typomorphic characteristics of pyrite: criteria for 3d ... · typomorphic characteristics of...

Journal of Geochemical Exploration 164 (2016) 136–163

Contents lists available at ScienceDirect

Journal of Geochemical Exploration

j ourna l homepage: www.e lsev ie r .com/ locate /gexp lo

Typomorphic characteristics of pyrite: Criteria for 3D explorationtargeting in the xishan gold deposit, China

Gongwen Wang a,⁎, Yuan Feng a,e,⁎, Emmanuel John M. Carranza b, Ruixi Li a, Zonglie Li a, Zhankui Feng c,Xiandong Zhao d, Daojun Wang a, Liang Kong a, Wenjuan Jia a, Botao Wen a

a China University of Geosciences, Beijing, 100083, Chinab Department of Earth and Oceans, James Cook University, Townsville, Queensland, Australiac Luanchuan County Xin Chuan Mining Co., Ltd., Henan, 471500, Chinad Yantai Zhengyuan Geological Exploration Institute, Shandong, 264003, Chinae Northwest Geological Institute of Nonferrous Metals, 710000, China

⁎ Corresponding authors.E-mail addresses: [email protected] (G. Wang), 3

http://dx.doi.org/10.1016/j.gexplo.2016.01.0030375-6742/© 2016 Elsevier B.V. All rights reserved.

a b s t r a c t
a r t i c l e i n f o
Article history:Received 20 April 2015Revised 3 January 2016Accepted 5 January 2016Available online 13 January 2016

This paper describes 3Dmodeling of exploration criteria derived from typomorphic characteristics of pyrite in theXishan quartz vein-type gold deposit (China). The methodology consists of five steps: (1) modeling of orebodythickness and grade using ordinary kriging in longitudinal section; (2) analysis of major/trace element contentof Au-bearing pyrite from each ore paragenetic stage; (3) analysis of thermoelectric parameters of Au-bearingpyrites and estimation of ore-forming temperatures and comparison with homogenization temperatures fromfluid inclusion analysis; (4) 3Dmodeling of orebodies using surface geological mapping,mining tunnels in differ-ent levels, and a borehole dataset; and (5) 3Dmodeling of thermoelectricity coefficients and estimated temper-atures from Au-bearing pyrites for exploration targeting via discrete smooth interpolation and concentration-volume fractalmodeling. The results indicate that: (1) Au-bearing pyrites from four ore paragenetic stages recordgradually decreasing temperatures from the earliest to the latest stages, and the frequencies of occurrence of py-rite crystal combination forms and element components are closely correlated with P-type values of pyrite;(2) orebodies Nos. 108–1 and 107 are continuous at depth and potential exploration targets of their continua-tions extend more than 700 m downward from their present mining levels; whereas orebodies Nos. 55 and108–2 discontinue at depth.

© 2016 Elsevier B.V. All rights reserved.

Keywords:Typomorphic characteristics of pyriteDSIOrdinary kriging interpolationC-V fractal3D modelingLinglong gold deposit

1. Introduction

Mineral deposit models, which contain information on geological,geochemical, geophysical, genesis, grade and tonnage of specific deposittypes are useful in regional-scale exploration and resource assessmentsas well as in three-dimensional (3D) district-scale geological modelingfor exploration targeting (Cox and Singer, 1986; Dagbert and Harfi,2002; Fallara et al., 2006; De Kemp et al., 2011; Wang et al., 2011,2013, 2015; Mejía-Herrera et al., 2014; Vollgger et al., 2015). Deposit-scale exploration targeting has also benefitted from 3Dmodeling of sur-face and subsurface geochemical datasets (Jackson, 2010). The citedcase studies of 3D geological modeling are mostly concerned withtargeting for porphyry–skarn or VMS deposits across a range of scalesfrom regional- to deposit-scales and commonly have sufficientmultiplegeoscience datasets. However, 3D geological modeling for explorationtargeting of vein-type mineral deposits is even more challenging be-cause uncertainty in 3D modeling of complex geological systems, such

[email protected] (Y. Feng).

as mineral deposits, is difficult to eliminate (Fallara et al., 2006;Lindsay et al., 2012, 2013a, 2013b).

In this paper, we address the question “How can we construct a 3Dexploration targeting model using typomorphic features of pyrites incomplex and irregular quartz-vein hosted gold in the Xishan deposit,China”? The targeting criteria that we use are derived from hydrother-mal pyrite, which contains significant amounts of minor and trace ele-ments, and provides a record of hydrothermal fluid evolution in anore system (e.g., Deditius et al., 2011; Reich et al., 2013). For example,pyrite in gold-bearing quartz veins provides information on the poten-tial source and timing of gold and related fluid processes responsiblefor mineralization (Large et al., 2009; Thomas et al., 2011; Cook et al.,2013). Typomorphic features of pyrite in different ore-forming stagesor different ore bodies in a mineral deposit have been analyzed inmetallogenic studies to support mineral prospecting (e.g., Chen et al.,1989; Meng et al., 2001; Shen et al., 2013; Xue et al., 2014).Typomorphic characteristics of pyrites, such as crystal type, chemicalcomposition, and thermoelectricity, have been shown to be closely re-lated to the physicochemical environment of their formation (e.g., Liet al., 1994, 2012; Shen et al., 2013).

http://crossmark.crossref.org/dialog/?doi=10.1016/j.gexplo.2016.01.003&domain=pdf

mailto:[email protected]

mailto:[email protected]

http://dx.doi.org/10.1016/j.gexplo.2016.01.003

www.elsevier.com/locate/gexplo

137G. Wang et al. / Journal of Geochemical Exploration 164 (2016) 136–163

In this paper, we describe 3D deposit-scale modeling for explorationtargeting based on typomorphic thermoelectric property of pyrite (Liet al., 2012), 3D models of orebody based on Au grade and vein geome-try, and knowledge of metallogenesis of the Xishan gold deposit.

Fig. 1. Gold deposits and geological map of Jiaodong Peninsula, China (A: modified

2. Geological setting

The Jiaodong Peninsula along the southeastern margin of the NorthChina Craton (NCC) is the most important gold producing province in

from WenWen et al., 2015; B: modified from Goldfarb and Santosh, 2014).

Fig. 2. Faults and quartz-vein gold deposits in the Linglong gold district.

138 G. Wang et al. / Journal of Geochemical Exploration 164 (2016) 136–163

China, with an explored gold reserve of 3000 t (Guo et al., 2013). It isbounded by the NE- to NNE-trending Tan-Lu fault zone to the west andby the Su-Lu ultrahigh pressure metamorphic belt to the south (Fig. 1).

Fig. 3. Gold orebodies and engineering constraints cross-section in the Xisha

The gold deposits in the Jiaodong Peninsula are generally classified asthe Linglong-, Jiaojia-, and Pengjiakuang-type (Tan et al., 2012). Despitethe three types of deposits having different modes of occurrence and

n gold deposit. The arrow indicates sample (grade/thickness) location.

Fig. 4. The No. 108–1 orebody at the top of mining channel (−185 m level) includes potassic, sericitic, pyritic, and silicic alterations (four paragenetic stages: I, II, III, and IV).


ore textures, all orebodies are fault-controlled (Guo et al., 2013). In addi-tion, the gold deposits in thepeninsula have a consistent spatial–temporalassociation with the Late Jurassic-Early Cretaceous magmatism (130 to110 Ma), which is considered to be related to NCC lithospheric thinningand an NCC destruction event (Tam et al., 2011).

The Linglong gold district is situated at the northern tip of the Zhao-Ping fault zone to the east of the Zhao-Ye gold belt (Fig. 1). This gold dis-trict is bordered to the southeast by the Potouqing fault and the north-ern segments of the Zhao-Ping fault (the Linglong Fault zone is part ofthe Zhao-Ping fault) (Fig. 2). The Linglong Fault is a NE-trending shearzone, with widths varying from 50 to 150 m, strike of N25°–30°E anddips of 65° to 85° to the NW. The lithology along the Linglong Fault

Fig. 5. A. No. 55 orebody vein thickness OK interpolation map. Semi-variogram of vein thickmineralization in the Xishan deposit. B. No. 55 orebody Au grade OK interpolation map. SemN60°E main trend of mineralization in the Xishan deposit.

zone ismonzonite granite with different types of alteration (sericite, sil-ica, silica + sericite, and K-feldspar) and mineralization. The K-feldsparalteration ranges in age from 110.3 ± 1.0 Ma to 69.47 ± 0.9 Ma (Shenet al., 2013), and the geochronological data show that several eventshave occurred along the Linglong Fault during and after the period ofmetallogenesis.

3. Deposit features

The Xishan gold deposit is located in the western part of the Jiaodonggold province, approximately 40 km north of Zhaoyuan City (Fig. 1). Thegold deposit is a Linglong-type (quartz vein-type), which is characterized

ness in orebody No. 55 along a longitudinal section following the N60°E main trend ofi-variogram of Au grade in orebody No. 55 along a longitudinal section following the

Fig. 6. A. No. 107 orebody vein thickness OK interpolation map. Semi-variogram of vein thickness in orebody No. 107 along a longitudinal section following the N60°E main trend ofmineralization in the Xishan deposit. Red points are pair points of semi-variogram values using other measured points which are the overall spatial autocorrelation of the measuredpoints in the database (107 orebody in the longitudinal section), which are the empirical semi-variogram values, and blue dots are averaged points according to the lag and width anddirection for the model of semi-variogram. B. No. 107 orebody Au grade OK interpolation map. Semi-variogram of Au grade in orebody No. 55 along a longitudinal section followingthe N60°E main trend of mineralization in the Xishan deposit.


by massive auriferous quartz veins occurring in second- or third-orderfaults of the NNE-trending Linglong Fault zone and the NEE- to NNE-trending Potouqing Fault zone (Fig. 2). These two major fault zones andtheir branches control hundreds of auriferous quartz veins, with a generalNNE-NE trend but with local abrupt changes in attitudes of associatedfractures (Shen et al., 2013; Yang et al., 2013;Wen et al., 2015). Goldmin-eralization occurs in single ormultiple relatively continuous quartz veins,which can be as long as 5 km. The veins range froma few centimeters to afewmeters in width; they extend downwards for at least 700 m (Fig. 3).More than 50 auriferous quartz veins occur within the Linglong district(Fig. 2). Mineralization appears associated with pyrite-sericite-silica al-tered rocks or fine pyrite veins. The types of alteration associated withthe Au-bearing veins are mainly K-feldspathization, hematitization,sericitization, silicification, pyritization, and calcitization, with minorchloritization (Fig. 4).

Themajor Au-bearing quartz veins are Nos. 55, 107, 108–1 and 108–2 (Figs. 3, 5, 6, 7, 8) and the smaller Au-bearing quartz veins of No. 71vein system (Fig. 3). At the 198 m level, vein No. 108 branches intotwo veins (Nos. 108–1 and 108–2). Vein No. 107 is one of the branchesof No. 71 vein at the surface; it has a continuous length of 1400 m, itstrikes N40°–65°E and dips to the NW at angles of 65°–75°. Vein No.10 is concealed and itmay be one branch of vein No. 71, with an averageof 0.782 m, and its Au grades are 1.09–13.55 g/t, with an average of4.19 g/t. Vein No. 108 is exposed at the surface along a length of1600 m, strike of N60°–65°E, and it dips to the NW at angles of 72°–78° with an average of 75°; it has a thickness of 0.30–3.0 m, with an

average of 0.782 m, and its Au grades are 1.75–16.04 g/t, with an aver-age of 4.93 g/t. and is located in the central part of the study area. VeinNo. 55 has a continuous length of 1200 m, it strikes N15°–25°E anddips steeply to the NW at angles of 62°–73° with an average of 68°. Itis laminated, and it is strictly controlled by a compressive-shear frac-tures. It has a thickness of 0.30–1.90 m, with an average of 0.693 m,and its Au grades are 0.61–14.18 g/t, with an average of 5.93 g/t.

In most of the veins of the Xishan gold deposit, the major ore min-erals include native gold, electrum and pyrite, whereas the minor oreminerals are chalcopyrite, galena, and sphalerite (Fig. 9). Native goldgrains occur mainly as crystal gap gold in fissures of pyrite and quartzor as inclusions in pyrite crystals and gangue minerals, with secondaryoccurrences as inter-crystal gold (Fig. 9). The main gangue mineralsconsist of quartz, plagioclase and sericite, and the minor gangue min-erals are chlorite and calcite.

