Keywords

1 Geological Setting of the Liaoji Belt

The Liaoji orogenic belt is one of the three Paleoproterozoic mobile belts on the NCC, located in the eastern part of the Eastern Block (Fig. 7.1; Luo et al. 2008; Zhao et al. 2005). The belt consists of greenschist to lower amphibolite facies metasedimentary, metavolcanic successions, meta-granitic and mafic intrusions with metamorphic age of ca. 1.9 Ga (Yin and Nie 1996; Luo et al. 2004; Lu et al. 2006). These rocks are called the Liaohe Group in the eastern Liaoning Peninsula, which are further divided into the North Liaohe Group and the South Liaohe Group (Zhang 1984) by the Qinglongshan-Zaoerling ductile shear zones and faults (Li et al. 2005). The North Liaohe Group is characterized by the abundance of clastic and carbonate rocks, whereas the South Liaohe Group contains much more volcanic rocks (Fig. 7.2; Zhang and Yang 1988; Bai 1993; Lu et al. 1996; Bai and Dai 1998), including a volcanic-rich sequence in the lower part (the Lieryu and Gaojiayu Formations), a carbonate-rich sequence in the middle (the Dashiqiao Formation), and a pelitic sequence in the upper part (the Gaixian Formation). The well-known Paleoproterozoic boron deposits are hosted in the Lieryu Formation (~1300 m thick) that comprises metamorphosed boron-bearing volcano-sedimentary successions (Zhang 1984; Peng and Palmer 1995).

Fig. 7.1
figure 1

The tectonic subdivision of the North China Craton (after Zhao et al. 2005); also shown is the Jiao-Liao-Ji Paleoproterozoic belt

Fig. 7.2
figure 2

Sketch geological map of the Paleoproterozoic granites in the Jiao-Liao-Ji Belt (after Hao et al. 2004), with sampling localities

There are two groups of Paleoproterozoic granites in the Liaoji Belt, including dominantly metamorphosed gneissic granodiorite-granite with zircon ages from 2.2 to 2.1 Ga (Hao et al. 2004; Lu et al. 2006), and subordinate un-metamorphosed syenite-granite with zircon ages of ca. 1.84–1.87 Ga (Li et al. 2003; Cai et al. 2002; Lu et al. 2004). The latter are post-collisional, showing cross-cut contact with the metasedimentary and volcanic successions (the Liaohe Group). The gneissic granites typically contain foliated minerals such as hornblende and magnetite due to the regional metamorphism (Hao et al. 2004). In this paper, we select five typical gneissic granite plutons for petrological and geochemical studies, trying to clarify their origin and geodynamic setting (Fig. 7.2). Petrological descriptions of these plutons are as below.

2 Petrological Descriptions of the Gneissic Granites

The Hupiyu gneissic pluton is composed mainly of 30 % microcline, 15 % perthite, 15 % plagioclase, 25 % quartz, and small amounts of biotite and hornblende, and accessory epidote, magnetite, zircon, apatite, and sphene. Lineation of hornblende, magnetite, and sphene forms the gneissic structure. No obvious compositional zoning is observed in plagioclase grains that have An contents of 0.47–1.14. Mafic dikes are present in the pluton, which have been metamorphosed into amphibolites (Fig. 7.3a). In addition, a few mafic microgranular enclaves (MME) of dioritic composition can be seen in the pluton.

Fig. 7.3
figure 3

a Microphotograph of amphibolite intruded into the Hupiyu gneissic granite pluton; b field photo of a mafic enclave from the Hadabei pluton; c inner structure of MME, showing hornblende and needle-like apatites; d plagioclases included in the hornblende from the MME; e textural disequilibrium of plagioclase near the MME; f compositional and textural disequilibrium of plagioclase from the gneissic granite; also shown are the An contents; g microphotographs of the Dafangshen granite; h microphotograph of tourmaline from the Dafangshen pluton; i accessory minerals of the Hupiyu pluton. Kfs potash feldspar; Pl plagioclase; Q quartz; Hb hornblende; Mag magnetite; Ttn titanite; Tur tourmaline; Ap apatite

