ORIGINAL PAPER

Tschermak fractionation in calc-alkaline magmas: the EoceneSabzevar volcanism (NE Iran)

Federico Lucci1 & Federico Rossetti1 & John Charles White2 & Hadi Shafaii Moghadam3&

Alireza Shirzadi4 & Mohsen Nasrabady5

Received: 21 March 2016 /Accepted: 12 July 2016# Saudi Society for Geosciences 2016

Abstract Calc-alkaline arc magmatism at convergent platemargins is volumetrically dominated by metaluminous andes-ites. Many studies highlighted the importance of differentiationvia fractionation processes of arc magmas, but only in the lastdecades, it has been demonstrated that not all rock-formingminerals may affect the evolution of calc-alkaline suites. Inparticular, a major role exerted by Al-rich hornblende amphi-bole as fractionating mineral phase has been documented inmany volcanic arc settings. The aim of this work is to under-stand the role of the Tschermak molecule (CaAlAlSiO6) hostedin the hornblende and plagioclase fractionation assemblage indriving magma differentiation in calc-alkaline magmatic suites.We explore this issue by applying replenishment–fractionalcrystallization (RFC) and rare earth element–Rayleigh fraction-al crystallization (REE-FC) modeling to the Sabzevar Eocene(ca. 45–47 Ma) calc-alkaline volcanism of NE Central Iran,

where hornblende-controlled fractionation has been demon-strated. Major element mass balance modeling indicates RFCdominated by a fractionating assemblage made of Hbl52.0–52.5 + Pl44.1–44.2 + Ttn3.3–3.9 (phases are expressed on total crys-tallized assemblage). REE-FC modeling shows, instead, a low-er degree of fractionation with respect to RFC models that isinterpreted as due to hornblende and plagioclase resorption bythe residual melt. Calculations demonstrate that fractionation ofthe Tschermak molecule can readily produce dacite and rhyo-lite magmas starting from a calc-alkaline andesite source(FC = ca. 30%). In particular, the Tschermakmolecule controlsboth the heavy rare earth elements (HREE) and light rare earthelement (LREE) budgets in calc-alkaline differentiation trends.

Keywords Calc-alkaline magmatism . Hornblendefractionation . Tschermak . Fractional crystallizationmodeling . Iran

Introduction

Convergent plate margins are sites of intensive magmatism asa part of the subduction factory processes at volcanic arcs(e.g., Tatsumi 2005; Pichavant and Macdonald 2007; Zhanget al. 2011). Andesite is the volumetrically dominant magmat-ic product at convergent settings, and therefore, it is consid-ered to exert a primary role in the growth, evolution, anddifferentiation of the continental crust (Stern 2002, 2011;Kelemen et al. 2003; Rudnick and Gao 2003; Tatsumi 2005;Kimura et al. 2010; Larocque and Canil 2010; Stern 2011;Takahashi et al. 2013).

Petrological and geochemical studies on arc magmas havedocumented the importance of fractionation (FC and RFC)and assimilation/fractionation (AFC) processes in magma dif-ferentiation (e.g., Grove et al. 1982; Brophy 1987, 1990; Kerr

Electronic supplementary material The online version of this article(doi:10.1007/s12517-016-2598-0) contains supplementary material,which is available to authorized users.

* Federico Luccifederico.lucci@uniroma3.it

1 Dipartimento di Scienze, Sezione Scienze Geologiche, UniversitàBRoma Tre^, Largo S. L. Murialdo, 1-00146 Rome, Italy

2 Department of Geography and Geology, Eastern KentuckyUniversity, Richmond, KY 40475, USA

3 ARC Centre of Excellence for Core to Crust Fluid Systems andGEMOC ARC National Key Centre, Department of Earth andPlanetary Sciences, Macquarie University, Sydney, NSW 2109,Australia

4 Science and Research Branch, Islamic Azad University, Tehran, Iran5 Department of Geology, Imam Khomeini International University,

Qazvin, Iran

Arab J Geosci (2016) 9:573 DOI 10.1007/s12517-016-2598-0

et al. 1995; Santo et al. 2004; Macpherson et al. 2006;Davidson et al. 2007; Dessimoz et al. 2012; Gao et al. 2012;Rossetti et al. 2014). However, not all mineral phases found incalc-alkaline rocks may be involved in FC processes, andtherefore, this may have impact in the evolution and differen-tiation trends of calc-alkaline suites. Indeed, the pristine gab-broic assemblage (plagioclase + cl inopyroxene,orthopyroxene ± olivine ± magnetite; e.g., Shand 1943,1947; Gill 1981) seems to have limited effect on the REEfractionation budget (Davidson et al. 2007; Larocque andCanil 2010), with the anorthite plagioclase controlling theEu budget (e.g., Davidson et al. 2013) and clinopyroxenediscriminating significantly among the REE (Davidson et al.2007; Davidson et al. 2013; Larocque and Canil 2010).However, even if clinopyroxene and hornblende fraction-ations show similar effects on whole REE budget, the Cpx-KD REE ≪ Hbl-KD REE (Davidson et al. 2007). Therefore,hornblende fractionation is expected to enhance light rareearth element (LREE) enrichment (monitored by increasingLa/Yb and decreasing Dy/Yb with respect to clinopyroxenefractionation (Macpherson et al. 2006; Davidson et al. 2007;Dessimoz et al. 2012; Davidson et al. 2013). Hornblende frac-tionation is indeed recognized as the key factor in calc-alkaline magma evolution and differentiation in magmaticarc setting (Cawthorn et al. 1973; Arculus and Wills 1980;Defant and Drummond 1990; Muntener et al. 2001; Groveet al. 2002; Sisson et al. 2005; Macpherson et al. 2006,Davidson et al. 2007; Ulmer 2007; Macpherson 2008;Chiaradia et al. 2009; Moyen 2009; Bryant et al. 2011;Dessimoz et al. 2012; Nandedkar et al. 2014; Rossetti et al.2014; Moghadam et al. 2016).

With the aim to contribute to the comprehension of andes-ite production and evolution in arc settings, this paper is fo-cused on the study of the effect of the Tschermak molecule(CaAlAlSiO6) in the fractionating Al-rich hornblende–plagio-clase assemblage during arc magma differentiation. In partic-ular, this paper is intended as a companion paper ofMoghadam et al. (2016) and uses part of its geochemicaldataset from the Sabzevar middle Eocene calc-alkaline volca-nism of Central Iran (Fig. 1), where the dominant role ofamphibole fractionation in magma differentiation has beendemonstrated (Jamshidi et al. 2015; Moghadam et al. 2016).Results document that the Tschermak molecule fractionationcontributes for more than 30 % to the evolution of the mag-matic suite, controlling the Al, Ca, and REE budget duringdifferentiation of calc-alkaline melts.

