on the lithosphere.

Meanwhile, other parts of the SCLM may have been affected by fluids from eclogite dehydration during subduction with presumably heavy δ⁹⁸Mo, which could explain some of the heavier δ⁹⁸Mo signatures and possibly the sample with the heaviest δ⁹⁸Mo (0.07‰) and highest Mo/Ce (0.049) (Figure 6, white dot within arc-lava field). Such fluids come from a relatively young eclogite so that they could retain relatively radiogenic Nd and Hf isotopes consistent with the Gudchikhinsky-formation values. In addition, the documented crustal contamination en route to the surface of ST-LIP magmas may have added to the Mo isotope heterogeneity towards heavier δ⁹⁸Mo than eclogites (Fig. 6). The general lack of correlation between δ⁹⁸Mo and the radiogenic isotope tracers (except the Nadezhdinsky-formation which exhibits a trajectory consistent with crustal contamination in δ⁹⁸Mo vs. ⁸⁷Sr/⁸⁶Sr and εNd (Figure 5) is more difficult to assess, as we do not know the isotopic composition of the unmodified plume, nor the overall compositional variance of the eclogite-bearing and metasomatized SCLM, nor the δ⁹⁸Mo variance of the crustal rocks in the vicinity of STLIP. In δ⁹⁸Mo vs. εNd space (Figure. 5D), for example, we could envision a mixing line emanating from the MORB field towards heavy δ⁹⁸Mo(fluids) and light δ⁹⁸Mo (eclogite), with variably low (enriched) εNd. Our hypotheses above can be better constrained by additional data on eclogites and samples of the SCLM, as well as crustal rocks from the Siberian platform.

Our study suggests that the subduction-modified SCLM may have played a key role in shaping the composition of the erupting ST-LIP magmas. Mo isotopes can provide new insights into the processes and sources of some of the earth’s more voluminous volcanic eruptions.

References