The Alpha–Mendeleev Rise in the Arctic Ocean, which consists of the Mendeleev Rise and the Alpha Ridge and which is part of the East Arctic uplifts along with the Chukchi Plateau, Lomonosov Ridge, and Northwind Ridge (Fig. 1), is of key importance for understanding the history of the geological development of the Arctic as a whole. A wealth of geological information has been accrued in recent years about this structure by integrated geological and geophysical expeditions sent to collect data for the evidence basis of the Russian Federation’s submissions to the UN Commission on the Law of the Sea to extend the outer limit of the Russian continental shelf in the Arctic. An important aspect of this research is studying the role of magmatic phenomena in the process of formation of the Alpha–Mendeleev Rise. A variety of magmatic rocks was obtained during the Russian Arktika-2012 expedition [1], the time of formation being determined for some of them. For rocks of the basic composition, the age of samples from Trukshina seamount was estimated at 127 Ma [1] and that of those from the southern part of Mendeleev Rise at 260 and 498–500 Ma [2] using the U/Pb geochronology of zircons, while samples from Mendeleev Rise were in the age range of 436 to 471 Ma using 40Ar/39Ar isotope geochronology [3]. Based on the data of other expeditions for rocks from the northern part of Alpha Ridge, the 40Ar/39Ar ages of two basalts were measured at 82 ± 2 and 89 ± 1 Ma [4] and that of tuff, at 90.40 ± 0.26 Ma [5]. The present review shows that data on the timing of magmatism in the Alpha–Mendeleev Rise are extremely scarce and contradictory because among other things (1) they are sometimes insufficiently substantiated and (2) the position of the studied rocks in the geological section has not been established.

Fig. 1.
figure 1

The East Arctic seabed structure scheme. The quadrangles show the location of sampling sites; next to them are their numbers. The inset shows the position of the study area (black rectangle) in the Arctic region.

Full size image

We have at our disposal magmatic rocks collected in different areas of the Alpha–Mendeleev Rise during two Mendeleev-2014 and Mendeleev-2016 deep-sea geological expeditions on a research submarine (RS) organized by the Main Directorate of Deep-Sea Research of the Russian Navy, Defense Ministry; AO Geosluzhba GIN RAN; and the Geological Institute of the Russian Academy of Sciences (RAS) [6]. The samples were collected by the submarine manipulators from bedrock outcrops of the slope. This paper discusses the results of isotope geochronology U/Pb dating of magmatic rocks sampled by these two expeditions. Three strata were identified earlier based on the results of the study of sedimentary rocks taken by the same expeditions in the visible section of the Alpha–Mendeleev Rise basement in the study areas. They sequentially succeeded each other upwards along the section including the lower stratum (O3-S) composed of dolomites, limestones, and quartzite sandstones, the middle stratum (D2-D3) formed by limestones and sandstones, and the upper stratum (K1) consisting of sandstones, tuffs, and lavas of basic composition [7].

The samples analyzed were collected at three sites located in the southwestern (site 1) and central (site 3) parts of the Mendeleev Rise and on Trukshina seamount (Alpha Ridge) (site 2) (see Fig. 1). The sampling method is described in [6, 7], and the coordinates and depths of the sampling points are given in Table 1.

Table 1. The list of analyzed samples and their characteristics
Full size table

The identification of the rocks was based on microscopic and petrochemical studies. Two groups of rocks were identified including (1) basalts and microgabbro of normal alkalinity and (2) trachybasalts, trachyandesites, trachyandesibasalts, and their tuffs, which together formed a single series of differentiation of moderate-alkaline rocks. The basalts (samples 14-18, 22) were found at site 1 within cliffs complicating the main slope of the Rise where these rocks make up low (up to 1 m) ledges. Microgabbro (samples 1602/3, 11, 23) form massive outcrops on the slopes of Trukshina. Trachyandesites (samples 14-02, 05, 06) are widespread at site 1. Judging by their outcrops in the form of stratum bodies stretching along the slope, they form sills. There are also vitroclastic tuffs (14 to 19) that occur at the foot of the slope in the form of a flat-shaped body up to 10 m thick. Trachyandesibasalts (samples 1601/14, 16, 21, 24) and crystalline lithoclastic tuffs (samples 1601/11, 25) were sampled at site 3 where they make up separate blocks located at different levels of the slope composed of sedimentary Paleozoic rocks. These volcanites are porous rocks, and sample 1601/16 is a volcanic bomb indicating the formation of trachyandesibasalts under subaerial conditions. Both trachybasalts (1602/4, 5) were taken from the same outcrop at site 2.

The zircons for U/Pb-dating were separated from rocks by the standard method at the Geological Institute, Russian Academy of Sciences. U–Th–Pb–isotope analysis was done using a SHRIMP-RG ion microscope at the Research School of Earth Sciences, Australian National University, Canberra, following the procedure described in [8, 9]. The measurements were processed with the SQUID Excel Macro [10] and ISOPLOT/EX Excel Macro software [11].

