Petrology and geochemistry of Plio-Quaternary high-Nb basalts from Shahr-e-Babak area: Insights into post-collision magmatic processes in the Kerman Cenozoic Magmatic Arc

Post-collision Pliocene-Quaternary basaltic rocks outcrop in the Kerman Cenozoic Magmatic Arc (KCMA) to the northwest and east of Shahr-e-Babak city. These porphyritic and vesicular basaltic rocks are composed essentially of clinopyroxene, olivine, and plagioclase. These basalts display alkaline affinity and negative Ta, Zr, Rb anomaly, but slightly negative Nb anomaly, relative to elements with similar compatibility, and positive Ba, K, Sr anomaly, suggesting their magma source related to subduction-accretion with implication of subducted slab derived components to the source. In the primitive mantle and chondrite normalized diagrams, these rocks show trace elements (except depletion in Nb, Ta) and Rare Earth Element (REE) patterns similar to the Ocean Island Basalts (OIB) and share trace and major element characteristics similar to High-Nb Basalts (HNBs). Geochemical analyses for major and trace elements suggest that the Shahr-e-Babak HNBs have undergone insignificant crustal contamination and minor olivine + Fe-Ti oxide ±clinopyroxene fractional crystallization. These HNBs derived from a partial melting (~5%) of garnet-peridotite mantle wedge, which have already metasomatized by overlying sediments, fluids, and adakitic (slab-derived) melts as major metasomatic agents in post-collision setting in the KCMA. We conclude that asthenospheric upwelling arising from slab break-off followed by the roll-back of subducting Neotethys slab also triggered metasomatized peridotite mantle wedge and caused its partial melting in the subduction zone.

showing location of study area; C) Simplified geological map of the study area, northwest and east of Shahr-e-Babak (modified from Geological map of Iran, 1:100000 Sardic et al., 1971).
The study of post-collisional magmatic mafic rocks especially HNBs provides a unique opportunity to identify the asthenospheric-lithospheric properties (Carlson et al., 2005), geodynamic processes, and tectonic settings (e.g. Castillo, 2012;Dilek and Furnes, 2011). The tectonic and magma evolution of the UDMA has been studied widely (e.g. Khaksar et al., 2020;Moradi et al., 2021;Nazarnia et al., 2018). However, very few investigations on the petrology and geochemistry of the alkaline basalts and HNBs, with or without associated adakites, have been published. The aim of this study is to review the petrographical and chemical composition of the Shahr-e-Babak HNBs, in the Kerman Cenozoic Magmatic Arc (KCMA), in order to provide new constraints on the conditions of their genesis and source region. The study herein suggests a tectono-magmatic model to explain the evolution of the post-subduction arc magmatism in the Shahr-e-Babak area.

Urumieh-Dokhtar Magmatic Arc
The Zagros Orogenic Belt (ZOB), in the central part of the Alpine-Himalayan orogenic belt, generated as a result of the NE-dipping subduction of the Neo-Tethys oceanic lithosphere, accretion and subsequent collision of the Arabian plate with the CIMB (e.g. Alavi, 1994;Berberian and King, 1981). The Zagros Fold and Thrust Belt (ZFTB) (Berberian and King, 1981;Mohajjel and Fergusson, 2000), the Sanandaj-Sirjan Zone (SSZ), and the UDMA are the three major sub-parallel tectono-stratigraphic structures in the ZOB (Fig. 1).
The SSZ is separated from the ZFTB to the southwest and from the UDMA to the northeast by the main Zagros thrust fault and fore-arc depressions respectively (Alavi,1994;Glennie, 2000).
The UDMA is considered to be an active continental margin (Moin Vaziri, 1991;Takin, 1972;Verdel et al., 2011) about 1000km long, 50km wide and 4km thick , extending from NW to SE in central Iran. The magmatism in the UDMA is generally composed of subduction-related voluminous calc-alkaline, and locally tholeiitic rocks (e.g. Azizi and Jahangiri, 2008;Omrani et al., 2008) from Cretaceous to Quaternary age. However, the peak of magmatic activity is thought to have been in the Eocene with a notable magmatic flare-up from ~55Ma to ~37Ma (Verdel et al., 2011).