4. Methodology

The methodologies used in this study include geochemical measure-ments using an election probe micro-analyzer (EPMA) and 3D spatialmodeling techniques (i.e., ordinary kriging (OK) of orebody Au gradeand vein thickness, discrete smooth interpolation (DSI) modeling of themetallogenic environment, and concentration-volume (C-V) modeling).The OK was implemented in ArcGIS, DSI in GoCAD software (Mallet,1989) and C-V modeling in GeoCube software (Wang et al., 2015).

Fig. 7. A. No. 108–2 orebody vein thickness OK interpolation map. Semi-variogram of vein thickness in orebody No. 108–2 along a longitudinal section following the N60°Emain trend ofmineralization in the Xishan deposit. B. No. 108–2 orebody Au grade OK interpolation map. Semi-variogram of Au grade in orebody No. 108–2 along a longitudinal section following theN60°E main trend of mineralization in the Xishan deposit.


4.1. Spatial analysis of orebodies

Variation in vein thickness is strongest in orebody No. 55, indicatingpinch-and-swell character of veins (Table 1);whereas variations in veinthickness of the other three orebodies are lower and more-or-less sim-ilar, suggesting more-or-less uniform vein thickness. Orebody No. 107has the strongest variation in Au grade, indicating it has themost erraticdistribution of gold (Table 1). Variation in Au grade is also quite strongin orebody No. 55. Variations in Au grade are lowest and more-or-lesssimilar in orebodies Nos. 108–1 and 108–2, indicating that Au is more-or-less evenly distributed in these two orebodies. These spatial featuresof the orebodies based on statistics and variogram analysis are key

constrains for exploration targeting by 3Dmodeling using interpolationmethod in GoCAD software (see 4.3.1 (1) above). Variations in veinthickness and Au grade in the orebodies (linked with veins Nos. 108–1, 108–2, 107 and 55) weremodeled using semi-variograms and spatialdistributions of these variables based on the semi-variograms weremodeled using OK in ArcGIS (version 10.1) software. Results are sum-marized in Table 2 and shown on Figs. 5, 6, 7, and 8.

The semi-variograms of Au grade and vein thickness in orebody No.55 suggest that the spatial variations of these variables can depicted byan exponentialmodel (Table 2), and themain trends of theirmajor con-tinuity along the N60°E longitudinal section (359° and 9°, respectively)suggest that the spatial variations of these variables persistmainly along

Fig. 8. A. No. 108–1 orebody vein thickness OK interpolation map. Semi-variogram of vein thickness in orebody No. 108–1 along a longitudinal section following the N60°Emain trend ofmineralization in the Xishan deposit. B. No. 108–1 orebody Au grade OK interpolation map. Semi-variogram of Au grade in orebody No. 108–1 along a longitudinal section following theN60°E main trend of mineralization in the Xishan deposit.


the vertical direction. However, the strong nugget effects despite thelong ranges of the spatial variations of these variables (Table 2, Fig. 5) in-dicate the poor continuity of orebodyNo. 55 along the vertical direction.Accordingly, the results of OK of vein thickness and Au grade (Fig. 5Aand B, respectively), based on the semi-variograms of these variablesalong a longitudinal section following the N60°E major trend of miner-alization in the Xishan deposit, suggest that the No. 55 quartz-veinorebody is likely discontinuous and possibly a low priority target veinsystem at depth (from 0 m to −300 m level).

The semi-variograms of Au grade and vein thickness in orebodyNo. 107 show that the spatial variations of these variables can be

depicted by a spherical model (Table 2), and the major trend oftheir continuity along the N60°E longitudinal section (83° and100°, respectively) suggest that the spatial variations of these vari-ables persist mainly along the horizontal direction. However, thenil to small nugget effects regardless of the short ranges of the spatialvariations of these variables (Table 2, Fig. 6) indicate the good verti-cal continuity of orebody No. 107. Accordingly, the results of OK ofvein thickness and Au grade (Fig. 6A and B, respectively), based onsemi-variograms of these variables along a longitudinal section fol-lowing the N60°E major trend of mineralization in the Xishan depos-it, suggest that the No. 107 quartz-vein orebody is likely continuous

Fig. 9. A: Crystal types of Au-barren and Au-bearing pyrites of Xishan gold deposit. 9B: Crystal types of Au-bearing pyrites of Xishan gold deposit.


and possibly a high priority target vein system at depth (from 100 mto −300 m level).

The semi-variograms of Au grade and vein thickness of No. 108–2orebody show that the spatial variations of these variables can be

depicted by spherical and exponential models, respectively (Table 2),and the major trends of their continuity along the N60°E longitudinalsection (0° and 80°, respectively) suggest that Au grade of this orebodypersists mainly along the vertical direction while its vein thickness

Table 1Statistics of Au grade and vein thickness in different orebodies in the Xishan gold deposit.

Number of samples Min Max Mean Std. Dev. Coefficient of variation

No. 55 – vein thickness (m) 236 0.10 11.4 1.12 1.16 1.04No. 55 – Au grade (g/t) 236 0.10 84.5 4.06 8.21 2.02No. 107 – vein thickness (m) 417 0.10 2.50 0.75 0.43 0.57No. 107 – Au grade (g/t) 417 0.01 4.23 0.37 4.39 11.86No. 108–2 – vein thickness (m) 198 0.10 3.00 1.06 0.49 0.46No. 108–2 – Au grade (g/t) 198 0.16 17.30 4.04 3.43 0.85No. 108–1 – vein thickness (m) 177 0.20 2.00 1.03 0.49 0.48No. 108–1 – Au grade (g/t) 177 0.22 22.57 4.57 3.62 0.79


persists mainly along the horizontal direction. The persistence ofvein thickness along the horizontal direction and the strong nuggeteffect of Au grade despite its long range (Table 2, Fig. 7) indicatethe poor continuity of orebody No. 108–2 along the vertical direc-tion. Accordingly, the results of OK of vein thickness and Au grade(Fig. 7A and B, respectively), based on semi-variograms of these var-iables along a longitudinal section following the N60°E major trendof mineralization in the Xishan deposit, suggest that the No. 108–2quartz-vein orebody is likely discontinuous and possibly a low prior-ity target vein system at depth.

The semi-variogram of Au grade and vein thickness of orebodyNo. 108–1 show that the spatial variations of these variables can bedepicted by an exponential model (Table 2), and the major trendsof their continuity along the N60°E longitudinal section (150° and125°, respectively) suggest that the spatial distributions of these var-iables persist along both vertical and horizontal directions. The poornugget effects and long ranges of the spatial variations of these var-iables (Table 2, Fig. 8) indicate the good continuity orebody No.108–1 along both the horizontal and vertical directions. Accordingly,the results of OK of vein thickness and Au grade (Fig. 8A and B, re-spectively), based on semi-variograms of these variables along a lon-gitudinal section following the N60°E main trend of mineralizationin the Xishan deposit, suggest that the No. 108–1 quartz-veinorebody is likely continuous and possibly a high priority target veinsystem at depth.

Althoughwe analyzed the above four orebody features spatially, it ishard to visualize the spatial correlation between geological features andorebodies. Thus, as the limited grade and thickness datasets are insuffi-cient for inferring the trend of ore distribution at depth, we constructed3D orebody trend models using thermoelectric parameters of pyrite tounderstand ore-forming environment/condition (e.g., environment,pressure) and aid in the inference of potential targets at depth (see4.3 section below).

4.2. Genetic mineralogy analysis

4.2.1. Pyrite in ore paragenesisBased on field observations (e.g., Fig. 4) and microscopic obser-

vations (e.g., Fig. 9), ore paragenesis is divided into four stages.

Table 2Semi-variogram analysis of Au grade and vein thickness in different orebodies along SW (240°

Orebody - variable Semi-variogram model Data transformation

No. 55 – vein thickness (m) Exponential NoneNo. 55 – Au grade (g/t) Exponential LogNo. 107 – vein thickness (m) Spherical NoneNo. 107 – Au grade (g/t) Spherical LogNo. 108–2 – vein thickness (m) Spherical NoneNo. 108–2 – Au grade (g/t) Exponential LogNo. 108–1 – vein thickness (m) Exponential NoneNo. 108–1 – Au grade (g/t) Exponential Log

Stage I – a pyrite-quartz stage – comprises quartz, sericite, and py-rite. In this stage, pyrite occurs as blocks or disseminated precipi-tates, which are euhedral granular and broken patchy, with agranularity of 1–2 mm and makes up N10% of ore specimen. StageII – a quartz-pyrite stage – comprises mainly quartz and smallamounts of pyrite (b10%) occurring as veins or stockworks,where pyrite is euhedral-subhedral granular, with a granularityof 1–5 mm. Stage III – a pyrite and polymetallic sulfide stage – isthe main Au-ore stage and comprises essentially the same mineralsas stage II, but are mostly pyrite and secondarily quartz in additionto small amounts of pyrrhotite. In stage III, pyrite also occurs asveins or stockworks, where pyrite is euhedral to subhedral andbroken patchy, with a granularity of 0.1–2 mm and makes upN80% of ore specimen. Stage IV – a quartz carbonate stage – com-prises iron calcite, ankerite, siderite, quartz, and pyrite. In thisstage pyrite occurs in veins, where it exhibits euhedral or filledstructure, with a granularity of 1–5 mm and makes up 5%–10% ofthe ore specimen.

Pyrite is the primary Au-bearing mineral in the Xishan deposit.It accounts for more than 90% of the total amount of metal, and~85% of gold is distributed in the lattices, micro-fissure andinter-granular fractures of pyrite (Shen et al., 2013). Samples ofore dominated by stage III pyrites are the richest in Au content,followed by samples of ore dominated by stage II-III pyrites(Table 3, Figs. 9 and 10). These data support the hypothesis thatAu enrichment in the Xishan deposit started during stage II andpeaked during stage III.

4.2.2. Types of pyrite crystalsSystematic statistical analysis of the shape of pyrite crystals in

the Xishan deposit provides evidence for a diverse range of crystalhabits of pyrites in this deposit. The chief crystal habits of pyrite are{100}, {111}, {210} (Fig. 9), with combined habits (e.g., mostly{100} + {210}, and the secondary crystal habits are {100} +{210} + {111}, {210} + {111}). The temporal evolution of pyritecrystal habits follows the ore paragenesis: the number of cubic{100} pyrite is highest (N320°) in stage I, it decreases in the mainAu-ore stage (III) and falls to its lowest (b170°) in stage IV. Thenumbers of pentagonal dodecahedral {210} pyrite and combined

) – NE (60°) longitudinal sections in the Xishan gold deposit.

Major direction of continuity(°) Nugget effect Sill Range (m)

9 0.70 0.80 55.8359 0.37 1.12 96.6100 0.06 0.12 39.783 0.00 0.76 7.580 0.02 0.22 19.20 0.42 0.63 65.1125 0.12 0.16 244.1150 0.24 0.72 61.6

Table 3Whole-rock assay data of ore samples from the Xishan deposit.