As follows from Table 1, coherent groups of zircon ages were obtained for two trachyandesibasalts (1601/14 and 1601/16) and for one tuff (1601/25), which allows us to estimate their ages at 112.5 ± 0.9, 110.2 ± 0.6, and 114.3 ± 0.7 Ma, respectively (Figs. 2a, 2b, 2c) (Early Cretaceous). The coherent age group in each sample included 22 or 23 zircons typical of volcanic rocks, which were shaped like elongated prismatic crystals (average cross size of 200 × 80 µm) with developed pyramids (Fig. 3a) and with slightly pronounced oscillatory CL zonation. Two older zircons of 298 ± 3.0 and 465 ± 7 Ma were also found in sample 1601/16.

Fig. 2.
figure 2

The concordant diagrams of the samples where coherent zircon ages were identified: (а) 1601/25, (b) 1601/14, (c) 1601/16, and (d) 1601/11. Blue lines are concordia and the numbers along them are time values in million years. N is the number of grains measured.

Full size image
Fig. 3.
figure 3

Microphotographs of zircon grains separated from tuffs: (a) sample 1601/25, (b) sample 1601/11. The scale bars in each figure correspond to 100 µm.

Full size image

A coherent group of ages with a mean value of 2675.1 ± 11.6 Ma (Late Archean) was formed by 22 grains separated from tuff 1601/11 (Fig. 2d). These are short prismatic or barrel-shaped grains with an average cross size of about 150 μm, completely rounded or with smoothed faces and apexes (Fig. 3b). Their CL zonation is irregular and disturbed.

Two other trachyandesibasalts studied (1601/21, 24) did not yield coherent groups of zircon ages. Sample 1601/21 had the two youngest grains aged 87 ± 0.7 and 83 ± 15 Ma (Late Cretaceous) and the ages of other grains in these samples covered a wide range from 134 to 2475 Ma. The Late Cretaceous zircons differed greatly from all other grains by the small size (not more than 50 μm), cathodic luminescence revealing nuclei of different tones not conformal to the grain outlines.

The measured trachyandesite zircons in 14-02, 14-05, and 14-06 from site 1 did not form coherent age groups, but had the youngest zircons of 111 ± 5, 93 ± 0.7, and 111 ± 0.8 Ma, respectively, was found in each of them. The age of these grains was close to the age of trachyandesibasalts in 1601/14, 16. The ages of the other zircons formed a wide time interval from 141 to 1830 Ma. Vitroclastic tuff in the sample 14-09 from site 1 contained the five youngest grains (97 to 124 Ma) close in age to samples 1601/14, 16. The other grains were older in the range from 131 to 1864 Ma.

Two trachybasalts (1602/4, 5) from site 2 are parts of the same body. Out of 88 grains measured in them, only four had ages close to those of trachyandesibasalts in 1601/14, 16 ranging from 101 to 119 Ma (Table 1). One younger grain of 85 ± 2 Ma was also found, the rest of the grains being between 131 and 2641 Ma.

The age of microgabbro and basalt rocks of normal alkalinity could not be determined by the method of isotope U/Pb geochronology. A coherent group of zircon ages with a mean value of 428.2 ± 2.8 Ma was identified only in microgabbro sample 1602/3, but this sample also contained four younger grains aged between 233 and 254 Ma with a mean value of 248 Ma. The remaining grains were more ancient, up to 1295 Ma old. There was one very young grain aged 104 ± 1 Ma in microgabbro 1602/11 among measured grains, the ages of other grains being scattered in the range of 196 to 1862 Ma. The concordant age was obtained in sample 1602/23 for only 4 grains, all of them had different ages ranging from 250 to 1181 Ma.

The above studies allowed us to ascertain reliably the age of trachyandesibasalts and of their tuffs from the central part of the Mendeleev Rise. For three of them in samples 1601/14, 1601/16, and 1601/25, it was 112.5 ± 0.9, 110.2 ± 0.6, and 114.3 ± 0.7 Ma, respectively. Obviously, other trachyandesibasalts in which coherent zircon ages could not be obtained nevertheless had similar ages. This conclusion is based on the petrochemical similarity of all trachyandesibasalts, the same way of their occurrence, and the presence of single grains of Cretaceous age therein.

U/Pb dating of the tuff in 1601/11, where the Late Archean zircons had formed a coherent group of 2675.1 ± 11.6 Ma ages, produced unexpected results. In addition to fragments of volcanic rocks similar to those in the Early Cretaceous tuff of 1601/25, this tuff contained quartz and potassium feldspar grains and their aggregates. The reason for the hybrid composition of tuff in 1601/11 may be that the Early Cretaceous magma assimilated Late Archean rocks fragments of which are represented in the tuff by quartz, potassium feldspar, and zircon.

The age of other moderately alkaline rocks including trachyandesites and tuffs from site 1 and trachybasalts from site 2 could not be determined. Nevertheless, there is every reason to believe that they were also formed in the Early Cretaceous. First, they form a single petrochemical series with trachyandesibasalts, and second, they all contain single generally young zircons with an age close to that of the trachyandesibasalts.

