Temporal–Spatial Separation of Cu from Mo in the Jiama Porphyry Copper–Polymetallic Deposit, Southern Tibet, China

: Jiama is a super-large porphyry copper–polymetallic deposit located in the Gangdese metallogenic belt of southern Tibet. The deposit consists of a combination of a polymetallic skarn, Cu–Mo mineralization at the contact between the Jiama Porphyry and hornfels, and distal Au mineralization in fault. The current metal reserves are 7.4 Mt Cu, 0.6 Mt Mo, 1.8 Mt Pb–Zn, 6.65 Moz Au, and 360.32 Moz Ag, with a skarn to porphyry–hornfels host-rock ratio of ~3:1. Based on detailed ﬁeld and laboratory investigations, this paper indicates that: (i) the skarn and porphyry–hornfels orebodies are almost entirely preserved; (ii) the emplacement age of the Cu-bearing plutonic rocks is earlier than the plutons containing elevated Mo assays; (iii) the permeability of the wall rocks gradually decreases in an upward direction; (iv) the ﬂuid temperature during the precipitation of Cu was higher than that of the Mo mineralization; (v) the lithospheric pressure during the precipitation of Cu and Mo was the same; (vi) the laser Raman spectroscopy shows that the ﬂuid carrying the Cu was rich in magnetite, hematite, and anhydrite, and the ﬂuid carrying Mo was rich in pyrite, CO 2 , and CH 4 ; and (vii) the SR–XRF mapping shows that the concentration of Cu in the mineralizing ﬂuid was high and that of Mo was low when Cu was deposited. Conversely, the concentration of Cu was relatively low and the concentration of Mo was relatively high during deposition of the Mo. This study also shows that the temporal and spatial separation of Cu and Mo in the porphyry copper–polymetallic deposit at Jiama was associated with the emplacement of the Jiama Porphyry. Transportation of mineralized ﬂuid was controlled by the permeability of the wall rocks, and deposition of the metals related to changes along a redox front and pressure releases during hydrothermal brecciation at the roof of the Jiama Porphyry.

Zheng et al. [10,13] discussed the distribution of metals in skarn at Jiama, and recognized a metal zonation that extends from a central Cu-Mo zone outward to Cu (-Mo), Pb-Zn (-Cu), and Au zones. They proposed that the zonation is related to lateral passages of fluids where metals are precipitated with drops in temperature or pressure in the skarn. However, continued mining and drilling has delineated the zonation in a vertical orientation from an early upper-level of Cu mineralization to a late lower-level Mo mineralization at the contact of the Jiama Porphyry with hornfels. This indicates that the mineralizing fluids migrated in a vertical direction.
There are limited exposures of plutonic rocks in the Jiama area, and those intersected in drill holes are stocks and dykes intruding hornfels in the Copper Mountain-Zegulang area (Figure 1b). The plutonic rocks are mafic to felsic and include dolerite, porphyritic quartz diorite, porphyritic granodiorite, porphyritic monzonite to quartz monzonite, and porphyritic monzogranite. The mafic rocks are the oldest and are locally mineralized in  [37]; and (b) generalized geological map of the Jiama ore district (modified from [10]). Abbreviations: JSS = Jinsha suture; BNS = Bangong-Nujiang suture; IYZS = Indus-Yarlung Zangbo suture. gently (10°-30°) to the NE (Figures 2 and 3). The orebody is ~3 km long, 2.5 km wide, and is open to the NE at depth of around 1000 m.
The orebody at the contact between the porphyry and hornfels is a cylindrical body controlled by multiple stages of magmatism. The orebody strikes NW-SE with a subvertical dip for over 1200 m, and is hosted by hornfelsed clastic units in the Linbuzong Formation (Figures 2 and 3; [10]). The orebody at the contact between the porphyry and hornfels is a cylindrical body controlled by multiple stages of magmatism. The orebody strikes NW-SE with a subvertical dip for over 1200 m, and is hosted by hornfelsed clastic units in the Linbuzong Formation (Figures 2 and 3; [10]).

