Most crust in Iran and Turkey is part of the Alpine–Himalaya orogenic belt, which is situated at the junction of Gondwana (in the south) and the Eurasian plate (in the north) (Dewey et al., 1973; An and Harrison, 2000; Göncüoglu, 2010; Prelević and Seghedi, 2013; Tian et al., 2017). This crust includes a number of microcontinents, such as the Taurides–Anatolides and Biston–Avoramanm, Bilitis, and Van microcontinents, with the latter being added by the authors in this research (Figure 1). These microcontinents were distributed across the Tethys Ocean in the Mesozoic to early Cenozoic eras but are now surrounded by oceanic-crust remnants (Jassim and Goff, 2006). Nevertheless, there is no well-defined information to confirm that these microcontinents completely drifted from the Arabian plate (Wrobel-Daveau et al., 2010; Azizi et al., 2013; Nouri et al., 2016).

The northern fossil remnant of oceanic crust is known as the Paleo–Tethys trace, which extends from northern Iran to northern Turkey (Kroner and Romer, 2013; Manafi et al., 2013; Shafaii Moghadam and Stern, 2014). The central and southern parts (Figure 2) are known as the Neotethys remnant (Stöcklin and Nabavi, 1973; Golonka, 2004; Ghasemi and Talbot, 2006; Moghadam and Stern, 2015; Hassanzadeh and Wernicke, 2016; Nouri et al., 2019).

Magmatic activity increased in both northwestern Iran and Turkey in the Late Cretaceous and has continued through the Quaternary (Alavi, 1994). Magmatism in both regions was concentrated in four main periods: The Late Cretaceous, Eocene, Miocene, and Quaternary. According to many published research studies, most of the magmatic activity was related to the Tethys (Neotethys) orogeny, which stemmed from the subduction and collision of the Arabian plate with the Turkey and Iran plates, as well as from postcollisional tectonic regimes (Agard et al., 2006; Agard et al., 2011; Ali et al., 2013; Whitechurch et al., 2013; Nouri et al., 2016; Nouri et al., 2017; Nouri et al., 2020; Azizi et al., 2018b; Shabanian et al., 2018). Certainly, in Turkey and the western and central parts of Iran, magmatic rocks show clear similarities to the typical calc-alkaline, magmatic series common to an active continental margin (Parlak et al., 2000; Aldanmaz, 2006; Dilek and Altunkaynak, 2009; Topuz et al., 2004; Topuz et al., 2017).

The focus of this study is on magmatic activity in the region between the Black and Caspian Seas, which includes northwestern Iran, eastern Turkey, Georgia, Armenia, and Azerbaijan (Figure 2). This magmatism is of interest, as it has traditionally been considered to be a result of the subduction of oceanic crust of the Neotethys beneath Iran and southern Turkey. The site of this margin is considered to be represented by the Bitlis–Zagros–Fault (BZF) suture zone (Figures 2, 3). As the BZF suture zone is distant from the southern and western Caspian Sea (>300 km), it is problematic for the highly, calc-alkaline magmatism to be sourced from subduction coming from that area. Furthermore, the location of the Khoy ophiolite sequence would appear to be inconsistent with subduction focused solely along the BZF. The large size of the Khoy ophiolite and its composition of pelagic sediment, a large thickness of pillow lava, sheeted dikes, both layered and massive gabbro, and cumulate and tectonite peridotites, are highly consistent with typical ophiolite sequences (Hassanipak and Ghazi, 2000; Ghazi et al., 2003; Khalatbari-Jafari et al., 2003) that have been generated in a mature, oceanic basin. These seeming inconsistencies in the geological record led us to propose the existence of a new microcontinent or continental fragment that sits between the BZF suture zone to the south and the Khoy ophiolite sequence to the northeast.

In this study, we introduce a new microplate, the Van microplate, which has an area of approximately 50,000 km² and is situated between northwestern and central Iran, the Anatolian block, and the Arabian plate (Figure 3). We outline and explain the proposed, microplate boundaries, followed by an explanation of how the northern branch of the Neotethys has controlled the magmatic activity of northwestern Iran and the Caucasus, based on this new microplate, despite the southern branch.

