Baddeleyite and zircon, including zircon overgrowths on baddeleyite, co-occur in a granophyric monzonite porphyry intruding volcanic rocks of the Majiahe Formation of the Xiong’er Group in the Xiaoshan area of the western Henan Province (China). It is inferred that the low silica content of the initial magma resulted in the formation of euhedral baddeleyite. Sensitive high-resolution ion microprobe (SHRIMP) U-Pb dating of the baddeleyite yields a weighted mean 207Pb/206Pb age of 1779 ± 8 Ma, representing the crystallization of the porphyry. Survival of baddeleyite as the magma became silica saturated implies a rapid cooling process.

Zircon grains in the porphyry are anhedral to needle shaped, and are often clustered together within K-feldspar stringers or along the interface of quartz and K-feldspar in the granophyric groundmass, suggesting undercooling likely due to rapid emplacement. These observations indicate that zircon did not crystallize until final emplacement. Our U-Pb analyses of zircon yield a weighed mean 207Pb/206Pb age of 1777 ± 8 Ma, similar to that of baddeleyite. Subsequent medium- to high-temperature hydrothermal alteration affected most minerals at subsolidus conditions. Amphibole-plagioclase thermobarometry indicates pressure of 2.4–4.3 kbar and temperature of 470–580 °C for this alteration stage. Related alteration, such as total albitization of plagioclase, crystallization of secondary apatite, and sporadic secondary zircon overgrowths on baddeleyite, indicates that the fluid phase was enriched in Si, Na, and halogens (e.g., F, Cl). In addition to direct replacement of baddeleyite, zirconium required for the development of zircon overgrowths may also have been available through the alteration of Zr-bearing matrix phases such as amphibole and ilmenite.


U-Pb geochronology has been improved dramatically through the recent development of in situ techniques. Zircon is common in silica-saturated rocks, whereas the mineral baddeleyite (ZrO2) occurs primarily as a U-bearing accessory phase in igneous rocks with low silica content (Kamo et al., 1989; Heaman and LeCheminant, 1993; Scoates and Chamberlain, 1995; Klemme and Meyer, 2003; Rajesh and Arai, 2006; Xiang et al., 2012). Baddeleyite has relatively high concentrations of U, negligible common Pb, and is highly resistant to Pb loss because it has a high closure temperature for the diffusion of Pb (Wingate and Compston, 2000). In addition, baddeleyite is typically magmatic in origin and does not commonly occur as xenocrysts (Wingate and Compston, 2000), and thus baddeleyite geochronology can be used to constrain the timing of igneous crystallization. For example, baddeleyite has been extensively applied for dating of alkaline syenites (Heaman and Machado, 1992; Teixeira et al., 1997), carbonatites (Reischmann et al., 1995), and mafic igneous rocks such as kimberlites, layered mafic intrusions, gabbro, and diabase sills (Krogh et al., 1987; Heaman and Grotzinger, 1992; Heaman and LeCheminant, 1993; Schärer et al., 1997; Kerschhofer et al., 2000; Wingate and Compston, 2000; Li et al., 2006; Li et al., 2013). In contrast, baddeleyite in intermediate igneous rocks is rarely reported.

Polycrystalline zircon aggregates often develop as rims on preexisting baddeleyite during high-grade metamorphism (Davidson and van Breemen, 1988; Patterson and Heaman, 1991; Heaman and LeCheminant, 1993; Wingate et al., 1998; Lumpkin, 1999; Rioux et al., 2010). Zircon overgrowths can also be formed during hydrothermal alteration (Heaman and Grotzinger, 1992; Heaman and LeCheminant, 1993; Wingate, 2001). In some cases, baddeleyite may be completely replaced by polycrystalline zircon aggregates (Davidson and van Breemen, 1988). As a high field strength element (HFSE), zirconium (Zr) is usually stable in host minerals, but when fluid is present, it can be easily transported as halogen or hydroxyl complexes (Rasmussen, 2005; Migdisov et al., 2011; Ayers et al., 2012; Bernini et al., 2013). It is noteworthy that baddeleyite and zircon are not the sole Zr-rich phases. Hornblende and ilmenite are also the major carriers of Zr (Rubin et al., 1993; Fraser et al., 1997; Bingen et al., 2001; Charlier et al., 2007).