The younger volcanic activity in the UDMA was mainly alkaline (Amidi et al., 1984;Moradian, 1997) and associated with post-collisional tectono-magmatic processes (Richards, 2003). The timing of collision is controversial, some authors suggested it was before or during the late Miocene (e.g. Allen et al., 2004;Mohajjel et al. 2003,) while other, (Hassanzadeh,1993), proposed the late Miocene and/or the late Neogene. According to Berberian and King (1981) and Jahangiri (2007), the onset of the alkaline volcanic activity in the UDMA (6-5 Ma) was due to the sinking of the final broken pieces of oceanic slab to a depth where alkaline melts were generated.
The distinctive southeastern part of the UDMA is known as the Dehaj-Sarduiyeh volcano-sedimentary belt (Dimitrijevic, 1973) or KCMA (Fig. 1A). The KCMA is about 450km long and 80km wide (Dewey et al., 1973) and mainly composed of calc-alkaline volcano-plutonic rocks (Zarasvandi et al., 2007). It host porphyry copper deposits, several of them of large size and many of small-medium size (Taghipour, 2007;Zarasvandi et al., 2011). The volcanic and plutonic activities in the KCMA are considered to have reached their peaks in the middle Eocene and Oligocene-Miocene, respectively (Hassanzadeh, 1993).
There are also a series of northeast-southwest-trending faults, with small inferred displacements, developed in shear zone associated with the Nain-Baft and Rasanjan faults ( Fig. 1A) that likely have controlled ascent and emplacement of magmas and ore-deposit formation in the Shahr-e-Babak area.
As a whole, it seems that the Tertiary volcanic rocks in Shahr-e-Babak area formed during three stages: i) an explosive volcanic event including pyroclastic material (ash and spinel bombs), ii) an efusive basaltic lava distributed around cones and vesicular basaltic rocks (scoria) and iii) an explosive event with vesicular basaltic rocks and bombs erupted in the final stage of the volcanism.

FIELD DESCRIPTIONS AND PETROGRAPHY
The Plio-Quaternary volcanism in Shahr-e-Babak area is represented by small basaltic scoria cones and lava flows. These volcanic rocks are exposed in two areas, where they overlie unconformably Neogene volcanic rocks (Saric and Djordjevic, 1971) and are usually covered by thick layers of Quaternary sediments ( Fig. 2A). They include vesicular basaltic rocks, lapilli, tuff, volcanic bombs and basaltic lavas.
The Shahr-e-Babak basaltic rocks are generally highly porphyritic. They show vesicular texture and are dark gray or brown to black colour with a red leaching patch, that most probably formed due to the presence of iron oxides/hydroxide phases during evolution of these rocks (Fig. 2B). Different types of volcanic bombs are seen in the studied area, e.g. bread-crust bombs, cylindric bombs, cannonball bombs, spindle bombs, pear and toothpaste shaped bombs ). Bombs show relatively low porosity (~20%) as compared with other basaltic rocks, and sometimes include partly epidotized fragments from the Naein-Baft ophiolite mélange and particles of rock that are probably from the cryptodome intruded immediately before the volcanic eruption (Fig. 2F). The Shahr-e-Babak basaltic rocks have been classified as olivine basalt and tracky basalt. Porphyry texture is the dominant texture of these rocks, though glomeroporphyry and sieve textures in plagioclase and, rarely, in clinopyroxene are also observed ( Insights into source and magmatic processes post-collisional high Nb-basalts 5 plagioclase, small olivine, clinopyroxene, and opaque minerals with amygdaloidal texture, rarely filled by calcite and quartz (Fig. 3C). The vesicles show primary porosity of the rocks which are generated via degassing during lava extrusion (Navarro et al., 2020).
Olivine phenocrysts are homogeneously distributed in the groundmass; however, in some places, they form glomeroporphyry texture along with clinopyroxen minerals. They are subhedral to euhedral, mostly fresh but some are altered to iddingsite along the margins and cracks (Fig. 3D). Clinopyroxene phenocrysts (augite) are the second most abundant mafic mineral in these basaltic rocks. They are mostly euhedral, some subhedral, up to 1-5mm long, and fresh. The plagioclase is confined to the groundmass with length <0.5mm as small laths, fresh and show albite and Carlsbad twinning with weakly to strongly develop trachytic texture (Fig. 3A).

