Explosive microfractures induced by K-metasomatism

Journal of Asian Earth Sciences 23 (2004) 307–319 www.elsevier.com/locate/jseaes

Explosive microfractures induced by K-metasomatism Xing-Wang Xu*, Xin-Ping Cai, Bao-Lin Zhang, Jie Wang Institute of Geology and Geophysics, CAS, P.O. Box 9825, Beijing 100029, China Received 19 November 2002; revised 2 April 2003; accepted 3 April 2003

Abstract Ultra-fine mineral aggregates in K-metasomatic rocks of the Hougou area, northwestern Hebei Province, China contain peculiar and distinct intergranular and intragranular geometries, compositions, and textures. These features indicate solidification of expanded and enclosed relict fluids within tensile microfractures. Two basic morphologic types of textures are present: saw-toothed and wheel-shaped, and several composite patterns also are present, such as X-shaped, grid, and network. The appearance of these features indicate explosion from an instantaneous force. These microscopic explosive microfractures are directly related to the enclosed relict fluids. Theoretical estimates show that volume expansion induced by mineral replacements during K-metasomatism may have caused the K-metasomatic fluids to be confined and strongly compressed in order to build up powerful forces that produced the ultra-fine mineral aggregates and explosive microfractures. The thick-walled texture of K-metasomatic rocks confined fractures that propagated only in the replaced rocks with the lowest strength. Both pumping pressures and the propagation of the K-metasomatism were self-governed and controlled by introduced chemical elements, specially Kþ. q 2003 Elsevier Ltd. All rights reserved. Keywords: Explosive; Microfractures; Ultra-fine mineral aggregates; K-metasomatism

1. Introduction Mineral replacement in rocks usually involves volume changes, such as volume expansion during replacement of plagioclase by K-feldspar (e.g. Wadsworth, 1968; Collins, 1994) and volume reduction during the transformation of granulites to eclogite (Austrheim et al., 1999). Such volume changes have potential to generate stress capable of breaking rocks. The patterns and mechanics of microfracturing related to mineral replacement and associated volume changes are unknown, but evidence presented in this paper suggests that microfracturing associated with K-metasomatism is potentially significant. We observed a variety of microstructures in K-metasomatic rocks in the Hougou area, northwestern Hebei Province, China. They are present as ultra-fine mineral aggregates (UMA) and associated microfractures. Their patterns indicate that they were related to relict pore fluids that were enclosed during K-metasomatism. This paper * Corresponding author. Tel.: þ 86-10-62007331; fax: þ 86-1062010846. E-mail address: [email protected] (X.-W. Xu). 1367-9120/03/$ - see front matter q 2003 Elsevier Ltd. All rights reserved. doi:10.1016/S1367-9120(03)00135-4

relates these micofractures and mineral aggregates to K-metasomatism and probable rock volume changes.

2. Geological setting The K-metasomatized rocks in the Hougou area are concentrated along the upper or north sides of a secondorder fault (F3) of the Xiaozhangjiakou –Shuiquangou thrust fault (F2) (Fig. 1). This fault is on the southern part of the east – west striking Chicheng – Chongli fault (F1), in the northern boundary of the North China craton (Ye and Xiang, 1989). It is present in fine-grained alkali granite that yields Rb – Sr isochron ages of about 260.0 ^ 9.2 Ma (Li, 1992) and is part of the Shuiquangou – Dananshan Permian alkali complex. This alkali complex is intruded into Achaean high-grade metamorphic rocks and Proterozoic ultra-basic rocks. There are many gold ore deposits and occurrences in the K-metasomatized rocks. The K-metasomatism is overprinted on fine-grained alkali granite and some small plagiogenesis xenoliths. The corresponding K-metasomatized rocks contain a large quantity of coarse-grained microcline (averaging 1 mm in diameter, the largest up to 10 mm). On the basis of

308

X.-W. Xu et al. / Journal of Asian Earth Sciences 23 (2004) 307–319

Fig. 1. Location (A) and distribution (B) of the K-metasomatic rocks in the Hougou area, northwestern Hebei Province, China. Line A–B shows location of profile A –B in Fig. 9. The isochron ages of the metamorphic and intrusive rocks are after Gao and Gao (1988), Li (1992), and Pen and Ma (1992).

metasomatic stages and the content of microcline determined by microscopic examination and chemical analysis, the K-metasomatic rocks can be divided into three types: (1) weakly mono-generation K-metasomatic rocks (WKMR) with the microcline content from 10 to 85 vol% and recording one stage of K-metasomatism; (2) intensive mono-generation K-metasomatic rocks (IKMR) with microcline content more than 85 vol% recording one stage of Kmetasomatism; and (3) polycyclic K-metasomatic rocks (PKMR) containing up 90 vol% microcline indicating more than two stages of K-metasomatism, which are regularly zoned with orderly outer, middle, and central zones (Fig. 1). The ages of metasomatic microcline range from 176.7 ^ 0.6 to 202.6 ^ 1.0 Ma (laser probe 40Ar/39Ar age; Xu et al., 2002). Analyses of fluid inclusions in the microcline reveal that the metasomatic fluid could be rich in K and sulfate, with a cooling temperature ðTh Þ of 300 – 500 8C and a pressure of 300– 400 MPa. The high content of sulfate (2.14 mg/g) indicates a high O2 fugacity in the metasomatic fluid. The largest K-metasomatic body is the shape of an irregular ear, about 3 km long and 1.5 km wide, with a narrow WKMR and IKMR on the outside and middle and larger polycyclic K-metasomatic rocks in the center. The open part of the K-metasomatic body is connecting to fault F3, and this suggests that fault F3 may be a principle and important pathway for K-bearing fluids (Fig. 1).

