Zircon U-Pb geochronology and emplacement history of intrusive rocks in the Ardestan section, central Iran

This work is licensed under Creative Commons Attribution-ShareAlike 4.0 International.

The NW-SE trending Zagros Mountains in Iran are a member of the Alpine-Himalayan orogenic belt and represent one of the youngest continental collision zones on Earth. The Zagros orogeny consists of three NW-SE trending parallel subdivisions from Northwest to Southeast: the Zagros Fold-Thrust Belt (ZFTB), the adjacent Sanandaj-Sirjan Zone (SSZ), and the Urumieh- Dokhtar Magmatic Arc (UDMA) (Alavi, 2004). The tectonic history of the Tethyan region has been studied by many authors (e.g. Stöcklin, 1974; Berberian and King, 1981; Mohajjel et al., 2003; Agard et al., 2011). From the Late Precambrian up to the Permian, part of Gondwana was separated from the Eurasian Plate by the Paleo-Tethys. During the Middle to Late Triassic, during the closure of the Paleo-Tethys in the North, rifting in the continental plate along the Zagros thrust zone occurred, resulting in the opening of a new ocean, the Neo-Tethys. During the Triassic–Jurassic, following the closure of Paleo-Tethys, the oceanic crust of Neo-Tethys started to subduct beneath the Eurasian Plate. This subduction progressively closed the Neo-Tethys ocean and during the terminal collision formed the Zagros orogenic belt of Iran (Berberian and King, 1981; Alavi, 2004).

Volcanism in the UDMA initiated during the Eocene and continued throughout this period (Berberian and King, 1981). Its composition has a common calc-alkaline to shoshonitic and alkaline affinity with geochemical and petrological features similar to those of Andean-type magmatism (cf. Berberian et al., 1982). The oldest igneous rocks in the UDMA are the calc-alkaline Shir-Kuh granitic complex, which cut across Upper Jurassic formations and are overlain unconformably by Lower Cretaceous fossiliferous limestone exposed in the southeastern margin of central Iran (Nabavi, 1972; Hajmolla-Ali et al., 2000). The youngest rocks in the UDMA consist of lava flows and pyroclastics which form Pliocene to Quaternary volcanic cones with alkaline to calc-alkaline compositions (Berberian and Berberian, 1981). It is suggested that the Plio-Quaternary volcanism formed as a result of: i) the modification of the geothermal gradients due to uplift and erosion, ii) the strike-slip shearing motion created by differential movement of fault blocks due to the continued convergence of Arabia and Eurasia, and iii) the existence of large strike-slip faults, which developed a region of tension at both ends of these blocks (Berberian and King, 1981).

The collision has been recently studied in terms of geophysics, kinematics, and neotectonics. It is generally admitted that the Neo-Tethyan subduction started during the Triassic in Iran (Berberian and Berberian, 1981; Wilmsen et al., 2009; Chiu et al., 2013). The timing of its ending, and the onset of the Arabia-Eurasia collision, has long been a subject of debate. Timing estimates of the collision varies from ~65 to ~5Ma, based on a wide variety of presumed geologic responses to collision, such as:

• ophiolite obduction in the Late Cretaceous and coeval development of a foreland basin on the Arabian margin (Berberian and King, 1981; Alavi and Mahdavi, 1994);

• late Eocene as based on structural, lithological, and palaeobiogeographical evidences from both sides of the original Arabia-Eurasia suture (Allen and Armstrong, 2008; Allen, 2009);

• early to Mid- Miocene transition from marine to non-marine sedimentation (McQuarrie et al., 2003);

• late Miocene/Pliocene influx of coarse clastics into the foreland basin, rapid cooling in the Alborz Mountain, and rapid subsidence of the Caspian Sea (McQuarrie et al., 2003);

• absence of sedimentation during late Eocene- Oligocene in western part of Central Iran, suggesting that the collision began between 35-25Ma, whereas oceanic subduction is still active (Agard et al., 2005);

• emplacement of the Kermanshah ophiolite, that was probably an ancient oceanic core produced by large oceanic detachment faults, indicates that final closure of the Neo-Tethys Ocean occurred in the Late Miocene (Ao et al., 2016).

More recently, Ballato et al. (2011) proposed a two- stage collision model that involves an initial collision in the Late Eocene and an acceleration of the regional deformation in the Early Miocene.