ANALYTICAL METHOD
Based on microscopic observations, we selected 15 samples of basalt for whole-rock chemical analysis to be performed in the analytical laboratories of the SGS Minerals, Toronto, Canada. Major oxide abundances were determined by X-Ray Fluorescence (XRF) and recalculated to 100%, volatile free (Table 1). Approximately a 50g split of each sample was pulverized to fine powder in an agate ring mill. The powdered samples were dried at 75-90°C to eliminate adsorbed water, ignited at 950±50ºC and then fused with 50% lithium metaborate (LiBO 2) and 50% lithium tetraborate (Li 2 B 4 O 7 ) in a fluxer to produce a glass disc. The glass disc was analyzed on the sequential XRF spectrometer. Quantitative determination was made through previously prepared calibration standards. Data reduction was done using laboratory information management system software that performs all necessary calculations automatically to calculate the percent oxide for  Insights into source and magmatic processes post-collisional high Nb-basalts 6 each element and the percent total major-element content of the sample, including Loss On Ignition (LOI). The detection limit was 0.01wt% of the oxide.
Trace elements of the bulk rock were analyzed with Inductively Coupled Plasma-Mass Spectrometry (ICP-MS). Approximately 0.1g of powdered rock samples were fused by Na-peroxide in graphite crucibles and dissolved using dilute HNO 3 . The fused solution was aspirated into the ICP-MS where the ions were measured and quantified according to their unique mass. The used detection limit was vary during analysis for different elements. It was 5ppm for Cu, Ni, V; 2ppm for Mo; 1ppm for Ag, Ga, Hf, Nb, Sn, W; 0.5ppm for Co, Ta, Tl, Y, Zr; 0.2ppm for Rb; 0.1ppm for Ce, Cs, La, Nd, Sm, Th, Yb and 0.05ppm for Dy, Eu, Er, Gd, Ho, Lu, Pr, Tb, Tm.

Whole-rock geochemistry
The major and trace element concentrations in the samples are presented in Table 1    The LOI values are 0.9 to 3.4wt%, except in one sample with 4wt% (Table 1). These rocks are classified as basalt and trachybasalt in the SiO 2 versus total alkali (Na 2 O+K 2 O) diagram (Fig. 4A). In the Zr/Ti versus Nb/Y diagram (Fig.  4B), the studied samples plot in the alkali basalt field. The samples plot in the alkaline series in the SiO 2 versus K 2 O diagram (Fig. 4C) and slightly fall within the transition field between potassic and sodic series in the Na 2 O versus K 2 O diagram (Fig. 4D).
Primitive mantle normalized (Sun and McDonough, 1989) spider diagram of trace-element abundances for the Plio-Quaternary basalts in the studied area show that these rocks are enriched in LILEs compare to HFSEs (Fig. 5B). Most of the samples show positive Sr, K, Ba anomaly and negative Zr, Rb, and Ta, but slightly negative Nb anomaly. The studied basaltic rocks are considered as HNB with Nb concentration from 20 to 30ppm.
The plot of REE normalized by chondrite (Sun and McDonough, 1989; Fig. 5A) shows an enrichment of LREE than HREEs for the Shahr-e-Babak HNBs with La/ Yb CN = 9.7-13 and around 100 to 200 times chondrite for the LREE and about 10-20 times chondrite for the HREE.
The REE content in these samples is high, ranging from 135.2 to 194.7ppm (average= 150ppm). The average values of Eu*/Eu and Ce/Ce* are 0.97 and ~1, respectively, which basically do not show a Eu anomaly. All samples show similar trends, suggesting similar REE fractionation degrees and magma source. They are well correlated with Ocean Island Basalt (OIB) (Fig. 5A)

Role of fractional crystallization
The Mg# values of the Shahr-e-Babak HNBs (57,1-66,8) do not show a significant difference from those of the original basaltic magma (65-70) (Frey et al., 1978), suggesting that the basaltic magma did not experienced significant fractional crystallization.