Sample Number Dominant stage of pyrite in sample Au Co Ni Cu Zn Ag As Se Sb Te

044–1 I 0.21 1.81 3.85 10.57 17.93 0.07 9.5 0.03 0.22 0.04D012–1 I 1.13 2.07 2.79 157.70 57.42 3.09 40.0 0.02 1.07 0.01D012–2 I 0.03 0.92 1.87 72.73 3664.00 1.59 10.6 0.02 0.59 0.0175–1 I 0.08 3.25 4.15 14.41 66.52 0.06 6.2 0.05 0.37 0.02019–3 I 0.02 2.52 3.78 6.77 47.99 0.02 2.4 0.02 0.15 0.0179–4 I 0.01 1.90 2.65 7.41 30.50 0.03 3.4 0.03 0.31 0.0118 I 0.02 7.78 32.87 8.02 117.80 0.04 1.8 0.02 0.17 0.019 I 0.02 2.05 4.74 9.46 75.16 0.08 17.0 0.03 0.84 0.0113 I 0.03 23.11 184.80 8.62 231.40 0.05 1.6 0.02 0.09 0.014 I 1.91 8.08 2.80 84.10 272.80 0.52 39.0 0.06 0.46 0.13017–2 I 0.02 2.37 5.34 36.07 39.36 0.15 12.4 0.07 10.80 0.02052–5 I 0.02 2.76 3.58 16.55 128.00 0.28 7.2 0.03 0.22 0.01017–1 I 0.20 4.33 3.31 116.10 35.02 0.28 22.0 0.03 20.90 0.05007–1 I 1.11 5.19 3.26 19.01 27.35 1.07 12.0 0.03 0.27 0.10020–1 I 0.17 2.40 2.95 8.34 11.64 0.02 4.0 0.02 0.21 0.01020–2 I 11.28 55.55 11.26 191.70 65.28 4.96 780.0 0.05 2.52 1.70007–3 I 0.15 2.57 2.93 127.90 63.74 0.21 5.6 0.04 0.62 0.04D010 I-II 0.05 1.64 3.59 10.89 29.11 0.21 8.2 0.06 0.74 0.06055–3 I-II 2.18 5.72 3.10 182.80 13,320.00 9.91 42.0 0.05 1.86 0.05055–1 II 0.39 2.70 6.19 446.25 10,740.00 8.99 41.0 0.03 0.74 0.01052–1 II 0.22 9.98 3.79 6.53 38.29 0.10 38.0 0.07 0.14 0.10D006 II 0.03 0.98 2.59 8.55 46.30 0.14 2.4 0.04 0.38 0.02022–3 II 0.02 1.47 2.86 9.11 36.78 0.03 3.1 0.02 0.32 0.0173–1 II 2.16 11.24 7.31 34.94 23.28 1.31 108.0 0.14 7.90 0.4875–2 II 0.84 7.79 2.70 10.46 63.04 0.15 23.0 0.07 0.32 0.1178 II 1.74 6.35 8.35 288.60 22.58 1.97 230.0 0.03 0.92 0.02056–2 II 10.23 6.40 2.45 28.54 215.90 2.37 44.0 0.02 0.31 0.02007–4 II 0.01 1.83 2.97 9.77 36.67 0.05 4.8 0.03 0.21 0.01008–1 II 0.07 12.73 5.41 9.94 107.90 0.24 44.0 0.09 0.39 0.18518–3 II 0.18 1.27 2.98 20.65 44.47 3.82 108.0 0.05 0.50 0.01008–2 II-III 2.70 18.62 16.36 1261.05 1369.00 11.70 1200.0 0.07 2.45 0.19019–2 II-III 0.03 3.89 3.45 19.66 30.08 0.08 7.2 0.03 0.35 0.06007–2 II-III 7.41 6.82 3.46 7910.70 274.00 35.19 86.0 0.02 66.00 0.20022–2 II-III 7.92 4.13 4.25 44.05 265.20 0.25 40.0 0.04 0.52 0.0879–2 II-III 2.28 19.82 11.90 70.63 955.40 5.60 810.0 0.05 11.20 0.2014 II-III 365.94 38.40 18.23 7745.85 73.80 98.12 310.0 0.02 0.54 2.1979–3 II-III 0.05 3.06 4.87 12.18 1816.00 0.16 114.0 0.04 0.99 0.02019–1 II-III 0.58 31.88 5.55 45.23 16.39 0.38 168.0 0.22 0.51 0.67052–2 II-III 1.11 9.68 11.59 11.33 137.60 0.51 20.5 0.22 0.62 0.19008–3 II-III 0.52 28.45 119.57 28.17 76.78 0.34 3.3 0.14 1.01 0.0479–1 II-III 0.11 10.49 5.67 22.59 131.40 0.76 136.0 0.08 1.84 0.1732 II-III 0.32 1.95 4.28 13.53 65.23 0.15 4.4 0.03 0.53 0.01056–1 II-III 0.31 1.61 2.85 29.07 146.40 0.37 22.0 0.02 0.24 0.013 III 0.01 2.16 2.93 12.91 35.11 0.03 3.2 0.02 0.29 0.01022–4 III 70.87 14.73 24.13 5257.35 42,510.00 205.50 360.0 0.41 232.00 4.46D014–1 III 0.92 1.17 2.90 28.44 170.20 0.75 22.0 0.04 0.40 0.01502 III 0.02 2.05 2.46 19.18 70.48 0.35 21.0 0.02 0.27 0.01022–1 III 70.18 22.73 21.51 1916.25 117,095.00 30.96 440.0 0.03 2.06 0.74D015-1 III 4.33 3.48 9.85 495.10 167.20 2.08 9.2 0.04 0.49 0.14052–3 III 0.41 8.69 3.24 11.59 41.68 0.18 88.0 0.03 0.16 0.05052–4 III 18.17 8.82 4.13 911.60 18.91 8.05 41.0 0.02 0.26 0.08044–2 IV 0.01 2.15 3.65 7.04 20.56 0.03 1.9 0.03 0.16 0.01D014-2 IV 9.83 7.58 2.79 165.00 3384.00 12.86 63.0 0.07 4.90 0.19D015-2 IV 0.17 2.17 3.06 12.52 29.54 0.21 4.6 0.10 0.43 0.21518–2 IV 0.02 1.86 3.43 8.51 22.47 0.42 10.6 0.04 0.37 0.01518–4 IV 0.01 2.23 2.81 11.50 62.61 0.06 4.9 0.02 0.13 0.01518–5 IV 0.03 3.07 4.63 14.94 19.99 0.16 29.0 0.05 0.40 0.0129 IV 0.01 1.39 3.30 7.46 21.86 0.06 2.3 0.05 0.25 0.0130 IV 0.00 2.19 2.61 7.34 32.71 0.05 2.3 0.05 0.24 0.0161 IV 0.06 1.82 2.89 10.78 18.30 1.56 12.0 0.03 0.63 0.0163 IV 0.01 41.45 199.10 17.20 84.67 0.08 1.7 0.03 0.11 0.0165 IV 0.01 1.32 2.81 6.49 100.50 0.08 2.7 0.02 0.26 0.00034–3 IV 0.01 3.24 2.89 6.38 37.49 0.02 2.1 0.04 0.17 0.00035–2 IV 0.01 1.91 2.79 6.64 24.81 0.03 2.0 0.02 0.16 0.01505 IV 0.14 1.54 2.99 826.60 162.20 0.68 8.0 0.02 0.21 0.01508 IV 0.01 1.55 2.32 6.68 18.66 0.05 3.4 0.04 0.18 0.0075–3 IV 0.01 1.74 3.24 9.27 77.27 0.04 2.7 0.03 0.44 0.01021–2 IV 0.07 2.75 3.13 16.56 45.96 0.09 4.8 0.03 0.44 0.01021–1 IV 0.26 3.66 4.37 17.98 21.49 0.09 12.4 0.03 0.45 0.06024–2 IV 0.01 4.43 3.16 9.67 19.70 0.12 38.0 0.28 0.47 0.15


habits {100} + {210} tend to vary inversely with the number ofcubic {100} pyrite, but are noticeably largest in the main Au-orestage (III). Pentagonal dodecahedral {210} pyrite reflects sulfur-

rich mineralizing fluid, and so it is usually regarded as evidencefor identifying the main Au-ore stage (Shao, 1988; Shen et al.,2013).

Fig. 10. Gold concentrations and dominant paragenetic stage pyrites in ore samples fromthe Xishan deposit.


4.2.3. Chemical composition of pyriteThe chemical composition of pyrite is closely related to physico-

chemical conditions that influenced its formation and transforma-tion, and which have important effects on various minerals in anore deposit. Therefore, studying the chemical composition of pyrite,as one of its most important typomorphic characteristics, is impor-tant in subsurface exploration for gold in the Jiaodong Peninsula,China (Chen et al., 1989; Shen et al., 2013). In this study, 62 pyritesamples were collected from primary ore and alteration zones onthe surface as well as subsurface locations (elevations from 280 mto−225m) formajor and trace element analyses by EPMA. The anal-yses (using a JXA-8100 electron microprobe) were carried out at theChina University of Geosciences (State Key Laboratory of GeologicalProcesses and Mineral Resources Laboratory) in Beijing. Analyticconditions were a 20 kV accelerating voltage and a 10 nA beam cur-rent with the diameter of 1 μm and calibration by American SPI ref-erence materials. The chemical composition of the pyrite sampleswas quantitatively analyzed, and the results are summarized inTable 4 and Fig. 11.

The S/Fe molar ratio of pyrite in a gold deposit is complicated by thepresence of a variety of trace elements such as Co and Ni, whichisomorphously replace Fe, and As, Se, and Te, which in turn canisomorphously replace S. These isomorphous substitutions of trace ele-ments can promote the distribution of gold in the crystal lattice of pyriteand enhance the Au-bearing capacity of pyrite. A molar S/Fe ratio of b2is attributed to sulfur loss, N2 to iron loss (Doyle and Mirza, 1996;Oberthur et al., 1997). From the EPMA analyses (Table 4) of S and Fecontents (inmolecular formula,w(S) andw(Fe), respectively) of pyritesfrom the Xishan deposit, the S/Fe molar ratios [n(S)/n(Fe)] vary from1.95 to 2.05. Pyrites from the different paragenetic stages of the Xishandeposit have generally statistically similar S/Fe molar ratios (Fig. 11A).However, compared to the theoretical chemical composition of pyrite[i.e., w(S) = 53.45%, w(Fe) = 46.55%, and n(S)/n(Fe) = 2], most sam-ples of stage I and IV pyrites in the Xishan deposit exhibit sulfur losswhereas some samples of stage II and III pyrites have slightly higher S/Fe molar ratios and exhibit iron loss (Table 4; Fig. 11A).

Cobalt to nickel (Co/Ni) ratios of N1 are typical of intrusive-hydrothermal pyrites (Bralia et al., 1979; Cook, 1996). The Colo/Ni ratiosof all pyrite samples from the Xishan deposit vary strongly from 0.05 to10.5,with average and standard deviation of 1.56 and 2.02, respectively.The data indicate that 18 samples have Co/Ni of N1 and 30 samples have

Co/Ni of ≤1. In particular, most samples of stage I pyrites have Co/Ni ra-tios of N1 (Fig. 11B), implying that they were associated with anintrusive-hydrothermal geological setting, whereas most samples oflate stage (II – IV) pyrites have Co/Ni ratios of ≤1, implying that theylikely precipitated as a result of mixing of early-stage intrusive-relatedhydrothermalfluidswith late-stagemeteoricfluids. Thus, the above fea-tures of Co/Ni values shows the complicated geological setting in studyareawhich are associatedwithmultiplemagma-hydrothermal functionto ore-forming fluids.

The As element is usually held in a volatile, hydrothermal solution,and it often occurs in isomorphous form with S in the pyrite lattice.Therefore, the w(Fe)/w(S + As) ratio has been used to characterizethe correlation between the pyrite features and the formation depthof a gold orebody (Shen et al., 2013). Thew(Fe)/w(S+ As) ratios of py-rites from the Xishan deposit are uniform (range of 0.85–0.89, with av-erage and standard deviation of 0.88 and 0.01, respectively) reflecting apossible shallow environment for the gold mineralization (cf. Shenet al., 2013).

The pyrite EPMA geochemical data suggest that (a) that pyrites inthe main Au-ore stage III tend to have slightly higher S/Fe molar ratiosthan pyrites in the other paragenetic stages of the Xishan deposit(Fig. 11A), (b) samples of stage III pyrites have the lowest Co/Ni ratiosof ≤1 (Fig. 11B), (c) samples of stage III pyrites have the lowest w(Fe)/w(S + As) ratios (Fig. 11C), (d) Au is richest in stage III pyrites(Table 4, Fig. 11D), which is consistent with the Au concentration datafrom ore samples (Table 3, Fig. 10) and (e) pyrites from the differentparagenetic stages of theXishan deposit have generally statistically sim-ilarw(Co+Ni) andw(Cu+Zn) (Fig. 11E and F). However, Au in pyritesdoes not show significant correlation with other trace elements(Fig. 12A, Table 4), whereas Au in ore samples shows significant corre-lationswith Cu, Te, Ag, As and Sb (Fig. 12B, Table 3). Therefore, the com-plexmajor/trace element geochemistry of pyrite (Fig. 11) can aid in theinterpretation of ore genesis, but it does not provide useful criteria for3D exploration targeting in the study area.