Geological background

The closure of the Neotethyan oceanic realm along the Zagroscollisional zone of the Eurasia plate was accompanied by dif-fuse arc magmatism during development of a cordilleran-type

continental margin along the Mesozoic Sanandaj–Sirjan andthe Tertiary Urumieh–Dokhtar structural zones of western Iran(e.g., Berberian and King 1981; Mohajjel et al. 2003;Baharifar et al. 2004; Agard et al. 2005; Ghasemi and Talbot2006; Verdel et al. 2007, 2011; Dargahi et al. 2010; Azizi et al.2011; Moghadam and Stern., 2011; Allen et al. 2013; Chiuet al. 2013; Pang et al. 2013). A major magmatic flare-up isdocumented during Eocene times, as documented by distribu-tion of magmatic products both along the collisional zone andintraplate domains of Central Iran (Omrani et al. 2008; Agardet al. 2011; Moghadam and Stern 2011; Verdel et al. 2011;Allen et al. 2013; Chiu et al. 2013; Pang et al. 2013; Jamshidiet al. 2015; Moghadam et al. 2015, 2016). Verdel et al. (2011)attributed the Eocene magmatism to decompression meltingof the lithospheric mantle hydrated by fluids released from thesubducted Tethyan slab. Agard et al. (2011) refer the Eocenemagmatic pulse to break off of the Neotethyan slab, whereasPang et al. (2013) propose a post-collisional convective re-moval of the thickened lithosphere.

The NW–SE trending Sabzevar structural zone of NE cen-tral Iran (Fig. 1) consists of a polyphase tectono-metamorphicophiolite domain developed during oceanic subduction andsuturing of a Late Cretaceous back-arc oceanic domain(Sabzevar Ocean; McCall 1997; Besse et al. 1998; Stampfliand Borel 2002; Bagheri and Stampfli 2008) formed in theupper plate of the Neotethyan subduction (Rossetti et al. 2010;2014; Omrani et al. 2013; Moghadam et al. 2015). Thetectono-metamorphic evolution was accompanied and follow-ed by generation, extraction, migration, and emplacement ofcalc-alkaline magma (e.g., Rossetti et al. 2014; Moghadamet al. 2016). Geochronological evidence documents two majormagmatic phases: (i) a Late Paleocene slab melting event andassociate adakite magmatism at ca. 57 Ma, based on U–Pbzircon and amphibole and muscovite Ar–Ar geochronology(Rossetti et al. 2014), and (ii) a post-orogenic Eocene event atca. 45–47, based on U–Pb zircon and micas Ar–Ar geochro-nology (Moghadam et al. 2016). Both Paleocene and Eocenemagmatic events show a clear metaluminous calc-alkaline af-finity (Rossetti et al. 2014; Moghadam et al. 2016), with meltevolution and differentiation dominantly controlled by amphi-bole fractionation (Rossetti et al. 2014; Jamshidi et al. 2015;Moghadam et al. 2016).

Petrography

Tertiary volcanic rocks are distributed along all the Sabzevarophiolite belt and occur as lava flows, pyroclastic deposits, andfelsic domes (Jamshidi et al. 2015; Moghadam et al. 2016).

Volcanic sequences are characterized by pyroclastic de-posits and lava flows (andesitic to intermediate). Locally,subvolcanic dacitic and rhyolitic domes are also observed tointrude the ophiolite belt. Albitization due to secondary

573 Page 2 of 12 Arab J Geosci (2016) 9:573

carbonate alteration, in both mafic and intermediate–felsicrocks, is locally observed. In the following, mineral abbrevi-ations are after Whitney and Evans (2010).

Andesitic rocks

Andesitic rocks occur both as lava flows and as feederdykes intruding the ophiolite successions. Lavas anddykes show isotropic to fluidal fabric and are medium tohigh porphyritic with phenocrystals made of Hbl + Pl +Ttn associated with a microcrystalline groundmass ofcomparable composition (Fig. 2a–d). Rare augitic Cpxphenocrystal (either preserved or strongly resorbed bymelt) is observed only in most mafic samples (i.e.,SZD7C, Fig. 2c). Hornblende amphibole shows oscillato-ry compositional zoning accompanied by variations inpleochroism (from brown to greenish brown); in manyinstances, amphibole shows strongly resorbed rims(Fig. 2a). Also, the plagioclase crystals, generally, presentresorbed rims; sieve texture is evident where it is in con-tact with resorbed amphiboles (Fig. 2b). Titanite occurs asinclusion in amphibole (Fig. 2d), as an allotriomorphicspecimen (Fig. 2b), or as a microcrystalline aggregate as-sociated with amphibole and plagioclase (Fig. 2c).

Dacites and rhyolites

Felsic lavas and domes show comparable texture, withmedium-to-low porphyritic texture and fluidal fabric.Groundmass is mainly constituted by sanidine and quartz.However, phenocrystals are still made of hornblende amphi-bole and plagioclase, both zoned and resorbed by melt.Titanite is very rare and often recognized as inclusion in am-phibole phenocrystals (Fig. 2d). Most evolved felsic domes(high-silica rhyolites) show aphyric texture with absence ofCa–Al–Ti-rich mineral phases. Groundmass is always micro-crystalline to holoyaline, and it is characterized by white micaassociated with alkali feldspar and quartz.

Workflow

Major element mass balance models (Bryan et al. 1969) havebeen used preliminary in Moghadam et al. (2016) to test thehypothesis of replenishment–fractional crystallization (RFC;White et al. 2009) of the Sabzevar calc-alkaline rocks.

In RFC modeling, parental melt composition is assumed asmatrix b; if all equations are solved for b, then b = Liquid(Daughter or Melt) + Minerals. When compositions ofLiquid and Minerals are known (expressed in matrix A), it is

Fig. 1 Regional geological context. a Satellite image of the Iran region.The yellow rectangle indicates the Sabzevar structural zone (ImageLandsat from Google Earth Pro, Map Data: AND, Google; courtesy ofGoogle). b Localization of the Eocene volcanic calc-alkaline rocks

selected for this study (Image Landsat from Google Earth Pro, MapData: Google; courtesy of Google). c Simplified geological map of theSabzevar structural zone (modified after Moghadam et al. 2016)

Arab J Geosci (2016) 9:573 Page 3 of 12 573

possible to estimate, by least square approximation, their pro-portion (in matrix c).

The similarity of b’ (matrix c multiplied with matrix A) to b(real value) is quantified with the sum of the square of theresiduals (Σr2) as follows:

∑r2 ¼ ∑n

i−1b

0i−bi

� �2ð1Þ

RFC models are considered acceptable when Σr2 < 1.0.Proportion of daughter melt is expressed with F in matrix c.

Mass balance modeling can be used also to test crystalaccumulation (BCumulate^ in White et al. 2009). If theCumulate is assigned to the matrix b and all equationsare solved for b, then Cumulate = Liquid + Minerals.Cumulate models are considered acceptable whenΣr2 < 1.0. Proportion of starting melt is expressed withF in matrix c.

Starting from the andesite melt source (SZD19C) andusing the same fractionating mineral assemblage (Hbl +Pl + Ttn) adopted in Moghadam et al. (2016), in thiscompanion study, we explore the RFC mass balance mod-el to better constraint the magma evolution (forrepresentative mineral chemistry, see SupplementaryDocument 1 in Moghadam et al. 2016). The mineral com-positional dataset is here used to test the major elementmass balance modeling and the rare earth element–Rayleigh fractional crystallization (REE-FC) modelingbased on the avai lable whole-rock dataset (seeSupplementary Document 2 in Moghadam et al. 2016).

Improved REE-FC models, using refined and weighted KD

REE coefficients, are produced for LREE, heavy rare earthelements (HREE), and Eu.