We also failed to determine the age of rocks of the normally alkaline series. Only one grain was measured at a time in basalts. There was no clustering of zircon ages in microgabbros of 1602/11 and 1602/23 from smt. Trukshina; they were dispersed over a wide age range. At the same time, one Early Cretaceous grain aged 104 ± 1 Ma was measured in sample 1602/11, but nevertheless we did not lump this microgabbro in the same association with subalkaline Early Cretaceous volcanites because petrochemically it differed significantly from the latter. Two age clusters were encountered among the microgabbro 1602/3 zircons including a coherent age group of 428.2 ± 2.8 Ma and a less significant group (four grains) with a mean value of about 248 Ma. The value of which group shall we assume to be the age of the rock? Is it the age of the coherent group (428.2 ± 2.8 Ma) and then that the Early Triassic zircons (248 Ma) are the product of later transformation of Silurian zircons or is it Early Triassic and then the Silurian zircons are xenogenic? This is a question we leave open for further study.

Thus, Early Cretaceous basic rocks of moderate alkalinity in the form of lavas, tuffs, and sills, which formed either over 110–114 million years or at the turn of 112 Ma, are widespread among magmatic rocks within the study area, the latter value corresponding to the mean value of the dates obtained.

Volcanic rocks of Cretaceous age were also found in other structures of the East Arctic uplifts. Ar/Ar ages were determined in the range from 85.9 ± 3.9 to 75.5 ± 3.9 Ma for alkaline and subalkaline basalts of a seamount northwest of the Northwind Ridge [12]. The Ar/Ar age of the Northwind Ridge basalts was estimated at 112 ± 1 Ma [13]. The above ages of basalts from the East Arctic uplifts and the rock ages that we obtained are close to those of Cretaceous volcanic rocks widespread in the Arctic islands such as Spitsbergen, Franz Josef Land, North Greenland, the Canadian Archipelago, and Novosibirsk Islands (Benetta Island). These terrestrial magmatic manifestations are included by researchers in the HALIP magmatic province, the origin of which is attributed to the uplift of the deep mantle plume [14]. In the work [15] also was included the magmatic rocks of the East Arctic uplifts described above in the HALIP province, which was highly active between 160 and 60 million years ago. Following these authors, we also believe that the volcanites of the moderately alkaline series from the Alpha–Mendeleev Rise analyzed by us represent the HALIP province.

Three phases are distinguished in the formation history of the HALIP province: Late Jurassic, Early Cretaceous, and Late Cretaceous [16]. According to U/Pb dating, the rocks that we studied were formed during the Early Cretaceous phase. We did not find rocks of the Late Cretaceous phase, but zircons of this age (83 to 87 Ma) were found in some Early Cretaceous volcanites (samples 1601/24 and 1602/5) (Table 1). Apparently, the Late Cretaceous zircons are a product of recrystallization of Early Cretaceous zircons, which probably occurred under the influence of hydrothermal solutions associated with Late Cretaceous magmatism. The morphology of these grains and their internal structure do not contradict this conclusion. Products of Late Cretaceous volcanism are present in Alpha Ridge [5].

U/Pb zircon dating often discovers grains older than the rock among zircons separated from the same sample. Most researchers believe that these are xenogenic zircons captured by the melt from host rocks on its way to the surface. As can be seen in Table 1, ancient xenogenic grains were also found in the Early Cretaceous volcanites that we studied. The age composition of xenogenic zircons is extremely diverse. It is hard to imagine that the Early Cretaceous melts rising upward consistently “tried out” the entire crustal section. Rather, the supplier of ancient zircons of such a wide age diversity could be a single source containing all these zircons. These can only be sandstones, especially Early Cretaceous sandstones of graywacke composition [7], which compose the upper strata in the acoustic basement of Alpha–Mendeleev Rise. There are also sandstones in the Paleozoic strata, but clearly they cannot contain Mesozoic and Late Paleozoic zircons (Table 1).

The tuff 1601/11 contains only ancient zircons, and these are part of only one Late Archean cluster with its peak at 2675.1 ± 11.6 Ma. Their very narrow age interval indicates that (1) the melt–substrate interaction was pointwise and (2) the source of the zircons is Late Archean rock that is most likely magmatic in nature. This is a very important conclusion indicating the presence of an ancient crystalline basement at Alpha–Mendeleev Rise

Thus, the most widespread among the magmatic rocks of Alpha–Mendeleev Rise are the Early Cretaceous rocks of the moderately alkaline series, which represent the middle formation stage of the HALIP magmatic province the products of which are common in the Arctic islands. They include trachyandesibasalts and tuffs in the central part of Mendeleev Rise, which were formed either over 110–114 million years or at the turn of 112 Ma, and most likely trachyandesites and tuffs from its southwestern part and trachybasalts from smt. Trukshina. These rocks contain a large amount of older xenogenic zircons the age analysis of which shows that subalkaline melts interacted with crustal rocks at two levels including the Early Cretaceous sandstone horizon and rocks of the Late Archean crystalline basement. There is also a small amount of younger zircons aged 83 to 87 Ma, which were probably formed during recrystallization of the Early Cretaceous zircons under the influence of Late Cretaceous magmatism also characteristic of the HALIP province.