4.2.4. Thermoelectricity of pyrites

4.2.4.1. (1) Thermoelectric theory. Pyrite is a common strongly authigenictransition mineral in gold deposits. Several studies have suggested thatgold is often associated with sulfides, as the crystallization of pyritefrom Au-bearing hydrothermal fluids rich in Fe, Cu and S can create fa-vorable physical and chemical conditions for the co-crystallization ofgold in many quartz vein-type deposits in orogenic settings (Groveset al., 1998; Qiu et al., 2002; Large et al., 2009; Deditius et al., 2011;Reich et al., 2013).

Pyrite is a semiconductor mineral. Under different temperature con-ditions, the non-equilibrium carriers in a semiconductor diffuse fromhigh-temperature zones to low-temperature zones, resulting in thegeneration of a thermal electromotive force (E). This thermoelectricphenomenon, known as the Seebeck effect, is strongly dependent onthe chemical composition of the semiconductor. The thermoelectric co-efficient (α) of a semiconductor like pyrite can be defined as (Shao et al.,1990):

a ¼ �EtH−tC

; �μv=∘C�� ð1Þ

where tH denotes the hot-end temperature and tC represents the cold-end temperature. The positive and negative sign of E is correlated withthe positive and negative properties of the carriers with a certain con-duction type (Shao et al., 1990). The conduction type can be either elec-tronic type (N-type) or hole type (P-type). The values of α and theconduction type of pyrite are influenced by isomorphous impurities inthe composition of pyrite, defects in crystal structure, density, and ex-ternal excitation conditions (e.g., temperature and pressure gradients).Because these factors are influenced by conditions at the depth of ore

Table 4Major/trace element content (%) determined by EPMA for pyrites from the Xishan deposit. Blank cells mean trace element was not detected.*

Samplenumber

Parageneticstage of pyrite

S Fe Co Ni Cu Zn Ag As Se Sb Te Au* Total n(S)/n(Fe) w(Fe)/w(S + As) w(As + Se + Sb + Te) w(Co)/w(Ni) w(Co + Ni) w(Cu + Zn)

044–1 I 52.51 46.01 0.27 0.06 0.16 0.32 0.08 0.04 0.06 0.01 99.52 1.99 0.87 0.18 4.50 0.33 0.48D012–1 I 52.70 46.46 0.12 0.06 99.34 1.98 0.88 0.12 0.00D012–2 I 52.69 46.69 0.39 0.17 0.14 0.21 0.02 0.02 100.33 1.97 0.89 0.02 2.29 0.56 0.1475–1-1 I 52.98 46.46 0.07 0.11 0.02 0.04 0.06 0.10 0.28 100.12 1.99 0.88 0.48 0.64 0.18 0.0275–1-2 I 52.83 46.46 0.24 0.11 0.32 0.03 0.02 100.01 1.98 0.88 0.03 0.24 0.43007–3-1 I 52.09 46.47 0.22 0.25 0.04 0.10 0.26 0.39 0.09 99.91 1.95 0.89 0.48 0.88 0.47 0.14007–3-2 I 53.01 46.41 0.36 0.10 0.03 0.08 0.08 0.11 0.03 100.21 1.99 0.88 0.27 3.60 0.46 0.00007–1-1 I 52.79 46.66 0.21 0.02 0.06 0.15 0.06 0.01 99.96 1.97 0.88 0.21 10.50 0.23 0.00007–1-2 I 52.56 46.30 0.41 0.09 0.01 0.11 0.02 0.02 99.52 1.98 0.88 0.14 0.41 0.09004–1 I 52.83 46.12 0.21 0.21 0.10 0.06 0.09 0.10 0.01 0.05 99.78 2.00 0.87 0.11 1.00 0.42 0.16004–2 I 52.79 46.11 0.41 0.13 0.35 0.09 0.02 0.21 100.11 1.99 0.87 0.23 3.15 0.54 0.35020–2-1 I 52.89 47.24 0.01 0.07 0.12 0.18 0.03 0.01 100.55 1.95 0.89 0.21 0.01 0.07020–2-2 I 52.93 46.39 0.13 0.14 0.18 0.22 0.01 0.35 0.04 0.03 100.42 1.99 0.87 0.42 0.93 0.27 0.40017–1 I 52.79 46.94 0.19 0.10 0.14 0.06 0.06 0.09 0.10 0.1 100.57 1.96 0.89 0.25 0.19 0.24020–1 I 52.33 45.60 0.48 0.22 0.02 0.27 0.07 0.30 0.18 0.02 99.49 2.00 0.87 0.48 2.18 0.70 0.29017–2-1 I 52.77 45.93 0.08 0.05 0.14 0.08 99.05 2.00 0.87 0.22 0.08 0.05017–2-2 I 52.78 46.77 0.05 0.27 0.22 0.04 0.15 0.01 100.29 1.97 0.89 0.01 0.19 0.32 0.26055–3 I-II 52.44 46.46 0.26 0.03 0.03 0.08 0.44 0.06 0.03 99.83 1.97 0.88 0.50 8.67 0.29 0.03009–1 I-II 52.62 47.04 0.10 0.02 0.19 0.07 0.08 0.04 0.17 0.24 0.01 100.58 1.95 0.89 0.45 5.00 0.12 0.26518–3 II 53.39 45.42 0.19 0.27 0.13 0.23 0.03 0.04 0.02 0.01 99.73 2.05 0.85 0.09 0.70 0.46 0.13052–1-1 II 52.64 45.48 0.13 0.14 0.05 0.16 0.10 0.16 0.04 0.02 98.92 2.02 0.86 0.22 0.93 0.27 0.21052–1-2 II 53.05 46.54 0.06 0.14 0.06 0.05 0.02 99.92 1.99 0.88 0.43 0.20 0.06056–2 II 52.81 46.16 0.07 0.41 0.04 0.12 0.03 0.02 99.66 1.99 0.87 0.03 0.17 0.48 0.04055–1-1 II 52.57 46.19 0.07 0.28 0.02 0.13 0.03 0.07 0.09 0.02 99.47 1.98 0.88 0.16 0.25 0.35 0.15055–1-2 II 52.48 46.19 0.24 0.34 0.12 0.09 0.14 0.03 0.06 0.03 99.72 1.98 0.88 0.09 0.71 0.58 0.2178–1 II 53.32 46.24 0.32 0.07 0.02 99.97 2.01 0.87 0.09 0.32 0.0078–2 II 52.35 46.28 0.05 0.19 0.04 0.01 0.01 0.11 0.07 0.01 99.12 1.97 0.88 0.20 0.26 0.24 0.0475–2-2 II 52.05 45.82 0.19 0.32 0.28 0.19 0.09 0.08 0.07 0.01 99.1 1.98 0.88 0.25 0.59 0.51 0.28

(continued on next page)

147G.W

angetal./JournalofG

eochemicalExploration

164(2016)

136–163

Table 4 (continued)

Samplenumber

Parageneticstage of pyrite

S Fe Co Ni Cu Zn Ag As Se Sb Te Au* Total n(S)/n(Fe) w(Fe)/w(S + As) w(As + Se + Sb + Te) w(Co)/w(Ni) w(Co + Ni) w(Cu + Zn)

73–1 II 52.57 46.32 0.38 0.22 0.05 0.31 0.03 0.07 0.04 99.99 1.98 0.88 0.07 1.73 0.60 0.36008–1-1 II 52.08 46.61 0.15 0.22 0.21 0.19 0.05 0.08 0.01 99.6 1.95 0.89 0.13 0.68 0.37 0.21008–1-2 II 52.40 46.69 0.12 0.21 0.03 0.18 0.05 0.04 0.12 0.01 0.02 99.87 1.95 0.89 0.22 0.57 0.33 0.03007–4 II 52.79 47.15 0.28 0.45 100.67 1.95 0.89 0.62 0.73 0.00052–2 II-III 52.52 46.33 0.24 0.22 0.06 99.37 1.97 0.88 1.09 0.46 0.0079–2-1 II-III 52.53 46.29 0.15 0.33 0.16 0.03 0.05 99.54 1.98 0.88 0.45 0.48 0.1679–3-1 II-III 52.53 45.90 0.14 0.35 0.07 0.03 0.18 0.95 0.01 0.05 0.1 100.31 1.99 0.86 1.01 0.40 0.49 0.1079–1-1 II-III 52.85 45.96 0.15 0.07 0.14 0.02 0.01 0.08 99.28 2.00 0.87 0.03 0.15 0.0779–1-2 II-III 52.74 46.17 0.23 0.18 0.18 0.05 0.28 0.08 0.02 99.93 1.99 0.87 0.36 1.28 0.41 0.18014–1 II-III 52.31 46.24 0.01 0.19 0.08 0.36 0.36 0.26 0.04 0.15 0.11 0.02 100.13 1.97 0.88 0.56 0.05 0.20 0.44008–2-1 II-III 52.67 46.47 0.21 0.09 0.18 0.00 0.08 0.20 0.43 0.03 100.36 1.97 0.88 0.71 2.33 0.30 0.18008–2-2 II-III 52.38 45.79 0.29 0.13 0.45 0.13 0.06 0.07 0.06 99.36 1.99 0.87 0.07 2.23 0.42 0.58007–2-1 II-III 52.12 45.92 0.24 0.16 0.07 0.13 0.03 0.01 0.04 0.01 0.01 0.05 98.79 1.98 0.88 0.07 1.50 0.40 0.20007–2-2 II-III 52.86 46.84 0.09 0.15 0.03 0.06 0.1 100.13 1.97 0.89 0.60 0.24 0.03022–2-1 II-III 53.12 46.92 0.02 0.10 0.25 0.02 0.05 0.03 100.51 1.97 0.88 0.20 0.12 0.27022–2-2 II-III 52.98 47.03 0.27 0.11 0.38 0.33 0.14 0.07 0.09 101.4 1.96 0.88 0.54 2.45 0.38 0.38022–2-3 II-III 53.17 46.19 0.20 0.01 0.02 99.59 2.00 0.87 0.01 0.20 0.00019–2 II-III 52.88 46.27 0.08 0.46 0.02 0.05 0.05 0.07 0.10 99.98 1.99 0.87 0.12 0.08 0.48019–1 II-III 53.14 45.84 0.20 0.23 0.03 0.01 0.30 0.08 0.09 99.92 2.02 0.86 0.39 0.87 0.43 0.03502 III 53.68 46.23 0.11 0.22 0.11 0.03 0.07 0.05 0.04 0.02 0.07 100.63 2.02 0.86 0.11 0.50 0.33 0.14052–3-1 III 52.74 46.45 0.10 0.15 0.13 0.07 0.02 0.12 0.04 99.82 1.98 0.88 0.21 0.67 0.25 0.13052–3-2 III 52.37 45.88 0.40 0.23 0.17 0.09 0.21 0.03 0.14 0.10 99.62 1.99 0.87 0.38 1.74 0.63 0.17D015-1 III 52.56 46.23 0.15 0.03 0.02 0.14 0.05 0.10 99.28 1.98 0.88 0.05 0.15 0.05D014–1-1 III 52.44 45.96 0.15 0.18 0.01 0.20 0.01 0.10 99.5 1.99 0.88 0.21 0.83 0.33 0.01D014–1-2 III 52.62 46.57 0.17 0.25 0.01 0.11 0.03 0.21 0.01 99.98 1.97 0.88 0.36 0.17 0.25003–1 III 52.95 45.91 0.11 0.36 0.27 0.01 0.03 99.64 2.01 0.87 0.01 0.31 0.47 0.27022–4-1 III 52.60 46.51 0.24 0.11 0.08 0.01 0.02 99.57 1.97 0.88 0.09 0.24 0.11044–2 IV 53.29 46.44 0.03 0.02 0.04 99.82 2.00 0.87 0.04 1.50 0.05 0.00D015-2-1 IV 52.91 46.70 0.14 0.19 0.04 0.28 0.08 0.02 0.01 100.37 1.97 0.88 0.02 0.74 0.33 0.32D015-2-2 IV 52.67 46.09 0.11 0.19 99.06 1.99 0.88 0.58 0.30 0.00D014-2 IV 52.69 46.08 0.28 0.13 0.18 0.19 0.01 0.01 99.57 1.99 0.87 0.20 2.15 0.41 0.0075–3 IV 53.13 45.90 0.01 0.03 0.26 0.03 0.03 0.02 99.41 2.02 0.86 0.03 0.33 0.04 0.26021–1-1 IV 52.75 46.76 0.31 0.32 0.04 0.13 0.01 0.06 0.11 100.49 1.96 0.89 0.18 0.97 0.63 0.17021–1-2 IV 53.22 46.61 0.38 0.06 0.01 100.28 1.99 0.88 0.38 0.06

⁎ Our EPMA of pyrites in the Xishan gold deposit did not measure Au concentrations, but we estimated Au content by double eyepiece observation of Au-bearing pyrite samples (see Fig. 9B).