The equation chosen to test REE–fractional crystallizationis that of Rayleigh (1896):

CL ¼ C0 � Fð Þ D−1ð Þ ð2Þ

where C0 represents the element concentration in parental melt,D is the bulk mineral/melt partition coefficient, F represents theresidual melt fraction (where F = 1 − FC, and FC is the degree offractionation, with 0 < FC < 1), and CL is the element calculatedconcentration in the residual melt for a specific degree of frac-tionation (after Rossetti et al. 2014; Moghadam et al. 2016).

Comparison of results is then used to assess the role of theTschermak molecule (hosted in Ca-, Al-rich minerals) in theevolution of Sabzevar calc-alkaline magmas and, in particular,to quantify the amount of the Tschermakmolecule involved inthe fractional crystallization process.

Mineral chemistry of the fractionation assemblage

Plagioclase

Feldspar phenocrystals consist of labradoritic plagioclase(mean XAn ≈66 wt%), typical of basaltic melts. Normal zoningis observed in the most of the analyzed samples. Orthoclase isalways less than 1 wt%.

Fig. 2 Thin section and back-scattered electron (BSE) imagesof the Hbl + Pl + Ttn fractionationassemblage. a Resorbed androunded phenocrystal ofTschermakite Hbl in andesite(sample SZD19C). b Example ofresorbed phenocrystals of Pl andHbl with minor Ttn, commonlyrecognized in dacites and low-silica rhyolites (samples I10–35,I10–32). c Example of highlyporphyritic mafic andesite char-acterized by the assemblages Hbl,Pl, and Ttn in association withCpx phenocrystals (sampleSZD7C). d BSE images showingHbl + Pl + Ttn assemblage. Notethe Ttn inclusions in Hblphenocrystals

573 Page 4 of 12 Arab J Geosci (2016) 9:573

Hornblende

Hornblende phenocrystals belong to the tschermakite group(Leake et al. 2004) with a mean atomic Mg/(Mg + Fe2+ +Fe3+) = 0.53. Most of the hornblende crystals are zoned,rounded, and resorbed by the melt, with rims characterizedby exsolution of spinels (Fig. 2a).

Clinopyroxene

Clinopyroxene phenocrystals show Al2O3 higher than 2 wt%and a meanMg# [molar MgO/(MgO + FeOTOT)] = 0.62. TiO2

(wt%) is always higher than alkali content (Na2O + K2Owt%). Representative clinopyroxene analyses show major di-opside with respect to hedenbergite and orthopyroxene com-ponents (54.4, 14.2, and 20.8 %, respectively, reported per-centages are expressed on total calculated components); alu-mina content is controlled by Tschermak molecule (6 7 wt%,on total calculated components). These chemical characteris-tics attest for a typical augitic pyroxene from basaltic calc-alkaline melt.

Titanite

Titanite is ubiquitous. Titanite shows enrichments in FeOTOT

and Al2O3 (2.45 wt% and 3.53 wt%, respectively).

Whole rock geochemistry of selected samples

Andesitic samples show SiO2 ranging from 52.84 to 54.84wt%, with higher MgO content (3.5–4.32 wt%), high Al2O3

(17.52–17.89 wt%), and lower Na2O (3.35–3.4 wt%).Felsic rocks show intermediate to acid composition

(SiO2 = 63.23–74.28 wt%) with low MgO content (0.17–1.84 wt%) and enrichment of Na2O (3.96–6.10 wt%) withrespect to K2O (1.14–2.79 wt%).

Altered mafic and felsic rocks present progressive enrich-ments in Na2O content (up to 7.7 wt%), in K2O (up to 3.81wt%) opposite to a strong CaO depletion (down to 4.63 and1.74 wt%, in andesites and rhyolites, respectively).

Collectively, the Sabzevar magmatic suite shows asubalkaline affinity and a distribution from basalt–andesites torhyolites on a TAS diagram (Le Maitre et al. 2002). Accordingto Peccerillo and Taylor (1976), a medium-K calc-alkaline sig-nature can be recognized. Andesites show a clearmetaluminous signature with A/CNK (molar [Al2O3/(CaO +Na2O + K2O)]) in the gap 0.79–0.85. Dacites and rhyolitesinstead have A/CNK values in the range 0.87–1.21, with amean of 1.05 indicating a weakly peraluminous character.

Samples show fractionated REE patterns (Table 1)enriched in LREE with respect to HREE, with (La/Yb)N and(Eu/Yb)N values ranging 1.58–7.47 and 1.32–6.43, respec-tively (chondrite normalization after Sun and McDonough1989). Moreover, REE-normalized content progressively

Table 1 Bulk REE parametersfor Sabzevar Eocene volcanicsuite

Sample Rock type Eu* YbN (La/Yb)N (Eu/Yb)N (Dy/Yb)N

SZD7C 0.94 12.71 2.52 1.34 1.17

SZD18D AlbitizedMafic

0.95 8.35 5.71 1.86 1.30

SZD19C 0.98 11.82 4.92 1.74 1.21

D1 Albitized 1.05 12.53 6.16 1.91 1.22

SZD29A Albitized 0.90 9.00 3.16 1.43 1.18

l10-35 1.09 4.18 14.45 2.32 1.13

SZD11C 0.96 8.24 8.30 1.84 1.20

SZD1A Intermediate 1.10 3.76 16.48 2.58 1.33

SZD2C 0.78 14.29 6.38 1.79 1.27

D21 1.10 7.65 11.31 1.74 0.82

SZD27A 0.98 3.94 13.38 2.31 1.44

SZD38A Albitized 1.00 1.18 19.26 4.82 1.77

l10-32 1.02 6.41 6.91 1.54 0.98

SZD20C 1.10 1.12 10.27 2.65 1.48

SZD31AAcid

0.92 4.65 14.07 2.24 1.33

SZD21C 0.98 1.59 13.58 2.54 1.29

SZD34A 0.89 3.18 24.44 2.43 1.20

SZD33A 0.66 3.76 10.10 1.32 1.28

SZD26A 0.98 0.88 13.77 6.43 3.48

Eu*, EuN / [SmN × GdN]1/2 . Chondrite-normalized REE values after Sun and McDonough (1989)

Arab J Geosci (2016) 9:573 Page 5 of 12 573

decreases from andesites to rhyolites. Europium anomaly [EuN /(SmN × GdN)

1/2] is absent (mean value of 0.97) or slightly neg-ative in few felsic rocks (0.66–0.89). HREE shows flat profile ashighlighted also by (Dy/Yb)N ratio ranging 0.82–1.77 (with amean of 1.25; up to 3.48 only for aphyric high-silica rhyolites).In a (La/Yb)N vs. YbN diagram (after Martin et al. 2005; Moyen2009), the Eocene Sabzevar volcanism shows a major calc-alkaline arc setting signature (Moghadam et al. 2016).

Results

RFC modeling: dacites and rhyolites

The RFC modeling has been applied to all selected dacites(see Table RFC1 in the supplementary material). It wasverified that a fractionation of a Hbl + Pl + Ttn assem-blage in the range of 45.1–55.3 wt% (Σr2 0.04–0.87) isnecessary to produce the Sabzevar dacite melts. RFCmodeling for high-silica rhyolites (see Table RFC2 inthe supplementary material) indicates a major degree offractionation of the Hbl + Pl + Ttn assemblage up to59.5–65 wt% (Σr2 0.11–0.30).