The precise timing of collision between Arabia and Eurasia is still controversial and the geological history of the magmatic arcs in the Zagros orogenic belts in Iran is still poorly known. The reconstruction of tectonic and magmatic episodes requires a precise knowledge of their radiometric ages and the characterization of the petrogenetic processes. An increasing number of geochronological results has been published in the last decade for various rock units in Iran, whereas radiometric age constraints on the timing of plutonism in the UDMA are scarce. The aim of this study is to constrain the timing of the emplacement of the Kuh-e Dom, Mehrabad, Nasrand, Zafarghand, and Feshark intrusive rocks located in the middle part of UDMA. Achieving this we seek to provide ing clues about the geodynamic history of the Ardestan magmatic segment, to address some of the most important questions regarding the genesis of the intrusive rocks in the UDMA.

Sixty representative samples from the Kuh-e Dom, Mehrabad, Nasrand, Zafarghand, and Feshark were collected on the basis of their mineralogy, freshness, and geographic distribution. The selected samples were analyzed for major and trace elements. Major elements were analyzed by X-ray fluorescence (XRF) (Rigaku RIX 2000 with a Rh end window tube) using fused glass disks at Naruto University in Japan. Glass beads from finely grounded samples were prepared with a sample-to-flux (Li.B.O.) ratio of 1:10 and analyzed for major elements using fundamental parameter method spectrometry with analytical errors <1%.

Six representative samples for U-Pb zircon dating were collected from the Kuh-e Dom (granodiorite and diorite), Mehrabad (granodiorite), Nasrand (granodiorite), Zafarghand (granodiorite), and Feshark (granodiorite) intrusions. The samples were 3–7kg in weight. Whole rock samples were crushed using jaw crushers. The crushed samples were then sieved to obtain the size fraction above 53 mesh (~300µm). All samples were washed with water to remove dust. After washing and drying, magnetic minerals, such as magnetite, were removed with a Franz magnetic separator. More than 100 zircon grains were separated from the samples by using standard mineral-separation techniques at University of Tehran, specifically heavy liquid (bromoform and methylene iodide), and finally purified by handpicking under a binocular microscope.

Zircon separates were mounted by using epoxy resin and polished. Prior to age determination, all zircon grains were investigated by cathodoluminescence (CL) to highlight zoning and possible inherited cores. U-Pb isotopic analyses of zircon were conducted by laser ablation inductively coupled plasma-mass spectrometry (LA ICP-MS) at the State Key Laboratory of Geological Processes and Mineral Resources, China University of Geosciences, Wuhan (see Simonetti et al., 2006). Detailed operating conditions for the LA system and the ICP-MS instrument and data reduction were the same described by Liu et al. (2008, 2010a, 2010b).

Zircon 91500 was used as external standard for U-Pb dating, and was analyzed twice every five analyses. Time- dependent drifts of U-Th-Pb isotopic ratios were corrected using a linear interpolation (with time) for every five analyses according to the variations of zircon 91500 (i.e. 2 zircon 91500 + 5 samples + 2 zircon 91500) (Liu et al., 2010a). Preferred U-Th-Pb isotopic ratios used for zircon 91500 were from Wiedenbeck et al. (1995). Uncertainty of preferred values for the external standard zircon 91500 was propagated to the ultimate results of the samples. Concordia diagrams and weighted mean calculations were made using Isoplot/Ex_ver3 (Ludwig, 2003).

Felsic rocks in the Kuh-e Dom, Mehrabad, Nasrand, Zafarghand, and Feshark intrusions, are mostly granular, medium grained and composed mainly of quartz, K-feldspar, plagioclase, biotite, and hornblende, with small amounts of accessory minerals, including zircon, apatite, titanite, as well as opaque minerals, such as magnetite, hematite, pyrite, and chalcopyrite. Quartz and K-feldspar are typically anhedral and occur as interstitial phases. Occasionally, K-feldspar phenocrysts are an present as large, irregular, poikilitic patches. They enclose fine-grained early, dominant phase. Green hornblende and brown biotite are common mafic minerals and form euhedral to subhedral prisms of variable size.

The mafic-intermediate rocks in these intrusions are granular and medium grained; they contain more mafic minerals and plagioclase, and less quartz and K-feldspar compared to the granodiorite rocks. Quartz and orthoclase are present in variable proportions. Both minerals are commonly late interstitial phases. Plagioclase occurs as euhedral to subhedral crystals, commonly with polysynthetic twining and normal and oscillatory zoning. Hornblende and biotite are the most common mafic minerals, whereas in the mafic units orthopyroxene occurs as rare grains in the Kuh-e Dom and Zafarghand intrusions and forms euhedral to anhedral crystals and usually contains small inclusions of plagioclase. Apatite, titanite, and opaque minerals (mostly titanomagnetite, hematite, pyrite, and chalcopyrite) appear as accessory phases.