148G.W

angetal./JournalofG

eochemicalExploration

164(2016)

136–163

Fig. 11.Major/trace element geochemistry of pyrites in the Xishan deposit: (A) S/Fe molar ratios; (B) Co/Ni ratios; (C) w(Fe)/w(S + As) ratios; (D) Au concentrations; (E) w(Co + Ni);(F) w(Cu + Zn).


formation, it follows that pyrite thermoelectricity can be used as an in-dicator of the depth of the ore-forming processes (Abraitis et al., 2004;Shen et al., 2013; Zhang, 2010).

4.2.4.2. (2) Measurements of thermoelectric properties of pyrite.Measure-ments of the thermoelectric properties of the pyrite samplesweremadeat the China University of Geosciences (Beijing) Genetic MineralogyLaboratory. The measurement apparatus is a BHTE-06 ThermoelectricCoefficient Measuring Instrument. The temperature of the freezingend was set to 20 °C, and the temperature of the hot end was set to80 °C. Fifty grains of pyrite per sample were selected for measurement,and each grain size was larger than 0.18 mm. The measured data aresummarized in Table 5.

In this study, thermoelectricity measurements were made from1600 grains of pyrites from 32 samples from different levels of theXishan deposit (from the surface down to−225m along six boreholes;Figs. 13 and 14, Table 5). Existing pyrite thermoelectric data are fromeight samples in Chen et al. (1989) and three samples from Shen et al.(2013) (Table 5). From the 1600 grains of pyrites, the minimum andmaximum values of measured α were −331.1 and 342.5 μV•°C −1, re-spectively, indicating that the pyrites were either N- or P-type. The per-centages of N- and P-types pyrites were determined per sample(Fig. 13B, Table 5), and the data indicate that N- and P-type pyrites com-prise on average 26.3% (with 22.1% standard deviation) and 74.0% (with22.1% standard deviation), respectively, of every sample. Therefore,most of the pyrites in the Xishan deposit are P-type. The P-type pyrites

Fig. 12. Trace element correlations in the Xishan deposit depicted by R-type cluster analysis of: (A) EPMA data of pyrites; and (B) whole-rock assay data.


have α values ranging from 1.6 to 342.5 μV•°C −1 whereas the N-typepyrites have α values ranging from −331.1 to−1.6 μV•°C -1.

4.2.4.3. (3) Variations in pyrite thermoelectric coefficients with parageneticstages. Figs. 13B and 14 show the variations of pyrite thermoelectric co-efficients in the different paragenetic stages of the Xishan deposit. Thecombinations of pyrite conduction types are mainly P N N for stages Ito III whereas stage IV shows a P b N combination (Fig. 14A). Most

high Au grade zones are generally associated with stage III pyrites inthe study area (Figs. 10 and 11D) but some high Au grade zones arealso associated with stages II-III pyrites in orebodies Nos. 108–1 and107 (sample numbers 4 and 008–2 in Table 3).

4.2.4.4. (4) Pyrite thermoelectric coefficients in different levels. Previousstudies (e.g., Karpov, 1981; Large et al., 2009; Shen et al., 2013) suggesta zonal distribution of pyrite thermoelectricity in gold deposit: with

Table 5Thermoelectricity data and analysis of pyrites from the Xishan deposit.

Sample/Orebody Paragenetic stage ofpyrite in sample

Level(m)

N-type thermoelectricity, α(μV/°C))

N (%) Temp.(°C)

P-type thermoelectricity, α(μV/°C))

P (%) Temp.(°C)

XnP γ

Maximum Minimum Average Maximum Minimum Average

D012–1 I 230 −29.8 −269.4 −139.93 22 310.55 342.5 4.9 198.23 78 192.27 16 46.5D012–2 I 210 −59.34 18 354.88 332.8 3.3 154.05 82 165.76 10 47.5044–1 I 150 −72.47 6 347.66 334.5 4.9 222.23 94 206.67 −8 5275–1 I −145 −33.6 −232.4 −120.9 10 321.02 322 52.6 191.49 90 188.23 −22 45.54 I −185 −20.4 −13.4 −86.3 4 355.4 320.8 11.6 179.16 96 180.83 34 41.5007–1 I −185 −1.6 −36.8 −112.9 30 336.86 312.6 15 139.66 70 157.13 16 46007–3 I −185 −6.7 −227.6 −112.4 24 325.69 307.3 10.2 170.77 76 175.79 −2 50.5017–1 I −225 −50.3 −117.5 −82.3 6 342.25 290.9 14.9 160.31 94 169.52 28 4355–1 II 138 −40.8 −253.3 −142.02 10 309.4 337.3 28.2 216.58 90 203.28 44 3956–1 II 130 −6.8 −236.2 −95.59 14 334.94 312 6.6 200.57 86 193.67 36 4156–2 II 125 −10.1 −210.8 −75.55 20 345.96 327.2 16.8 196.97 80 191.51 22 44.573–1 II −145 −81.9 −302.5 −148.72 10 305.72 341.3 15.2 202.75 90 194.98 34 41.575–2 II −145 −13.2 −173.1 −61.62 10 353.62 327.1 6.7 184.62 90 184.1 36 4178 II −145 −3.4 −272.7 −128.06 30 317.08 335.6 11.9 211.05 70 199.96 12 47007–4 II −185 - - - 0 - 342.5 32.2 212.66 100 200.93 58 35.5008–1 II −185 −41.2 −60.8 −49.13 8 360.5 327 16.8 155.92 92 166.88 16 46

108 II 220Data from Chen et al.(1989)

−325.4 37.5Data from Chen et al.(1989)

325.6 62.5

79–1 II-III −145 −9.8 −40.5 −25.15 4 373.69 317.7 8.2 189.26 96 186.89 34 41.5007–2 II-III −185 −14.5 −285.7 −144.27 36 308.16 331 5.1 180.56 74 181.67 2 49.5008–2 II-III −185 −13.5 −325.4 −148.08 36 306.07 340.8 15 178.06 64 180.17 −14 53.514 II-III −185 −1.7 −285 −91.97 22 336.93 323.8 20.2 190.87 78 187.85 16 46019–1 II-III −225 −3.3 −189.9 −80.33 18 343.33 319.1 6.7 160.28 82 169.5 10 47.5022–2 II-III −225 −48.1 −324.9 −186.95 20 284.69 330 11.1 186.36 80 185.15 14 46.555 II-III −10 Data from Shen et al.

(2013)−94.1 21.21 Data from Shen et al.

(2013)171.2 78.79 47.73

55 II-III −90 −65.4 3.33 187.8 97.67 38.33D014–1 III 190 −16.8 −180.3 −78.68 12 344.24 327.1 13 217.3 88 203.71 46 38.53 III −185 −5.1 −130.4 −84.33 12 341.13 332.8 9.9 166.26 88 173.09 −22 45.5022–1 III −225 −11.5 −262.1 −118.67 14 322.24 336.2 23.4 194.97 86 190.31 32 42108 III 380

Data from Chen et al.(1989)

−387.9 25

Data from Chen et al.(1989)

278.2 75108 III 300 −195.3 37.5 342.4 62.5108 III 230 −369.1 22.22 294.5 77.78108 III 190 −244.1 20 275.5 8055 III 230 −240 36 223.2 6455 III 190 −157.1 37.1 305 62.9022–4 III −225 −1.6 252 −106.34 26 329.03 326.1 5.1 154.71 74 166.16 −4 51D015-1 III 170 −3.5 −295.7 −112.13 80 325.84 258.6 3.3 129.88 20 151.26 −82 70.5D014-2 IV 180 −3.3 −266.3 −85.92 44 340.26 323.9 1.6 154.71 56 166.16 −32 58D015-2 IV 165 −3.3 −331.1 −127.57 72 317.35 287.6 5 122.93 28 147.09 −80 70044–2 IV 140 −68.6 −241.7 −107.82 100 328.21 - - - 0 - −102 75.575–3 IV −145 −1.7 −210.7 −61.67 58 353.6 201.4 8.4 78.22 42 120.26 −58 64.5021–1 IV −225 −1.7 −182 −81.48 50 342.7 290.3 10 110.88 50 139.86 −21 60.5

108 IV 340Data from Chen et al.(1989)

0 0Data from Chen et al.(1989)

287.9 100

55 IV −50Data from Shen et al.(2013)

−85.2 63.33Data from Shen et al.(2013)

137.6 36.67 36.67

Blank cells mean not variable detected. P (%) is the occurrence rate of P-type pyrite. N (%) is the occurrence of N-type pyrite. Xnp is thermoelectric parameter. γ is denudation percentage.Instrumental unit: Mineral Typomorphy Laboratory of Chinese University of Geosciences (Beijing). Measuring instrument: BHTE-06 Thermoelectric Coefficient Measuring Instrument.


mostly or entirely P-type pyrite in the shallow portions of the orebody,and with deepening N-type pyrite increasingwhereas P-type pyrite de-creases, gradually changing to mostly or entirely N-type pyrite in thedeepest portions. Accordingly, the top-to-bottom zonation is P-type(shallow section) → P-type + N-type (central section) → N-type(deep section) (Chen et al., 1989; Shao et al., 1990). Thus, variations inthe conduction type and thermoelectric coefficients of pyrite along thevertical direction can be an important indicator of orebody depth, thestatistical plot feature is as similar as Fig. 14C (Shen et al., 2013).

From Fig. 14B it can be deduced that the P-type pyrites tend to beprevalent at shallower levels whereas N-type pyrites tend to be preva-lent at deeper levels. The occurrence rate (percentage) of P-type pyritedecreases (50% of total) and N-type pyrite increases (50% of total) nearlevels 140m,−50mand−145m, and these levels are in torsional faultzones, which are regarded as small orebody zones, whereas higher P-type percentage zones in the other levels are usually in tension-shearfault zones, which generally are regarded as large orebody zones.

For No. 107 orebody (from the surface to−225m level), pyrite ther-moelectricity coefficients range from 200 to 300 μV•°C−1 and the valuerange is small. From the surface to−145m level of theNo. 108 orebody,pyrite thermoelectricity coefficients vary from 250 to 350 μV•°C −1;whereas at the−225 m level pyrite thermoelectricity coefficients varyfrom −150 to 0 μV•°C −1. Based on 11 samples obtained by Chenet al. (1989) and Shen et al. (2013) fromNo. 108 orebody (from the sur-face to−90m level) and ourmeasurements fromNos. 108, 107, and 55orebodies (from −105 m level to −225 m level) (Table 5), the P-typepyrite has three cyclicity distribution features in the vertical direction(from the surface to−225 m level). This indicates that the pyrites wereformed in the same period of mineralization and there was similar Au-bearing fluid evolution and crystallization among pyrites in the differentAu-bearing fracture-quartz veins (orebodies) in the Xishan deposit.Fig. 13B shows that the thermoelectric coefficients of pyrite in the Xishandeposit have steady high percentage of P-type distribution in ore-formingstages I, II, and III, and they have similar feature both at the surface and

Fig. 13. 3Dmodels of gold orebodies in the Xishan gold deposit. (A) Effective samples of Au-bearing pyrites (green and red solid 3D grid cells). (B) Histogram of thermoelectric coefficientsof pyrites in samples 73–1 and 78.


depth (Fig. 13A)thus,we canuse the trendmodel of P-type distribution in3D space to indicate potential prospects at depth. Figs. 14D and E showthe variation of thermoelectric features for P-type and N-type pyrites,Fig. 14F and G show the variation of estimated temperatures for P-typeand N-type pyrites respectively. The P-type pyrites has stable features invertical direction (from 200 m to -200 m level) compared with the fea-tures of N-type pyrites in study area.