Eocene volcanic rocks are widely distributed from the southern Caspian Sea to the southern Black Sea and are almost parallel to the Cretaceous volcanic rocks (Figures 2, 3). Figures 2, 3 were compiled based on more than 20 geology maps of Iran and the surrounding area (See the Supplementary S1 for the map references). Cretaceous volcanic rocks are also extensive, although less so in comparison to the Eocene rocks (Figures 2, 3), and are limited to the south of the Zagros fault in the northern Sanandaj–Sirjan zone (SaSZ). Cretaceous magmatism primarily has calc-alkaline affinity (Azizi and Jahangiri, 2008; Azizi and Moinevaziri, 2009). Ophiolite members, known as Neotethys remnants, are widely distributed along major faults, such as the BZF and Zagros thrust-fault (ZTF) suture in southern Turkey and western Iran. The Bitlis–Zagros, which includes the BZF and ZTF here (Figures 2, 3), is known as a major suture zone between the Arabian and Iranian–Anatolian plates. It was opened in the late Paleozoic Era and closed in the Cenozoic Era (Agard et al., 2011; Alirezaei and Hassanzadeh, 2012; Abdulzahra et al., 2016; Azizi et al., 2019).

Some dismembered mafic and ultramafic rocks are exposed throughout the entire Bitlis-Zagros suture zone in the border between Iran and Iraq along the BZF (Figures 2, 3). Some researchers considered this complex as ophiolites (Ismail et al., 2010; Ali et al., 2013; Ali et al., 2016; Mohammad, 2013; Mohammad and Cornell, 2017; Al Humadi et al., 2019) and found that these rocks are more consistent with plume-type magma and barren rifts (Wrobel-Daveau et al., 2010; Azizi et al., 2013; Azizi and Stern, 2019; Karim and Al-Bidry, 2020). In this case, a barren rift between the Van microplate and Arabian plate is more likely than mature oceanic crust. These mafic rocks define the Sanandaj–Sirjan zone (SaSZ) from the Arabian plate in northwestern Iran to southern Turkey. The distribution of the magmatic rocks with the major faults from eastern Turkey to northern Iran is detailed in Figures 4A–E.

Azizi and Stern (2019) divided the SaSZ in western Iran (Figure 2) into three main parts: northern, central, and southern. In the northern part, which is separated from the Van microplate in this study and the junction, some dismembered ophiolite has been exposed. We call this the Oshnavieh-Salmas-Khoy ophiolite belt (OSK-Ophiolites) here. The OSK-Ophiolites, which has a N-S trend, is the main suture zone between the Van microplate and northwestern Iran block. We describe the OSK-Ophiolites below.

We have compiled whole-rock, chemical data on 900 volcanic rocks (Agostini et al., 2007; Jahangiri, 2007; Azizi and Jahangiri, 2008; Haghnazar and Malakotian, 2009; Mazhari et al., 2009, 2011; Ahmadzadeh et al., 2010; Dilek et al., 2010; Ranin et al., 2010; Aghazadeh et al., 2011; Aghazadeh et al., 2015; Ansari et al., 2011; Dabiri et al., 2011; Hajialioghli et al., 2011; Mahmoudi et al., 2011; Asiabanha and Foden, 2012; Alishah et al., 2013; Ansari, 2013; Castro et al., 2013; Ghaffari et al., 2013; Kheirkhah et al., 2013; Moghadam et al., 2014; Neill et al., 2015; Doroozi et al., 2016; Haghnazar et al., 2016; Moritz et al., 2016; Sahakyan et al., 2017; Taki, 2017; Ashrafi et al., 2018; Haghighi Bardineh et al., 2018; Jafari Sough et al., 2018; Lechmann et al., 2018; Shafaii Moghadam et al., 2018; Zhang et al., 2018; Azizi et al., 2019; Ghalamghash et al., 2019; Temizel et al., 2019) from Cretaceous to Quaternary magmatism in northwestern Iran, southern Armenia, and eastern Turkey (Supplementary S2). The sample locations and references for each area are shown in Figures 4A–E.