Average dacite and rhyolite (Table 2) RFCmodeling attestsfor intermediate and acid melts being produced by two differ-ent degrees of fractionation (50 and 60 %, respectively; Σr2

0.06–0.07) of the same fractionating assemblage (Hbl52.0–52.5 + Pl44.1–44.2 + Ttn3.3–3.9; percentages are expressed on totalcrystallized assemblage).

A Harker-like graphic solution for the RFC models, withmajor element compositions recalculated on anhydrous basisand normalized to 100 wt%, is presented in Fig. 3. All majorelements show the same fractionation trend for both dacitesand rhyolites, with minor variation in Al2O3 and TiO2 con-tents, mainly related to the different amounts of Ttn (3.9 and3.3 % in dacite and rhyolite, respectively) involved in the FCprocess.

Cumulus modeling: mafic andesites

We approach the calculation considering SZD19C as Liquidand the Hbl + Pl + Ttn assemblage as the cumulate assemblageleading to production of the mafic, highly porphyritic, andes-ites (i.e., SZD7C, Fig. 2c).

A crystal accumulation of 21.8 %, mainly controlled byHbl fractionation (77.5 % of the total crystallized assemblage;

Table 2 Major element RFCmodeling for average dacite and rhyolite. Sum of the square of the residuals (Σr2) and proportion of daughter melt (F) arereported in italic

b Liquid and mineral phases (matrix A) b’ b–b’ c

Model A: SZD19C to average dacite (RFC modeling)

SZD19C Av. Dacite Hbl Pl Ttn

SiO2 54.84 65.75 42.48 50.90 32.08 54.82 0.02 Av. dacite 0.477 = F

TiO2 1.01 0.37 0.89 0.05 31.09 1.04 -0.03 Hbl 0.266 52.0 %

Al2O3 17.89 16.15 12.62 29.98 3.53 17.91 -0.02 Pl 0.226 44.1 %

FeOTot 6.44 3.44 17.28 0.69 2.45 6.45 0.00 Ttn 0.020 3.9 %

MnO 0.13 0.08 0.34 0.01 0.04 0.13 -0.01 Sum 0.99

MgO 3.50 1.32 10.81 0.10 0.41 3.54 -0.04

CaO 8.03 3.59 10.07 13.70 26.78 8.01 0.02

Na2O 3.40 4.90 1.72 3.71 0.08 3.64 -0.24

K2O 0.84 1.56 0.33 0.12 0.03 0.86 -0.02

Sum 96.39 0.06 = Σr2

Model B: SZD19C to average rhyolite (RFC modeling)

SZD19C Av. Rhyolite Hbl Pl Ttn

SiO2 54.84 71.87 42.48 50.90 32.08 54.83 0.01 Av. rhyolite 0.373 = F

TiO2 1.01 0.14 0.89 0.05 31.09 0.97 0.03 Hbl 0.321 52.5 %

Al2O3 17.89 15.13 12.62 29.98 3.53 17.87 0.02 Pl 0.270 44.2 %

FeOTot 6.44 1.51 17.28 0.69 2.45 6.34 0.10 Ttn 0.020 3.3 %

MnO 0.13 0.05 0.34 0.01 0.04 0.13 −0.01 Sum 0.98

MgO 3.50 0.37 10.81 0.10 0.41 3.64 −0.14CaO 8.03 1.63 10.07 13.70 26.78 8.08 −0.05Na2O 3.40 5.10 1.72 3.71 0.08 3.46 −0.06K2O 0.84 2.36 0.33 0.12 0.03 1.02 −0.18

Sum 96.35 0.07 = Σr2

573 Page 6 of 12 Arab J Geosci (2016) 9:573

Σr2 0.44), is able to produce mafic andesites (model 14,Table RFC3 in the supplementary material).

Due to the presence of augitic Cpx phenocrystals inmafic andesites (Fig. 2c), we repeat the model addingthe Cpx composition to the mineral matrix. Results (mod-el 15, Table RFC3 in the supplementary material) againpoint to a crystal accumulation of 22 % (Σr2 0.43), chief-ly controlled by Hbl fractionation (76.4 % of the totalcrystallized assemblage), and demonstrate the irrelevantrole of augitic Cpx fractionation (1.0 % of the total crys-tallized assemblage).

Rayleigh fractional crystallization applied to REE

Results from RFC modeling do not completely agreewith whole-rock REE geochemistry: a progressive extrac-tion of Hbl should lower the Dy/Yb value with

differentiation (Davidson et al. 2007; Rossetti et al.2014), that is at odds with the REE signature of theEocene Sabzevar magmatic rocks, which instead showflat HREE patterns and a near constant value of (Dy/Yb)N ratio (Moghadam et al. 2016). The latter evidence,in particular, suggests permanence rather than extraction/fractionation of Hbl during magma differentiation.

Based on RFC model results and considering the REEgeochemistry, we approach the REE–Rayleigh FC model-ing by calculating differentiation curves in three differentsystems: (La/Yb)N vs. YbN, (Eu/Yb)N vs. YbN, and (Dy/Yb)N vs. YbN. The REE-FC models were calculated formineral phase weight percentage, obtained through aver-age rhyolite RFC model (Hbl 52.5 %, Pl 44.2 %, Ttn3.3 %) after normalization to phase density (tschermakitehornblende 3.44 g/cm3 in Deer et al. 1997; labradoriteplagioclase 2.68 g/cm3 in Deer et al. 2001; titanite3.53 g/cm3 in Deer et al. 1982).

Fig. 3 Harker diagrams as derived frommajor element RFCmodeling ofthe Sabzevar Eocene volcanic suite. Major elements have beenrecalculated to 100 % on anhydrous basis. Black arrows (straight,dashed, and dotted are Hbl, Pl, and Ttn, respectively) representFC = 20 % of pure phases. Blue and pink stars represent average

compositions of dacite and rhyolite, respectively (see Table 2). Blue andpink lines represent fractionation trends as calculated from average daciteand rhyolite compositions (see Table 2), respectively. Samples colorsrefer to Fig. 1b

Arab J Geosci (2016) 9:573 Page 7 of 12 573

Phase/melt partition coefficients for basaltic–andesitic/dioritic source were selected from literature (Schnetzlerand Philpotts 1970; Nagasawa and Schnetzler 1971;Matsui 1977; Luhr and Carmichael 1980; Green andPearson, 1983; Green and Pearson 1985; Fujimaki et al.1984; Bacon and Druitt 1988) through the GERM KD

database (https://earthref.org/GERM/).Refined calculated partition coefficients are as follows:DLa

1.899, DEu 2.257, DDy 2.355, and DYb 3.646 (with maximumvalues tested of 2.292, 8.639, 16.033, and 9.245, respective-ly). The high variance for Eu, Dy, and Yb is a consequence ofREE–element diffusivities in hornblende structure (Cherniakand Dimanov 2010) and of high REE–KD values for titanite(Luhr and Carmichael 1980).

Results as obtained by application of Eq. (2) are presentedin Table 3. These results show that major element and REE FCmodels converge only in reproducing the aphyric high-silicarhyolites, with a fraction of ca. 60 %, of the Hbl + Pl + Ttnassemblage. Instead, average dacite and rhyolite compositionsare obtained for ca. 20 and 40 % fractionation of the sameassemblage (Figs. 3 and 4).