In this study coexisting baddeleyite and zircon grains as well as zircon overgrowths on baddeleyite were found in a monzonite porphyry intruding volcanic rocks of the Majiahe Formation of the Xiong’er Group in the Xiaoshan area (western Henan Province, China). The monzonite porphyry shows granophyric texture and underwent wide and intense hydrothermal alteration, which also influenced the Xiong’er Group widely in the Waifangshan, Xiong’ershan, and Xiaoshan areas (Zhao et al., 2001; Han et al., 2006). We present petrography, amphibole-plagioclase thermobarometry, and sensitive high-resolution ion microprobe (SHRIMP) U-Pb geochronology of baddeleyite and zircon to better evaluate the processes of baddeleyite and zircon crystallization and zircon overgrowth on baddeleyite.


The Xiong’er volcanic belt occurs at the southern margin of the North China craton covering an area of >60,000 km2 (Fig. 1A; Zhao et al., 2002). The belt is bound to the northwest by the Jiangxian-Lintong fault, to the northeast by the Luoyang-Baofeng fault, and to the south by the Luonan-Luanchuan fault, which separates it from the North Qinling orogenic belt (Fig. 1B; Zhao et al., 2009). The Xiong’er volcanic rocks are mainly distributed in the Zhongtiaoshan, Xiaoshan, Xiong’ershan, and Waifangshan areas. The rocks unconformably overlie Archean to Paleoproterozoic crystalline basement, and are unconformably overlain by Mesoproterozoic–Neoproterozoic terrigenous sandstones, limestones, and calc-silicate rocks (Zhao et al., 2009). The Xiong’er Group has been divided, from bottom to top, into the Dagushi, Xushan, Jidanping, and Majiahe Formations, which are composed of basaltic andesite, andesite, dacite, and rhyolite pyroclastic units interlayered with minor sedimentary interbeds (Zhao et al., 2002). Previous age data for the volcanic rocks based on K-Ar, Rb-Sr analyses of minerals or whole-rock samples, conventional multigrain U-Pb zircon analyses, and several Sm-Nd analyses (Qiao et al., 1985; Zhang et al., 1994; Ren and Li, 1996) indicated that these rocks formed from 1700 to 1400 Ma (Bai, 1993; Ren et al., 2000 and references therein). However, such ages are considered unreliable because of large errors and mobilization of K-Ar and Rb-Sr during later alteration and mixing of different components. In recent years, high-precision geochronology of Xiong’er volcanic rocks, using SHRIMP and laser ablation–inductively coupled plasma–mass spectrometry (LA-ICP-MS) analyses, indicates that most of the volcanic rocks formed at 1.80–1.75 Ga (Zhao et al., 2004; Peng et al., 2008; He et al., 2009). Coeval mafic to felsic subvolcanic rocks intruded the Xiong’er volcanic rocks and the metamorphic basement in this area (Pang and Yan, 2004).

The monzonite porphyry of this study intruded the Majiahe Formation and is locally covered by Cenozoic sediments in Duoyang area, the southeastern part of Xiaoshan area (Fig. 1C). Three drill-core samples of the porphyry at different depths were collected (long 111°22′45.9″E, lat 34°24′47.4″N).

The monzonite porphyry samples are granophyric (Figs. 2A, 2B) and have phenocrysts consisting of plagioclase (30%), hornblende (22%), K-feldspar (13%), and biotite (5%). The groundmass is dominated by granophyric intergrowth of K-feldspar (15%) and quartz (12%). The accessory minerals include ilmenite (3%), baddeleyite, zircon, apatite, rutile, and zirconolite.

Plagioclase is euhedral and tabular, and has undergone epidotization. Thin quartz veins containing fine-grained epidote cut through the cleavage of plagioclase phenocrysts (Figs. 2C and 3A). Hornblende is usually replaced by chlorite and biotite from the core to rim. Widely distributed ilmenite grains occur closely associated with hornblende (Figs. 2D–2F and 3B–3F) and are intensively altered to titanite along crystal boundaries and fractures (Figs. 3B–3F). Apatite usually occurs at the margin of altered hornblende (Figs. 2D, 3B, and 3D) and also shows close spatial relationship with secondary titanite (Figs. 3B–3E); some grains cut through hornblende, ilmenite, and the granophyric groundmass (Figs. 3E, 3F). K-feldspar is also often altered by kaolinization and sericitization.