Cretaceous volcanic rocks include basalt to rhyolite but are mostly classified as andesite. Cretaceous intrusive bodies also show some variation from gabbro to granite, with less exposure compared to volcanic rocks. Whole-rock composition shows that the Cretaceous, magmatic rocks have a large variation in TiO₂ from 0.19 to 3.82 wt% (average = 1.52 wt%, n = 125), Al₂O₃ from 11.2 to 22.7 wt% (average = 15.9 wt%, n = 125), and MgO from 0.03 to 11.8 wt% (average = 5.27 wt%, n = 125). In addition, these rocks have a high concentration of Ba (average = 470 ppm, n = 125), U (average = 2.82 ppm), and Th (average = 8.9 ppm). Eocene–Miocene, magmatic rocks have a large variation in SiO₂ (43–77 wt%), high content of Rb (average = 100 ppm; n = 389), Sr (average = 618 ppm, n = 429), Th (average = 11.7 ppm, n = 331) and U (3.47 ppm, n = 341). The upper Miocene-Quaternary magmatic rocks are mostly categorized into two main groups: adakite, with higher ratios of Sr/Y and La/Yb; and alkali basalts, with a higher content of Nb (average = 27 ppm, n = 328). See Supplementary S2 for more information about the whole-rock chemistry of each group. For our discussion, the samples are categorized into three main groups: Cretaceous, Eocene-Miocene, and upper Miocene-Quaternary. All of the groups show variances from the basalt to the rhyolites (Figure 5) in the K₂O + Na₂O diagrams (Le Bas et al., 1986). Nevertheless, the alkali elements for the Cretaceous and Eocene magmatic rocks largely varied compared to the upper Miocene and Quaternary rocks.

For understanding tectonic setting, some key and general petrological diagrams are briefly used here. The Ti/V ratio diagram (Shervais, 1982) shows a wide distribution of the Cretaceous volcanic rocks from the arc to extensional basins (Figure 6A). Eocene rocks have more affinities for the arc-magma series. The late Miocene-Quaternary arc affinity was decreased and extended to alkaline and within-plate series (Figure 6A). The Th/Yb vs. Nb/Yb diagram (Pearce, 2008) shows that most of the mafic rocks plot in the volcanic-arc field because of the higher ratios of Th/Yb (Figure 6B). The enrichment of Th occurred during the fluids released from the subduction of the oceanic crust and/or crustal contamination with some assimilation-fractionation and contamination (AFC) processes. The Cretaceous basaltic rocks show a dichotomy for both intraoceanic arc and active continental-tectonic regimes. Eocene basaltic rocks show a clear relation to the typical active continental margin. Upper Miocene-Quaternary basaltic rocks also plot in the volcanic-arc field, with some similarity to oceanic-island basaltic rocks (OIB) near the mantle array, with less effect on crustal components (Figure 6B). To avoid the mobility of some trace elements, all Eocene and younger basaltic rocks, other than Cretaceous volcanic rocks because of the wide ranges, were normalized using rare-earth elements (REEs) of the chondrite value (Sun and McDonough, 1989). The normalized pattern shows that both the Eocene and upper Miocene-Quaternary rocks have higher ratios of light REEs than heavy REEs, and negative slopes from LREEs to HREEs (Figures 7A,B). Furthermore, the REE_S contained in the upper Miocene-Quaternary rocks are higher than those in the Eocene basaltic rocks (Figures 7A,B).