Discussion

Comparing major element and REE-FC results, a gap inthe degree of fractionation in the order of 20–30 %(compare Figs. 3 and 4) to reproduce the EoceneSabzevar volcanic suite is evident. Based on the wide-spread textural evidence at microscale showing stronglyresorbed Hbl and Pl phenocrystals in the groundmass ofthe Sabzevar volcanics, we therefore hypothesize that suchgap corresponds to fractionated, not extracted Tschermak-rich (Ca–Al-rich) phenocrystals (hornblende and plagio-clase) that suffered interaction with residual melt (Fig. 2a,b). This hypothesis is supported by nearly constant Eu*anomaly and (Dy/Yb)N vs. YbN differentiation trends

(Table 1 and Fig. 4). In this view, taking into account thegeneral incompatible behavior of REE elements (e.g.,Philpotts and Ague 2009), the fractionation gap may

Table 3 REE-FC modeling results

YbN (La/Yb)N (Eu/Yb)N (Dy/Yb)N

C0 values

Andesite SZD19C 11.82 4.92 1.74 1.21

FC degree CL values

0.1 8.95 5.92 2.01 1.38

0.2 6.55 7.27 2.37 1.61

0.3 4.60 9.18 2.85 1.92

0.4 3.06 12.02 3.53 2.34

0.5 1.89 16.53 4.55 2.96

0.6 1.05 24.41 6.20 3.95

0.7 0.49 40.35 9.24 5.72

Fig. 4 Rayleigh REE-FC modeling for the Sabzevar Eocene volcanicsuite. The cyan fractionation curves represent solutions as derived inTable 3. The red dotted lines represent calculated solutions formaximum KD values calculated from literature (see text for furtherdetails)

573 Page 8 of 12 Arab J Geosci (2016) 9:573

represent the REE budget re-distributed in the melt by re-sorption of not extracted, Tschermak-rich phenocrystals.

Convergence of both major and REE fractionation modelsonly for aphyric high-silica rhyolites (hence characterized bythe absence of the Hbl + Pl + Ttn assemblage) supports thehypothesis of the REE budget fully controlled by theTschermak-bearing phases.

Validity and limitations of the RFC modeling

Three albitized samples (a high-altered andesite, amedium-altered dacite, and a poorly altered rhyolite, re-spectively) have been used to test the validity and limita-tions of the presented RFC modeling. Andesite and dacitemodels (Table RFC4 in the supplementary material) fail toreproduce the differentiation trend (Σr2 5.66 and 5.68,respectively), mainly due to the large r2 for Al, Ca, andalkalines. Rhyolite model, instead, produces a nearly ac-ceptable result (Σr2 1.09). Since the rhyolite rocks showaphyric textures devoid of Ca–Al-rich (Tschermak-bearing) mineral phases, these results emphasize the dom-inant role exerted by Tschermak-bearing mineral phases incontrolling magma differentiation.

The Tschermak effect

Major element RFC and REE-FC modeling of the EoceneSabzevar volcanic suite highlights the major roles exerted byCa–Al-rich (Tschermakmolecule-bearing) phases such as Hbland Ca–Pl.

Throughout, a simple stoichiometry calculation is pos-sible to define the amount of Tschermak molecule in-volved in RFC processes. Considering the simplified gen-eral mineral formulas for the following: (i) TschermakiteHbl [Ca2(Mg,Fe)2MgAl4Si6O22(OH)2], (ii) labradoritic Pl[XAn 66 wt%; 0.66 (CaAl2Si2O8) + 0.34 (NaAlSi3O8)],and (iii) Ttn [CaTiSiO5], it is possible to write the frac-tionating assemblage as follows:

Ca2 Mg; Feð Þ2MgAl4Si6O22 OHð Þ2� þ ½0:66 CaAl2Si2O8ð Þ

þ 0:34 NaAlSi3O8ð Þ� þ ½CaTiSiO5�

ð3Þ

Factoring in minor terms:

2 CaAl2SiO6ð Þ þ Mg; Feð Þ2Si2O6 þ 2SiO2 þMg OHð Þ2� �

þ 0:66 CaAl2SiO6ð Þ þ 0:66 SiO2ð Þ þ 0:34 NaAlSi3O8ð Þ½ �þ CaTiSiO5½ �

ð4Þ

with Tschermakite Hbl expressed as 2 Tschermak mol. + 1Opx + 2 Qtz + 1 Brc and labradoritic plagioclase as 0.66Tschermak mol. + 0.66 Qtz + 0.34 Ab, respectively.

Using phase percentage obtained through major elementRFC modeling, we can rewrite Eqs. (3) and (4) as follows:

Average dacite can be defined as

1:331 CaAl2SiO6ð Þ þ 1:331 SiO2ð Þ þ 0:52 Mg; Feð Þ2Si2O6

þ 0:52Mg OHð Þ2 þ 0:15 NaAlSi3O8ð Þ þ 0:039 CaTiSiO5ð Þ

ð5Þ

Average rhyolite can be defined as

1:342 CaAl2SiO6ð Þ þ 1:342 SiO2ð Þ þ 0:525 Mg; Feð Þ2Si2O6

þ 0:525Mg OHð Þ2 þ 0:15 NaAlSi3O8ð Þ þ 0:033 CaTiSiO5ð Þ

ð6Þ

Normalizing to 100, the major role exerted by theTschermak molecule on the fractionating assemblage for bothdacite and rhyolite models, up to 34.2 and 34.3 %, respective-ly, is evident. Furthermore, stoichiometric factors indicate thatthe Tschermak molecule is mainly controlled by Hbl (78 %)with respect to Pl (22 %).

Based on the stoichiometry, the FC degree gap betweenmajorRFC and REE-FC models shows that 6 to 10 % of theTschermak molecule suffered interaction with fractionated melt.

The Tschermak role in the calc-alkaline magmatism

It is widely accepted that Hbl is the main mineral phase con-trolling differentiation and major and REE fractionation inandesites and in calc-alkaline magmas (e.g., Cawthorn andO’hara 1976; Anderson 1980; Defant and Drummond 1990;Romick et al. 1992; Grove et al. 2002; Davidson et al. 2007,2013; Pichavant and Macdonald 2007; Moyen 2009;Larocque and Canil 2010; Bryant et al. 2011; Deering et al.2011; Zhang et al. 2011; Dessimoz et al. 2012; Takahashi et al.2013; Rossetti et al. 2014; Moghadam et al. 2016). Majorelement RFC modeling presented in this study confirms theprimary role of Hbl and Pl fractionation (52 and 44 %, respec-tively), in magma differentiation.

For Hbl and Pl fractionation, stoichiometry shows the follow-ing fractionating components (percentages are on total fraction-ated assemblage): Tschermak (34.2–34.3%), Qtz (34.2–34.3%),Opx (13.4%), Brc (13.4%), andAb (3.9%). Tschermak +Qtz +Brc provides the 81.8–81.9%of the total crystallized assemblagefor both dacite and rhyolite melts. Due to (i) the marginal effectof Opx in REE budget (e.g., Davidson et al. 2007; Larocque andCanil 2010), (ii) the late crystallization of Ab, enriched in theresidual melt, and (iii) the general REE-poor/neutral behaviors ofBrc (e.g., Kodolányi et al. 2011) and Qtz (e.g., Monecke et al.2002; Götze et al. 2004), the Tschermak molecule is the onlyefficient component whose fractionation is able to produce daciteand rhyolite magmas starting from a calc-alkaline andesitesource. In particular, the Tschermak molecule controls both theHREE and LREE,when it is part of double-chain inosilicate (i.e.,Hbl) and Eu and LREE in framework silicate (i.e., Ca-rich Pl).