Spatial Relationships
The porphyry-hornfels orebody formed during fracturing related to fluid overpressure at the top of the magma chamber during the emplacement of the porphyry. This resulting in the development of a cylindrical-shaped fracture zone allowing passage of overpressured mineralized fluids [9,39].

Spatial Relationships
The porphyry-hornfels orebody formed during fracturing related to fluid overpressure at the top of the magma chamber during the emplacement of the porphyry. This resulting in the development of a cylindrical-shaped fracture zone allowing passage of overpressured mineralized fluids [9,39].
The maximum vertical thickness of the orebody is 940 m. At an elevation >4800 m, the orebody is 200-400 m thick, contains 0.3-0.5% Cu and <0.03% Mo, and is hosted by biotite hornfels and the porphyry (Figure 4). The orebody is 400-570 m thick at an elevation of The maximum vertical thickness of the orebody is 940 m. At an elevation the orebody is 200-400 m thick, contains 0.3-0.5% Cu and <0.03% Mo, and is h biotite hornfels and the porphyry (Figure 4). The orebody is 400-570 m thick at a tion of <4800 m, with an average grade >0.06% Mo and <0.2% Cu (Figure 2b,c an 3b,c). The host rocks are felsic and diopside hornfels and the porphyry (Figure 4)

Temporal Relationships
Detailed logging and over 100 samples of drill hole were collected from t area for petrological studies. The study indicates that the porphyry-hornfels oreb be divided into two relatively separate an early Cu and a later Mo mineralizing e mentioned above). This includes a progression of: Cu-I → Cu-II → Mo-I → M Cu-III → Mo-III stages. These are discussed below, starting from the relative stage progressing to the youngest.

Characteristics of the Cu Stages
The Cu-I and Cu-II stages in the progression represent the main Cu min event at Jiama. The Cu-I stage consists of fine-grained disseminated chalcopyrit and a small amount of quartz, hosted by banded biotite hornfels associated with dal albite. The Cu-II stage consists of crosscutting <1 mm thick pyrite-chalcopyrit veins, which cuts the disseminated sulfides of the first stage (Figure 5a-d,j-l).

Temporal Relationships
Detailed logging and over 100 samples of drill hole were collected from the Jiama area for petrological studies. The study indicates that the porphyry-hornfels orebody can be divided into two relatively separate an early Cu and a later Mo mineralizing events (as mentioned above). This includes a progression of: Cu-I → Cu-II → Mo-I → Mo-II → Cu-III → Mo-III stages. These are discussed below, starting from the relatively oldest stage progressing to the youngest.

Characteristics of the Cu Stages
The Cu-I and Cu-II stages in the progression represent the main Cu mineralizing event at Jiama. The Cu-I stage consists of fine-grained disseminated chalcopyrite, pyrite, and a small amount of quartz, hosted by banded biotite hornfels associated with framboidal albite. The Cu-II stage consists of crosscutting <1 mm thick pyrite-chalcopyrite-quartz veins, which cuts the disseminated sulfides of the first stage (Figure 5a-d,j-l).
The Cu-III stage crosscuts the main Mo stages and is earlier than the last Mo stage. This stage is characterized by subvertical quartz-pyrite-chalcopyrite veins that are up to 10 mm wide, contain medium-grained chalcopyrite, and are commonly characterized by a miarolitic cavities (Figure 5e-f).

Characteristics of the Mo Stages
The first Mo stage is represented by <10 mm thick quartz-molybdenite veins that are straight and dip <40 • NE. The molybdenite is fine-grained and commonly concentrated at the contact between the veins and wall rocks (Figure 5g).
The second (Mo-II) stage is the main Mo mineralizing event forming a stockwork or conjugate set of <10 mm thick quartz-molybdenite veins, which dip 40 • -70 • N and ENE ( Figure 5h). The veins are irregular, and mineralization is richest where veins are narrow or at their intersections containing medium-sized molybdenite. The Cu-III stage crosscuts the main Mo stages and is earlier than the last Mo stage. This stage is characterized by subvertical quartz-pyrite-chalcopyrite veins that are up to 10 mm wide, contain medium-grained chalcopyrite, and are commonly characterized by a miarolitic cavities (Figure 5e-f).