The low SiO 2 content (45.4-49.5wt%), high MgO content (≥6.6wt%), and total FeO content (>11wt%) are also distinctive features that indicate slight fractionation nature of these samples. Thus, the Shahr-e-Babak HNBs probably are not representative of primary melts (Wilson, 1989), but their compositions do not reveal significant modification through fractionation of mantle-derived melts. Considering of the relatively high Ni (155-244ppm) content in the studied HNBs than unfractionated mafic magmas (Ni= 200-450ppm) ( Table 1) also rule out the highly fractionated olivine as the important mafic phases in basaltic rocks. Moreover, the studied HNBs show a high content of Cr (234-374ppm) compared to the reported values for the primary magmas (142ppm; Hughes, 1982), which provide a good geochemical indicator of slight clinopyroxene fractionation during magma evolution. These evidence show that they can considered as enough primary to investigate the magma source and genesis in this region.
The elevated concentrations of Ba and Sr together in the samples compared to the average OIB (Fig. 5B) is probably attributed to incompatible behavior of these elements during fractionation of mafic phases (i.e. olivine and clinopyroxene). The absence of a Eu anomaly also suggests that they probably crystallized much less than about 18% plagioclase prior to extrusion (John et al., 1968). Geochemical modeling of the studied HNBs (Fig.  6A) suggests a clear olivine and magnetite fractionation during magma evolution. However, these modifications are minor and insignificant.
The La versus La/Yb diagram (Fig. 6B) show that most of the samples follow a positive correlation with the partial melting trend, while three samples are plotted along both partial melting and fractional crystallisation trends and two samples (from KH1) are plotted along fractional crystalization. The studied samples also are plotted as a cluster possibly affected by partial melting, in the high La/Sm ratio versus La concentration (Fig. 6C), imply that these trace elements are more likely controlled by partial melting than fractional crystallization during evolution of magma in this area.

Effects of crustal contamination
It is well known that when hot basaltic magma traverse through continental crust some chemical components from the crust may play a significant role in the petrogenesis of the magma-derived rocks (Ashwal, 2021). The studied HNBs exhibit enrichment in LILE and LREE but depletion in Nb, Ta and HREE, as characteristic feature of magmas in active continental margin volcanism (Nagudi et al., 2003) and subduction zones (Wilson, 2007). Thus, these HNBs may derived from a metasomatized mantle wedge above the subducting oceanic crust with negative anomalies of HFSEs (i.e. Nb and Ti Wilson, 2007) and/or affected via the contamination and magma mixing with crustal material during magma ascent and emplacement (Wilson, 2007).  Sun and McDonough, 1989). Symbols are the same as in Figure 4.
Trace element ratios have been routinely used to investigate crustal contamination in basaltic rocks. Th/ Yb and Ta/Yb ratios are almost independent of fractional crystallization and/or partial melting, thus they can reveal source variations and crustal contamination. Source region metasomatism caused by subduction processes, results enrichment in Th and higher Th/Yb ratios than Ta and Ta/Yb respectively. Crustal contamination also increases Th/Yb ratios relative to Ta/Yb ratios due to the higher abundance of Th relative to Ta in the crustal rocks (Fadaeian et al., 2022). In the Th/Yb versus Ta/Yb diagram (Pearce, 1983) (Fig. 6D), the HNBs are plotted with Th/Yb ratios higher than the mantle array. However, it is difficult to distinguish the effects of crustal contamination on the magma composition from metasomatism caused by subduction components. Thus, the high Th/Yb ratio in most samples is unlikely to be the result of crustal contamination solely.
On the other hand, Th/Ta ratio ranges from 4 to 7, which fall between the original mantle ratio (Th/Ta= 2.3) and continental crustal (Th/Ta= 10) ratio (Sun and McDonough, 1989), which indicates insignificant crustal contamination in the evolution of the HNBs.