The plotting of the intermediate-to-acidic rocks (SiO₂ > 52 wt%) in FeO/(FeO + MgO) and Na₂O + K₂O–CaO vs. SiO₂ (Frost et al., 2001) shows that the lower ratios of FeO/(FeO + MgO) for most of the samples from the three groups plot in the magnesian and/or I–type granite fields. Some of the Cretaceous and Eocene samples extend to the ferroan granite field. The upper Miocene-Quaternary rocks mainly plot in the magnesian field and/or I-type granite area (Figure 8A). In the Na₂O + K₂O–CaO vs. SiO₂ (wt%), Cretaceous and Eocene samples are scattered from the calcic to alkalic fields. Nevertheless, most of the samples were concentrated in the alkali-calcic field. The upper Miocene-Quaternary rocks plot in the alkali-calcic field, with less variation than the Cretaceous and Eocene samples (Figure 8B).

Intermediate-to-acidic rocks (SiO₂ > 52 wt%) have higher ratios of Rb; lower ratios of Y, Ta, and Nb (Figures 9A,B); and almost plot in the active margin to within-plate domains (Pearce et al., 1984). Cretaceous intermediate-to-acidic volcanic rocks mainly plot in the active margin and within-plate tectonic fields. The variation in the Ta/Yb vs. Th/Yb ratios (Pearce, 1982) shows a wide range for the Eocene and Cretaceous intermediate-to-acidic magmatic rocks and more affinity with the upper Miocene-Quaternary for the within-plate granite field (Figure 9C).

The scattering of the Cretaceous volcanic rocks probably supports different tectonic regimes during the Cretaceous and requires more detailed work because the Cretaceous volcanic rocks were strongly altered by seawater reactions. Some parts of intermediate-to-acidic rocks have higher ratios of Sr/Y and La/Yb and, based on the adakite key diagrams (Defant and Drummond, 1990), are classified as adakite (Figure 10A). The high affinity of the upper Miocene-–Quaternary volcanic rocks for adakite groups, as well as the accompanying high-Nb basaltic rocks, means that there is a correspondence to a postcollisional tectonic regime in northwestern Iran since the Miocene (Azizi et al., 2014a; Asiabanha et al., 2018; Torkian et al., 2019).

We plotted 203 sample data points (Supplementary S2) in the ⁸⁷Sr/⁸⁶Sr vs. ¹⁴³Nd/¹⁴⁴Nd diagram collected from published works (Alpaslan et al., 2004; Agostini et al., 2007; Haghnazar and Malakotian, 2009; Mazhari et al., 2009, Mazhari et al., 2011; Aghazadeh et al., 2011, Aghazadeh et al., 2015; Dabiri et al., 2011; Allen et al., 2013a, Allen et al., 2013b; Ghaffari et al., 2013; Moghadam et al., 2014; Neill et al., 2015; Sahakyan et al., 2017; Haghighi Bardineh et al., 2018; Lechmann et al., 2018; Shafaii Moghadam et al., 2018; Azizi et al., 2019). Most samples plot near the mantle array with some variation for the Cretaceous and Eocene igneous rocks. Cretaceous samples plot in two main areas: the first group plots near the cross lines of the chondrite uniform reservoir (CHUR) and bulk silicate earth (BSE); and the second group, with lower ratios of ¹⁴³Nd/¹⁴⁴Nd and higher ratios of the ⁸⁷Sr/⁸⁶Sr, plots with the combination of both mantle and crustal components for the evolution of these rocks (Figure 11).

Eocene igneous rocks mainly plot in the depleted-mantle area and extend to the lower ¹⁴³Nd/¹⁴⁴Nd ratio and higher ⁸⁷Sr/⁸⁶Sr ratio fields, which show heterogeneous sources such as depleted mantle and primitive and metasomatized mantle for the sources of these rocks. Furthermore, the scattering of Eocene samples in the ⁸⁷Sr/⁸⁶Sr-¹⁴³Nd/¹⁴⁴Nd diagram suggests the involvement of both mantle material and crustal components in the magmatic activity during the Eocene. The scattering of the Sr–Nd isotope ratios shows three main trends for the evolution of the Eocene magmatic rocks, which we call EO.1, EO.2, and EO.3 (Figure 11). These three trends show depleted mantle with higher ratios of ¹⁴³Nd/¹⁴⁴Nd (>0.51290), such as from the subcontinental-lithospheric mantle, as the primary source. During magmatic evolution and/or later, they have been affected by crustal components (EO.2 and EO.3 trend) and/or alteration (EO.1 trend). Due to the wide distribution of Eocene volcanism and the variation in Sr-Nd isotope ratios, metasomatized mantle from the oceanic slab is also the main alternative for the genesis of these rocks, which has been suggested by many researchers for the sources of these rocks.