3. Ultra-fine mineral aggregates as microfratures In the K-metasomatic rocks studied, microfractures appear as intergranular and intragranular UMAs. These aggregates consist of quartz, sericite, chlorite, calcite, sylvite, hematite, and magnetite, and irregular voids. They have two basic patterns: saw-toothed and wheel-like, and several composite patterns: X-shaped, grid-like, and network-like. 3.1. Saw-toothed ultra-fine mineral aggregates Saw-toothed UMA have shapes that resemble blades and teeth, occurring intergranularly or intragranularly (Fig. 2A). Most teeth have an irregular funnel-form (Fig. 2Bb and C; their locations are shown in Fig. 2A). In such aggregates, some minerals are cryptocrystalline and less than 5 mm. Some teeth are ultra-fine minerals without voids (T1 in Fig. 2B); but others are partially filled with ultra-fine minerals (T2 in Fig. 2C). For example, T3 in Fig. 2C is filled with two mineral grains (sylvite and K-feldspar) and the rest is void. Similarly, not all of the blades are full of UMA, and voids can also be seen in them. The adjacent three teeth of T1, T2 and T3 (Fig. 2B) are connected to the same blade, suggesting a tensile microfracture system. This system most likely formed instantly and synchronically because UMAs fill them. The inhomogeneous distribution of voids indicates that the tensile microfractures would be different from those

X.-W. Xu et al. / Journal of Asian Earth Sciences 23 (2004) 307–319

309

Fig. 2. Microphotographs of saw-toothed ultra-fine mineral aggregates as microfractures (A) and their details (B –D) in thin section GH02 from PKMR. Ab: albite, Calc: calcite, KF: potassium feldspar, S: sericite, IeUMA/B: intergranular ultra-fine mineral aggregate and blade, Mm1 and Mm2: microcline megacryst, Sy: sylvite, Q: quartz, T1, T2 and T3: teeth.

liquid-filled cracks caused directly by injection of fluids (Takada, 1990). Sericite oriented parallel to the wall of the blades and teeth (Fig. 2D) appears to have been injected, implying that UMA-microfractures were forced aside after formation of the microfracture. Another phenomenon associated with saw-toothed UMA is the saw fracture T3 in Fig. 2C that cuts through replaced Albite (Ab), showing a displacement of about 10 mm almost equal to the width of the blade. The fracture T3 entered the microcline Mm1 (Fig. 2A and C), but no displacement accompanied it. This implies that when the microfracture system formed, Mm1 was incompletely crystallized and in a semi-plastic state, because normal microcline has the same property of brittle deformation as albite under the same and lower temperature conditions. 3.2. Wheel-like ultra-fine mineral aggregates Wheel-like UMA generally are composed of irregular cores surrounded by radial spokes, and cores are present intergranularly or intragranularly (Fig. 3A and B; the location of B is shown in A). Both the core and spokes of the wheel-like structures are fractures and microfracture

systems that contain both ultra-fine and cryptocrystalline textures. Cores include three parts: (1) the void central zone, (2) the middle zone, consisting of brecciated ultra-fine quartz and magnetite cemented by microcataclastic gouge of quartz, and (3) the outer zone, consisting of hematite rind with suspended solidified ferric fluids (Fig. 3E). Voids are irregular and common both in the core and spokes. The shapes and brecciated textures in wheel-like UMAs suggest that they are products of explosive expansion. Phenomena compatible with explosion are as follows: (1) The ferric membrane (FM) was broken by cataclastic breccia flows (CBF) that poured out of the middle of the core (Fig. 3C). The FM behaved plastically when it was broken through. Similar breakthroughs can be seen on the left of Fig. 3B. (2) Ring-like fractures, as fractures associated with wheellike fractures, are observed around and parallel to the outline of the wheel-like fractures (Fig. 3C). This implies that significant expansion of the pore fluids produced powerful compressive stress around the IeUMA. (3) Dendrite microfractures at the tips of some spokes (Figs. 3D and 4) usually have ‘Y’ and ‘c’ shapes

310

X.-W. Xu et al. / Journal of Asian Earth Sciences 23 (2004) 307–319

Fig. 3. Microphotographs of wheel-like intergranular ultra-fine mineral aggregates (A) and their details (B–D) in thin section G001 from IKMR. BB: quartz breccia body, BC: quartz breccia crust, CB: microcline collapsed breccia, CBF: cataclastic breccia flow, DF: dendrite like fractures, Hem: hematite, FM: ferric membrane, FO: ferric outer, IeUMA: intergranularly UMA, MM: microcline megacryst, Mt: magnetite, RF: ring like fracture, Q: quartz.