The Hadabei monzonite pluton (Fig. 7.2) shows similar mineralogy to the Hupiyu pluton, with 40 % plagioclase, 25 % perthite, 30 % quartz, small amounts of hornblende and biotite, and accessory pyrrhotite, zircon, titanite, epidote, magnetite, and apatite. The pluton is also intruded by mafic dykes that have been transformed into amphibolites during the metamorphism. In addition, gabbroic diorite enclaves are common in the granite (Fig. 7.3b, c) and some plagioclase crystals are included by the hornblende in the MME (Fig. 7.3d). Plagioclases are characterized by compositional zoning (Fig. 7.3e, f): those far away from the MMEs show normal compositional zoning with An contents decreasing from core to rim, and those near the MMEs show reversal and complicated zoning textures (Fig. 7.3f).

The Muniuhe pluton (Fig. 7.2) is a gneissic hornblende biotite granite pluton, composed mainly of 35 % plagioclase, 15 % K-feldspar (perthite and small amounts of microline), 30 % quartz, 8 % biotite (partially altered to chlorite), and 7 % hornblende. Accessory minerals are zircon, sphene, and magnetite. The hornblende is euhedral. The biotite is foliated, and the plagioclase is kaolinised. Zonal texture of plagioclase is rare in the Muniuhe granite. Again, some dioritic dykes intruded into the granite.

The Dafangshen pluton is a gneissic K-feldspar granite pluton with metamorphic equilibrium textures (Fig. 7.3g), composed of 25 % microline, 15 % perthite, 5 % plagioclase, 40 % quartz, 10 % hornblende, and small amounts of biotite. Accessory minerals are tourmaline, zircon, zoisite, titanite, and magnetite. Banded distribution of hornblende, magnetite, and sphene forms the gneissic structure. The tourmaline is euhedral with obvious zonal structure (Fig. 7.3h).

The Simenzi pluton is a hornblende monzogranite pluton, composed mainly of 35 % microline, 25 % plagioclase, 25 % quartz and 10 % hornblende, and accessory zircon, sphene, and epidote. The hornblende is subhedral to euhedral. There are some dioritic dykes intruding into the granite.

3 Geochronology

The Liaoji Belt consists of Paleoproterozoic meta-volcano-sedimentary rocks, and granitic intrusions (Li et al. 2011). Previous studies have shown that the metasedimentary and volcanic rocks were formed at 2.2–2.0 Ga (Luo et al. 2004, 2008; Lu et al. 2006; Wan et al. 2006) and underwent greenschist to amphibolite facies metamorphism at ca. 1.9–1.93 Ga (Luo et al. 2004, 2008; Lu et al. 2006; Wan et al. 2006; Tam et al. 2011, 2012a, b). A few age data have been reported for the Paleoproterozoic gneissic granites. The gneissic granites in Haicheng and Kuandian have been dated at 2140 and 2070 Ma, respectively (Zhang and Yang 1988; Sun et al. 1993). The Qianzhuogou gneissic granite in the south of Liaonan was dated at 2160 Ma by Lu et al. (2005). Lu et al. (2004) reported a zircon U-Pb age of 2160 Ma for the Hupiyu gneissic granite.

Our new experimental data gained by LA-ICP-MS U-Pb zircon dating method shows that the Hupiyu gneissic granite was emplaced at 2183 Ma and the Hadabei pluton at 2173 Ma (Table 7.1), consistent with previous age data (Wu and Zheng 2004; Li and Zhao 2007). The Muniuhe pluton was formed at 2201 Ma and the Simenzi pluton at 2203 Ma. In summary, the gneissic granites in the JLJB were formed at ca. 2.17–2.20 Ga (Fig. 7.4). It should be stressed that we found inherited Archean zircons both in the gneissic granites and the amphibolite dykes intruding into the granites, such zircon inheritance was scarcely reported by previous researchers (Lu et al. 2004, 2005; Luo et al. 2004; Zhao et al. 2005; Li and Zhao 2007; Li et al. 2011). The ages of those inherited Archean zircons cluster at around 2.5 Ga and 2.7–2.8 Ga, in accordance with the ages of the two main Archean volcanisms within the NCC (Peuct et al. 1986; Liu et al. 1992; Wang et al. 1997; Geng et al. 2002; Zhao et al. 2002; Lu et al. 2004). These zircons of Archean ages suggest an Archean basement for the main source of the gneissic granites. The metamorphosed mafic dykes that intruded into the gneissic granites have been dated at 2159 Ma, indicating the intrusion happened shortly after the emplacement of the granites. U-Pb dating of the zircon overgrowth rims from the amphibolite yields an age of ca. 1.9 Ga, which coincides with the published metamorphic ages (Luo et al. 2004, 2008; Lu et al. 2006; Li et al. 2005; Li and Zhao 2007; Zhou et al. 2008; Tam et al. 2011) and is interpreted as the metamorphism age of the gneissic granites.

