Early Proterozoic UPb Zircon Ages from Basement Gneiss at the Solovetsky Archipelago , White Sea , Russia

The central region of the Neoarchaean Belomorian Mobile Belt (BMB) is, except for the Solovetsky Archipelago, largely covered by the White Sea. A newly discovered granitic gneiss outcrop on Solovetsky Island, Russia, enables a first age determination of the archipelago and evaluation of the hitherto poorly constrained central BMB. Zircons separated from the orthogneiss were analysed with SIMS-SHRIMP and LA-ICP-MS techniques. Both techniques yield a concordant U-Pb age of ca. 2.430 Ga, coinciding with ages of granitic intrusions in the BMB ca. 50 km west of the Solovetsky Islands.


Introduction
The Neoarchaean to Palaeoproterozoic Belomorian Mobile Belt (BMB) forms a SW-NE trending belt of mostly metamorphic rocks, which extends from northern Finland to the southern region of the Arkhangelsk Oblast in the Russian Federation (see e.g.[1], and inset of Figure 1).Located to the NE is the Neoarchaean Murmansk Craton (comprising granodioritic and trondhjemitic orthogneisses), and the Kola Province (TTG-type orthogneisses of Neoarchaean age).In the SW, the BMB borders the greenstone-granite domain of the Neoarchaean Karelian Craton (e.g.[2]).The central area of the BMB is, except for the Solovetsky Archipelago, largely covered by the White Sea, making access to rocks a difficult task.The BMB consists of abundant TTG-type rocks (age ca.2.9 Ga), and metavolcanic rocks of an age of ca.1.9 Ga (e.g.[2][3][4]).The latter are associated with metasediments and a younger suite of granitic and gabbroic intrusions.Moreover, it was inferred that ca.2.4 Ga old plutonic bodies are present in the BMB in significant proportions [4].The Archaean units of the BMB were partially reworked during the Palaeoproterozoic Lapland-Kola collisional event ca.1.9 Gyrs ago (e.g.[1,2], and references therein).Rifting at the Kandalaksha-Dvina aulacogen crossing the western White Sea in SE-NW direction from the eastern Onega Peninsula to the Kandalaksha Bay started possibly in the late Palaeoproterozoic [1].The aulacogen is segmented into grabens and horsts by several mostly WSW-ENE trending fault systems [5].The oldest deposits in the grabens are volcanoclastic sediments, basic flows, and tuffs with a suggested age of ca.1.8 -1.4 Ga [5].A simplified geological map is illustrated in the inset in Figure 1.A more detailed map and description of the BMB and Karelia is given by [6].
The Solovetsky Archipelago is located in the western White Sea, about 50 km east of the city of Kem' in Karelia and ca.30 km NW of the Onega Peninsula (Figure 1).The island group consists of six large and several smaller islands with Bolshoi (Big) Solovetsky Island being the largest.About 50 km east of the archipelago, the Kandalaksha-Dvina aulacogen traverses the White Sea from SE to NW.
The first geological map (scale 1:200,000) of the archipelago was produced by [7], a section showing lithologies of the islands and the western coast of the White Sea is given in Figure 1.The islands are covered by glacial deposits derived most likely from the Kola Peninsula and the north-western areas of the BMB.The last glaciation of the archipelago occurred in the late Pleistocene during the Valdai glaciation (e.g.[8]).A Neoarchaean basement (labelled "Bottom Strata" in Figure 1) with an age of ca.2.6 -2.2 Ga was suggested by [7] to be present beneath the islands.Moreover, the archipelago is directly located on two WSW-ENE trending zones marked by local late Archaean granitic intrusions [7].In addition, gravimetric, seismic, and geoelectrical data recorded on Bolshoi Solovetsky by [9] revealed that the crust is thinnest (ca.28 km) in comparison to the surrounding area, and a mantle diapir of unkown age is situated beneath the islands.Devonian magmatic activity in the western White Sea region was likely initiated by an upwelling mantle plume, as suggested by [10], and [11] on the base of tectonic interpretation of satellite images.This triggered an uplift of the area and may still support present-day weak local positive thermal anomalies observed in lakes and soils, and formation of an extrazonal thermophilic ecosystem on Bolshoi Solovetsky Island [12,13].
In July 2009, a group of Russian and German geolo-gists and geophysicists discovered on this island a gneiss outcrop which is probably related to the BMB.The outcrop (ca.7.5 × 5 m) is located at the lower flanks of Sekirnaya Hill in the north-western part of the island (Figure 1).It consists of relatively fresh granitic gneiss.Other lithologies were not found in the outcrop.This provides an opportunity of a first age determination of the basement of the archipelago and, in a broader context, may help to elucidate the geological evolution of the central region of the BMB.The orthogneiss was dated via the U-Pb zircon dating method employed to two different instrumental techniques (SIMS-SHRIMP and LA-ICP-MS).