(Fig. 4A): a trunk fracture and one or more pairs of branch fractures. The trunk fracture starts from the tips of spokes and branch fractures are symmetrical with respect to the trunk fracture. The angle between two branch fractures may range from 60 to 908. This

fracture pattern is due to the shear stress concentration at the fracture tip (Clemens and Mawer, 1992). Each dendrite microfracture system has sub-dendrite microfractures (Fig. 4B) that form dendrite microfracture networks. Dendritic microfracture networks may be

X.-W. Xu et al. / Journal of Asian Earth Sciences 23 (2004) 307–319

311

Fig. 4. Microphotographs of typical dendrite microfractures in thin sections G114 (A) and G115-3 (B) from IKMR. Chl: chlorite, IrUMA: intragranular ultrafine mineral aggregates, Q: quartz.

related to fluid filling (Cook and Gordon, 1964; Pollard, 1977). The complexity of dendrite microfractures reflects the strength of the force that produced the wheel-like microfracture system. 3.3. Composite patterns of ultra-fine mineral aggregates Three composite patterns of UMA are present in the Kmetasomatic rocks: (1) X-shaped UMAs, (2) grid UMA, and (3) network UMA. Fig. 5 shows an X-shaped UMA, present in a metasomatic microcline megacryst, randomly oriented and surrounded by fine-grained quartz, anorthoclase and albite (Fig. 5A). The angles between two branches of the X-shaped UMA (Fig. 5C) are about 608, with their bisector parallel to, and their obtuse bisector perpendicular to the elongation of the microcline megacryst. The branches vary in width (Fig. 5B) and some of them have irregular ‘V’-shapes. A small amount of displacement occurred at the crossover point of the X-shaped UMA (Fig. 5D). Taking the ‘X’shaped as an X-shaped microfracture system, the principal axes of compressive stress should be parallel to the obtuse bisector, and hence perpendicular to the elongation of the microcline megacryst in which exists this microfracture system. This suggests that these microfractures were formed in a semi-plastic state, and the metasomatic microcline megacryst was likely viscoelastic when the microfractures were forming. Elongation of microcline megacrysts also could have been induced by a semi-ductile deformation. Grid UMA are present between metasomatic crystals (Fig. 6A), as combinations of saw-toothed UMA. Two parallel blades extend along the long boundaries of microcline megacrysts and teeth parallel and connected to each other and perpendicular to the blades.

Network UMAs are a combination of wheel-like and saw-toothed UMA (Fig. 6B). Intergranular UMA are associated with intragranular fractures. In addition, spoke fractures may parallel planar cleavage and rarely cut the cleavage suggesting that the two formed synchronically. Composite patterns of UMA are difficult to mechanically analyze as single fracture systems. They may be either a combination of synchronically formed fracture systems or a combination of fracture systems formed at different times. So it is reasonable to think about their origin from their components: saw-toothed UMA and wheel-like UMA.

4. Interpretation of textures The absences of UMA-microfractures in the adjacent unaltered granite in the Hougou area indicate that these peculiar UMA-microfractures were not caused by regional tectonic processes. Textures, patterns and distributions of these UMA-microfractures, as mentioned above, indicate that the UMA-microfractures in the K-metasomatic rocks are hydraulic fractures and likely controlled by the occurrence and expansion of the pore fluids in a finite volume. Expansive textures suggest that the pressures of these pore fluids had been pumped. Since fluids within enclosed and confined conditions can be pumped under pressure, naturally, these pore fluids are a type of enclosed fluid. 4.1. Enclosed fluid Intergranular and intragranular occurrences of UMAmicrofractures in the K-metasomatic rocks are consistent with formation from enclosed relict fluids, because these fluids would have provided the material to form

312

X.-W. Xu et al. / Journal of Asian Earth Sciences 23 (2004) 307–319

Fig. 5. Microphotographs of X-shaped composite ultra-fine mineral aggregates within a microcline megacryst (Mm1) (A) and their details (B– D) in thin section G173 from WKMR. Ab: albite, Chl: chlorite, Hem: hematite, IrUMA: intragranular ultra-fine mineral aggregates, Mm1, Mm2 and Mm3: microcline megacrysts, SCP: stress corroded pits, Ser: sericite, Q: quartz, s1 and s3 : axis of maximum and minimum stress.

the UMA-microfractures and the hydraulic force to induce microfracturing into which the UMA-microfractures grew. Consideration of the UMA-microfractures as the products of the enclosed relict fluids strongly implies that such fluids existed throughout the K-metasomatism system. In Fig. 7 there are three generations of metasomatic microcline (McI, McII, and McIII), the last of which (McIII) replaced

part of a wheel-like UMA. Therefore, formation of UMA would have continued throughout the K-metasomatism. A relationship between the content of UMA and Kmetasomatic strength is demonstrated by intragranular UMA that are present in weakly K-metasomatic rocks (WKMR) (volume content , 5%). In contrast, UMA are more abundant in intensively K-metasomatic rocks (IKMR),

Fig. 6. Microphotographs of composite ultra-fine mineral aggregates. (A) Grid ultra-fine mineral aggregates in thin section GH02 from PKMR. (B) Network ultra-fine mineral aggregates in thin section G114 from IKMR. Ab: albite, DF: dendrite like fracture, IeUMA: intergranular ultra-fine mineral aggregates, IrUMA: intragranular ultra-fine mineral aggregates, C: cleavage, Mm1 and Mm2: microcline megacrysts, GF: grid microfracture, TF: tensile microfracture.