Moreover, low Nb/La ratios (<1.0) in the basaltic rocks are a key index and reflect crustal contamination in the magma evolution (Kieffer et al., 2004). The studied HNBs from KH1 location have Nb/La ratios from 1 to 1.  Sun and McDonough (1989); UC (Upper crust)is from Taylor and McLennan (1985); Fields for Karacadag and Hatay are from Parlak et al. (2000), Field for the basaltic volcanism of the Arabian plate is from Lustrino and Wilson (2007). Western Anatolia field is from Aldanmaz et al. (2007). Symbols are the same as in Figure 4. o l o g i c a A c t a , 2 0 . 8 , 1 -1 9 ( 2 0 2 2  Insights into source and magmatic processes post-collisional high Nb-basalts 10 indicating that crustal signature is insignificant in these rocks, but those basalts from KH2 location with Nb/La ratios from 0.6 to 0.9 reflect some crustal contamination. In the Nb/Th versus Nb/La (Fig. 7A), and Rb/La versus Th diagrams (Fig. 7B), all the samples plotted close to the mantle composition, ruling out significant crustal contamination. Zr/Nb ratio can be used to determine the influence of continental crust on the mantle derived magma. The Zr/Nb ratio ranges from 3.5 to 5.6, in the HNBs being this range similar to the range observed in the OIB mantle sources (3.2-5), lower than those of continental crust (16.2) and primitive mantle (14.8), and far lower than normal-MORB (30) (Saunders et al., 1988;Weaver, 1991a), which ruled out significant crustal contamination. Therefore, more likely the composition of the studied HNBs was mainly controlled by compositional differences in the source region and by mantle partial melting.

Role of the slab components
The Shahr-e-Babak HNBs formed in a continental margin environment and are different from continental intra-plate basalts in term of trace elements (Fig. 6D). Subduction-related magmas are characterized by high Ba/ Nb (>28) and Ba/Ta (>450) ratios (Fitton and Dunlop, 1985;Gill, 1981). These ratios range from 23 to 109 and 720 to 3633, respectively in the studied samples which, indicate that the subduction components play an important role in the source region.
In order to determine the influence of subduction materials on the source of the basaltic samples, we used the Th/Nb versus Ba/Th diagram (Orozco Esquivel et al., 2007) (Fig. 8A). The variable Th/Nb along with variable Ba/Th ratios is resulted from the addition of both hydrous fluid and melt to the source. Similar behaviors are also inferred from Th/Nb versus La/Nb plot (Fig. 8B). The Th/ Nb versus La/Nb diagram is used to distinguish between melt (high Th/Nb and low La/Nb) and fluid (low Th/Nb and high La/Nb) components as slab-derived material in the subduction zones. In this diagram, the studied HNBs indicate a transitional nature between the melt and fluid trends (Fig. 8B). Hastie et al. (2013) also developed the Th/La versus (Ce/Ce*)N diagram (Fig. 8C) to determine the slabderived components in the petrogenesis of arc related magmatic rocks. The studied samples are plotted in the volcanic detritus field on a subvertical trend corresponding to a mixture between continental detritus/GLOSS II and N-MORB. This confirms that the mantle source region has been enriched by melts derived from subducted sediments in the studied area.
According to Ayers (1998), the presence of longlasting phases containing Nb, Ta, Ti elements such as rutile, sphene, apatite and ilmenite in eclogite-facies rocks in subducted oceanic crust or non-melted mantle wedges can be considered a reason for negative anomalies of compatible elements in the mantle wedge derived magmatism. Therefore, these evidences suggest slab melting also played a critical role in the genesis of the Shahr-e-Babak HNBs.

Source of the Shahr-e-Babak High-Nb basalts
Trace element compositions of Shahr-e-Babak HNBs, especially their enrichment in the LILEs and strong fractionation between LREE and HREE are similar to OIB (Fig. 5B), but they show no obvious negative Nb anomalies or even positive Nb anomalies similar to OIB.
Two different types of HNBs have been proposed in the volcanic arcs: i) HNBs originated from the OIB source Th versus Rb/La diagram (Taylor and McLennan, 1981). LC= Lower crust, UC= Upper crust, PM= Primitive mantle Symbols are the same as in Figure 4. mantle (similair enriched, OIB-like isotopic and OIBlike trace element signature; Castillo, 2008;Gazel et al., 2011) and ii) HNBs from enriched mantle wedge that was metasomatized by slab-derived adakites in amphibolite or eclogite facies (similair N-MORB like isotopes and OIB-like trace elements; Defant et al., 1992;Imaoka et al., 2014;Straub et al., 2013). The enriched (plume-type) mantle source (first HNB type) is achieved either through influx of asthenospheric mantle through slab windows (Castilli et al., 2008) or mixing between enriched (plumetype) and depleted components within mantle wedge (e.g. Macpherson et al., 2010).