Characteristics of the Mo Stages
The first Mo stage is represented by <10 mm thick quartz-molybdenite veins that are The third (Mo-III) stage consists of irregular quartz-molybdenite veins with dips of generally >70 • with widths of <10 mm (Figure 5f,i). The molybdenite grains are mediumsized and either disseminated in the veins or in a linear distribution along the center of the veins. This is also an important Mo stage that crosscuts the Mo-II stage.

Sample Selection
The focus of the fluid inclusion study is the paragenesis of the Cu-II, Cu-III, Mo-I, Mo-II, and Mo-III stages using quartz crystals from each stage. Representative samples of the porphyry-hornfels orebody were uniformly selected from the No. 16 cross-section through the center of the orebody and includes 13 drillhole core samples from the Cu stages and 15 from the Mo stages (Figure 2a).

Petrography and Microthermometry
Petrographic observations of the rock samples were completed using 0.3-mm-thick double-sided polished sections and a LEIKA DM2500P microscope with a magnification of 100-1000 times. The inclusion type, shape, size, filling percentage, and other characteristics were recorded.
Fluid inclusion phase changes during heating and freezing cycles were determined on a dual-purpose heating-freezing stage at the Department of Geology, Chengdu University of Technology, Chengdu, China. Firstly, the fluid inclusions were classified under a LEIKA DM2500P microscope, and the inclusions with a clear gas-liquid boundary that were suitable for microthermometry were selected. Then, the temperature was measured on a Linkam-THMS600 heating-freezing microscope (temperature range: −196 • C-+600 • C), produced in England. Between −120 • C and −70 • C, the determination accuracy is 0.5 • C. The accuracy of determination at −70 • C-+100 • C is 0.2 • C. The accuracy of determination at 100 • C-500 • C is 2 • C. During the test, the rate of temperature rise/fall is between 0.5 • C-15 • C min −1 , and the rate near the phase transition point decreases to

Laser Micro-Raman Spectra Analyses
Laser Raman spectroscopy analysis of fluid inclusions was conducted at the Chengdu Institute of Geology and Mineral Resources, Chinese Geological Survey, Chengdu, China. The experimental instrument is a Renishaw RM2000 laser Raman probe, made in England. Experimental conditions: temperature: 23 • C, Ar ion laser (514.5 nm), laser power: 40 mw, scanning speed: 10 s/3 times superposition, spectrometer slit: 10 um.

SR-XRF Analyses
Trace elements were analyzed by SR-XRF mapping and in situ sampling of single inclusions at the Institute of High Energy Physics, Chinese Academy of Sciences, Shanghai, China. The analytical instrument is synchrotron radiation X-ray fluorescence microprobe (SR-XRF), and the experimental method is micro-beam X-ray fluorescence analysis (µ-XRF). In situ nondestructive analysis is carried out to test the element composition and its two-dimensional distribution characteristics in the sample. The X-ray source used in the experiment comes from a 4W1B beam line focused by a K-B mirror in the Shanghai Synchrotron Radiation Facility (SSRF). The electron energy of BEPC storage ring is 2.2 GeV, the beam intensity is 60-120 mA, the photon energy is 14.04 keV, the spot size is about 2 µm × 2 µm, and its spatial resolution is up to 2 µm. The detection limit is 10 −10 -10 −12 g, and the relative concentration is 10 −6 . The magnification of the microscope in the microscopic observation system is 140 times, and the energy resolution of the detector is 140 eV. The NIST SRM612 and NIST SRM614 glass samples synthesized by the National Bureau of Standards were used to correct the test results.

Petrography and Microthermometry
Fluid inclusions in quartz from the porphyry-hornfels include liquid (L), vapor (V), and solids (S) based on their phase state and percentage of vapor-liquid filling at room temperature. The L-type inclusions contain vapor and liquid phases at room temperature with 10-50% vapor. The inclusions are 5-40 µm in diameter, irregular or oval, and negative crystals (Figure 6a-d,f,g,i). The temperature at which Cu-stage inclusions convert to homogeneous as a liquid upon heating is 229 • -478 • . The temperature at which Mo-stage inclusions convert to homogeneous as a liquid upon heating is 189 • -427 • (Table 1).