X.-W. Xu et al. / Journal of Asian Earth Sciences 23 (2004) 307–319

Fig. 7. Microphotograph showing replacements of previous microcline megacrysts and ultra-fine mineral aggregates by sequent microcline megacrysts in thin section GH03 from PKMR. Ab: albite, DF: dendrite fracture, F: fracture, IeUMAI: intergranular ultra-fine mineral aggregates of the first K-metasomatic generation, IrMAIII: intragranular ultra-fine mineral aggregates of the third K-metasomatic generation, C: cleavage, Mm: microcline megacryst, MmI, MmII and MmIII: three generations of microcline.

both intragranular and intergranular, with a volume content of 5 – 10 vol%. UMA in PKMR are smaller with a volume content of 5 vol%. The content of some minerals in UMAs is related to the protolith of K-metasomatic rocks and their metasomatic strength. The volume content of hematite sharply varies between UMAs in intensively K-metasomatic rocks replaced from granite, and from plagioclase genesis: hematite content up to 80% in plagiogneiss protolith and , 10% in granite protolith. In addition, contacts between UMA and their host or adjacent microcline megacrysts are clean and sharp, suggesting that the UMA were not derived from microcline megacrysts. All of these can be considered as the indirect evidence that enclosed relict fluids existed throughout the K-metasomatism. These observations are consistent with the wheel-like UMA being products of explosion of an enclosed and pressurized fluid. Features suggesting rapid formation of the saw-toothed UMA-microfractures also support a fluidexplosion mechanism. The suspending structure shown by UMA, that is, a structure made up of ultra-fine mineral grains and voids, is typically explosive. 4.2. Explosive force Microcataclastic breccia in the core of a wheel-like UMA (Fig. 3B) was cemented by cataclastic quartz gouge rather

313

than by hematite. This implies that the ferric fluids were not injected into the quartz breccia during the explosive process. Formation of these microcataclastic breccia are different from those in common breccia related to explosions of fluids from magmas (Norton and Cathles, 1973; Xu et al., 2000), and may not be directly caused by explosion of the ferric fluids. Centrifugal escape of microcataclastic breccia (Fig. 3) may have been related to intensive dilatation, and therefore the explosive force of the ferric fluids was not hereditary. Microcataclastic quartz breccia and floc-like cryptocrystalline sericite indicate that deformation temperatures were lower, and thermal expansion, as suggested by Norris and Henley (1976), did not play a major role in this case. Various microfracture patterns suggest that the amount of pressure build-up in different enclosed relict fluids also varied. That is to say, the pumping of pressure to the relict fluids was local rather than regional and uniform. Replacement of metasomatic microcline megacrysts and some associated UMA-microfractures by new generation microcline suggests that these UMA-microfractures formed during K-metasomatism and therefore the explosive force also was related to K-metasomatism. K-metasomatism could provide this force in the following ways. When microcline replaces albite, anorthoclase, and anorthite and biotite breaks down into K-feldspar and magnetite. This results in volume expansion in the rock. For example, the lattice expansion space necessary for crystallization of K-feldspar would have been gained by compressing adjacent pre-crystallized minerals and/or fluid (Collins, 1996). This expansion would have occupied parts of the fluid passages so that some fluid was confined. Lattice expansion by sequent crystallization and replacement would further compress the enclosed fluid, which would make the stress accumulated in the enclosed fluid increase gradually. When the accumulated stress achieved strength greater than the enclosing minerals, failure and explosive fractures were produced (Hubbert and Willis, 1959; Hubbert and Rubey, 1959; Norris and Henley, 1976; Etheridge et al., 1983, 1984; Atkinson, 1984; Xu, et al., 2000). The differential stress ðDPÞ accumulated in the enclosed fluid can be estimated by using the following formula: DP ¼ B dDV

ð1Þ

where dDV is the volume compressibility of the unit enclosed fluid due to the volume expansion of enclosing minerals during metasomatism, and B is the bulk modulus of the enclosed fluid. The dDV can be determined by using the effective unit volume expansion of replaced material ðDVe Þ and the volume content of enclosed fluid ðVrf Þ: The effective unit volume expansion ðDVe Þ is related to four factors: (1) the volume content of enclosed fluid in metasomatic rocks ðVrf Þ; (2) the volume expansion of a monomolecular lattice produced by microcline replacing albite, anorthoclase, and anorthite or by biotite breaking down into K-feldspar and magnetite ðdDVm Þ; (3) the volume content of albite, anorthoclase, anorthite, and biotite in the protolith ðVm Þ

314

X.-W. Xu et al. / Journal of Asian Earth Sciences 23 (2004) 307–319

Fig. 8. Volume changes in the process of K-metasomatism.

and the volume expansion ðDVÞ produced by their complete replacement; and (4) the content of resolved minerals during metasomatism ðVrm Þ: Thus we may have the following formula: dDV ¼ DVe =Vrf