All the pyrites of stage-I and stage-II and stage III have complexcrystal shape (Fig. 9B), high P-type percentage (Fig. 14A), and theirhistograms of the thermoelectric coefficient have similar distribu-tions (Fig. 13B). The thermoelectric coefficients of stages I and IIpyrites are discontinuous whereas those of stage III pyrites are con-tinuous (Fig. 13B), and the latter characteristics are likely related tothe metallogenesis of the gold deposits in the study area becausethe period of ore-forming conditions in stage III is continuous andlonger compared to those in stages I and II (Shen et al., 2013). There-fore, the thermoelectric coefficients of P-type pyrites have importantimplication for ore-forming stage and gold concentration in theXishan deposit, and it can be used to interpret and delineate vein-type gold mineralization in the study area by using it as a proxy forore-environment conditions (e.g., ore-forming temperature).

4.2.4.5. (5) Characterization of mineralization temperature4.2.4.5.1. (a) Temperature estimation using fluid inclusions. To validate

temperature estimation using pyrite thermoelectricity, we measuredtemperatures from fluid inclusions in quartz, which is a metallogenicmineral like pyrite in all the four paragenetic stages of the Xishan

deposit. Through petrography, microthermometry, and Raman spec-troscopy, two types of fluid inclusions trapped in Au-bearing quartzveins were identified to be related to the ore-forming stage (Figs. 15,16 and 17). Type I inclusions contain two (liquid H2O + CO2-richvapor) or three (liquid H2O + liquid CO2 + CO2-rich vapor) phasesat room temperature (Fig. 16a). They are generally 4–9 μm in diam-eter and appear as clusters (Fig. 16b) or are found along pseudo-secondary trails (Fig. 16c). Type II aqueous inclusions are small(2–7 μm) and comprise two phases (liquid H2O + vapor H2O)(Fig. 16d). They exist in clusters and along pseudo-secondary trails(Fig. 16e), coexisting with type I inclusions (Fig. 16f). They are dis-tributed within CO2–H2O inclusions clusters (Fig. 16h). In the fluidinclusions of samples subjected to microthermometric analysis, theliquid phase homogenization temperatures of fluid inclusions fromstage I vary in a large range, mostly between 140 and 300 °C(Fig. 17A). The main ore-forming stage homogenization tempera-tures are concentrated mostly in the 160–250 °C range.

4.2.4.5.2. (b) Temperature estimation using pyrite thermoelectricity.Pyrite thermoelectricity and conduction types vary depending on differ-ent temperatures during ore formation. The conduction types aremost-ly N-type at high temperatures, more-or-less equal proportions of N-and P-types occur at moderate temperatures, and mostly P-typesoccur at low temperatures (Chen et al., 1989). This supports theobservation that P-type pyrites tend to be prevalent at shallower levelswhereas N-type pyrites tend to be prevalent at deeper levels (Fig. 14B).Therefore, it appears that pyrite thermoelectricity is related to its crys-tallization temperature. The relationships between thermoelectricity

Fig. 14. Thermoelectricity data and analysis of pyrites from the Xishan deposit (Table 5). B: the occurrence rate of P-type and N-type pyrites in Xishan gold deposit in Linglong district; C:The occurrence rate of P-type pyrites in Dongshan gold deposit in Linglong district (Shen et al., 2013).


and temperature have been defined by equations derived from experi-mental data as (Zhang, 2010; Xue et al., 2014):

T ∘Cð ÞN‐type ¼ 704:51−jαjð Þ=1:818 ð2Þ

T ∘Cð ÞP‐type ¼ 3 122:22þαð Þ=5:0 ð3Þ

The measured pyrite thermoelectric coefficients obtained fromour samples were used in the above respective formulas (Fig. 18),

and it was concluded that the formation temperatures of pyrites inthe Xishan deposits vary between 120.26 and 373.69 °C, with N-type pyrites forming in the temperature range of 284.69–373.69 °Cand P-type pyrites forming in the range of 120.26–206.67 °C(Table 5). The histogram of values in the pyrite thermoelectriccoefficient–temperature map (Fig. 19A) shows that the formationtemperatures are concentrated mostly between 180 and 260 °C.This is consistent with our fluid inclusion data, which show a domi-nant range of 140–280 °C homogenization temperatures (Fig. 17A).

Fig. 15. Fluid inclusion micrographs of the Xishan gold deposit. (a) and (b) Type I inclusions consisting of three phases (liquid H2O+ liquid CO2 + CO2-rich vapor). (c) Type II inclusionsconsisting of two (liquid H2O + CO2-rich vapor) phases. (d) Type II inclusions occurring along pseudo secondary trails. (e) Type II inclusions consisting of two (liquid H2O + CO2-richvapor) phases. (f) Type II inclusions occurring in clusters and along pseudo secondary trails.


Thus, the Xishan deposit is a low–medium temperature deposit, as itfalls within the classification of mineral deposits based on homoge-nization temperatures of quartz (156–365 °C) and decrepitatingtemperatures of pyrite (193–343 °C) described in the literature(Chen et al., 1989; Shen et al., 2013). Thus, pyrite thermoelectric co-efficients can be used as a proxy for mineralization temperatures andfor characterizing the features of temperature change in theorebodies being modeled.

4.2.4.5.3. (6) Thermoelectric parameter of pyrite. The pyrite thermo-electric parameter (Xnp) can be calculated based on the thermoelectriccoefficient (Yang and Zhang, 1991; Yang and Meng, 1991; Shen et al.,2013; Xue et al., 2014), thus:

Xnp ¼ 2 f I þ f IIð Þ‐ f IV þ 2 f Vð Þ ð4Þ

where f corresponds to levels of thermoelectric coefficients of pyrite inthe samples: fI is α N 400 μV/°C, fII is α in the range 200 – 400 μV/°C,

fIII is α in the range 0 – 200 μV/°C, fIV is α in the range 0 – -200 μV/°C,and fV is α b −200 μV/°C. The values of Xnp of pyrites in the Xishangold deposit vary from −102 to 58 (Table 5), with high values (i.e., inthe fIII range) pertaining to stage II and III pyrites collected from the sub-surface (mostly in the -145 m to -225 m depth interval), and with lowvalues (i.e., in the fIV range) pertaining to stage IV pyrites collectedfrom the surface.

Values of Xnp have been used to estimate the erosional or exhu-mation level (γ) of gold deposits as γ = 50 – Xnp/4 (Yang andZhang, 1991; Yang andMeng, 1991; Xue et al., 2014). The Xnp valuesand corresponding γ values from pyrites (Table 5) suggest that theXishan gold deposit is not yet significantly eroded. For example,values of Xnp from pyrites in the No. 108 orebody vary between 35and 75, with an average of 56, suggesting that the No. 108 orebodylikely represents the upper part, below the paleo-top, of the Xishandeposit. On the other hand, values of γ from pyrites in the No. 108orebody, calculated from thermoelectric parameters from Shen

Fig. 16. Laser Raman spectra for fluid inclusions representative of the various types described from Xishan gold deposit.


et al. (2013), vary between 35.5% and 75.5%, with an average value of51%, suggesting that parts of the Xishan deposit above the No. 108orebody have already been eroded. These interpretations of theXnp and γ data are consistent with the n(S)/n(Fe) values of 1.95–2.05 (Table 4), with average of 1.92 and standard deviation of 0.02,which suggest that the No. 108 orebody likely represents the origi-nal upper to middle parts of the Xishan deposit (cf. Karpov, 1981;Xue et al., 2014). With the earlier interpretation of shallow environ-ment for gold mineralization based onw(Fe)/w(S + As) ratios of py-rites (cf. Shen et al., 2013) and the interpretation that the Xishangold deposit is not yet significantly eroded based on the Xnp and γdata from pyrites, it is likely that significant economic mineraliza-tion still likely exists beneath the present level of erosion of the

Xishan deposit. Therefore, 3D modeling becomes important in thecontext of exploration targeting.

4.3. Exploration targeting in 3D

4.3.1. Methodology of 3D modelingThe datasets available for 3D modeling in the Xishan area include

1:2000 scale geologic data, a digital elevationmodel (DEM), an explorato-ry engineering layout plan per level at 100m, 50m, 0m− 45m,−90m,−145m,−185m, and-225m, and related prospecting line profile maps.These datasets were used to model the orebodies and geology in theXishan area in 3D. The X coordinates of the model area range from40,543,920 m to 40,545,420 m, the Y coordinates from 4,148,500 m to

Fig. 17. A: Fluid inclusions temperature, B: salinity frequencies analysis of ore-formingstages in Xishan gold deposit.


4,150,000m, and the Z coordinates from−400mto570m(Fig. 13A). Themodel area at the surface measures 2.4 km2. The 3D orebody model re-gion contains the orebodies Nos. 108, 55, 71, and 10, which represent12% volume of the 3D block model of the study area.

The thermoelectric coefficient data from the Xishan deposit(Table 5) are discrete and sparse, and thus we cannot calculate thesemi-variograms in the horizontal and vertical directions in 3D space.Thus, the DSI (discrete smooth interpolation) technique in GoCAD wasused to construct 3D trend models of potential targets from the com-bined 3D orebody model and mining engineering data sets (e.g., gradeand thickness of different levels, history research data of orebody atthe subsurface).

4.3.1.1. (1) DSI method. The DSI method is the core interpolation algo-rithm based on the SGrid model used by GoCAD software (Mallet, 1989,2002). The basic element of the DSImethod is the establishment of a con-nected network among discretization points in 3D space. If the values ofnetwork points meet the constraint criteria, the values at unsampledpoints can be obtained by computing linear equations (Mallet, 2002).DSI is particularly suitable for 3D implicit spatial modeling of natural ob-jects combining explicit spatial distribution of geological properties(e.g., 3D orebodymodel and 3D fracture model of gold deposit) and usu-ally sampled at a few discrete locations (e.g., drillholes). In this paper, ad-ditional constraints like orientation data (e.g., the strike and dip of anorebody at different levels) can be included in the DSI method, and the3D shape/geometry modeling capabilities can be combined with interac-tive construction techniques.

4.3.1.2. (2) C-V fractal method for cutoff calculation. The C-V (concentra-tion-volume) fractal model in 3D space was proposed by Afzal et al.(2011), based on the concentration-area fractal model proposed byCheng et al. (1994), to separate geochemical anomalies from back-ground in 2D space. Afzal et al. (2011) demonstrated the C-V fractalmodel in 3D space could be used for distinguishing between supergeneand hypogene mineralized zones as well as barren host rocks in theChah-Firouzeh and Sungun porphyry Cu deposits in Iran. Wang et al.(2013) used the C-V fractal model to identify in 3D space primary andsecondary Cu orebodies in the Tongshan deposit in China. The C-V

fractal model can be expressed as (Afzal et al., 2011):

V ρ≤νð Þ∞ρ−a1 ;V ρ≥νð Þ∞ρ−a2 ð5Þ

where V(ρ ≤ ν) and V(ρ ≥ ν) denote two volumes with concentrationvalues (ρ) less than or equal to and greater than or equal to, respective-ly, the contour value (ν), which represents the cutoff/threshold value ofa zone (or volume); and a1 and a2 are characteristic exponents.

The C-V fractal method has also been used to calculate cutoffs anddelineate mineralized zones, using the distribution of major, medium,minor or paragenetic elemental concentrations in different types of al-teration associated with magma-skarn Cu deposits (e.g., Wang et al.,2013, 2015). Cutoff values from the C-V fractal model are likely tomap the boundaries between different ore zones (Afzal et al., 2011,2013;Wang et al., 2013). In this paper, the C-V fractalmodelwas imple-mented to calculate cutoffs of thermoelectric data pyrite using GeoCubesoftware (Wang et al., 2015).

4.3.2. Modeling the distribution of the thermoelectric coefficient of pyrite in3D

The P-type thermoelectric coefficients of pyrite were interpolatedwith the DSI technique for 3D modeling of potential targets, using a3D-grid cell (17 × 17 × 10 (m3)). C-V cutoffs and their classificationnumbers were calculated using the GeoCube software (Fig. 18 andTable 6). We constructed the 3D model of pyrite temperature basedon the linear relationship between P-type thermoelectric coefficientsand temperatures (Eq. (3)), and using DSI and C-V fractal methods(Fig. 19, Table 7). The dominant range of estimated pyrite formationtemperatures (i.e., 180–260 °C) in the 3D model derived using the DSIand C-V methods (Fig. 19A) is consistent with the dominant range of140–280 °C homogenization temperatures measured from quartz fluidinclusions (Fig. 17A). Therefore, we can validate the 3D thermoelectriccoefficients model by using other thermometric methods includingfluid inclusions analysis of quartz in ore-forming stages.