Upper Miocene to Quaternary samples plot near the mantle array from the depleted mantle to enriched mantle in the ⁸⁷Sr/⁸⁶Sr–¹⁴³Nd/¹⁴⁴Nd ratio diagram (Figure 11). Most samples are concentrated near the cross line of the CHUR, which mainly shows the combination of depleted (¹⁴³Nd/¹⁴⁴Nd > 0.512638) to primitive mantles (¹⁴³Nd/¹⁴⁴Nd = 0.512638) as the main sources of these rocks. Crustal-component interactions, for example, assimilation with the magma, are lower than those of the Eocene magmatic rocks (Figure 11), which may suggest the thinning of the continental crust after collision. Therefore, depleted mantle, oceanic-island basalt (OIB), source-like, and enriched mantle (EM2) contributed to the genesis of Eocene and upper Miocene-Quaternary magmatic activity, with less contribution from continental crust for the upper Miocene-Quaternary rocks.

In most of the traditionally published research, the northern and eastern Bitlis–Zagros suture zone has been considered as an active continental margin since the Jurassic Period (Davoudian et al., 2008; Agard et al., 2011; Hassanzadeh and Wernicke, 2016). Recently published research has confirmed that the initiation of Neotethys subduction beneath the SaSZ (Azizi and Stern, 2019) occurred in the Cretaceous Period, and the collision of the Arabian plate and Van microplate in northwestern Iran occurred in the late Eocene epoch (Azizi et al., 2019). Nevertheless, there are some weak points concerning the relationship of ultramafic rocks to the ophiolite complex in northeastern Iraq within the suture of the Arabian plate and Van microplate. There is also no strong evidence for a typical ophiolite complex here (Azizi et al., 2013). In line with our suggestion, the new OSK zone shows that northwestern Iran-eastern Turkey magmatism was mainly controlled by Neotethys oceanic subduction, which was situated on the northern side of the Van microplate (Figures 2, 3). The Khoy ophiolites mainly occurred during the Late Cretaceous Period (Ghazi et al., 2003; Khalatbari-Jafari et al., 2003, 2004; Moghadam and Stern, 2015) and are in contact with upper Neoproterozoic/lower Paleozoic granites (Azizi et al., 2011) in the east. Additionally, the Precambrian basements mainly include intrusive granite and metamorphic rocks, which were exposed in the Khoy, Urmia (Urmiah), and Tabriz areas. Northwestern Iran was obducted onto the east side of the OSK suture zone (Figure 3).

Tertiary volcanism surrounded the Van microplate (Figures 2, 3). Meanwhile, distinct depression basins, such as the Urmia Lake basin (Figures 2, 3), were formed near the OSK suture zone. All evidence shows a typical arc-type tectonic regime here. Nevertheless, our information on the Cretaceous magmatic rocks in northwestern Iran is not sufficient. The Neotethys ophiolite or ophiolite-like belt in northwestern Iran is divided into two branches in the area around Piranshahr (Figure 3). The southern (southwestern) branch extends to northeastern Iraq and continues to southern Bitlis, and it is known as the BZF suture zone here. This branch does not show a clear relationship to typical ophiolites, but it has some correlation with plume-type (Azizi et al., 2013) sources with some affinity for immature oceanic crust and/or continental rifts. The second branch is the OSK suture zone, which extends in a NW-SE direction in western Iran and joins the Sevan-Akera suture zone (SAS) in its northern part (Figures 2, 3). The southern branch on the Iran side is named the Kermanshah-Nayriz ophiolites and is exposed along the Zagros fault in western Iran (Figure 3).