ð2Þ

dDV ¼ ðDVm þ Vrf 2 Vrm Þ=Vrf X  dDV ¼ dDVm £ Vm þ Vrf 2 Vrm =Vrf

ð3Þ ð4Þ

The volume changes mentioned above and their relationships are diagrammatically sketched in Fig. 8. The volume expansion coefficient of molecular lattices produced by K-metasomatism can be roughly estimated through the following metasomatic reactions (Orville, 1963; Deng, 1986; Collins, 1996): NaAlSi3 O8 ðalbiteÞ þ Kþ ¼ KAlSi3 O8 ðmicroclineÞ þ Naþ ð5Þ Na0:66 K0:34 AlSi3 O8 ðanorthoclaseÞ þ 0:66Kþ ¼ KAlSi3 O8 ðmicroclineÞ þ 0:66Naþ

ð6Þ

CaAl2 Si2 O8 ðanorthiteÞ þ 4SiO2 þ 2Kþ ¼ 2KAlSi3 O8 ðmicroclineÞ þ Ca2þ KFe3 ðAlSi3 O10 ÞðOHÞ2 ðbiotiteÞ þ

1 2

ð7Þ

O2

¼ KAlSi3 O8 ðmicroclineÞ þ Fe3 O4 ðmagnetiteÞ þ H2 O ð8Þ

Table 1 gives the estimated results. The volume expansion rate of a monomolecular lattice is 8.6% for microcline replacing albite, 7.8% for microcline replacing anorthoclase, 13.4% for K-feldspar replacing anorthite and quartz, and 2.2% for biotite breakdown into K feldspar and magnetite. If albite, anorthoclase, anorthite, and biotite in alkali granite and plagiogneiss are completely replaced by microcline, the volume expansion of unit volume (approx. 1 m3) is 7.2– 7.6% for alkali granite and 7.2– 7.8% for plagiogneiss. From the volume content of UMAs in the studied Kmetasomatic rocks, it can be inferred that the volume expansion rate by K-metasomatism was about 50%, and that the volume content ðVrf Þ of the enclosed fluid could have been in a range of 2.5– 5%. The volume contents of quartz in IKMR and PKMR from alkali granite are in a range of 0.5 – 1%, this indicates that most quartz would take part in formation of microcline with introduced Kþ and Al3þ or be dissolved and displaced partly during the K-metasomatic process. Quartz in plagiogneiss would instead take part in a reaction to form microcline directly. The volume decrease related to quartz dissolution would provide part of the space for volume expansion. Other lesser volume changes in the K-metasomatism process, such as volume decrease caused by elastic deformation of microcline, are not significant and therefore the effective volume expansion ðDVe Þ by K-metasomatism

Table 1 Monomolecular volume and volume expansion rations related to K-metasomatism Mineral

Molecule unit ˚ 3)a (A

Microcline Anorthite Albite Anorthoclase Biotite Quartz Magnetite P Total ð dDVm £ Vm Þ

180.46 167.54 166.1 167.45 248.9 37.67 73.85

a

dDVm (%)

13.4 8.6 7.8 2.2

Content in granite ðVmg Þ (vol%)

dDVm £ Vmg (%)

55 –60 30 5 10 –5

4.8–5.2 2.3 0.1

7.2–7.6

Content in plagioclase genesis ðVmp Þ (vol%)

dDVm £ Vmp (%)

50–55

6.7–7.4

25–20 25

0.5–0.4

7.2–7.8

Molecule unit parameters are cited from ‘The lattice parameters of high-temperature triclinic sodic feldspars’ (Carmichael and Machenzie, 1964) and ‘The structure of maximum microcline’ (Brown and Bailey, 1964).

X.-W. Xu et al. / Journal of Asian Earth Sciences 23 (2004) 307–319 Table 2 Strength and elastic modulus of protolith and K-metasomatic rocks in the Hougou area Sample number

G174

G173

G001

GH02

Rock type

Protolith (granite) #5 85.68 38.47 0.257 4.491

WKMR

IKMR

PKMR

15 77.6 36.61 0.187 4.067

85 46.84 32.5 0.273 3.668

95 32.74 20.44 0.308 3.133

Content of microcline (vol%) Compressive strength sc (MPa) Elastic modulus E (GPa) Poisson’s ratiom Tensile strength st (MPa)

The experiments were operated by single-axis loading at room temperature and pressure in an Electronic Compression-Testing Machine (YED-200) in the laboratory of rock mechanics, University of Science and Technology Beijing. Six specimens of each type rock were tested with the size of F2 £ 6 cm, and the weighted arithmetic mean of the six testing result data was calculated and presented here.

could range between 9.7 and 12.8% for plagiogneiss protolith and between 0.2 and 8.6% for alkali granite protolith. Therefore, if only 0.1% of the effective volume expansion ðDVe Þ was applied to compress the enclosed fluid, the volume compressibility rate could be 2%. For the enclosed fluid dominated by water (the bulk modulus of water is about 2 GPa), this volume compressibility rate could induce a stress of 40 MPa. This value is about 10 times that of the tensile strength of the alkali granite and K-metasomatic rocks in the Hougou area (Table 2), and four times the average tensile strength of the granitic rocks (Table 2, see also Etheridge et al., 1983, 1984; Atkinson, 1984).

315

The observations and interpretations above suggest that enclosed fluid under compression has the potential to produce explosive UMA-microfractures. The explosive force is simply the differential stress accumulated in the enclosed fluid by volume expansion due to mineral replacement during K-metasomatism.