Table 7.1 In situ zircon U-Pb isotopic dating for the gneissic granites
Fig. 7.4
figure 4

Zircon U-Pb isotope concordia diagrams; a the Hupiyu granite; b the Hadabei granite; c the Muniuhe granite; d the Simenzi granite; e mafic dyke

4 A-type Granites?

The gneissic granites have been regarded as A-type granite in previous studies (Wu and Zheng 2004; Hao et al. 2004; Zhao et al. 2005), mainly based on limited data plotted in some geochemical diagrams such as Zr + Nb + Ce + Y versus FeO/MgO.

In this paper, we suggest that the gneissic granites are calc-alkaline I-type granites based on the following lines of evidences. (1) A-type granite commonly formed in an anorogenic setting and shows water-deficient and alkaline characteristics (e.g., Whalen et al. 1987). Hornblendes in A-type granites, if any, are generally Na-rich alkaline ones (Whalen et al. 1987). We saw no alkali hornblende or sodium hornblende in these gneissic granite samples, rather, the hornblendes within these rocks are mainly euhedral calcium hornblendes and magnesiohornblendes (Figs. 7.3i, c and 7.5; Table 7.2). (2) Lots of titanites and magnetites occur in these granites (Fig. 7.3i), which, along with the abundance of calcium hornblende and some MMEs (Fig. 7.3b), suggest a calc-alkaline, H2O-rich, and high fO2 affinity of the parental magma, typical of I-type granites. (3) The major elements show medium potassium calc-alkaline features (Fig. 7.6). These granites are enriched in LILEs, such as K, Rb, Sr, and Cs, and depleted in some HFSEs that are otherwise enriched in A-type granites, such as Nb, Ta, and Th. The chondrite-normalized REE patterns of these granites are different from those of A-type granites which are commonly characterized by significant negative Eu anomalies and tetrad effects. (4) Negative Ba anomalies are not so significant as in many A-types (Whalen et al. 1987), which may be caused by the enrichment of H2O in the magma system. Because water-saturated felsic magma tends to show high oxygen fugacity and high internal water pressure due to exsolution of discrete vapor phases and the thermal breakdown of H2O (Bonin 1990). As a consequence, alkali feldspar fractionation and, thus, negative Ba anomalies, are reduced. In addition, the contents of Fe and Ti are also lower than those of typical A-type granites, which may also be caused by the H2O-enrichment in the magma system, because titanomagnetite and Nb-Mn ilmenite precipitate in an early stage, decreasing the contents of Fe and Ti in residual magma (Bonin 1990). These data suggest that the Paleoproterozoic gneissic granites could be basically water-enriched I-type granites (Tables 7.3 and 7.4).

Fig. 7.5
figure 5

Classfication diagram of hornblende from the Hadabei and Muniuhe granites (after Leake et al. 1997)

Table 7.2 Microprobe analysis for representative hornblende from the Liao-Ji gneissic granites
Fig. 7.6
figure 6

a K2O-SiO2 diagram (after Peccerillo and Taylor 1976) and b A/NK-A/CNK diagram for the granites

Table 7.3 Microprobe analysis for representative plagioclase from the gneissic granites
Table 7.4 Whole-rock chemical compositions of the gneissic granites in the Jiao-Liao-Ji Belt