Arab J Geosci (2016) 9:573 Page 9 of 12 573

Finally, cumulation and permanence of Tschermak-richphenocrystals in the melt are the key factor to understand bulkREE budget in calc-alkaline differentiation trends.

Conclusion

The main results of this study can be summarized in the fol-lowing points:

1. Major element RFC modeling outlines that the SabzevarEocene calc-alkaline volcanic rocks are genetically linkedmelts due to fractional crystallization dominated by theTschermak component hosted in hornblende + Ca–plagio-clase + titanite (Hbl52.0–52.5 + Pl44.1–44.2 + Ttn3.3–3.9;phases are expressed on total crystallized assemblage).Furthermore, dacites and rhyolites represent two differentdegrees of fractionation (FC values 45–55 and 59–65 %,respectively) starting from the same andesitic source.

2. Mafic andesites, considering their isotropic phenocrystal-richfabric and taking in account mass balance results (Hbl77.5 +Pl21.7 + Ttn0.8), can be interpreted as Hbl cumulates. RFCmodeling evidences a negligible role of clinopyroxene frac-tionation (≈1 % with respect to total crystal assemblage).

3. REE-FC modeling confirms a co-genetic magmatic suite,spanning from andesites to dacites and rhyolites. The lowerdegree of fractionation for dacites and rhyolites (approxi-mately 20 and 40 %, respectively) is interpreted as a due tore-equilibration of the bulk REE budget between fractionat-ed phases and the residual melts. This hypothesis is in agree-ment with (i) the general incompatible behavior of REE; (ii)the presence of resorbed Hbl and Pl phenocrystals, (iii) theREE patterns (in particular, Eu* anomaly and flat HREEpatterns), and (iv) convergence at FC = 60 % for both majorand REE-FC models only for aphyric high-silica rhyolites.

4. Fractional crystallization of the Tschermak molecule(hosted in Hbl and Pl) dominantly controls magma differ-entiation in calc-alkaline magmatic suites.

Acknowledgments This manuscript benefitted from constructive com-ments and advice from an anonymous reviewer.

References

Agard P, Omrani J, Jolivet L, Mouthereau F (2005) Convergence historyacross Zagros (Iran): constraints from collisional and earlier defor-mation. Int J Earth Sci 94(3):401–419

Agard P, Omrani J, Jolivet L, Whitechurch H, Vrielynck B, Spakman W,Monié P, Bertrand M, Wortel R (2011) Zagros orogeny: asubduction-dominated process. Geol Mag 148(5–6):692–725

Allen MB, Kheirkhah M, Neill I, Emami MH, McLeod CL (2013)Generation of arc and within-plate chemical signatures in collisionzone magmatism: Quaternary lavas fromKurdistan Province, Iran. JPetrol 54:887–911

Anderson AT (1980) Significance of hornblende in calc-alkaline andes-ites and basalts. Am Mineral 65(9–10):837–851

Arculus RJ, Wills KJ (1980) The petrology of plutonic blocks and inclu-sions from the Lesser Antilles island arc. J Petrol 21(4):743–799

Azizi H, Asahara Y, Mehrabi B, Chung SL (2011) Geochronological andgeochemical constraints on the petrogenesis of high-K granite fromthe Suffi abad area , Sanandaj-Sirjan Zone, NW Iran. Chem Erde-Geochem 71(4):363–376

Bacon CR, Druitt TH (1988) Compositional evolution of the zonedcalcalkaline magma chamber of Mount Mazama, Crater Lake,Oregon. Contrib Mineral Petrol 98(2):224–256

Bagheri S, Stampfli GM (2008) The Anarak, Jandaq and Posht-e-Badammetamorphic complexes in central Iran: new geological data, rela-tionships and tectonic implications. Tectonophysics 451(1):123–155

Baharifar A, Moinevaziri H, Bellon H, Piqué A (2004) The crystallinecomplexes of Hamadan (Sanandaj–Sirjan zone, western Iran):metasedimentary Mesozoic sequences affected by Late Cretaceoustectono-metamorphic and plutonic events. Compt Rendus Geosci336(16):1443–1452

Berberian M, King GCP (1981) Towards a paleogeography and tectonicevolution of Iran. Can J Earth Sci 18(2):210–265

Besse J, Torcq F, Gallet Y, Ricou LE, Krystyn L, Saidi A (1998) LatePermian to Late Triassic palaeomagnetic data from Iran: constraintson the migration of the Iranian block through the Tethyan Ocean andinitial destruction of Pangaea. Geophys J Int 135(1):77–92

Brophy JG (1987) The Cold Bay volcanic center, Aleutian volcanic arc.Contrib Mineral Petrol 97(3):378–388

Brophy JG (1990) Andesites from northeastern Kanaga island, Aleutians.Contrib Mineral Petrol 104(5):568–581

Bryan WB, Finger LW, Chayes F (1969) Estimating proportions inpetrographic mixing equations by least-square approximations.Science 163:926–927

Bryant JA, Yogodzinski GM, Churikova TG (2011) High-Mg# andesiticlavas of the Shisheisky Complex, Northern Kamchatka: implica-tions for primitive calc-alkaline magmatism. Contrib MineralPetrol 161(5):791–810

Cawthorn RG, O’hara MJ (1976) Amphibole fractionation in calc-alkaline magma genesis. Am J Sci 276(3):309–329

Cawthorn RG, Curran EB, Arculus RJ (1973) A petrogenetic model forthe origin of the calc-alkaline suite of Grenada, Lesser Antilles. JPetrol 14(2):327–337

Cherniak DJ, Dimanov A (2010) Diffusion in pyroxene, mica and am-phibole. Rev Mineral Geochem 72(1):641–690

Chiaradia M, Müntener O, Beate B, Fontignie D (2009) Adakite-likevolcanism of Ecuador: lower crust magmatic evolution andrecycling. Contrib Mineral Petrol 158(5):563–588

Chiu HY, Chung SL, Zarrinkoub MH, Mohammadi SS, Khatib MM,Iizuka Y (2013) Zircon U–Pb age constraints from Iran on the mag-matic evolution related to Neotethyan subduction and Zagros orog-eny. Lithos 162:70–87

Dargahi S, Arvin M, Pan Y, Babaei A (2010) Petrogenesis of post-collisional A-type granitoids from the Urumieh–Dokhtar magmaticassemblage, southwestern Kerman, Iran: constraints on theArabian–Eurasian continental collision. Lithos 115(1):190–204

Davidson J, Turner S, Handley H, Macpherson C, Dosseto A (2007)Amphibole Bsponge^ in arc crust? Geology 35(9):787–790

Davidson J, Turner S, Plank T (2013) Dy/Dy*: variations arising frommantle sources and petrogenetic processes. J Petrol 54(3):525–537

DeerWA, Howie RA, Zussman J (1982) Rock-forming minerals, volume1A: orthosilicates. The Geological Society, London, p. 912

DeerWA, Howie RA, Zussman J (1997) Rock-forming minerals, volume2B, double-chain silicates. The Geological Society, London, p. 764