5. Evolution of K-metasomatism and UMAmicrofractures pulsation of K-metasomatism The compositions of metasomatic microcline vary regularly from WKMR, to IKMR and PKMR with an increase in K and a decrease in Na. For example, in profile A –B0 (Fig. 9), cation numbers for K in microcline from the three kinds K-metasomatic rocks vary from value between 0.931 and 0.943, 0.945 and 0.951, and 0.975 and 0.984, while the cation number of corresponding Na varies between 0.069 and 0.057, 0.055 and 0.049, and 0.025 and 0.016, respectively. This indicates that more than 90% of the Na was replaced by K during the first generation K-metasomatism, and more Na was replaced during formation of microcline in the IKMR than in the WKMR. The content variations of microcline in different generations of PKMR (Table 3) suggest that microcline in each successive generation showed an increase in K and corresponding decrease in Na as metasomatism proceeded and Na was gradually consumed. The number of microfractures apparently increases with intensity of K-metasomatism. For example, there are usually less than 10 microfractures per mm2 in WKMR,

Fig. 9. Major element cation of microcline from WKMR, IKMR and PKMR in profile A–B (location is showed in Fig. 1).

316

X.-W. Xu et al. / Journal of Asian Earth Sciences 23 (2004) 307–319

Table 3 Compositions of different generations of microcline in the thin section GH02 from PKMR K-metasomatic generation

I

Grain and testing number

cc37

cc34

cc41

cc39

cc40

cc42

cc35

cc36

cc38

64.9 0.0 18.1 0.1 0.0 0.0 0.0 0.7 16.0 99.8

64.2 0.0 17.8 0.0 0.0 0.0 0.0 0.7 16.0 98.7

64.4 0.0 17.8 0.1 0.0 0.0 0.0 0.7 16.0 99.0

64.0 0.0 17.9 0.1 0.0 0.0 0.0 0.7 16.0 98.7

63.5 0.0 17.8 0.1 0.0 0.0 0.0 0.6 16.1 98.1

64.8 0.0 18.0 0.1 0.0 0.0 0.0 0.6 16.3 99.8

63.5 0.0 17.5 0.1 0.0 0.0 0.0 0.6 16.2 97.9

64.5 0.0 17.9 0.2 0.0 0.0 0.0 0.6 16.3 99.5

64.7 0.0 17.9 0.1 0.0 0.0 0.0 0.5 16.5 99.8

Oxide

SiO2 TiO2 Al2O3 FeO MnO MgO CaO Na2O K2 O Total

Cation

Si Ti Al Fe Mn Mg Ca Na K

II

3.005 0.000 0.988 0.004 0.000 0.000 0.000 0.065 0.944

3.001 0.000 0.985 0.002 0.000 0.000 0.000 0.062 0.953

3.022 0.000 0.982 0.004 0.001 0.000 0.000 0.055 0.953

but the number increases to about 35 – 100 mm2 in the IKMR. Also, X-shaped UMA-microfractures are mostly observed in the WKMR, whereas wheel-like, ring-like and dendrite UMA-microfractures are present in the IKMR, and saw-toothed and grid-like UMA-microfractures are best developed in the PKMR. 5.1. Strength weakening and thick-wall texture Mineral replacement was influenced by both chemical composition and the rock strength. Experimental data (Table 2) show that the tensile strength, compressive strength, and elastic modulus of rocks are decreased as the content of metasomatic microcline megacrysts increase. The compressive strength of PKMR is half that of the protolith, because the strength of rocks in the replaced areas was weakened during the metasomatic process. Competent protolith (granite) was harder and provided a confined, thick-wall (Fig. 10) for the K-metasomatic rocks during microfracturing, because these thick-wall textures caused the K-metasomatic fractures to propagate only in the replaced rocks. 5.2. Self-organized pumping pressure and evolution of K-metasomatism and UMA-microfractures The characteristics of K-metasomatic rocks, UMA and peculiar microfractures in the Hougou area, suggest that volume expansions produced different patterns during the formation of multiple kinds of K-metasomatic rocks and microfractures. A dynamic model of this K-metasomatic process and the formation and extension of metasomatic