Whole-rock analyses of major and trace elements of all samples are reported in Table I (Electronic Appendix, available at www.geologica-acta.com). We use the classification diagram of Middlemost (1994) for characterizing the rock types. The samples plot in the fields of gabbro to granite (Fig. 2). The majority of the samples are subalkaline in nature and display a calc- alkaline trend (Irvine and Baragar, 1971; Fig. 3A). In the FeO/(FeO+MgO) vs. SiO. diagram (Frost et al., 2001), all samples plot in the magnesium field (Fig. 3B). The A/CNK [Al.O./(CaO+Na.O+K.O) molecular ratio, or Aluminum Saturation Index (ASI)] values of Shand (1943), are the most useful chemical discriminant between metaluminous (ASI<1) and peraluminous (ASI>1) intrusive rocks; the vast majority of the studied intrusion are metaluminous to slightly peraluminous (Fig. 3C). The least-evolved members are predominantly metaluminous, although some more evolved samples exhibit slightly peraluminous signatures with A/CNK ratios ranging from 1.0 to 1.1. The

FIGURE 2
Silica vs. alkali diagram (Middlemost, 1994), showing the studied samples plotted in the fields from granite to gabbro.

FIGURE 3
A) Alkaline ternary diagram (Na2O+K2O-FeO-MgO; wt.%). Line separating the tholeiitic and calc-alkaline fields is from Irvine and Baragar (1971), B) SiO2 vs. Fe/(FeO+ MgO) diagram (Frost et al., 2001), and C) [Al2O3/(CaO+Na2O+K2O)]mol vs. [Al2O3/(K2O+Na2O)]mol diagram (Shand, 1943) for the granites of the Ardestan area.

ASI value of the studied samples is less than 1.1, which is associated mainly with I-type granites (Chappell and White, 1974; Chappell, 1999). Low ASI is consistent with the mafic mineral assemblage of amphibole±pyroxene±biotite in the studied samples (Chappell and White, 1974; White and Chappell, 1988). The peraluminous nature of a few samples may be attributed to hornblende differentiation (Zen, 1986) or heterogeneity of water content in the protolith (Waight et al., 1998).

A discrimination plot of Zr+Ce+Y+Nb vs. (Na.O+K.O)/ CaO and FeOt/MgO (Whalen et al., 1987) shows that the studied samples plot within the field of I- and S-type intrusive rocks and are distinct from the A-type granites of alkali nature (Fig. 4). The sharply linear negative trend between P.O. and SiO. (Chappell, 1999) and sharply linear positive trend between Th and SiO. from the most mafic to the most felsic compositions (Chappell et al., 1998) can be considered typical of I-type granite (Fig. 5). The oxidized I-type geochemical signatures of these intrusive rocks are also supported by their mineralogy, i.e. predominance of biotite and hornblende as mafic silicates, and the abundance of euhedral titanite and magnetite as accessory phases (Chappell and White, 1974). Lack of muscovite and monazite minerals is typical of I-type intrusive rocks. Apatite inclusions are common in biotite and hornblende, as are expected for I-type intrusive rocks and occur in larger individual crystals in the S-type varieties. The wide range of modal distribution from gabbro to granite, indicates the studied samples are I-type granitoids.

In the Mid Oceanic Ridge Basalt (MORB) normalized trace-element spider diagrams (Pearce, 1983), the five groups of intrusive rocks have similar trace-element patterns

FIGURE 4
Zr+Ce+Y+Nb vs. (Na2O+K2O)/CaO and FeOt/MgO discrimination diagrams of Whalen et al. (1987).

FIGURE 5
SiO2 vs. P2O5 and Th diagrams indicating typical trends of I-type granite.