The 3D subsurface prospectivity of the Xishan deposit is modeledbased on 3D P-type thermoelectric coefficients of pyrite (Figs. 18D, 20A,22). For this 3D modeling, we added the shallow subsurface samples ofChen et al. (1989) and Shen et al. (2013) to our surface samples fromthe Xishan deposit (Fig. 13A). Fig. 20B and C are the longitudinal sectionmaps of gold veins, which show their NE60° trends modeled by ordinaryinterpolation in ArcGIS. The vertical projection mapwas used to estimatethe trend of gold concentrations at depth (Fig. 20B and C). The distribu-tion of pyrite thermoelectric coefficients on the surface (N120 m) of theXishan deposit has a zone of high values (with a maximum of190.90 μV•°C −1) in the northwestern part of the model area (Fig. 21a),a zone of low values (with minimum of 153.25 μV•°C −1) is present inthe southeastern part of themodel area (Fig. 21b), and a zone of interme-diate values in the central part of the model area (Fig. 21c). The distribu-tion of pyrite thermoelectric coefficients at depth (from 0 to −400 m)changes in shape from the northeast to the northwest of the area(Fig.21d), as controlled by NE-trending faults, presenting intermediatevalues (from100 to 150 μV•°C−1); the distribution shape in the southeastpart of the area shows a trend of high values (from 120 to 150 μV•°C −1)(Fig.21e). Fig. 21c and d show the 3D thermoelectric coefficientsmodel ofhigher 3D grid values (i.e., N60 μV•°C −1) at -105 m and -225 m levels,respectively.

4.3.3. 3D exploration targets in the xishan gold depositFig. 22 is the 3D thermoelectric coefficient model for identification

potential targets using 3D grid cellsmodel bade on C-Vmodeling recog-nized and extracted in Xishan gold deposit. Fig. 23 shows the 3D geolog-ical models and mineralization targets model using thermoelectriccoefficientmodelingwith 3D grid cells modeling in Xishan gold deposit.The 3D models suggest that the northeast and northwest zones in thestudy area, where three NE-trending deposit-controlling faults exist(Fig. 23), are continuous exploration targets, whereas the southeast

Fig. 18. 3Dmodels of thermoelectric coefficients of pyrites in the Xishan gold deposit. (a) 3D grid cells histogram of thermoelectric coefficients of pyrites. (b) C-V fractal model of pyritethermoelectric coefficients. (c) GoCAD 3D model of pyrite thermoelectricity. (D) GeoCube 3D model of pyrite thermoelectricity.


zone is a discontinuous exploration target at depth (i.e., below level-260 m) because it is constrained by the regional NE Linglong fault.Thus, there is no potential target in the southeast zone at depth fromthe No. 55 orebody (Figs. 22f and 23).

When comparing the 3D model of the thermoelectric coefficient ofpyrites and 3D orebody maps in the Xishan gold deposit (Fig. 20A),most of the main Au orebodies are located in the medium value range(100–150 μV•°C −1), as confirmed by the results of C-V modeling(Fig. 18B). Apparently, the main Au orebodies such as Nos. 108, 55,and 107 are located in transition zones between high and low P-type

thermoelectric coefficients (Figs. 20, 21, and 22), where thermoelectriccoefficients vary from 100 to 160 μV•°C −1. The trend of thermoelectriccoefficients of pyrites of orebodies in 3D (Figs. 20A, 21) and the trend ofthe No. 108–1 orebody in longitudinal section (Fig. 20B, C) are the sameas the trend (N60°E) of the orebody at depth. The orebodies Nos. 108–1and 107 extendmore than 700 m in the vertical direction in the Xishandeposit, and the potential exploration target from these orebodies seemto continue toward the NE direction and they seem to formone orebodyat depthwhere the faults that control orebodies Nos. 108–1 and 107 arejoined. These spatial relationships suggest that the main exploration

Fig. 19. 3D models of temperature of pyrites in Xishan Mining Area. (a) 3D grid cells histogram of temperature of pyrites in Xishan gold deposit. (b) C-V fractal model of pyritetemperatures. (c) GoCAD 3D model of pyrite temperatures. (D) GeoCube 3D model of pyrite temperatures.

Table 6The classification of 3Dmodeling of thermoelectric coefficient of pyrite in Xishan gold de-posit. The 3D Voxet/SGrid cell is 17 × 17 × 10 (m3).

Classification Goodness of fit Fractal dimension Range Number

4 0.73 0.53 [366.28, 409.86] 84,5663 0.99 8.79 [267.85, 366.28] 842,9142 0.99 2.70 [75.77, 267.85] 70,6381 0.98 8.62 [46.74, 75.77] 1881

Table 7The classification of 3D modeling of temperature coefficient of pyrite in Xishan gold de-posit. The 3D Voxet/SGrid cell is 17 × 17 × 10 (m3).

Classification Goodness of fit Fractal dimension Range Number

4 0.96 0.34 [246.05, 311.17] 68,4523 0.95 1.61 [225.27, 246.05] 120,9992 0.99 19.02 [185.09, 225.27] 653,9721 0.98 73.75 [172.62, 185.09] 17,036


Fig. 20. Trendmodel of Aupotential in theXishangold deposit. (A) 3Dmodels of thermoelectric coefficients of pyrites and orebody inXishan gold deposit. (B) and (C) Longitudinal sectionsof gold veins and their trend using OK interpolation (position: NE 60°).


targets at depth are located and trend toward the NE, along the strike of3D model of orebodies Nos. 107 and 108–1 (Figs. 18C, D, 20A, 21, 22,23c, d, e, 23f).

5. Discussions and conclusions

The Linglong Fault is the regional-scale structural control on quartzvein-type gold mineralization in the Linglong district, and its subsidiaryfaults are the local-scale structural control on the Xishan quartz vein-

Fig. 21. The 3D model of thermoelectric coefficient of

type gold mineralization in the northeastern zone of the study area.Accordingly, structural analysis in 3D would be optimum forexploration targeting in this and similar ore-control settings(e.g., Murphy et al., 2006; Sprague et al., 2006; Lindsay et al.,2013a; Hill et al., 2014; Martin-Izard et al., 2015; Nielsen et al.,2015). However, in this study, our focus was on answering the ques-tion “How can we construct a 3D exploration targeting model usingtypomorphic features of pyrites in complex and irregular quartz-vein hosted gold in the Xishan deposit, China”? Answering this

pyrites and orebody in the Xishan gold deposit.

Fig. 22. The 3D model of thermoelectric coefficients for identification potential targets using 3D grid cells modeling in the Xishan gold deposit.


question is non-trivial because pyrite is a ubiquitous mineral inmany types of mineral deposits and it presents itself as a suitableand abundant sampling media for exploration-relevant physico-chemical data, although it is only during the last decade or so that py-rite characteristics (e.g., trace element geochemistry are given im-portant consideration in exploration for structurally-controlledmineral deposits in orogenic settings (Large et al., 2009).

In this paper, we analyzed the pyrite features on basis of crystal type,geochemical component, Au concentration, thermoelectric coefficient,and temperature of samples from different ore-forming stages in theXishan gold deposit. The research results show that: (1) pyrite crystaltypes can be easily described but they do not provide quantitative infor-mation on ore-forming stages, which can be used in building explora-tion criteria of 3D targeting; (2) trace element compositions of pyritesfrom paragenetic stages are strongly variable and are not correlatedwith Au concentrations in ores, and thus are not useful for 3D explora-tion targeting; (3) thermoelectric parameters of pyrite are useful for

3D exploration targeting in the study area, because (a) pyrites fromore-forming stages I–III show similar thermoelectric parameters com-pared to pyrites from stage IV (e.g., pyrites in ore-forming stages I–IIIare dominantly P-type whereas pyrites in stage IV are dominantly N-type) and (b) ores in ore-forming stages I–III have higher Au gradescompared to stage IV ores.

To assess the uncertainty of DSI results in this study, we used 3D OKinterpolation to estimate the trend of 3D potential targets between−105 m and −225 m in study area (Fig. 24). The 3D potential targetsbased on 3D modeling of thermoelectric coefficients can be cross-validated by the 3D temperature model in Fig. 23e. Therefore, thereare low temperature targets in the southwest parts of the study areaand there are high temperature targets in the northeast parts. The tem-perature gradientmodel (Fig. 23e) and the knownmineralized zones inthe southwest parts of the study area indicate that the main orebody iscontinuous at depth (from−105 m to−400 m) in the northeast parts(Fig. 23f).

Fig. 23. The 3D geological models and targets model using thermoelectric coefficient modelingwith 3D grid cells modeling in Xishan gold deposit: (a) 3D fault and orebodymodel, (b) 3Dweak mineralization zones using thermoelectric coefficient modeling, (c) 3D strong mineralization zones using thermoelectric coefficient modeling, (d) 3D strong mineralization zoneswith thermoelectric coefficient values color bar, (e) 3D strong mineralization zones with temperature values color bar, (f) 3D weak and strong mineralization zones model.


The Xishan deposit hasmany quartz vein-type gold orebodies, and itis difficult to identify which of the various small vein-type orebodies arelinked with which main orebody in the different levels/drifts of themine. Therefore, it is instructive to construct 3D model of typomorphiccharacteristics of pyrite in the Xishan gold deposit using sparse sampleswith DSI interpolation and C-V method. In this paper, we showed that:(1) a 3D trendmodel of pyrite typomorphic characteristics is consistentwith a 3D model of gold orebody, and in the 3D trend model of pyritetypomorphic characteristics the four orebodies in the Xishan gold de-posit can be correlated to indicate/infer the deep exploration targets;(2) 3D modeling of pyrite typomorphic characteristics using DSI inter-polation can be used to help in decision-making for explorationtargeting in the subsurface, whereby we can design exploration bore-hole location, depth and dip, and plan for deep exploration based onthe features of the main orebodies at depth; (3) C-V modeling was ap-plied to derive 3D thermoelectric and temperature models, fromwhich we can obtain orebody volumes (Tables 4 and 5), although

these cannot be used to estimate metal resources because of the use ofsparse data in DSI modeling. However, the methodology can be usedin other gold camps in the world if pyrite is the chief Au-bearingmineral.

The research results show that orebodies Nos. 108–1 and 107 ex-tend more than 700 m in the vertical direction Xishan gold deposit,and the potential exploration targets for the continuation of theseorebodies are mainly located toward the northeast. Orebody No.107 likely continues at depth without branching whereas orebodiesNos. 108–2 and 55 are likely discontinuous at depth. At depth, theexploration targets are toward the NW of orebody No. 107 (from 0to 400 m).

Acknowledgments

The authors would thank Dr. Laurent Ailleres and an anonymous re-viewer for their constructive reviews, which improved the quality of

Fig. 24. 3D grade kriging interpolation analysis of Xishan deposit (the level is from−105 m to−225 m) in study area. A: 3D orebody model and sample locations in 3D space of Xishandeposit (the level is from−105m to−225m). B: The semi-variogramplot inN60°E inhorizondirection (the range is 142.20m). C: The semi-variogramplot in S150°E inhorizon direction(the range is 8.65 m). D: The vertical semi-variogram plot (the range is 40.18 m). E: 3D garde model using Ordinary Kriging interpolation of Xishan deposit (the level is from -105 m to-225 m).


this paper. GongwenWang is grateful to Prof. Clayton (University of Al-berta, Canada) for his Geostatistics teaching in Alberta of University, andto Profs. Ruiqing Gong and Longzhu An (Zijin Mining Group Co.) fortheir cooperation of field survey and discussion about exploration tar-gets in Xishan deposit. This researchwas supported by the National Sci-ence and Technology Support Project of the 12th “Five-Year Plan”(Grant No. 2011BAB04B06), the Fundamental Research Funds for theCentral Universities, China University of Geosciences (Beijing) (GrantNo. 2-9-2012-143), and the National Natural Science Foundation ofChina (Grant No 41572318).

References

Abraitis, P.K., Pattrick, R.A.D., Vaughan, D.J., 2004. Variations in the compositionaltextural and electrical properties of natural pyrite: a review. Int. J. Miner. Pro-cess. 74, 41–59.

Afzal, P., Dadashzadeh Ahari, H., Rashidnejad Omran, N., Aliyari, F., 2013. Delineation ofgold mineralized zones using concentration-volume fractal model in qolqoleh golddeposit, NW Iran. Ore Geol. Rev. 55, 125–133.