3.001 0.000 0.989 0.003 0.000 0.001 0.000 0.059 0.959

2.999 0.000 0.989 0.004 0.001 0.000 0.000 0.057 0.971

III

3.005 0.000 0.985 0.003 0.001 0.000 0.000 0.050 0.965

3.007 0.000 0.977 0.004 0.000 0.001 0.000 0.057 0.976

3.003 0.000 0.983 0.005 0.000 0.000 0.000 0.058 0.971

3.012 0.001 0.981 0.003 0.000 0.000 0.000 0.047 0.985

microfractures illustrating this volume expansion is shown in Fig. 11. When the K-rich fluids were introduced along secondary faults into former fractures in fine-grained alkali granite adjacent to thrust fault F3, K-metasomatism first began to operate along voids and fissures (Fig. 11). In the initial Kmetasomatic stage (stage I), corrosion by intruded supercritical fluids induced extended and connected voids and fissures in intensely K-metasomatized areas (IKMA) near fluid-filled fractures (e.g. domain B1), and in weakly Kmetasomatized areas (WKMA) far removed these fractures (e.g. domain A1). As K-metasomatism proceeded, replacement of minerals provided the principal influence on stress and fracturing. As more metasomatic microcline crystallized and more space was needed, fluids were pushed and expelled from the crystallizing microcline as dissolved quartz was displaced. This process is similar to pore water expulsion during freezing in an open system (e.g. Taber, 1930; Arvidson and Morgenstern, 1977; Chen et al., 1980) and resulted in channels of fluids expelled from the metasomatic system. In IKMA, some crystallized microcline megacrysts were joined, and relict or stagnant fluids were enclosed, in intergranular voids and pockets. In WKMA, microcline megacrysts were joined directly to the enclosing fine-grained granite. In this stage (II), the metasomatic systems changed from open to closed. If the crystallization velocity of metasomatic microcline was high, some metasomatic fluids were enclosed intragranularly as well. Delayed crystallization of microcline crystals and new crystal replacements were supported by intergranular and

X.-W. Xu et al. / Journal of Asian Earth Sciences 23 (2004) 307–319

317

Fig. 10. Thick wall texture model of K-metasomatic rocks in the Hougou area.

intragranular relict fluids, which produced the third Kmetasomatic stage (III) (Fig. 11). New volume increases built up high pore pressures within enclosed relict fluid pockets in IKMA, along with confining pressure in the surrounding fine-grained granite in WKMA. This dynamic process produced a self-organized pressure pumping system controlled by the K-metasomatic volume expansion, and is different from seismic pumping proposed by Sibson et al. (1975). While microcline megacrysts in WKMA progressively crystallized and pushed away the surrounding fine-grained granite, the megacryst were also compressed by the confined coeval pressure pumping from the fine-grained granite. As the pump-generated confining pressures exceeded the compressive strength of the microcline-with the lowest strength in the K-metasomatic rocks—it was fractured and X-shaped UMA-microfractures formed. When the pump-generated pore pressures exceeded the tensile strength of metasomatic microcline, the microcline megacrysts in IKMA were broken and microfractures formed. As a result, the enclosed fluids were instantly dilated and depressurized, allowing partial fluids from the enclosed intragranular fluid pocket to be injected into the connecting cracks, and producing wheel-like and networklike UMA-microfractures during stage IV (Fig. 11). After completion of the first stage K-metasomatism that formed the WKMR and IMKMR, a large quantity of K-bearing fluids were introduced into new fractures near

the regional faults. These fluids flowed into the Kmetasomatic rocks of the first generation and also the unaltered granite to produce new, second stage, Kmetasomatism. This new K-metasomatism overprinted unaltered granite to form new WKMR and IKMR, and also overprinted the first stage WKMR and IKMR to form PKMR (Fig. 11). The replacement of minerals and development of pumping pressures and fractures in the new WKMA and IKMA were similar to the first Kmetasomatic stage. For example, K-metasomatic processes in domain A2 in the new WKMA repeated the process of A1, and similarly B2 repeated B1, etc. The replacement of minerals and formation of pumping pressures and fractures in the new and first polycyclic K-metasomatic area (PKMA) were also similar to those in the IKMA of the first K-metasomatic stage, and are characterized by fewer UMA. As K-metasomatism of the second stage finished, more microcline and fractures formed, zoned K-metasomatic rocks developed, and the aerial extent of metasomatism increased. If the K-bearing fluids were a continuous part of the Kmetasomatic processes, this would have resulted in replacement of minerals, enclosing of relict fluids, pumping of confined pressures and pore pressures, fracturing of microcline and creation of new fluid channels. This process had the capacity to automatically propagate over a long period of time and was therefore able to produce Kmetasomatic rocks on a large scale.

318 X.-W. Xu et al. / Journal of Asian Earth Sciences 23 (2004) 307–319 Fig. 11. Dynamic model of K-metasomatism and its related UMA-microfractures in the Hougou area. IF: intergranular fluids, KM: k-metasomatism, Mm: metasomatic microcline, RF: enclosed relict fluids.

X.-W. Xu et al. / Journal of Asian Earth Sciences 23 (2004) 307–319

6. Conclusions Microfractures appearing as UMAs are related to Kmetasomatism in K-metasomatic rocks in the Hougou area, and were produced by explosion or an instantaneous force exerted by the enclosed fluids. The explosive force resulted from volume expansion induced by mineral replacements. The pumping processes were self-organized and controlled by the introduced chemical composition, especially Kþ. The K-metasomatism weakened and reduced the strength of the rocks. As a result, a thick-wall texture formed, and fractures were confined to the interiors of the weak metasomatic microcline minerals.

Acknowledgements This study is supported by the National Natural Science Foundation of China (Grant 49802021 and 40272090) and in part by the Chinese Academy of Sciences (Grant KZ951A1-404-02). We would like to express our gratitude to Yihan Xie for inclusion analyses, Baoxue Wang and Tong Yang for strength testing of rocks. We also want to thank Yuanming Pan, Danian Ye, Zhengmin Jin, Bolin Chong, Shanfeng Yi and Yucheng Chai for their suggestions and discussions. Shaozong Zhang, Rui-Xun Liu, and Stephen G. Peters provided reviews that improved the manuscript.