Methods
The orthogneiss sample (labelled "R1") from Bolshoi Solovetsky Island was sawed into smaller pieces for thin section preparation, crushing in a jaw breaker and steel mortar.From the crushed sample, a part was ground to powder for X-ray Fluorescence (XRF) analysis employing an ARL 9800 instrument at VSEGEI.The XRF data were further processed via CIPW norm for sample classification.
Another representative portion of the crushed material was used for separation of accessory zircon which was required for comparative age determination in different laboratories using SIMS (Secondary Ion Mass-Spectrometry) by Sensitive High-Resolution Ion Micro Probe (SHRIMP-II, at the Centre of Isotopic Research of VSE-GEI, St. Petersburg, Russia) and Laser Ablation-Sector Field-Inductively Coupled Plasma-Mass Spectrometry technique (LA-SF-ICP-MS, at the Institute for Mineralogy, University of Münster, Germany), respectively.In addition of dating a hitherto unknown outcrop in a geologically complex region, the two involved institutions were able to compare their respective techniques by using well-defined standards and unknown zircons.The separation procedures employed in the Russian and German laboratories are given in Table 1.
For analyses, selected zircon grains were mounted in Pb-free epoxy resin together with the TEMORA 1, GJ-1, and/or 91500 reference zircons.The grains were sectioned approximately in half, polished and coated.Cathodoluminescence (CL) images were used to reveal the internal structures of the zircon grains and thereby define target areas within them.The technical methods applied in the different institutions are given in Table 1.

SHRIMP Analyses
The results were obtained with a secondary electron multiplier operated in peak-jumping mode, as outlined by [14].A primary beam of molecular oxygen was employed to bombard the zircons in order to extract secondary ions.A 70 μm Kohler aperture allowed focusing of the primary beam so that the ellipse-shaped analytical spot had a size ca 25 × 20 μm.The sputtered secondary ions were accelerated at 10 kV.A mass-resolution of M/ΔM ≥ 5000 (1% valley definition) was achieved via an 80 µm wide slit of the secondary ion source combined with a 100 µm multiplier slit; thus, all possible isobaric interferences were resolved.Rastering for one minute over a rectangular area of ca.65 × 50 μm before each analysis removed the gold coating and any possible surface common-Pb contamination.
The ion species of interest were measured in a sequence as follows: 196 (Zr 2 O)-204 Pb-background (ca.204 a.m.u.)-206 Pb-207 Pb-208 Pb-238 U-248 ThO-254 UO with integration times ranging from 2 s to 14 s.Seven cycles for each analyzed spot were acquired.Apart from "unknown" zircons, each fourth measurement was carried out on the zircon Pb/U standard TEMORA 1, which has an ac-  [16]) was applied as a U-concentration standard.The Pb-U ratios have been normalized relative to a value of 0.0668 for 206 Pb/ 238 U of the TEMORA 1 reference zircons.Error in TEMORA standard calibration was 0.62%.The results were then processed with the SQUID 1.02 (see [17]) and Isoplot/Ex 3.00 (see [18]) software, using the decay constants of [19].The common Pb correction was done on the basis of measured 204 Pb and present-day Pb isotope composition, according to the model of [20].