The granophyric groundmass is composed of intergrowths of K-feldspar and quartz on a 1 mm scale and occurs around euhedral feldspar phenocrysts. K-feldspar and quartz intergrowths coarsen as they emanate from phenocrysts and form fan-like splays. Quartz distributed in K-feldspar varies in habit from cuneiform to vermicular and has a parallel to plumose and radiating texture (Figs. 2F, 2G, 3G, and 3H). Some quartz grains are round or triangular (Figs. 2H and 3I), indicative of the β-quartz polymorph. Intergrown K-feldspar and quartz in the granophyric groundmass often have the same extinction direction, indicating concurrent growth.


We conducted scanning electron microscope analyses on these samples in order to better understand the occurrence of the Zr-rich accessory mineral phases including baddeleyite, zircon, and minor zirconolite (CaZrTi2O7). Baddeleyite occurs as a euhedral and platy mineral and also as inclusions in hornblende, plagioclase, and K-feldspar and along their boundaries (Figs. 4A–4C). Zircon occurs partially as isolated crystals and as sporadic rims (<5 μm) on baddeleyite. Isolated zircon grains are anhedral and small (<100 μm), and zircons with irregular or acicular shape tend to cluster together and occur continuously along the secondary chlorite (Fig. 4D) located in the boundary of plagioclase and quartz or distributed along K-feldspar stringers in the granophyric groundmass (Figs. 4E, 4F). Needle-shaped zircon grains are observed growing along the interface of intergrown K-feldspar and quartz (Fig. 4G). Zircon overgrowths are generally small, sporadic, and occur mostly in the fractures and margins of baddeleyite (Fig. 4H). Only a few baddeleyite grains included in altered hornblende have continuous zircon rims (Figs. 4I, 4J). Zirconolite is distributed in plagioclase and quartz and was replaced by titanite along their margins (Figs. 4K, 4L).


Analytical Methods

Baddeleyite and zircon were dated on the SHRIMP II at the Beijing SHRIMP Center, Institute of Geology, Chinese Academy of Geological Sciences. The analytical procedures were similar to those of Williams (1998) and data were processed using the Excel-based Squid and Isoplot programs (Ludwig, 2001). Mass resolution during the analytical sessions was ∼5000 (1% definition). Spot sizes were 25–30 μm and each site was rastered for 180 s prior to analysis to remove the gold coating. Five scans on 90Zr216O+, 204Pb+, background,206Pb+, 207Pb+, 208Pb+, 238U+, 232Th16O+, and 238U16O+ were made for both baddeleyite and zircon. Standard zircon M257, with an age of 561.3 Ma and U content of 840 ppm (Nasdala et al., 2008), was used to calibrate U and Th content. For baddeleyite and zircon dating, Pb+/U+ ratios were calibrated with the power law relationship between Pb+/U+ and UO+/U+ relative to the Phalaborwa baddeleyite standard and the TEMORA zircon standard, dated as 2059.6 Ma (Heaman, 2009) and 417 Ma (Black et al., 2003), respectively. The standard M257 was conventionally analyzed prior to the unknown samples on the instrumental test target. In this U-Pb analysis of baddeleyite, M257 was also pasted on the unknown target together with an unknown baddeleyite sample and Phalaborwa standard to eliminate the errors induced by the electrical conductivity difference between the test target and the unknown sample target. Common lead corrections were made using the measured 204Pb. The errors for individual analyses in the data table and figures are quoted at 1σ, whereas errors for weighted mean ages are quoted at 2σ.

Dating Results

The results of SHRIMP U-Pb analyses on baddeleyite and zircon from the monzonite porphyry sample are listed in Table 1.

Baddeleyite occurs mostly as euhedral, platy, and monoclinic crystals as much as 200 μm long with length:width ratios ranging from 1.2 to 2 (Fig. 5A). They commonly show weak parallel striations and occur as polysynthetically twinned crystals. U and Th concentrations of 15 baddeleyite grains vary from 101 to 241 ppm and from 4 to 12 ppm, respectively. Analyses on 15 baddeleyite grains are concordant and yield a weighed mean 207Pb/206Pb age of 1779 ± 8 Ma (Fig. 6A). The weighed mean 207Pb/206Pb age of 7 analyses of the Phalaborwa baddeleyite standard is 2038 ± 18 Ma.