DeerWA, Howie RA, Zussman J (2001) Rock-forming minerals, volume4B, framework silicates: feldspars. The Geological Society, London,p. 992

573 Page 10 of 12 Arab J Geosci (2016) 9:573

Deering CD, Cole JW, Vogel TA (2011) Extraction of crystal-poor rhyo-lite from a hornblende-bearing intermediate mush: a case study ofthe caldera-formingMatahina eruption, Okataina volcanic complex.Contrib Mineral Petrol 161(1):129–151

Defant MJ, Drummond MS (1990) Derivation of some modern arcmagmas by melting of young subducted lithosphere. Nature347(6294):662–665

Dessimoz M, Müntener O, Ulmer P (2012) A case for hornblende dom-inated fractionation of arc magmas: the Chelan Complex(Washington Cascades). Contrib Mineral Petrol 163(4):567–589

Fujimaki H, Tatsumoto M, Aoki K-i (1984) Partition coefficients of Hf,Zr, and REE between phenocrysts and groundmasses. J GeophysRes 89:662–672

Gao Y, SantoshM, Hou Z,Wei R, Ma G, Chen Z,Wu J (2012) High Sr/Ymagmas generated through crystal fractionation: evidence fromMesozoic volcanic rocks in the northern Taihang orogen, NorthChina Craton. Gondwana Res 22(1):152–168

Ghasemi A, Talbot CJ (2006) A new tectonic scenario for the Sanandaj–Sirjan Zone (Iran). J Asian Earth Sci 26(6):683–693

Gill J (1981) Orogenic andesites and plate tectonics. Springer, Berlin, p.390

Götze J, Plötze M, Graupner T, Hallbauer DK, Bray CJ (2004) Traceelement incorporation into quartz: a combined study by ICP-MS,electron spin resonance, cathodoluminescence, capillary ion analy-sis, and gas chromatography. Geochim Cosmochim Acta 68(18):3741–3759

Green TH, Pearson NJ (1983) Effect of pressure on rare earth elementpartition coefficients in common magmas. Nature 305:414–416.doi:10.10138/305414a0

Green TH, Pearson NJ (1985) Experimental determination of REE parti-tion coefficients between amphibole and basaltic to andesitic liquidsat high pressure. Geochim Cosmochim Acta 49(1):465–1–46468.doi:10.1016/0016-7037(85)90295-9.

Grove TL, Gerlach DC, Sando TW (1982) Origin of calc-alkaline serieslavas at medicine Lake volcano by fractionation, assimilation andmixing. Contrib Mineral Petrol 80(2):160–182

Grove T, Parman S, Bowring S, Price R, Baker M (2002) The role of anH2O-rich fluid component in the generation of primitive basalticandesites and andesites from the Mt. Shasta region, N California.Contrib Mineral Petrol 142(4):375–396

Jamshidi K, Ghasemi H, Troll VR, Sadeghian M, Dahren B (2015)Magma storage and plumbing of adakite-type post-ophiolite intru-sions in the Sabzevar ophiolitic zone, Northeast Iran. Solid Earth6(1):49

Kelemen PB, Hanghøj K, Greene AR (2003) One view of the geochem-istry of subduction-related magmatic arcs, with an emphasis onprimitive andesite and lower crust. Treatise Geochem 3:593–659

Kerr AC, Kempton PD, Thompson RN (1995) Crustal assimilation dur-ing turbulent magma ascent (ATA); new isotopic evidence from theMull Tertiary lava succession, NWScotland. Contrib Mineral Petrol119(2–3):142–154

Kimura J-I, Kent AJR, Rowe MC, Katakuse M, Nakano F, Hacker BR,van Keken PE, Kawabata H, Stern RJ (2010) Origin of cross-chaingeochemical variation in Quaternary lavas from the northern Izu arc:using a quantitative mass balance approach to identify mantlesources and mantle wedge processes. Geochem Geophys Geosyst11. doi:10.1029/2010GC003050.

Kodolányi J, Pettke T, Spandler C, Kamber BS, Gméling K (2011)Geochemistry of ocean floor and fore-arc sperpentinites: constraintson the ultramafic input to subduction zone. J Petrol egr058.doi:10.1093/petrology/egr058

Larocque J, Canil D (2010) The role of amphibole in the evolution of arcmagmas and crust: the case from the Jurassic Bonanza arc section,Vancouver Island, Canada. Contrib Mineral Petrol 159(4):475–492

LeMaitre RW, Streckeisen A, Zanettin B, Le Bas MJ, Bonin B, BatemanP, Bellieni G, Dudek A, Efremova S, Keller J, Lameyre J, Sabine

PA, Schmid R, Sorensen H, Woolley AR (2002) Igneous rocks. Aclassification and glossary of terms. Reccommendations of theIUGS Subcomission on the Systematics of Igneous Rocks.Cambridge University Press, pp. 236

Leake BE, Woolley AR, Birch WD, Burke EA, Ferraris G, Grice JD,Hawthorne FC, Kisch JH, Krivovichev VG, Schumacher JC,Stephenson NCN, Whittaker EJC (2004) Nomenclature of amphi-boles: additions and revisions to the International MineralogicalAssociation s amphibole nomenclature. Am Mineral 89:883–887

Luhr JF, Carmichael ISE (1980) The Colima volcanic complex, Mexico.I: post-caldera andesites from Volcan Colima. Contrib MineralPetrol 71:343–372

Macpherson CG (2008) Lithosphere erosion and crustal growth in sub-duction zones: insights from initiation of the nascent East PhilippineArc. Geology 36(4):311–314

Macpherson CG, Dreher ST, Thirlwall MF (2006) Adakites without slabmelting: high pressure differentiation of island arc magma,Mindanao, the Philippines. Earth Planet Sci Lett 243(3):581–593

Martin H, Smithies RH, Rapp R, Moyen JF, Champion D (2005) Anoverview of adakite, tonalite–trondhjemite–granodiorite (TTG),and sanukitoid: relationships and some implications for crustal evo-lution. Lithos 79(1):1–24

Matsui Y (1977) Crystal structure control in trace element partition betweencrystal and magma. Bull Soc Fr Mineral Crystallogr 100:315–324

McCall GJH (1997) The geotectonic history of the Makran and adjacentareas of southern Iran. J Asian Earth Sci 15(6):517–531

Moghadam HS, Stern RJ (2011) Geodynamic evolution of UpperCretaceous Zagros ophiolites: formation of oceanic lithosphereabove a nascent subduction zone. Geol Mag 148(5–6):762–801

MoghadamHS, Li XH, LingXX, Santos JF, Stern RJ, Li QL, Ghorbani G(2015) Eocene Kashmar granitoids (NE Iran): petrogenetic con-straints fromU–Pb zircon geochronology and isotope geochemistry.Lithos 216:118–135

Moghadam HS, Rossetti F, Lucci F, Chiaradia M, Gerdes A, MartinezML, Ghorbani G, Nasrabady M (2016) The calc-alkaline andadakitic volcanism of the Sabzevar structural zone (NE Iran): impli-cations for the Eocene magmatic flare-up in Central Iran. Lithos248:517–535

Mohajjel M, Fergusson CL, Sahandi MR (2003) Cretaceous–Tertiaryconvergence and continental collision, Sanandaj–Sirjan zone, west-ern Iran. J Asian Earth Sci 21(4):397–412

Monecke T, Kempe U, Götze J (2002) Genetic significance of the traceelement content in metamorphic and hydrothermal quartz: a recon-naissance study. Earth Planet Sci Lett 202(3):709–724

Moyen JF (2009) High Sr/Y and La/Yb ratios: the meaning of theBadakitic signature^. Lithos 112(3):556–574

Muntener O, Kelemen PB, Grove TL (2001) The role of H2O duringcrystallization of primitive arc magmas under uppermost mantleconditions and genesis of igneous pyroxenites: an experimentalstudy. Contrib Mineral Petrol 141(6):643–658

Nagasawa H, Schnetzler CC (1971) Partitioning of rare earth, alkali, andalkaline earth elements between phenocrysts and acidic igneousmagmas. Geochim Cosmochim Acta 35:953–968. doi:10.1016/0016-7037(71)90008-1.