Petrography and Microthermometry
Fluid inclusions in quartz from the porphyry-hornfels include liquid (L), vapor (V) and solids (S) based on their phase state and percentage of vapor-liquid filling at room temperature. The L-type inclusions contain vapor and liquid phases at room temperature with 10-50% vapor. The inclusions are 5-40 μm in diameter, irregular or oval, and negative crystals (Figure 6a-d,f,g,i). The temperature at which Cu-stage inclusions convert to homogeneous as a liquid upon heating is 229°-478°. The temperature at which Mo-stage inclusions convert to homogeneous as a liquid upon heating is 189°-427° (Table 1). L-and S1-type inclusions in a Cu-bearing vein, and S1-type inclusions containing hematite (Hem) and pyrite (Py) (ZK1614-158.2 m); (c) combination of L-, V-, S1-, and S2-type inclusions in a Cu-bearing vein, and an S1-type inclusion containing Hem and an S2-type inclusion containing halite (Hal) (ZK1615-438 m); (d) combination of L-and S2-type inclusions in a Cu-bearing vein, and an S2-type inclusion containing Hal and Py (ZK1615-219 m); (e) combination of L-, V-, S1-, and S2-type inclusions in a Cu-bearing vein, and an S1-type inclusion containing irregular chalcopyrite (Cp), magnetite (Mag), and red Hem, while an S2-type inclusion contains cubic transparent Hal and red Hem (ZK1617-133.9 m); (f) combination of L-and V-type inclusions in a Mo-bearing vein (ZK1616-167 m); (g) combination of L-and S1-type inclusions in a Mo-bearing vein (ZK1616-404 m); (h) combination of L-and S2-type inclusions in a Mo-bearing vein, and an S2-type inclusion containing irregular sylvite (Syl) and cubic Hal (ZK1615-364.8 m); (i) combination of L-, S1-, and S2-type inclusions in a Mo-bearing vein, an S1-type inclusion containing Cp, and an S2 type inclusion containing transparent Hal and Cp (ZK1616-742.8 m).
The V-type inclusions are two-phase vapor-liquid inclusions at room temperature with 60-80% vapor. The inclusions are 5-35 μm in size and mostly irregular, spindle-like The V-type inclusions are two-phase vapor-liquid inclusions at room temperature, with 60-80% vapor. The inclusions are 5-35 µm in size and mostly irregular, spindle-like, and rectangular (Figure 6c,f). The temperature at which V-type inclusions convert to homogeneous as vapor upon heating in Cu stage and Mo stage is 328 • -415 • (Table 1). The S-type inclusions are multi-phase fluid inclusions containing vapor-liquid-solid at room temperature. The S-type inclusions can be divided into S 1 and S 2 types based on their daughter minerals. The S 1 -type inclusions are 10-25 µm in size, generally irregular in shape, and negative crystal. Vapor constitutes 35-60% of the inclusion, which contains only opaque irregular, triangular, or cubic daughter minerals. The grain size of the daughter minerals is <8 µm (Figure 6b,c,e,g,i). The S 2 -type inclusions are <40 µm in size, generally irregular in shape, and negative crystal. The vapor constitutes <30% of the inclusion that contains halite, sylvite, hematite, anhydrite, pyrite, magnetite, and chalcopyrite (Figure 6c,d,h,i).
Fluid inclusions in the quartz veins from the Cu stage are mainly isolated or linearly distributed, with some veins showing a scattered distribution. The inclusions are irregular or spindle-like in shape and generally small (commonly 10 to 20 µm in size). The inclusions are predominantly L-type, which account for~65% of the total fluid inclusions, followed by the V-(~25%) and S-(10%) types. The S 1 -type inclusions account for~80% of the S-type, and contain daughter minerals such as hematite, magnetite, anhydrite, and chalcopyrite. The S 2 -type inclusions are rare, accounting for~3% of the total S-type inclusions (Figure 6b-e).
Fluid inclusions in quartz veins from the Mo stage are relatively large (~15 µm). Ltype inclusions (~60%) are the most common in this stage, with lesser S-types (~25%) and V-types (~15%). The S-type inclusions are mainly S 2 -type, accounting for~20% of the total, and contain daughter minerals such as in inclusions from the Cu-bearing veins, although daughter minerals such as magnetite, hematite, and anhydrite, as well as oxidized minerals, are markedly lower in the Mo-bearing veins than in the Cu-bearing ones (Figure 6g-i).
The temperature, pressure, and salinity estimates for the different fluid inclusion types are presented in Table 1.