However, some samples from the Dafangshen pluton show A-type features. As shown in Fig. 7.7, these samples have significant negative Eu anomalies (Eu/Eu* = 0.1), slight REE tetrad effect (TE1,3 = 0.12–0.21) and lower Zr/Hf ratios (30.5–33.7) than in normal granites (35–40, Jahn et al. 2001; Chen et al. 2014). We suggest that the A-type-like characteristics are attributable to the highly evolved nature of the pluton as shown by the high SiO2 contents (76.7–77.1 wt%), which could have been caused by the high boron contents of the pluton, because tourmaline (texturally in equilibrium with other rock-forming minerals) is common in this pluton (Fig. 7.3h). The B-rich feature of the Dafangshen granite (>1 wt% boron for the crystallization of tourmaline; Lukkari and Holtz 2007) could have lowered the solidus temperature and viscosity of the magma system (Chen et al. 2014), and thus prolong the process of magma evolution and strengthen the melt-rock interaction, leading to high silica, non-CHARAC characteristics (e.g., low Zr/Hf ratios) and the slight REE tetrad effect of the pluton (Bau 1996; Chen et al. 2014). This is also supported by the occurrence of extremely Ab-rich plagioclase in the Dafangshen pluton (with Ab = 71.1–98.3), a feature typical of highly evolved (A-type) granites (Whalen et al. 1987). The Ab-rich feature could have been related with addition of boron in the magma system, which was proved to shift the ternary minimum composition towards the Ab apex in the phase relations of Ab-Or-Qtz (Manning 1981).

Fig. 7.7
figure 7

a Chondrite-normalized REE patterns of ths gneissic granites. Normalization values of chondrite are from Sun and McDonough (1989). b Primitive mantle (PM)-normalized spidergrams of the granites. PM values are from Sun and McDonough (1989). DF Dafangshen; HD Hadabei; HP Hupiyu; MN Muniuhe; SM Simenzi

5 Role of Magma Mixing/Mingling

According to the petrography and geochemical characteristics of these gneissic granites, we suggest that they formed through a mingling/mixing process between dominant crustal melts and subordinate mantle-derived mafic magmas. This is supported by the following lines of evidence.

First, MMEs are common in the gneissic granites (Fig. 7.3b, c). Although MMEs have been suggested as early crystallized mineral cumulates from magma chamber (Noyes et al. 1983), or as residual phases after partial melting of source rocks (White et al. 1999; Chappell et al. 2000), we suggest that they are originally mantle-derived mafic magma and then modified by mixing/mingling with surrounding felsic melts (Holden et al. 1987; Vernon et al. 1988; Bonin 2004; Sklyarov and Fedorovsky 2006; Chen et al. 2008, 2009; Feeley et al. 2008; Ma et al. 2013) for evidence as below. (1) MMEs are elongated in shape without solid-stated deformation (Fig. 7.3b), due to stretching plastically within a partially crystallized, convective magma (Chen et al. 2008, 2009). (2) There are some plagioclase inclusions in the hornblende crystal from the MME (Fig. 7.3d), indicating that the mafic magma injected into the felsic magma and captured the plagioclase crystallized in the early period. (3) MMEs show more fine-grained and equigranular textures than host rocks (Fig. 7.3c) and many needle-like apatites can be seen in the MMEs (Fig. 7.3c), suggesting that the hot mafic magma has undergone a quenching process when injected into the felsic magma with lower temperatures. The existence of MMEs is typical of a mingling/mixing process for the genesis of the host magma (Clynne 1999; Kemp 2004; Chen et al. 2009).

Second, some plagioclases near the MMEs in the host rock show complicated compositional and textural disequilibrium (Fig. 7.3e, f), which commonly results from magma mixing between intermediate-mafic and felsic melts (Anderson 1976; Janoušek et al. 2004; Chen et al. 2009). As shown in Fig. 7.3h, the Na-rich plagioclase core (An18) was crystallized from felsic magma, and the relatively Ca-rich overgrowths (An35 and An39) may result from two pulses of input of mafic magma into the magma system and subsequent magma mixing (Chen et al. 2013).

Third, the whole-rock Nd isotopic compositions vary significantly (ε Nd(t) = −8.6 to 1.5; Fig. 7.8), even for samples from a single pluton. This argues against a process of closed system magma evolution (Griffin et al. 2002; Kemp and Hawkesworth 2006), rather, a magma mixing process between two end members with distinct Nd isotopic compositions is required. This is consistent with the large variation of ε Hf(t) values (from −1.27 to 5.58; Fig. 7.9) of zircons from each pluton. Large Hf isotopic variation is also considered as having been resulted from a mixing/mingling process between magmas with distinct sources by some other researchers (Kemp and Hawkesworth 2006; Yang et al. 2008; Zhang et al. 2011) (Tables 7.5 and 7.6).