Adakitic plutonism and volcanism in the vicinity of the studied HNBs have been reported in the Shahr-e-Babak area. The source of adakite series in the Dehaj area are considered as a partial melting product of eclogitized mafic lower crust . Adakitic plutonic rocks in northwest of Shahr-e-Babak (Ghadami and Nazarnia, 2022) and the Meiduk and Parkam (Alirezaei et al., 2017) are attributed to partial melting of garnet bearing to amphibolitic lower continental crust and lithospheric mantle respectively. Ghadami et al. (2008) suggested a slab melting mechanism for post-collisional Plio-Pleistocene adakitic volcanism from Javazm, northwest of Shahre-Babak. The Plio-Pleistocenic subvolcanic porphyritic andesitic-dacitic domes in Anar-Dehaj (Shaker ardakani, 2016) with adakite signitures also attributed to slab melting and underplating of basaltic magmas under thick Plio-Pleistocene continental crust.
Partial melting of subducting oceanic slab in the amphibolite-eclogite transition (i.e. in low-water fluid conditions) led to titanite stability in the residual slab (König et al., 2010). The residual titanites liberate Nb into a slab melt than Ta preferentially . Moreover, rutile also preferentially retains Ta over Nb, beacuse DNb is lower than DTa in rutile/melt pair during high pressure melting (Xiong et al., 2011). Thus Nb will be released into slab derived melt and subsequently, Nb will be enriched in mantle wedge after metasomatism by flux from hot oceanic lithosphere . Moreover, adakitic magmas (slab melt) are characterized by high Sr/Y Insights into source and magmatic processes post-collisional high Nb-basalts 12 and (La/Yb)N ratios and low YbN values. Thus, when the mantle wedge is metasomatized by adakitic magmatism, it also should show similar chemical compositions to adakite (Tang et al., 2010;Wang et al., 2007). Chemically, the studied HNBs also show similarity to a mantle wedge metasomatized by adakitic melts with high Sr/Y.
As discussed in the crustal contamination section, trace elements have not been affected by significant fractional crystallization. Therefore, we used trace and rare earth element ratios as key parameters to determine the studied HNBs source. These samples with low Zr/Nb coupled with high Zr/Y and Nb/Y ratios (3.5-5.6, 5.5-6.3 and 1.1-1.7 respectively), are different with those of Depleted Mantle (DM) and its related melts (e.g. N-MORB and oceanic arc basalt) but resemble the source of OIB.
Different sources have been proposed for OIB magmas. For example many researchers consider the fertile lower mantle peridotite (e.g. Woodhead, 1996), while others suggest a lithospheric mantle source contaminated by carbonatite/plume derived melts as source the OIB magmas (Mazahari, 2015;Nakamura and Tatsumoto, 1988). Plotting of the samples in the Zr/Y versus Nb/Y diagram (Fig. 9A) suggests that they are more likely derived from OIB-like mantle. However, other distinct geochemical index of the studied HNBs, i.e. Nb/U (14.9-32.3 with average 23.7), displays compositional differences with both N-MORB and OIB (Nb/U 50;Hofmann, 2004).
The similarity between this study and recent works (e.g. Castillo et al., 2007;Sajona et al., 1996) demonstrate that unlike OIB lavas, HNBs are not generated from a pure mantle plume source or metasomatism of the mantle by carbonatite/ mantle plume melts. Consequently, Shahre-Babak HNBs were generated within garnet lherzolite mantle wedge (in asthenospheric mantle) metasomatized by fluids, overlying sediments and adakitic melt derived from subducted Neotethys oceanic crust.
Ce/Yb ratio in basaltic rocks is very sensitive to the lithospheric thickness (Ellam, 1992) due to their high stability during fractional crystalization. High Ce/Yb ratios   (Aldanmaz et al., 2000). OIB and MORB values are from Sun and McDonough (1989); C) Zr/TiO 2 versus Ce/P 2 O 5 tectonic setting discrimination diagrams (Muller et al., 1997). Symbols are the same as in Figure 4..
(29.2-49.3) in the studied samples also correspond to smaller melt fractions and/or garnet control (depth higher than 110-120km in the studied area).
In summary, based on the combined interpretation of gravity, topography data sets (Monilaro et al., 2005), geophysical-petrological methodology (Tunnini et al., 2015), and trace elements geochemistry (Ce/Yb, and Nb/La (0.6 to1.1)), it appears that the mantle source for the HNBs in Shahr-e-Babak area is located in the upper portion of the mantle asthenosphere.