Zircon selected for U-Pb analysis is subhedral, lamellar to trigonal, and ranges from 30 to 100 μm in length with length:width ratios ranging from 1.2 to 2.5 (Fig. 5B). Some zircon crystals are striated and also show polysynthetic twinning. U and Th concentrations of 21 zircon grains vary from 173 to 1021 ppm and from 96 to 675 ppm, respectively. The alignment of the discordant data indicates that these zircon grains belong to the same age group as the concordant spots. All data points define an upper intercept age of 1777 ± 17 Ma. The 12 discordant spots plotted below the concordia line with a large discordance tend to have relatively high U contents and much younger ages ranging from 1772 to 1589 Ma, indicating that the discordant analyses were affected by Pb loss. Omitting the disconcordant data, the remaining 9 analyses yield a weighed mean 207Pb/206Pb age of 1777 ± 8 Ma (Fig. 6B). The weighed mean 206Pb/238U age of 6 analyses of the TEMORA zircon standard is 416.8 ± 3.5 Ma.


Mineral Chemistry

Amphibole and adjacent plagioclase pairs were measured rim on rim with a JEOL JXA-8800R at the Laboratory of Electronic Probe, Institute of Mineral Resources, Chinese Academy of Geological Sciences. All analyses were conducted under 20 kV accelerating voltage, 20 nA beam current with 5 μm beam size. As the porphyry samples were intensively altered by hydrothermal fluids, amphibole and plagioclase pairs with smooth surfaces were chosen for temperature and pressure estimation. The measured content of main oxides and calculated cation numbers are shown in Tables 2 and 3. Notably, Zr content in the altered hornblende ranges from 160 to 1810 ppm (Table 2), with an average value of ∼809 ppm. According to the classification of Leake et al. (1997), the hornblendes are classified as ferro-edenite (Fig. 7).

Pressure-Temperature Estimates

Previous Al-in-hornblende barometers only considered the influence of Al content in hornblende on pressure (P) estimation (Hammerstrom and Zen, 1986; Hollister et al., 1987; Johnson and Rutherford, 1989; Schmidt, 1992). A revised formula (Equation 1) proposed by Anderson and Smith (1995) incorporated the effect of temperature (T) using experimental data at ∼675 °C (Schmidt, 1992) and at ∼760 °C (Johnson and Rutherford, 1989) and was taken as the proper barometer for pressure estimation of the porphyry in this study. 
The initial amphibole and plagioclase thermometer proposed by Blundy and Holland (1990) was a semiempirical formula based on the edenite-tremolite reaction and often resulted in high temperatures (Poli and Schmidt, 1992). Holland and Blundy (1994) recalibrated the thermometer by extending the data set and taking nonideal mixing both in hornblende and plagioclase into account. The thermometer A proposed by Holland and Blundy (1994) is also based on the edenite-tremolite reaction and is applicable for quartz-bearing igneous rocks. Holland and Blundy (1994) also formulated the thermometer based on the edenite-richterite reaction to get thermometer B (Equation 2), which is also suitable for quartz-free assemblages. We calculated the pressures and temperatures by iteration of the barometer (Equation 1) proposed by Anderson and Smith (1995) with thermometer B (Equation 2; Holland and Blundy, 1994), because Anderson (1996) suggested that this calibration could yield more reliable results. Results are shown in Table 2. 
for Xab > 0.5, Yab-an = 3.0 kJ. Otherwise, Yab-an = 12.0(2Xab – 1)2 + 3.0 kJ.

Consequently, this calculation yielded subsolidus temperature estimates of 470–580 °C, as shown in the P-T plot (Fig. 8), indicating subsolidus reequilibration of amphibole and plagioclase in this porphyry during later hydrothermal alteration.


Formation of Granophyric Texture during Emplacement

Granophyric texture often occurs in epizonal granitic bodies (Barker, 1970), particularly those associated with extrusive rocks (Buddington, 1959; Dunham, 1965). The granophyric intergrown K-feldspar and quartz form the groundmass of the monzonite porphyry. These groundmass minerals appear to nucleate on preexisting feldspar phenocrysts (Figs. 2F, 2G, 3G, and 3H). In the granophyric intergrowths, K-feldspar stringers that emanate from the same seed porphyry show optical continuity with each other and have the same extinction direction. Quartz grains distributed between feldspar stringers have a parallel, plumose, and radiating arrangement; some are round and triangular in different sections, and they all have the same extinction direction. The granophyric textures indicate rapid and simultaneous crystallization of feldspar and quartz under magmatic undercooling conditions (Vogt, 1921; Dunham, 1965; Barker, 1970; Lofgren, 1971; Smith, 1974; Fenn, 1986; London et al., 1989; Lentz and Fowler, 1992; Candela, 1997; Lowenstern et al., 1997). Lowenstern et al. (1997) proposed that elongate K-feldspars grow as cellular, quasi-skeletal forms on preexisting feldspar phenocrysts during undercooling with quartz nucleating on the K-feldspar stringers and filling in the interstices, to form granophyric intergrowths.