Magmatic activity since the Late Cretaceous Period developed in the northern and eastern parts of the Van microplate (Figure 3). The northern branches of the Neotethys, which were situated between the Van microplate to the south and the northwestern Iran-Armenia block (NWIA) to the north, were subducted beneath the NWIA block during the middle to Late Cretaceous Period. To consider the magmatic evolution in the southern rim of the NWIA, we suggest the following schematic model for the evolution of magmatic rocks in this belt since the Mesozoic.

Based on our suggested model, Jurassic rocks developed in the Van microplate. NWIA also probably shows the fragmentation of the Van microplate that occurred in the Late Jurassic/Early Cretaceous periods. Some evidence shows that the Jurassic magmatism was related to the extensional regime and continental rifts in the Khoy area (Khalatbari-Jafari et al., 2003; Lechmann et al., 2018; Miyazaki et al., 2018; Shafaii Moghadam et al., 2019), and they probably confirm the extensional regime or drifting of the Van microplate during the Cimmerian orogeny. This evidence is also consistent with the SaSZ tectonic regime, which had an extensional regime in the Middle-Late Jurassic Period (Azizi et al., 2018a; Badr et al., 2018; Azizi and Stern, 2019).

The OSK ophiolites have been thrust onto the passive margin and are overlaid by Paleocene sedimentary rocks. The ophiolite obduction on the passive margin probably occurred in a continental-arc collision system and, following the Neotethys, it started to subduct beneath the NWIA in the Late Cretaceous (Figures 12A,B). During subduction, nonarc magmatism, arc magmatism, and forearc basins developed in an active margin in the NWIA. Precambrian basements were obducted at this time (Figures 12A,B), and Upper Cretaceous-Eocene, magmatic rocks were produced during this process.

After colliding with the Van microplate and NWIA in the late Oligocene-Miocene, the oceanic crust experienced rollback, and the NWIA plate was overridden and affected by the extensional regime; some subsiding basins developed among the Eocene fragmented volcanic belt (Figure 12C). Deposition in the sedimentary basins started unconformably with a basal conglomerate on top of the Cretaceous-Eocene complex (Figure 12C). The extensional regime after the collision, the thinning of the NWIA, and the high-angle subduction connected both the lithospheric and asthenospheric mantle for the genesis of the magmatic rocks in the Miocene Epoch (Figure 12C). In the late stage, the oceanic crust was dismembered, and hot magma was injected from the slab window (Figure 13). Some parts of the magma erupted directly and made basaltic rocks that had a high similarity to the OIB composition in the HTV belt from Ararat to northern Sanandaj (Kheirkhah et al., 2009; Asiabanha and Foden, 2012; Azizi et al., 2014a; Torkian et al., 2016; Asiabanha et al., 2018). Other parts crystallized inside the lower crust and increased the geothermal gradient in the continental crust. This process was the basis for the partial melting of the lower crust and the production of adakite-like magma in the HTV belt during the Pliocene-Quaternary (Azizi et al., 2014b; Torkian et al., 2019).

This research shows that the Van microplate drifted from the Arabian plate in the Jurassic or earlier and is now surrounded by dismembered Neotethys ophiolite. During the closure of the northern branch of the Neotethys and the incorporation of the Van microplate into the southern Eurasian plate, a long period of magmatic activity (from the Late Cretaceous Period to the present) occurred along the southern rim of Eurasia, northwestern Iran, and the Caucasus. Therefore, the main suture is not Bitlis–Zagros, which has been suggested as the junction of the Arabian and Iran-Armenia plates. Instead, it seems that Zagros–Khoy–Akera is the main suture zone, and it should be considered in future studies. The Zagers–Khoy–Akera suture can explain magmatism in the Caucasus and northwestern Iran more easily than the BZS. Furthermore, the northern SaSZ separated from the Anatolian plate via the Van microplate.

The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author.

All authors listed have made a substantial, direct, and intellectual contribution to the work and approved it for publication.

This research was supported by the Japan Society for the Promotion of Science (JSPS) KAKENHI Grant No. 17H01671.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.