Laser Micro-Raman Spectra
The Raman spectrum analysis was conducted at room temperature on daughter minerals in the Cu-and Mo-bearing quartz veins. Chalcopyrite, hematite, magnetite, anhydrite, and pyrite were identified by fitting the Raman shift curve (Figure 7a-e; [40]. In combination with detailed microscopic identifications, the daughter minerals were found to be predominantly distributed in the Cu-stage quartz veins, and rarely in the Mo-stage quartz veins. Carbon dioxide peaks were detected in 10 V-type inclusions of the Cu stage, but CH 4 peaks were not detected (Figure 7f). Carbon dioxide peaks were also detected in five V-type inclusions of the Mo stage, and three of these inclusions have CH 4 peaks (Figure 7g). In addition, broad H 2 O peaks were detected in L-type inclusions from the quartz veins of the Cu and Mo stages (Figure 7h).

SR-XRF Mapping
The fluid inclusions in the Cu-bearing quartz veins from the porphyritic quartz diorite and Mo-bearing quartz veins from the porphyritic granodiorite were analyzed using Synchrotron radiation X-ray fluorescence (SR-XRF). The fluid inclusions were analyzed for Au, Cu, Fe, Mn, Mo, S, and As to study the proportions and nature of ore-forming elements in them.
The relative concentrations of Cu and Fe in the Cu-bearing quartz veins are low in the vapor-liquid phase. The concentration of Mo in the liquid phase is relatively high and very low in the vapor phase. Relative concentrations of Au are high in the vapor-liquid phase. The relative concentration of S in the vapor phase is much higher than that in the liquid phase (Figure 8). In the Mo-bearing quartz veins, the relative concentrations of Mo are low in the vapor-liquid phase. In contrast, Cu and Fe are relatively high in the vapor phase and very In the Mo-bearing quartz veins, the relative concentrations of Mo are low in the vaporliquid phase. In contrast, Cu and Fe are relatively high in the vapor phase and very low in the liquid phase. The relative concentrations of Au are high in the vapor phase and low in the liquid phase, and the relative concentrations of S are much higher in the vapor phase than in the liquid phase (Figure 9).

Magmatic Metal Fractionation and Magma Evolution
Tang et al. [38] relate igneous rocks in the Jiama area with the genesis of the ore, and divided them into an early Cu mineralization associated with porphyritic quartz diorite and porphyritic granodiorite, and a later Mo mineralization associated with porphyritic monzonite to quartz monzonite and porphyritic granodiorite (Figures 2 and 3). The differences in the ore that formed during the two stages depend on the geochemical behavior of the metals during the crystallization of the magma. In a series of experiments conducted at 750 °C and 1.4 kbar, Candela and Holland [41] measured the extent to which Cu and Mo are partitioned between granitic melt and coexisting aqueous fluids containing both Cl − and F − . These experiments showed that the partitioning of Cu is strongly controlled by the concentration of Cl − , defined by the equation The different ages of plutonic rocks that are sources for Cu and Mo are an important factor in their temporal separation. In addition, the oxygen fugacity (fO2) of the magma plays an important role in elemental separation [42][43][44]. Fluid inclusions in the Cu-mineralized quartz veins contain hematite, magnetite, and anhydrite, indicating that the magma was oxidized during the early stages (Figure 6b-e and Figure 7b-d). The distribution ca-