References Arvidson, W.D., Morgenstern, N.R., 1977. Water flow induced by soil freezing. Canadian Geotechnical Journal 14, 237 –245. Atkinson, B.K., 1984. Subcritical crack growth in geological materials. Journal of Geophysical Research 89, 4077–4114. Austrheim, H., Malthe-Sorenssen, A., Jamtveit, B., 1999. Dynamics of fluid-induced contractional metamorphism: a model based on observation on the granulite to eclogite transition. The 12th Kongsberg Seminar, Processes at the Fluid–Rock Interface Abstracts (from the following web: http://www.fys.uio.no/faststoff/sup/Meetings/ AbsKongsberg99/index.html). Brown, B.E., Bailey, S.W., 1964. The structure of maximum microcline. Acta Crystallographica 17, 1391. Carmichael, I.S.E., Machenzie, W.S., 1964. The lattice parameters of hightemperature triclinic sodic feldspars. Mineralogical Magazine 33, 949–962. Chen, X., Jiang, P., Wang, Y., 1980. Pore water pressure of saturated gravel during freezing. Journal of Glaciology and Cryopedolgy 2 (4), 33–37 (in Chinese). Clemens, J.D., Mawer, C.K., 1992. Granitic magma transport by fracture propagation. Tectonophysics 204, 339 –360. Collins, L.G., 1994. The origin of granite and continental masses in an expanding Earth. In: Barone, M., Selleri, F. (Eds.), Frontiers of Fundamental Physics, Plenum Press, New York, pp. 309–314.

319

Collins, L.G., 1996. The Metasomatic origin of the Cooma Complex in southeastern Australia. Theophrastus Contributions to Advanced Studies in Geology 1, 105– 112. Cook, J., Gordon, J.E., 1964. A mechanism for the control of crack propagation in all brittle systems. Proceedings of the Royal Society of London (Series A) 282, 508– 520. Deng, J., 1986. Phase Equilibrium and Origin of Petrology, China University of Geosciences Press, Wuhan, p. 32 (in Chinese). Etheridge, M.A., Wall, V.J., Vernon, R.H., 1983. The role of the fluid phase during regional metamorphism and deformation. Journal of Metamorphic Geology 1, 205–226. Etheridge, M.A., Wall, V.J., Vernon, R.H., 1984. High fluid pressure during regional metamorphism and deformation: implications for mass transport and deformation mechanisms. Journal of Geophysical Research B89, 4344–4358. Gao, M., Gao, F., 1988. Study on the Rb –Sr isochron of granulites from Zhangjiaokou and Xuanhua region, Hebei province. Bulletin of the Tianjin Institute of Geology and Mineral Resources (CAGS) 19, 121– 127 (in Chinese with English abstract). Hubbert, M.K., Rubey, W.W., 1959. Role of fluid pressure in mechanics of overthrust faulting: I. Geological Society of American Bulletin 70, 115– 166. Hubbert, M.K., Willis, D.G., 1959. Mechanics of hydraulic fracturing. Journal of Petroleum Technology 9, 153– 168. Li, F., 1992. Geological characteristics and origin of the gold deposit in Kaltered rocks in the Hougou area. Journal of Precious Metallic Geology 2/3, 126–130 (in Chinese). Norris, R.J., Henley, R.W., 1976. Dewatering of a metamorphic pile. Geology 4, 333–336. Norton, D.L., Cathles, L.M., 1973. Breccia pipes, products of exsolved vapor from magmas. Economic Geology 68, 540 –546. Orville, P.M., 1963. Alkali ion exchange between vapor and feldspar phase. American Journal of Sciences 261, 201 –237. Pen, F., Ma, P., 1992. Origin of metallogenic elements of the Jinjiazhuang gold deposit in the Chicheng area, Hebei. Journal of Precious Metallic Geology 2/3, 115 –120 (in Chinese). Pollard, D.D., 1977. Derivation and evaluation of a mechanical model for sheet intrusions. Tectonophysics 19, 233–269. Sibson, R.H., McMoore, J., Rankin, R.H., 1975. Seismic pumping—a hydrothermal fluid transport mechanism. Journal of the Geological Society London 131, 653–659. Taber, S., 1930. The mechanics of frost heaving. Journal of Geology 38, 303– 317. Takada, A., 1990. Experimental study on propagation of liquid-filled crack in gelatin shape and velocity in hydrostatic stress condition. Journal of Geophysical Research 95 (B6), 8471–8481. Wadsworth, W.B., 1968. The Cornelia pluton, Ajo, Arizona. Economic Geology 63, 101– 115. Xu, X.W., Cai, X., Qun, D., Wang, J., Liang, G., Zhang, X., Liu, L., 2000. Fluids double-fracturing genetic mechanism and mineralization of gold-copper of the breccia pipe at Qibaoshan in Shandong Province. Science in China (Series D) 43, 113 –121. Xu, X.W., Cai, X., Liu, Y., Zhang, B.L., 2002. Laser probe Ar –Ar age of metasomatic K-feldspar from the Hougou gold deposit, Northwestern Hebei province, China. Science in China (Series D) 45 (6), 559– 564. Ye, J., Xiang, S., 1989. The characteristics and formation mechanism of thrust-nappe structure in magmatic body area in Chicheng, Hebei province. Earth Science—Journal of China University of Geosciences 1, 95 –100.