Afzal, P., Alghalandis, F.Y., Khakzad, A., Moarefvand, P., Rashidnejad Omran, N., 2011. De-lineation of mineralization zones in porphyry Cu deposits by fractal concentration–volume modeling. J. Geochem. Explor. 108, 220–232.

Bralia, A., Sabatini, G., Troja, F.A., 1979. Revaluation of the Co/Ni ratio in pyrite as geo-chemical tool in ore genesis problems. Mineral. Deposita 14, 353–374.

Chen, G., Shao, W., Sun, D., 1989. Genetic Mineralogy of Gold Deposits in Jiaodong RegionWith Emphasis on Gold Prospecting (in Chinese). Chongqing Publishing House,Chongqing.

Cheng, Q., Agterberg, F.P., Ballantyne, S.B., 1994. The separation of geochemical anomaliesfrom background by fractal methods. J. Geochem. Explor. 54, 109–130.

Cook, N.J., 1996. Mineralogy of the sulphide deposits at sulitjelma, northern Norway. OreGeol. Rev. 11, 303–338.

Cook, N.J., Ciobanu, C., Meria, D., Silcock, D., Wade, B., 2013. Arsenopyrite-pyrite associa-tion in an orogenic gold ore: tracing mineralization history from textures and traceelements. Econ. Geol. 108, 1273–1283.

Cox, D., Singer, D.A., 1986. Mineral Deposit Models. U.S. Geological Survey Bulletin 1693.United States Government Printing Office, Washington, p. 379.

http://refhub.elsevier.com/S0375-6742(16)30003-6/rf0005
























Dagbert, M., Harfi, M., 2002. Resource estimation for the draa sfar south polymetallic de-posit. Explor. Mining Geol. l11, 99–112.

De Kemp, E.A., Monecke, T., Sheshpart, M., 2011. 3D GIS as a support for mineral discov-ery. Geochemist. Explor. Environ. Anal. 11, 117–128.

Deditius, A.P., Utsunomiya, S., Reich, M., Kesler, S.E., Ewing, R.C., Hough, R., Walshe, J.,2011. Trace metal nanoparticles in pyrite. Ore Geol. Rev. 42, 32–46.

Doyle, F.M., Mirza, A.H., 1996. Electrochemical oxidation of pyrite samples with knowncomposition and electrical properties. In: Woods, R., Doyle, F.M., Richardson, P.(Eds.), Proceedings of the Fourth International Symposium on Electrochemistry inMineral and Metal Processing. Pennington: Electrochemical Society Series 96,pp. 203–214.

Fallara, F., Legualt, M., Rabeau, O., 2006. 3-D integrated geological modeling in the Abitibisubprovince (Québec Canada): techniques and applications. Explor. Min. Geol. 15,27–41.

Groves, D.I., Goldfarb, R.J., Gebre-Mariam, M., Hagemann, S.G., Robert, F., 1998. Orogenicgold deposits: a proposed classification in the context of their crustal distributionand relationship to other gold deposit types. Ore Geol. Rev. 13, 7–27.

Guo, P., Santosh, M., Li, S.R., 2013. Geodynamics of gold metallogeny in the Shandongprovince. NE China: an integrated geological, geophysical and geochemical perspec-tive Gondwana Research 24, 1172–1202.

Hill, E.J., Oliver, N.H.S., Cleverley, J.S., Nugus, M.J., Carswell, J., Clark, F., 2014. Characterisa-tion and 3Dmodelling of a nuggety, vein-hosted gold ore body, sunrise dam,WesternAustralia. J. Struct. Geol. 67, 222–234.

Karpov, I.K., 1981. Physicochemical Computer Modeling in Geochemistry [in Russian].Nauka, Novosibirsk.

Large, R.R., Danyushevsky, L., Hollit, C., Maslennikov, V., Meffre, S., Gilbert, S., Bull, S., Scott,R., Emsbo, P., Thomas, H., Singh, B., Foster, J., 2009. Gold and trace element zonation inpyrite using a laser imaging technique: implications for the timing of gold in orogenicand Carlin-style sediment-hosted deposits. Econ. Geol. 104, 635–668.

Li, S., Chen, G., Shao, W., 1994. Research on the chemical composition of pyrite in therushan gold deposit in jiaodong peninsula (in Chinese). Gold Science Technology 2,7–12.

Li, S., Zhao, G., Santosh, M., Liu, X., Dai, L., Suo, Y., Tam, P.Y., Song, M., Wang, P., 2012.Paleoproterozoic structural evolution of the southern segment of the Jiao–Liao–Jibelt, north China craton. Precambrian Res. 200–203, 59–73.

Lindsay, M.D., Aillères, L., Jessell, M.W., de Kemp, E., Betts, P.G., 2012. Locating and quan-tifying geological uncertainty in three-dimensional models: analysis of the gippslandbasin southeastern Australia. Tectonophysics 546-547, 10–27.

Lindsay, M.D., Jessell, M.W., Ailleres, L., Perrouty, S., de Kemp, E., Betts, P.G., 2013a.Geodiversity: exploration of 3D geological model space. Tectonophysics 594, 27–37.

Lindsay, M.D., Perrouty, S., Jessell, M.W., Aillères, L., 2013b. Making the link between geo-logical and geophysical uncertainty: geodiversity in the Ashanti greenstone belt.Geophys. J. Int. 195 (2), 903–922.

Mallet, J.L., 1989. Discrete smooth interpolation. ACM (association for computingmachin-ery). Transactions on Graphics 8 (2), 121–144.

Mallet, J.L., 2002. Geomodeling. Applied Geostatistics. Oxford University Press, New York,pp. 1–10.

Martin-Izard, A., Arias, D., Arias, M., Gumiel, P., Sanderson, D.J., Castañon, C., Lavandeira,A., Sanchez, J., 2015. A new 3D geological model and interpretation of structural evo-lution of the world-class Rio Tinto VMS deposit, Iberian pyrite belt (Spain). Ore Geol.Rev. 71, 457–476.

Mejía-Herrera, P., Royer, J.J., Caumon, G., Cheilletz, A., 2014. Curvature attribute fromsurface-restoration as predictor variable in kupferschiefer copper potentials. Nat.Resour. Res. http://dx.doi.org/10.1007/s11053-014-9247-7.

Meng, F., Sun, D., Li, S., 2001. The typomorphic characteristics of pyrite in the nanzhangjiagold mineral deposit, Shandong province (in Chinese). Geoscience 15, 231–237.

Murphy, F.C., Rawling, T.J., Wilson, C.J.L., Dugdale, L.J., Miller, M.L., 2006. 3D structuralmodelling and implications for targeting gold mineralisation in western Victoria.Aust. J. Earth Sci. 53, 875–889.

Nielsen, S.H.H., Cunningham, F., Hay, R., Partington, G., Stokes, M., 2015. 3D prospectivitymodelling of orogenic gold in the marymia inlier, Western Australia. Ore Geol. Rev.71, 578–591.

Oberthur, T., Cabri, L.J., Weiser, T.W., et al., 1997. Pt, Pd and other trace elements in sul-fides of the main sulfide zone, great dyke, Zimbabwe: a reconnaissance study. Can.Mineral. 35, 597–609.

Qiu, Y., Groves, D.I., McNaughton, N.J., Wang, L., Zhou, T., 2002. Nature, age, and tectonicsetting of granitoid-hosted, orogenic gold deposits of the jiaodong peninsula, easternnorth China craton, China. Mineral. Deposita 37, 283–305.

Reich, M., Dedititus, A., Chryssoulies, S., Li, J., Ma, C., Parada, M., Barra, F., Mittermayr, F.,2013. Pyrite as a record of hydrothermal fluid evolution in a porphyry copper system:a SIMS/EMPA trace element study. Geochim. Cosmochim. Acta 104, 42–62.

Shao, J., 1988. Exploration Mineralogy of Gold Deposits (in Chinese). China University ofGeosciences Press, Wuhan, pp. 38–45.

Shao, W., Chen, G., Sun, D., 1990. Method of investigating thermoelectricity of pyrite andits application to pyrites from gold deposits in jiaodong region (in Chinese). Geosci-ence 4, 46–57.

Shen, J., Li, S., Ma, G., Liu, Y., Yu, H., 2013. Typomorphic characteristics of pyrite from thelinglong gold deposit: its vertical variation and prospecting significance. Earth Sci.Front. 20, 55–75.

Sprague, K., de Kemp, E., Wong, W., McGaughey, J., Perron, G., Barrie, T., 2006. Spatialtargeting using queries in a 3-D GIS environment with application to mineral explo-ration. Comput. Geosci. 32, 396–418.

Tam, P.K., Zhao, G.C., Liu, F.L., Zhou, X.W., Sun, S., Li, S.Z., 2011. Timing of metamorphism inthe Paleoproterozoic Jiao–Liao–Ji belt: new SHRIMP U–Pb zircon dating of granulites,gneisses and marbles of the jiaobei massif in the north China craton. Gondwana Res.19, 150–162.

Tan, J., Wei, J.H., Audétat, A., Pettke, T., 2012. Source of metals in the guocheng gold depos-it, jiaodong peninsula, north China craton: link to early cretaceous mafic magmatismoriginating from Paleoproterozoic metasomatized lithospheric mantle. Ore Geol. Rev.48, 70–87.

Thomas, H.V., Large, R.R., Bull, S.W., Maslennikov, V., Berry, R.F., Frazer, R., Froud, F., Moye,R., 2011. Pyrite and pyrrhotite textures and composition in sediments, laminatedquartz veins, and reefs at Bendigo gold mine, Australia: insights for ore genesis.Econ. Geol. 106, 1–31.

Vollgger, S., Cruden, A., Ailleres, L., Cowan, E., 2015. Regional dome evolution and its con-trol on ore-grade distribution: insights from 3D implicit modelling of the navachabgold deposit, Namibia. Ore Geol. Rev. 69, 268–284.

Wang, G., Li, R., Carranza, E.J.M., Zhang, S., Yan, C., Zhu, Y., Qu, J., Hong, D., Song, Y., Han, J.,Ma, Z., 2015. 3D geological modeling for prediction of subsurface Mo targets in theluanchuan district, China. Ore Geol. Rev. 71, 592–610.

Wang, G., Pang, Z., Boisvert, J.B., Hao, Y., Cao, Y., Qu, J., 2013. Quantitative assessment ofmineral resources by combining geostatistics and fractal methods in the tongshanporphyry Cu deposit (China). J. Geochem. Explor. 134, 85–98.

Wang, G., Zhang, S., Yan, C., Song, Y., Sun, Y., Li, D., Xu, F., 2011. Mineral potential targetingand resource assessment based on 3D geological modeling in luanchuan region,China. Comput. Geosci. 37, 1976–1988.

Wen, B., Fan, H., Santosh, M., Hu, F., Pirajno, F., Yang, K., 2015. Genesis of two differenttypes of goldmineralization in the linglong gold field, China: constrains from geology,fluid inclusions and stable isotope. Ore Geol. Rev. 65, 643–658.

Xue, J., Li, S., Sun, W., Zhang, Y., Zhang, X., 2014. Characteristics of the genetic mineralogyof pyrite and its significance for prospecting in the denggezhuang gold deposit,jiaodong peninsula, China. Sci. Chin. Earth Sci. 57 (4), 644–661.

Yang, G., Zhang, Y.H., 1991. The relationship between pyrite thermoelectricity and oreforming temperature and its significance in gold deposits (in Chinese). Geol. Inf.Shandong 1, 20–25.

Yang, G., Meng, W.P., 1991. The mechanism of pyrite thermoelectricity in the applicationof prospecting mineralogy of gold ore (in Chinese). Proceedings of 3th National Con-ference of Placer Mineral, China.

Yang, Q., Shen, J., Li, S., 2013. Oxygen, boron, chromium and niobium enrichment in nativeAu and Ag grains: a case study from the linglong gold deposit, jiaodong, easternChina. J. Asian Earth Sci. 62, 537–546.

Zhang, Y., 2010. Genetic Mineralogy and Deep Prospects of Zhaodaoshan Gold Deposit inMuping, East Shandong Province (in Chinese)Master Dissertation Chinese Universityof Geosciences, Beijing, pp. 38––54.















































http://dx.doi.org/10.1007/s11053-014-9247-7




































































typomorphic characteristics of pyrite: criteria for 3d ... · typomorphic characteristics of...

Documents