Magmatic Metal Fractionation and Magma Evolution
Tang et al. [38] relate igneous rocks in the Jiama area with the genesis of the ore, and divided them into an early Cu mineralization associated with porphyritic quartz diorite and porphyritic granodiorite, and a later Mo mineralization associated with porphyritic monzonite to quartz monzonite and porphyritic granodiorite (Figures 2 and 3). The differences in the ore that formed during the two stages depend on the geochemical behavior of the metals during the crystallization of the magma. In a series of experiments conducted at 750 • C and 1.4 kbar, Candela and Holland [41]  Fluid/melt = 2.5. Using these criteria, Cu being a compatible element is more readily partitioned into the fluid during the early stages of magmatic crystallization. This results in Cu mineralization concentrating in early granitic stocks or dykes. In contrast, Mo being an incompatible element tends to remain in the melt, causing a sharp decrease in the Cu/Mo ratio. Mo becomes saturated during the later stages of magmatic crystallization and concentrates in more felsic granite or dykes.
The different ages of plutonic rocks that are sources for Cu and Mo are an important factor in their temporal separation. In addition, the oxygen fugacity (f O2 ) of the magma plays an important role in elemental separation [42][43][44]. Fluid inclusions in the Cu-mineralized quartz veins contain hematite, magnetite, and anhydrite, indicating that the magma was oxidized during the early stages (Figures 6b-e and 7b-d). The distribution capacity of Cu in fluids is strengthened with a high f O2 , whereas the fluid inclusions in the Mo-mineralized quartz veins contained an abundance of pyrite, CO 2 , and CH 4 (Figures 6g and 7e-g), which indicate that the magma was reducing and felsic in composition during deposition of Mo [45]. Mo-mineralized quartz veins contained an abundance of pyrite, CO2, and CH4 (Figures 6g  and 7e-g), which indicate that the magma was reducing and felsic in composition during deposition of Mo [45].

P-T Conditions and Phase Separation
A comparison of P-T conditions during the deposition of Cu and Mo indicate that the conditions for the L-type fluid inclusions are nearly the same during the Cu and Mo stages of mineralization, while the temperatures of the V-and S-type fluid inclusions in the Cu stage are ~15 °C higher than those in the Mo stage. The average pressure in the Cu stage was ~16 bar higher than those in the Mo stage. However, very small differences in P-T conditions are not sufficient to cause the obvious separation of Cu and Mo, indicating that the mineral-bearing fluid for both of the Cu and Mo stages had similar P-T conditions and salinity ( Figure 10). Furthermore, L-, V-, and S-type fluid inclusions commonly coexist in the Cu-and Mo-bearing quartz veins, and the similar P-T conditions of the Cu and Mo stages demonstrate that the ore-forming fluids were immiscible, causing the separation of high-salinity fluid and vapor phases ( Figure 10; [46]). Figure 10. NaCl-H2O P-T-X phase diagram [47]. The liquid + vapor coexistence surface is projected onto the P-X plane and contoured for temperature. The limit of liquid + halite coexistence at 400 °C is also shown.

Transport and Precipitation of Metals
Results of the SR-XRF mapping show that the concentration of Cu in fluid inclusions is relatively low during the Cu stage, and the Mo concentration is relatively high and present in the liquid phase. In contrast, the concentration of Mo in the Mo stage is relatively low and the Cu content is relatively high being present in the vapor phase (Figures 8 and  9). These characteristics indicate: (1) Cu was enriched in the vapor phase, suggesting a sudden release in pressure and Mo in the liquid phase. During precipitation of the Cu, residual Cu was present in the vapor phase with residual Mo in the liquid phase. The process continued until Mo was deposited, leaving residual Mo in the vapor phase and Cu in the liquid phase; (2) Cu was transported in solution in the vapor phase and Mo in the liquid phase. Therefore, the Cu in the vapor phase was transported across greater distances than the Mo in the liquid phase [48][49][50], which may explain the spatial separation of Cu and Mo. The changes in Fe are consistent with changes in Cu, thus providing clues Figure 10. NaCl-H 2 O P-T-X phase diagram [47]. The liquid + vapor coexistence surface is projected onto the P-X plane and contoured for temperature. The limit of liquid + halite coexistence at 400 • C is also shown.