Fig. 7.8
figure 8

Whole-rock Nd isotopic data of the gneissic granites

Fig. 7.9
figure 9

Zircon Hf isotopic compositions of the granites. Zircon ε Hf(t) values were calculated at the crystallization ages of these granites. Values of ~2500 Ma granulite are from Jiang et al. (2013)

Table 7.5 In-situ Hf isotopic data for zircons of the gneissic granites
Table 7.6 Sm–Nd isotopic data of gneissic granites in the Jiao-Liao-Ji Belt

6 Source Characteristics

Potential end members that may have contributed to the formation of these granites are felsic magma derived from partial melting of the basement and mafic magma from the mantle. The gneissic granites have a wide range of whole-rock ε Nd(t) values (−8.6 to 1.5). The very negative ε Nd(t) values of some samples (as low as −8) suggest that the felsic end member should have been originated from partial melting of Archean basement rocks. This is supported by the ages of the inherited zircons from these gneissic granites mentioned above (2.53–2.78 Ga), and is compatible with the protolith ages (2.5–2.6 Ga) obtained by Kröner et al. (1988) and Zhao et al. (2001) for the basement rocks beneath the East Block. The Archean basement of the North China Craton is composed mainly of intermediate-mafic granulite/amphibolite and TTG gneisses (Jahn and Ernst 1990; Liu et al. 1992; Zhao et al. 2005). The Dafangshen, Hupiyu and Simenzi granites are rich in potassium, with K2O/Na2O ratios in the ranges of 1.6–2.1, 1.3 and 0.5–1.3, respectively. They also have relatively low ε Nd(t) values (−2.8 to −0.8, −8.6 to −3.7, −4.8 to −0.41, respectively), which are lower than that of the late Archean mafic-ultra mafic amphibolites at 2.2 Ga (−1.19 to 10.51), but in the range of TTG’s ε Nd (2.2 Ga) (−16.41 to 1.94). Hence, we suggest that the source of the felsic end member of the Dafangshen, Hupiyu and Simenzi plutons is dominated by the Archean TTG gneisses, which is consistent with the high contents of SiO2 (70.4–77.1 wt%) of the these plutons. The Muniuhe and Hadabei plutons, however, show relatively low K2O/Na2O ratios (mostly <1) and high ε Nd(t) values (−3.6 to 1.46), suggesting that some amounts of the late Archean amphibolites could have been involved in the source of the felsic end member of the two granite plutons.

We suggest that the other end member having contributed to forming the gneissic granites is mafic magma derived from a mantle source that was previously metasomatized by subduction zone fluids/melts released from a down-going oceanic slab. This is first supported by the positive ε Nd(t) values (+1.5 for sample MN-6) and zircon ε Hf(t) values (+5.6 for sample HP-1) of some gneissic granite samples, which is indicative of involvement of mantle-derived magma in the source. The abundance of euhedral hornblende in MMEs (Fig. 7.3c) and in coeval meta-mafic rocks (Li et al. 2003) suggests that the parental magma of the MMEs should be a hydrous basaltic magma (Chen et al. 2009, 2013), possibly derived from a mantle source above a subduction zone. The coeval meta-mafic rocks show high LILEs, such as Sr, Ba, Th and depletion of HSFEs, such as Nb, Ta, and Ti, which, along with the typical calc-alkaline features of the basaltic rocks (Faure et al. 2004; Wang et al. 2012; Li and Chen 2014), suggest a metasomatized lithospheric mantle for the source of the basalts. In addition, the meta-mafic rocks have extremely high K/Ta (>20,000, data from Li and Chen 2014) and Ba/La ratios (1.3–1062). This is probably caused by the high mobility of K and Ba in subduction zone fluids (Huang et al. 2001).