GEODYNAMIC IMPLICATIONS
Post-collision alkaline magmatism was produced by a complex combination of geodynamic and petrogenetic processes during the evolution of the Alpine-Himalayan collision belt, from west-to-east though Turkey, Iran and into Pakistan Azizi et al., 2014;Mazahari, 2015;Neill et al., 2015).
The most efficient geodynamic mechanisms during generation of post-collisional alkaline and HNBs in subduction zones are mantle plumes and/or upwelling asthenospheric mantle (e.g. Liu et al., 2018;Niu et al., 2012;Wang et al., 2015) associated with local extension regime or lithospheric delamination mechanisms (Guo et al., 2007), and/or the opening of slab detachment (slab break-off) within modern arc-trench systems (Gazel et al., 2009;Gill, 1984;Keskin, 2003) Both proposed mechanisms provide a way to initiate partial melting of mantle wedge.
A mantle plume usually leads to a dynamic uplift over an area 1000-2000km in diameter (Hill et al., 1992;Ritter and Christensen, 2007), which is not seen in the studied area. Lithospheric delamination leads to the upwelling of the hot asthenosphere and melting of the asthenosphere in shallow depth close to the Moho, but the Shahr-e-Babak HNBs formed at considerable depth (>110km) and also the geophysical studies (Clark, 1993;Saric and Mijalkovic 1973) also show that the crust beneath the NW of the KCMA has a thickness of 45-55km. Hence, the lithospheric delamination mechanism is not also favoured for the genesis of the studied HNBs.
The occurrence of adakitic volcanism in the vicinity of the Plio-Quaternary HNBs in Shahr-e-Babak area shows that break-off episodes played an important role in their generation following a Paleogene slab rollback as suggested for the UDMA mamatism Moradi et al., 2021;Verdel et al., 2011). On the basis of the data presented herein and the occurrence of adakitic rocks in the NW of the KCMA, the following scenario could be suggested for HNBs in this area: After collision of the Arabian and Iranian blocks caused by slab pull force during subduction of the Neotethyan oceanic lithosphere, the slab break-off occurred due to an approximately 30% decrease in the velocity of the Arabian plate (Mouthereau et al., 2012;Verdel et al., 2011). A modeling from Molinaro et al. (2005) based on the gravity, geoid and topography data sets also highlights a thinned lithospheric mantle below the KCMA which is attributed to recent slab break-off. Oceanic slab break-off provides a reasonable explanation for the origin of post-collisional magmatism (Davies and von Blanckenburg, 1995). Slab break-off had led to thermal perturbation by upwelling of hot asthenosphere which has prepared the appropriate conditions for partial melting of amphibolite or eclogite from detached subducting slab Martin, 2005) to produce adakitic magmas (Ahmadzadeh et al., 2010;Azizi et al., 2014;Ghadami et al., 2008;Ghasemi and Talbot, 2006;Kouhestani et al., 2017;Omrani et al., 2008). Subsequently the adakitic liquids metasomatized mantle wedge for a long period of time. The thermal flux originating from the deep asthenosphere and uprising through the oceanic slab break-off, probably provided the excess heat, triggered the partial melting of metasomatized mantle wedge, and produced the HNB magmas in the Shahr-e-Babak during Plio-Quaternary (Fig. 10). This geodynamic model has been proposed already for other post-collisional alkaline magmatic rocks widely distributed throughout the Tethyan orogenic belt (e.g. Aguillon-Robles et al., 2001;Azizi et al., 2014;Kepezhinskas et al., 2022;Hastie at al., 2011;Hussain et al., 2020;Mazhari, 2015;Wang et al., 2008;Zhu et al., 2018).

CONCLUSIONS
i) The post-collision Plio-Quaternary Shahr-e-Babak HNBs contain mainly olivine, clinopyroxene and plagioclase, and generated in a continental margin (extensional) setting. These alkaline basalts, show a high Nb content (>20) and geochemical features comparable with those of OIB.
ii) The geochemical signatures of the samples suggest that the HNBs derived from a low degree partial melting (<5-7%) of the garnet peridotite mantle wedge metasomatized by adakitic melts at considerable depth (>110).
iii) A slab break-off model is suggested to explain direct asthenosphere heat flow for melting of subducting slab to produce adakitic slab melts and also trigger mantle source of the HNBs.