(Fig. 6). These rocks are enriched in Large Ion Lithophile Elements (LILEs: Cs, Rb, Ba, and K) and strongly depleted in High Field Strength Elements (HFSEs: Nb and Ti). These features are typical of subduction-related magmas (Gill, 1981; Pearce, 1983; Wilson, 1989). Strong depletion in Nb and Ti relative to other HFSE has been explained in terms of: i) retention of Nb-Ti-bearing refractory phases (i.e. rutile, titanite, amphibole) at the site of partial melting (e.g.Briqueu et al., 1984); ii) melting of Nb-Ti-depleted source material, be it continental crust, island arc, or continental margin volcanic rocks; iii) recycling of ancient crustal material back into the mantle sources of magmas (e.g.Hergt et al., 1989; Sage et al., 1996); iv) metasomatized mantle wedge of volcanic arcs (Taylor and McLennan, 1985); and v) advanced fractional crystallization of Nb-Ti- bearing phases (ilmenite, titanite, etc.), while the negative P anomalies should result from apatite fractionation. In the Rb versus Y + Nb diagram (Pearce et al., 1984; Fig. 7), all the samples display geochemical characteristics typical of I-type magmas, formed in a volcanic arc setting, due to their relatively low values of Nb and Y (see Aliashrafzadeh, 2013; Ghahramani, 2013; Hamzehie, 2013).

One diorite sample from Kuh-e Dom (DK), and five granodiorite samples from Kuh-e Dom (GK), Mehrabad (MH), Nasrand (NS), Zafarghand (ZF), and Feshark (FS) were selected for zircon U-Pb geochronology (see U-Pb data in Table II). Zircon is a common accessory mineral in the studied rocks and forms prismatic crystals with a maximum size close to 300μm. The zircon crystals are automorphic to sub-idiomorphic, and display a well- developed oscillatory zoning indicative of growth under magmatic conditions (Koschek, 1993; Gao et al., 2007). Zircon is a reliable recorder of magmatic conditions in which it grew, because of its resistance to isotopic resetting at magmatic conditions and its ubiquitous presence in silicic magmas. In a few samples the occurrence of xenocrystic cores can be observed. To determine the age of emplacement of the granites in these samples, analyses were mostly carried out on the external domains of the zircons grains with magmatic textures.

Zircon grains have variable Th and U concentrations (Table II). Generally, a zircon with a Th/U ratio larger than

0.5 and positive correlation between Th and U is regarded as being derived from magmatic crystallization, while if the ratio is less than 0.1 it can be interpreted as metamorphic in origin (Belousova et al., 2002). In this study, their Th/U ratios (Table II) are interpreted as being representative of magmatic zircon. This interpretation is supported by the zircon shapes and textures; and by the positive correlation between Th and U (Fig. 8), that indicate a magmatic origin for these zircons (Hoskin and Black, 2000; Belousova et al., 2002; Xie et al., 2006).

FIGURE 6
MORB-normalized trace patterns for the studied samples (values from Pearce, 1983).

FIGURE 7
Trace-element geotectonic discrimination diagram (Pearce et al., 1984) for the studied granites. The granites plot in the volcanic arc granite field.

The results of calculating the isotopic age of the Kuh-e Dom granodiorite (GK), Kuh-e Dom diorite (DK), Mehrabad granodiorite (MH), Nasrand granodiorite (NS), Zafarghand granodiorite (ZF), and Feshark granodiorite (FS) are presented as concordia and average age graphics (Fig. 9) and the results are given in Table III. Therefore, the U-Pb zircon ages indicate that the Kuh-e Dom diorite and granodiorite were formed during the early to middle Eocene (Ypersian), the Mehrabad and Nasrand granodiorite during the late Eocene (Priabonian), the Zafarghand granodiorite during the late Oligocene (Chattian), and the Feshark granodiorite during the early Miocene (Burdigalian). Therefore, the plutonic rocks in the Ardestan magmatic segment reveal three main episodes of plutonic activity in the early-middle-late Eocene, late Oligocene, and early Miocene.

This study reports zircon LA ICP-MS U-Pb ages for the Ardestan intrusive rocks, emplaced in the central part of the UDMA. It is generally believed that the UDMA was overlying the subducting slab of the Neo-Tethyan oceanic lithosphere beneath the Iranian Plate (Berberian and King, 1981; Mohajjel et al., 2003; Agard et al., 2005). Geochemical studies indicate that these intrusions are oxidized I-type, metaluminous to slightly peraluminous, commonly attributed to the magnesian series of calcalkaline nature, and formed in a subduction-related setting, in accordance with the geological history of this area and in good agreement with other UDMA signatures.