7 Geodynamic Setting

Most researchers believe that the Paleoproterozoic Liaoji Belt was formed in an intra-continental rifting setting on the NCC (Zhang and Yang 1988; Peng and Palmer 1995; Chen et al. 2003; Li et al. 2003, 2005; Luo et al. 2004, 2008; Zhao et al. 2005; Li and Zhao 2007). Some others, however, suggested that the Paleoproterozoic Belt formed as a result of arc-continent collision at ca. 1.9 Ga (Bai 1993; He and Ye 1998a, b; Faure et al. 2004; Lu et al. 2006; Wang et al. 2011; Meng et al. 2013). Our new data on the Paleoproterozoic gneissic granites suggest a continental arc setting for these granites and thus for the Paleoproterozoic belt, and the arguments are as below.

Most gneissic granites in the Paleoproterozoic belt show calc-alkaline affinity, although some of them show features of A-types, e.g., the Dafangshen pluton, due probably to addition of boron in magma system. The calc-alkaline affinity is manifested by both the chemical compositions of I-type and the common occurrence of hornblende, apatite, magnetite and sphene, typical features of arc magmas. In addition, the arc affinity of the gneissic granites are strongly supported by the hydrous nature and mixed source characteristics (e.g., the presence of MMEs). The hydrous nature, as shown by the abundance of hydrous phases such as hornblende in the plutons, could be attributable to mixing of basaltic magma derived from a mantle- wedge previously metasomatized by subduction fluid/melt. Actually, the coeval metabasaltic rocks also show typical calc-alkaline affinity such as enrichment of LILEs (La, Th, Sr, Rb, etc.) and depletion of HSFEs (e.g., Nb, Ti) (Faure et al. 2004; Li and Chen 2014). Importantly, the gneissic granites show large variation in isotopic compositions with ε Nd(t) = −8.6 to 1.5 and ε Hf(t) = −1.26 to 5.59, which is consistent with the model of magma mixing between mantle-wedge derived basic magma and crustal melts in an arc setting.

As for the polarity of subduction, Bai (1993) proposed a northward subduction beneath the Longgang Block, while passive continental margin-type clastic sedimentary rocks were formed in the northern margin of the Rangnim Block. Final closure of the Paleoproterozoic ocean at ca. 1.9 Ga led to a continent-arc-continent collision and subsequent thrusting, forming the Liaoji Belt. In contrast, Faure et al. (2004) proposed a southward subduction beneath the Rangnim Block based on the fact that the South Liaohe Group is dominated by volcanic rocks. Our new zircon U-Pb data on the gneissic granites reveal the presence of some inherited zircons with ages of ~2500 and ~2700 Ma. These inherited zircons are in agreement with the age data reported for the TTG gneisses from the Longgang block that has received lots of U-Pb dating with ages ranging from 2500 to 3800 Ma but clustering at ~2500 and ~2700 Ma, which are commonly accepted as two major Archean crustal growth periods in the North China Craton (Zhao et al. 2001; Gao et al. 2004, 2005). By contrast, published data suggest that the Rangnim block contain rare rocks older than 2500 Ma (mainly 2440–2500 Ma; Wu et al. 2007a, b). So we can conclude that the source of the arc magmatism is mainly the lower crust beneath the Longgang Block in the north, which suggests a northward subduction in the Paleoproterozoic.

8 Conclusions

Most Paleoproterozoic gneissic granites in the Liaoji Belt are I-type granites, rather than A-type granites. The A-type features shown by some plutons (e.g., the Dafangshen pluton) are attributable to extra addition of boron in the magma system, which have prolonged magma evolution, producing the highly evolved granites. The gneissic granites in the JLJB were formed at ca. 2.17–2.20 Ga, and were metamorphosed at ca. 1.9 Ga. These granites contain inherited zircons with Ahchean ages (ca. 2.5 Ga and 2.7–2.8 Ga). Based on the petrographical, mineralogical and geochemistry characteristics, we suggest that mixing/mingling of lower crust-derived felsic magma with enriched mantle-derived mafic magma might have resulted in formation of these gneissic granites. Overall, the petrology (common occurrence of hornblende, magnetite and sphene), geochemical data (the calc-alkaline affinity) and large variation of Nd–Hf isotopic data suggest that the Paleoproterozoic granites formed in a continental arc setting, rather than in a rifting setting as suggested by many other researchers.