LA-ICP-MS Analyses
The U-Pb age determinations were done using a LA-SF-ICP mass spectrometer (Element 2, ThermoFinnigan) and a New Wave ArF Excimer Laser system at Univer-sity of Münster, Germany.Forward power was 1330 W, gas flow rates were about 0.7 L/min for He (employed as carrier gas for ablated material), and 0.9 L/min and 1 L/min for the Ar auxiliary and sample gas, respectively.The cooling gas flow rate was set to 16 L/min.Before starting the analyses, the system has been tuned on the standard reference material SRM 612 from the NIST (National Institute of Standards and Technology) by measuring 139 La, 232 Th, and 232 Th 16 O to get stable signals and a high sensitivity on 139 La and 232 Th peaks, as well as low oxide rates ( 232 Th 16 O/ 232 Th ~ 0.1%) during ablation.
The repetition rate was 10 Hz at an energy of ~5 J/cm 2 .The typical spot size was 35 µm, in some cases also 25 µm.
For U-Pb analyses of the zircons, the masses 204 Pb, 206 Pb, 207 Pb, and 238 U were measured.In addition, 202 Hg was analyzed to correct the interference of 204 Hg on 204 Pb, which is important to apply, if necessary, for a common-Pb correction.The common Pb proportion for the individual spots is calculated from the 204 Hg-corrected 204 Pb signal, applying the two-stage model of [20].When the contribution of estimated common 206 Pb to the total 206 Pb exceeded 1%, which corresponds approximately to the analytical uncertainty of the measured Pb isotope ratios, a common Pb correction was applied.Ten unknown samples were bracketed with three calibration standards (GJ-1; [21]) to correct for instrumental mass bias.The sensitivity for measured Pb and U isotopes for a 35 μm spot was typically in the range of 4000 cps/ppm.Age calculations were done with an in-house Microsoft ® Excel spreadsheet using the intercept method (e.g.[22]) to correct for elemental fractionation during laser analyses.Along with the samples, the 91500 zircon standard was measured to monitor accuracy and precision of the analyses.A three-shot pre-ablation was applied to all spots to remove the carbon coating and potential common Pb contamination from the surface.Out of 111 embedded zircons, 98 were selected for analyses after inspection of CL images.Cracks and inclusions were avoided.In total, 115 measurements were carried out.

Petrography
Sample R1 is a medium-grained foliated granitic gneiss with abundant orthoclase and amphibole.Hand specimen and thin section inspection yielded a composition of ca.45 vol% of feldspar, 25 vol% of amphibole, 25 vol% of quartz, ca. 5 vol% biotite, and accessory phases like garnet, apatite, zircon, and opaque phases.Results of XRF analyses are given in Table 2, classifying the sample as an acid alkaline rock because of its high silica and alkali content.As illustrated in Figure 2, ternary classification schemes after [23] and [24] indicate (syeno-)granite  as a precursor rock for R1.Therefore, R1 is classified as an orthogneiss (metasyenogranite).In addition, low MgO-CaO (0.42) and high P 2 O 5 -TiO 2 (0.28) ratios suggest a magmatic origin of the precursor rock [25].
The zircons are largely present as short prismatic and prismatic crystals with a length-wide ratio of ca.1.5 to 3. Facets of prisms and pyramids are distinct, but show sometimes signs of chemical corrosion.Small, sporadic mineral and gas-liquid inclusions are observed under a binocular.In addition, the zircons of R1 often show well developed oscillatory zoning, which is sometimes present as sector zoning and is considered a typical feature of magmatic zircons (e.g.[26], and references therein).Some examples are shown as CL images in Figure 3.

SHRIMP Results
All ten in situ U-Pb isotopic analyses on eight typical zircon grains, free of cracks and inclusions, gave a welldefined concordant age.The results of the zircon analyses are shown in Table 3 and Figure 4(a).During the course of this study, the 91500 standard zircon yielded a 207 Pb /206 Pb age of 1066 ± 7 Ma (n = 6).The zircons are characterized by an average Th-U ratio scattering around 0.77, and low contents of uranium (average 47 ppm), and thorium (average 36 ppm).Concentrations of non-radio- genic lead are insignificant (<0.5%).Zircon cores as well as metamorphic overgrows or rims were not found.Such geochemical features are commonly observed for zircons in magmatic rocks (e.g.[26,27]), but are not always a positive proof (for further discussion, see [28]).For age calculation, all analytical results were used without discrimination.As a result, SHRIMP analyses of eight R1 zircons yield a concordant age of 2433 ± 13 Ma (2σ).).However, for the remaining results, loss of Pb by later processes seems to have affected some of the zircons because of their lower Pb-U isotope ratios.These discordant ages (n = 28) were obtained mostly for the rims of the analysed zircons.Of the 115 analyzed spots, 17 results were discarded because the signal intensity was too low or dropped too fast (i.e., the chosen spot did not offer a sufficient amount of zircon material for analy- sis).The complete LA-ICP-MS data set is available upon request.