According to our U–Pb results, the granodiorite and diorite of the Kuh-e Dom intrusion were emplaced at 51- 54Ma, during the early-middle Eocene, which is slightly older than previous dating by K-Ar (Technoexport, 1981), which yielded late Eocene ages. According to field observations, the intrusive rocks of Mehrabad, Nasrand, Zafarghand, and Feshark, in the Ardestan area, are intruded into early Eocene volcanic rocks (andesite, dacite latite, and tuffs), thus indicating a post-early Eocene age, in agreement with an Oligo-Miocene age estimated earlier (e.g.Aghanabati, 2004). The new isotopic data provide a revised age for these intrusive rocks, from the late Eocene (Mehrabad and Nasrand), in contrast with the earlier views, to late Oligocene (Zafarghand), to early Miocene (Feshark). Therefore, the Tertiary plutonism in the Ardestan area can be divided into three main episodes; i) early-middle-late Eocene, ii) late Oligocene and iii) early Miocene intrusions, which have been considered central part of the UDMA plutonism. The intrusive rocks in the Ardestan area were emplaced from early Eocene to early Miocene plutonic interval and the Zafarghand and Feshark intrusive rocks, which are closer to the trench, are the younger intrusive units in the Ardestan area. It is apparent that the plutonism was more intense during the Eocene for the central part of the UDMA belt, as shown also by Berberian and Berberian (1981).

The geochronologic data of the intrusive rocks studied here were compared with those available of plutonic rocks elsewhere in the UDMA. The overall data are listed in Table IV.

Chiu et al. (2013) suggested that magmatism in the UDMA was most active and widespread in the Eocene and Oligocene, and magmatic activity lasting from ca. 21 to 6Ma ending with small-volume intrusions. Cessation of arc magmatism took place progressively from the Northwest to the southeast. This southeastward termination of the UDMA magmatism is consistent with the notion of oblique and thus diachronous collision between Arabia and Eurasia, with the collision initiating in the Northwest and propagating progressively to the Southeast along the Zagros suture zone (see Chiu et al., 2013).

FIGURE 8
The positive correlation between Th and U, indicating a magma origin of zircons.

FIGURE 9
Concordia diagrams displaying U-Pb data from A) the Kuh-e Dom diorite, B) Kuh-e Dom granodiorite, C) Mehrabad granodiorite, D) Nasrand granodiorite, C) Zafarghand granodiorite and D) Feshark granodiorite.

From a geochronological point of view, the Aredestan intrusive rocks are broadly similar to the other intrusive rocks in the UDMA spanning from 81.0 to 5.4Ma. This semi-continuously activity all along the UDMA can be related to the persistent subduction of the Neo-Tethys oceanic crust below the Iranian plate. The Kuh-e Dom, Mehrabad, Nasrand, and Zafarghand were emplaced during the Eocene-Oligocene epoch during the maximum magmatic activity. Another phase of plutonism in the Ardestan area was responsible for the emplacement of the Feshark intrusive rocks. Considering most of the researchers agree that the timing of the continental closure is Miocene (e.g. Dewey and Sengör, 1979; Robertson, 2000; McQuarrie et al., 2003; Mohajjel et al., 2003; Homke et al., 2004; Guest et al., 2006;Omrani et al., 2008; Ao et al., 2016). It seems that the Eocene-Oligocene intrusive rocks (i.e.Kuh-e Dom, Mehrabad, Nasrand, and Zafarghand) in the Ardestan area were emplaced prior to the collision, showing a pre-collisional volcanic arc setting, and that the younger intrusion (e.g. Feshark) was formed relatively late in the subduction history and may represent post-collisional plutonism.

The petrological, geochemical, and geochronological study of the Ardestan intrusive rocks reveals important clues to decipher the complex magmatic processes in the central section of the Urumieh-Dokhtar magmatic arc belt. All magmatic phases forming the intrusive rocks of the Ardestan area were metaluminous to slightly peraluminous, typical of I-type granites and belong to magnesian series of calc-alkaline nature and display the geochemical characteristics typical of oxidized I-type volcanic arc granitoids.

The LA ICP-MS U-Pb zircon geochronology of the Aredestan plutonic rocks documents two major intrusive phases. An Eocene-Oligocene subduction- related magmatism formed the Kuh-e Dom, Mehrabad, Nasrand, and Zafarghand intrusive rocks. The youngest lower Miocene Feshark intrusion probably represents post-collision magmatism within the central part of the UDMA belt.

Thanks go to the University of Kurdistan for supporting this project by means of grants provided by the research council (Number: 4.64500; date: 3/11/2015). We are grateful to Prof.

K. Zong at China University of Geosciences, who prepared the U-Pb-Th isotopic data. We thank David Lentz for comments on earlier versions of this manuscript. Thanks are extended to Prof. H. Azizi (University of Kurdistan). We acknowledge Prof. Antonio Castro, Laura Rincón, and anonymous reviewers for their constructive comments leading to an important improvement on the manuscript.