Magmas that ascend close to the surface without eruption can be undercooled by rapid rising of magma and heat conduction with cold country rocks. In addition, highly evolved magmas often possess large amounts of dissolved volatiles (e.g., H2O, F, and Cl; Heimann et al., 2008). Undercooling of magma may be related to saturation and exsolution of these volatiles with depressurization during magma emplacement (Lowenstern et al., 1997; Candela, 1997). Formation of granophyric texture of the studied sample is indicated by stage II in Figure 9.

Hydrothermal Alteration

The monzonite porphyry of this study was intensively altered by hydrothermal fluids. Plagioclase was originally oligoclase to andesine in composition, but was retrograded to albite in the studied samples, as confirmed by electron microprobe analyses (Table 3). Albitization of oligoclase involves a fluid phase that introduces Na and Si and releases Ca and Al (Engvik et al., 2008), and could occur either during epidotization, which might cause emigration of Ca from plagioclase, or by metasomatism of intermediate rocks by siliceous fluids enriched in Na (Li et al., 1979; Allaby and Allaby, 1999). The introduced epidote grains distributed in late-stage quartz veins (Figs. 2C and 3A) further indicate that the hydrothermal fluids were enriched in silica.

Hornblende was intensively replaced by chlorite and biotite, and ilmenite associated with hornblende was generally replaced by titanite along its margins (Figs. 3B–3F). Fine-grained apatite crystals tend to be distributed on the margins of hornblende (Figs. 3B–3D) and show a close spatial relationship with the secondary titanite (Figs. 3B–3E). Some needle-like apatite grains cut through hornblende, ilmenite, and the granophyric groundmass (Figs. 3E, 3F), implying that it is a late-stage crystallized phase and part of a secondary assemblage of chlorite + biotite + apatite + titanites, as shown by stage III of Figure 9. Alteration of hornblende can release the necessary Ca for apatite and titanite growth during chloritization. As a significant host of halogens such as F and Cl, formation of apatite by subsolidus reaction indicates that the hydrothermal fluids may also contain mobile F and Cl.

The monzonite porphyry intruded volcanic rocks of the Majiahe Formation nearly simultaneously with their eruption. Hydrothermal alteration including albitization has also influenced Xiong’er volcanic rocks in wide areas, including Waifangshan, Xiong’ershan, and Xiaoshan (Zhao et al., 2001; Han et al., 2006). Han et al. (2006) proposed that albitization of the Xiong’er volcanic rocks occurred during eruption by comparing these volcanic rocks with the late and unaltered intermediate-mafic intrusions. Hydrothermal alteration on the studied monzonite porphyry and surrounding volcanic rocks of Xiong’er Group is similar and may be synchronous, and the fluids may be related to the volatiles released by magma when it ascended to the shallow levels (Han et al., 2006). Dissolved halogens (F, Cl) in the hydrothermal fluids that altered the studied porphyry may have been derived from such volatiles.

Albitization and epidotization commonly occur during medium- to high-temperature hydrothermal alteration (Li et al., 1979; Hu et al., 2004). Our calculations of subsolidus alteration at 470–580 °C on the studied porphyry samples using amphibole and plagioclase thermobarometry fit well with this hypothesis. Similar alteration and resetting of hornblende and plagioclase thermobarometry was described in Anderson et al. (2012). Hence, our conclusion is that the porphyry was altered soon after emplacement by medium- to high-temperature hydrothermal fluids enriched in Si, Na, F, and Cl.