Transport and Precipitation of Metals
Results of the SR-XRF mapping show that the concentration of Cu in fluid inclusions is relatively low during the Cu stage, and the Mo concentration is relatively high and present in the liquid phase. In contrast, the concentration of Mo in the Mo stage is relatively low and the Cu content is relatively high being present in the vapor phase (Figures 8 and 9). These characteristics indicate: (1) Cu was enriched in the vapor phase, suggesting a sudden release in pressure and Mo in the liquid phase. During precipitation of the Cu, residual Cu was present in the vapor phase with residual Mo in the liquid phase. The process continued until Mo was deposited, leaving residual Mo in the vapor phase and Cu in the liquid phase; (2) Cu was transported in solution in the vapor phase and Mo in the liquid phase. Therefore, the Cu in the vapor phase was transported across greater distances than the Mo in the liquid phase [48][49][50], which may explain the spatial separation of Cu and Mo. The changes in Fe are consistent with changes in Cu, thus providing clues to for the formation of a fine-grained chalcopyrite-pyrite mineral assemblage during the Cu stage (Figure 5a-d).
In addition, the Cl − content of the initial vapor phase was generally high in the porphyry Cu deposits [51][52][53][54]. This association between Cu and Cl − indicates that the carrying capacity of Cu in the vapor phase was far greater than that in the liquid phase.
Quartz dissolution and precipitation experiments performed by Steele-MacInnis et al. [55] showed that at pressures of <20 MPa at a hydrostatic paleo-depth of~2 km and at temperatures of 350 • C-500 • C, the solubility of quartz is low in vapor with a low salinity of <2 molal [56]. This observation explains the weak silicification observed in the Cu-I and Cu-II stages (Figure 5a-d).
The Au in the inclusions of the Cu and Mo stages is enriched in the vapor phase. Based on the hypothesis regarding Cu and Mo transport discussed above, it is considered that Au was transported in the vapor phase, but did not precipitate until after the precipitation of Cu or Mo, resulting in the enrichment of Au in the residual vapor phase. This inference is consistent with the average Au grade in the Jiama porphyry-hornfels ore of generally <0.1 ppm (Figure 4). Moreover, the remaining relatively Au-rich residual vapor phase was probably transported to distal zones of the porphyry system, where it formed epithermal Au deposits in faults [48,57,58].

Influence of Wall-Rock Permeability on Metal Transport
The mineralization and zonation patterns in skarn are controlled by factors such as the lateral escape of fluids, the distribution of limestone, and temperature gradients [13,[59][60][61]. Mineralization in the Jiama Porphyry and adjacent hornfels studied is the result of the vertical escape of fluid from the hornfels and structures within controlling the zonation of metals. The shallower part of the hornfels in contact with the Jiama Porphyry is massive, dominated by biotite alteration, and exhibits a moderate density of factors (Figure 5a-b). Deeper zones are characterized by coarse-grained altered felsic granitic rock that are also significantly fractured (Figure 5f-I; [62]). The permeability of the deeper zone is higher than that of the upper zone (Figure 4), and the difference in permeability would have been favorable for transport of Cu in the vapor phase, thus facilitating the movement of Cu from the lower to the upper member of the wall rock. Metasomatism associated with vapor resulted in the formation of the disseminated and thinly veined chalcopyrite-pyrite mineral assemblage. Mo that was transported in the liquid phase is readily precipitated in rocks with a relatively high permeability [63].

Conclusions
An unusual pattern of metal zonation is present in the Jiama deposit divided into an early upper-level containing Cu and a late lower-level containing Mo. The Cu mineralization proceeded in three stages and is dominated by disseminated and veinlet types containing an assemblage of chalcopyrite-pyrite accompanied by hydrothermal biotite alteration. The Mo mineralization also went through three stages, and was dominated by a vein stockwork of quartz-Mol accompanied by silicic alteration. From deep to shallow, the change of rock porosity/fracture rate and the migration of Cu and Mo in different phases are the main reasons for the spatial separation of elements, while the continuous evolution of ore-bearing magmatic rocks and the emplacement of porphyries in different stages may be the main reasons for the temporal separation of elements. It can be seen that the element zoning mechanism of porphyry deposits is complex, and this study is an important supplement to the traditional element zoning mechanism of porphyry deposits, and also has guiding significance for the mining planning of this type of deposits.