Nandedkar RH, Ulmer P, Müntener O (2014) Fractional crystallization ofprimitive, hydrous arc magmas: an experimental study at 0.7 GPa.Contrib Mineral Petrol 167(6):1–27

Omrani J, Agard P, Whitechurch H, Benoit M, Prouteau G, Jolivet L(2008) Arc-magmatism and subduction history beneath the ZagrosMountains, Iran: a new report of adakites and geodynamic conse-quences. Lithos 106(3):380–398

Omrani H, Moazzen M, Oberhänsli R, Altenberger U, Lange M (2013)The Sabzevar blueschists of the North-Central Iranian micro-continent as remnants of the Neotethys-related oceanic crust sub-duction. Int J Earth Sci. doi:10.1007/s00531-013-0881-9

Arab J Geosci (2016) 9:573 Page 11 of 12 573

Pang KN, Chung SL, Zarrinkoub MH, Khatib MM, Mohammadi SS,Chiu HY, Chu CH, Lee HY, Lo CH (2013) Eocene–Oligocenepost-collisional magmatism in the Lut–Sistan region, eastern Iran:magma genesis and tectonic implications. Lithos 180:234–251

Peccerillo A, Taylor SR (1976) Geochemistry of Eocene calc-alkalinevolcanic rocks from the Kastamonu area, northern Turkey. ContribMineral Petrol 58(1):63–81

Philpotts A, Ague J (2009) Principles of igneous and metamorphic pe-trology. Cambridge University Press

Pichavant M, Macdonald R (2007) Crystallization of primitive basalticmagmas at crustal pressures and genesis of the calc-alkaline igneoussuite: experimental evidence from St Vincent, Lesser Antilles arc.Contrib Mineral Petrol 154(5):535–558

Rayleigh Lord JWS (1896) Theoretical considerations respecting the sep-aration of gases by diffusion and similar processes. London,Edinburgh, Dublin Philos Mag J Sci 42(259):493–498

Romick JD, Kay SM, Kay RW (1992) The influence of amphibole fraction-ation on the evolution of calc-alkaline andesite and dacite tephra fromthe central Aleutians, Alaska. Contrib Mineral Petrol 112(1):101–118

Rossetti F, Nasrabady M, Vignaroli G, Theye T, Gerdes A, Razavi SMH,Moin Vaziri H (2010) Early Cretaceous migmatitic mafic granulitesfrom the Sabzevar range (NE Iran): implications for the closure of theMesozoic peri-Tethyan oceans in central Iran. Terra Nova 22:26–34.doi:10.111/j.1365-3121.2009.00912.x

Rossetti F, NasrabadyM, Theye T, Gerdes A, Monié P, Lucci F, VignaroliG (2014) Adakite differentiation and emplacement in a subductionchannel: the late Paleocene Sabzevar magmatism (NE Iran). GeolSoc Am Bull 126(3–4):317–343

Rudnick RL, Gao S (2003) Composition of the continental crust. Crust,Treatise Geochem 3:1–64

Santo AP, Jacobsen SB, Baker J (2004) Evolution and genesis of calc-alkaline magmas at Filicudi volcano, Aeolian Arc (SouthernTyrrhenian Sea, Italy). Lithos 72(1):73–96

Schnetzler CC, Philpotts JA (1970) Partition coefficients of rare-earthelements between igneousmatrix material and rock-formingmineralphenocrysts; II. Geochim Cosmochim Acta 34(3):331–340.doi:10.1016/0016-7037(70)90110-9.

Shand, S. J. (1943). Eruptive rocks: their genesis, composition, and clas-sification, with a chapter on meteorites. J. Wiley and Sons,Incorporated

Shand SJ (1947) The genesis of intrusive magnetite and related ores.Econ Geol 42(7):634–636

Sisson TW, Ratajeski K, Hankins WB, Glazner AF (2005) Voluminousgranitic magmas from common basaltic sources. Contrib MineralPetrol 148(6):635–661

Stampfli GM, Borel GD (2002) A plate tectonic model for the PaleozoicandMesozoic constrained by dynamic plate boundaries and restoredsynthetic oceanic isochrons. Earth Planet Sci Lett 196(1):17–33

Stern RJ (2002) Subduction zones. Rev Geophys 40(4):1–38Stern CR (2011) Subduction erosion: rates, mechanisms, and its role in

arc magmatism and the evolution of the continental crust and man-tle. Gondwana Res 20(2):284–308

Sun SS, McDonough WF (1989) Chemical and isotopic systematics ofoceanic basalts: implications for mantle composition and processes.Geol Soc Lond, Spec Publ 42(1):313–345

Takahashi T, Hirahara Y, Miyazaki T, Senda R, Chang Q, Kimura JI,Tatsumi Y (2013) Primary magmas at the volcanic front of the NEJapan arc: coeval eruption of crustal low-K tholeiitic and mantle-derived medium-K calc-alkaline basalts at Azuma Volcano. J Petrol54(1):103–148

Tatsumi, Y. (2005). The subduction factory: how it operates in the evolv-ing Earth GSA today, 15(7):4

Ulmer P (2007) Differentiation of mantle-derived calc-alkaline magmasat mid to lower crustal levels: experimental and petrologic con-straints. Periodico Mineral 76(2–3):309–325

Verdel C, Wernicke BP, Hassanzadeh J, Guest B (2011) A Paleogeneextensional arc flare-up in Iran. Tectonics 30(3)

Verdel C, Wernicke BP, Ramezani J, Hassanzadeh J, Renne PR, Spell TL(2007) Geology and thermochronology of Tertiary Cordilleran-stylemetamorphic core complexes in the Saghand region of central Iran.Geol Soc Am Bull 119(7–8):961–977

White JC, Parker DF, Ren M (2009) The origin of trachyte andpantellerite from Pantelleria, Italy: insights from major element,trace element, and thermodynamic modelling. J VolcanolGeotherm Res 179(1):33–55

Whitney DL, Evans BW (2010) Abbreviations for names of rock-formingminerals. Am Mineral 95(1):185

Zhang X, Mao Q, Zhang H, Zhai M, Yang Y, Hu Z (2011) Mafic andfelsic magma interaction during the construction of high-K calc-alkaline plutons within a metacratonic passive margin: the EarlyPermian Guyang batholith from the northern North China Craton.Lithos 125(1):569–591

573 Page 12 of 12 Arab J Geosci (2016) 9:573