Crystallization of Baddeleyite

Baddeleyite tends to crystallize under silica-undersaturated conditions, and thus is rarely seen in quartz-bearing intermediate to acidic igneous rocks. The occurrence of baddeleyite in the studied monzonite porphyry cannot be explained by a simple model of fractional crystallization. Schärer et al. (1997) and Kerschhofer et al. (2000) proposed that baddeleyite megacrysts in kimberlite were inherited from solid mantle, and later recrystallized. Heaman and LeCheminant (2001) suggested that similar U-Pb ages of mantle-derived xenocrystic baddeleyite and groundmass perovskite in an alnöite were due to the close temporal association between a precursor mantle metasomatic event leading to baddeleyite crystallization and generation of the alnöite magma. Baddeleyite has also been reported as a late-stage crystallization product in other magmatic settings (Williams, 1978; Lorand and Cottin, 1987; Heaman and LeCheminant, 1993; Dawson et al., 2001).

Baddeleyite grains in the studied monzonite porphyry occur as inclusions in most other minerals, indicating that they formed at the early stage of crystallization. Baddeleyite formed under such conditions is much larger than that of rapid cooling in mafic dikes (normally <100 μm; Wingate and Compston, 2000; French et al., 2002; Schmitt et al., 2010). An inheritance origin of baddeleyite is precluded in this study because these monoclinic baddeleyite grains are euhedral, showing similar characteristics of luminescence. Moreover, U-Pb analyses of baddeleyite are clustered together on concordia and defined a weighted mean 207Pb/206Pb age that is consistent with that of zircon within error.

Ilmenite crystallized closely associated with early-stage hornblende and occurs as ∼3% in the studied monzonite porphyry. Contemporaneously, baddeleyite crystallized early from the original magma at stage I in Figure 9, implying initial relatively low silica activity. Low silica activity during early-stage crystallization is also supported by the lack of inherited zircon. The weighted mean 207Pb/206Pb age (1779 ± 8 Ma) of baddeleyite is interpreted as the primary igneous crystallization age of the porphyry.

Crystallization of Isolated Zircon

Zircon occurs commonly in igneous rocks of intermediate to acidic composition and ranges in size from ∼20 to 200 μm (Silver and Deutsch, 1963). The grain size and elongation (length:width) ratio of zircon is commonly believed to reflect rate of crystallization (Corfu et al., 2003). Euhedral, large zircon crystals with length:width ratios of ∼2–4 usually form in early zircon-saturated melts (Hoskin and Schaltegger, 2003). Small, needle-shaped zircon grains with large elongation ratios (to ∼12) usually imply crystallization (Hoskin and Schaltegger, 2003) in rapidly cooled, porphyritic, subvolcanic, high-level granites and gabbros (Corfu et al., 2003). Late-crystallized zircon crystals are mostly anhedral because they tend to grow interstitially to earlier formed minerals (Scoates and Chamberlain, 1995).

The isolated zircon grains in the monzonite porphyry are commonly small (<100 μm) and anhedral with acicular shapes (Figs. 4D–4G). Some needle-shaped zircon grains are distributed along with elongate K-feldspar or on the interface of quartz and K-feldspar in the granophyric groundmass. The isolated zircon grains are best explained as a late-stage and rapidly crystallized phase, suggesting that the early-stage magma was zircon undersaturated. It is inferred that saturation was only reached late in the crystallization history, because of very low Zr contents and/or high Zr solubility (Hanchar and Watson, 2003).

Apart from baddeleyite and zircon, Zr also occurs in hornblende and ilmenite in the studied porphyry. The average residual Zr content in altered hornblende is ∼809 ppm, which is significantly higher than the Zr concentrations of whole-rock compositions (ranging from 235 to 399 ppm; Cui et al., 2011). After fractionation of early-stage Zr-rich minerals, including baddeleyite, ilmenite, and hornblende, Zr concentrations in the residual melt may have been far below the saturation value.

The granophyric texture formed by undercooling during emplacement, and the hydrothermal alteration could have taken place nearly at the same time (Han et al., 2006), as discussed herein. The fact that the early-forming baddeleyite was not later dissolved as the magma became silica saturated also indicates a rapid cooling process of the porphyry magma during emplacement. Considering that zircon solubility decreases with decreasing temperature (Wilke et al., 2012) and increasing silica activity (Watson and Harrison, 1984; Bernini et al., 2013), the isolated zircon grains must have formed at late stage of the monzonite porphyry magmatism during its final emplacement, as shown by stage II in Figure 9. The weighed mean 207Pb/206Pb age (1777 ± 8 Ma) of isolated zircon grains is interpreted as emplacement time of the porphyry and within error agrees with the dating of baddeleyite.