Discussion
The geological complexity of the Belomorian Mobile Belt (BMB) in NW Russia has been known for several years (e.g.[4,30]).However, little data with respect to lithologies and rock ages are so far available for the White Sea region that covers large parts of the central BMB section.We show here that a hitherto unknown gneiss outcrop on Solovetsky Island in the western White Sea is related to granitic intrusions in the central BMB and yields a U-Pb zircon age of ca.2.43 Ga.This age was achieved independently via SHRIMP and LA-ICP-MS, again confirming the reliability of these methods as already shown by previous studies (e.g.[31]).
The composition of the orthogneiss R1 (a metasyenogranite) and its U-Pb zircon age coincide with granitic intrusions observed for the western BMB at ca. 2.4 Ga (e.g.[4,32]).This new age information strongly suggests magmatic activity at 2.43 Ga in the central section of the BMB where it was unknown so far because of missing accessible outcrops.Moreover, the granitic gneiss of Solovetsky Island is located on the convergence of two SW-NE trending zones of variably sized granite intrusions as suggested by [7] in a pioneering geological study of the western BMB region around the city of Kem' (Figure 1).The nature, origin, and extension of these zones of granitic intrusions especially further east are still uncertain.To note, charnockites to the west and SW of Kem' yield a U-Pb zircon age of 2.4 to 2.45 Ga as well [32].Taken together, the central and western region of the BMB was affected by granitic intrusions at ca. 2.4 Ga, however, the magma compositions varied on a local scale (e.g., charnockites, syenogranites).Potassium-Ar ages of the granites reported by [7] range from 1.86 to 1.93 Ga.These age data coincide with the Lapland-Kola collisional event at ca. 1.9 Ga that has obviously reset the granite K-Ar age due to metamorphism (e.g.[1,2]).Interestingly, some zircons analysed with LA-ICP-MS yield lower discordant U-Pb ages at their rims (Figure 3(d)).Again, this is in agreement with the time of metamorphic events; however, we cannot exclude multiple episodes of Pb loss as is tentatively suggested by a few discordant zircon ages of lower than 2.4 Ga.The influence of the mantle diapir on the U-Pb isotope composition of the zircons (Pb loss due to re-heating, or younger zircon growth during ascent of hydrothermal fluids) is difficult to evaluate.The lower intercept age given in Figure 4(b) is poorly constrained at 268 Ma ± 170 Ma because of the scarcity of zircons with a discordant U-Pb age.However, it still coincides within analytical uncertainty with the time of Devonian magmatic activity in the western White Sea region (e.g.[10]).The offset of the U-Pb age of some R1 zircons may result from ascending hydrothermal fluids and subsequent Pb loss and/or overgrowth.In contrast, [11] suggested on the basis of geothermal data and geophysical modelling that the last tectono-magmatic activity in the central BMB took place during early Miocene.In summary, the mantle diapir has likely affected the central BMB region in Phanerozoic time, but the extension and intensity of magmatic activity still remains uncertain.
Composition and age of the rock units forming the BMB can vary on a scale of sometimes a few meters (e.g.[4,30,33]); hence it is possible that a complex basement is situated beneath the cover of glacial deposits at the Solovetsky archipelago.As the geological map given by [6] is of a scale of 1:4,000,000, and the map of [7] lacks detailed information about magmatic rocks, thorough geological mapping and dating of possible other rock units of outcropping basement material at the islands may uncover a possibly complex geology.The mantle diapir beneath the island group likely facilitated rise of the islands above sea level [9], and, in conjunction with an extrazonal ecosystem characterized by thermophilic vegetation close to the polar circle [12,13,34], highlights a special geotectonic setting and possible modern tectonic activation to be investigated further.

Figure 1 .
Figure 1.Simplified tectonic and geological overview of the White Sea region and the Solovetsky Archipelago.The inset shows the boundaries of the different provinces in NE Fennoscandia (modified after [2]).I. T.: Inari Terrane; K. U. T. B.: Kolvitsa-Umba-Tersk Belt; L. G. B.: Lapland Granulite Belt; M. C.: Murmansk Craton; S. T.: Strelna Terrane; K.-D.: Kandalaksha-Dvina aulacogen (solid grey line); S. I.: Solovetsky Islands.The map shows strata of the Belomorian Mobile Belt around the city of Kem' and the Solovetsky Islands (after [7]).Bolshoi (Big) Solovetsky Island is characterized by numerous lakes of different sizes and glacial deposits.The sample location at the slope of the Sekirnaya Hill is shown by the black dot.

Figure 3 .Figure 4 .
Figure 3. CL images of four zircons (a)-(d).Laser spots are highlighted as dashed white circles, ages ( 207 Pb/ 235 U in Ma, ±2σ) and the degree of concordance is given for comparison.Sector zoning is visible in (a) and (b); Less developed zoning and a non-zoned core region marks the crystal in (c); The zircon shown in (d) exhibits slight resorption at the boundary beneath the left laser spot.It also yielded different ages for core and rim, respectively.