Zircon Overgrowth

Zircon overgrowths on baddeleyite can form during lower greenschist to granulite facies metamorphism (Davidson and van Breemen, 1988; Patterson and Heaman, 1991; Heaman and LeCheminant, 1993; Wingate et al., 1998; Lumpkin, 1999; Rioux et al., 2010) and hydrothermal fluid alteration (Heaman and Grotzinger, 1992; Heaman and LeCheminant, 1993; Wingate, 2001). It has been reported that the width of zircon rims increases when baddeleyite is subjected to progressively higher grades of metamorphism (Heaman and LeCheminant, 1993). Zircon rims formed in medium greenschist or higher grade metamorphic mafic rocks are usually wider than 10 μm and may cause Pb loss for the host baddeleyite (Davidson and van Breemen, 1988; Patterson and Heaman, 1991; Heaman and LeCheminant, 1993). In contrast, lower greenschist facies metamorphism or low-temperature fluid alteration forms zircon overgrowths that are small (<5 μm) and discontinuous on baddeleyite (Heaman and Grotzinger, 1992; Heaman and LeCheminant, 1993).

Unlike the late-stage magmatic zircon, the secondary zircon rims in the studied porphyry (Figs. 4H–4J) are mostly sporadic, small (generally <5 μm), and caused no obvious Pb loss for the host baddeleyite given the concordant U-Pb analysis results. These secondary zircon rims on baddeleyite are considered to be of hydrothermal origin. Zr could be easily transported as halogen or hydroxyl complexes (Rasmussen, 2005; Migdisov et al., 2011; Ayers et al., 2012; Bernini et al., 2013), mainly in the form of hydroxyfluoride species ZrF(OH)30 and ZrF2(OH)20 as recommended by Migdisov et al. (2011). Therefore, dissolved F and Cl in the hydrothermal fluids could effectively promote the dissolution and mobility of Zr and other HFSEs from the host minerals in the monzonite porphyry. Many baddeleyite grains are corroded by secondary zircon overgrowths internally (Figs. 4I, 4J) and must serve as the dominant but not the sole source of mobile Zr for secondary zircon growth. Notably, zircon has a very different Th/U ratio compared to the parent baddeleyite.

In general, hornblende contains highly mobile Zr (Rubin et al., 1993; Fraser et al., 1997) and ilmenite is also a major carrier of Zr (Bingen et al., 2001; Charlier et al., 2007), both of which occur in the studied porphyry and were intensively altered. Only a few baddeleyite grains included in altered hornblende have complete zircon coronas. Bingen et al. (2001) documented the contribution of ilmenite to zircon corona formation during granulite facies metamorphism. Significant Zr must have been leached from hornblende and ilmenite altered by halogen (e.g., F, Cl) rich fluids. The complete subsolidus reaction related to secondary zircon overgrowth is suggested as follows: 

With the increasing Zr concentration and higher silica activity of the melt, we conclude that zircon began to precipitate and nucleate on preexisting baddeleyite crystals to form polycrystalline aggregates. Intergrowths and overgrowths of zircon occurred by replacement of baddeleyite accompanied by additional silica supplied by fluids and extra Zr from other Zr-rich minerals (dominantly hornblende and ilmenite).


SHRIMP U-Pb dating of baddeleyite and zircon constrain the initial crystallization and final emplacement age of the studied monzonite porphyry as 1779 ± 8 and 1777 ± 8 Ma, respectively, which agree within error. Euhedral, platy baddeleyite crystallized early from the primitive, least-fractionated porphyry magma with lower silica activity. Primary zircon formed later during rapid cooling of the late-stage, more fractionated melt enriched in silica during its final intrusion into volcanic rocks of the Majiahe Formation.

Subsequent medium- to high-temperature hydrothermal alteration produced secondary zircon as intergrowths and overgrowths on baddeleyite. The fluids were enriched in silica, Na, and halogens. Additional Zr from decomposition of amphibole and ilmenite by the halogen-bearing fluids also made significant contributions to secondary zircon overgrowth.

We appreciate the editorial patience and comments of Raymond M. Russo. The comments of two anonymous reviewers are gratefully acknowledged. This study was financially supported by the Chinese Ministry of Land and Natural Resources (grant 201311116), the National Natural Science Foundation of China (grant 41173065), Chinese Ministry of Science and Technology (grant 2012FY120100), and the Basic Outlay of Scientific Research Work from the Chinese Ministry of Science and Technology (grant J1403).