Trace Element Geochemistry of Chalcopyrites and Pyrites from Golpu and Nambonga North Porphyry Cu-Au Deposits, Waﬁ-Golpu Mineral District, Papua New Guinea

: Studying elemental geochemistry of hypogene sulphides can discriminate the hydrothermal ﬂuids responsible for ore formation. To determine whether Golpu porphyry Cu-Au deposits are related to the Nambonga North porphyry system which is located 2.5 km apart in the Waﬁ-Golpu Mineral District, Papua New Guinea, we compare the trace element compositions of drill core chalcopyrites and pyrites analysed using Laser Ablation Inductively Coupled Plasma Mass Spectrometry (LA-ICPMS). The results for the Golpu chalcopyrites revealed high concentrations of Au, As, Se, Mo, Sb, Te and Bi and lower concentrations of Ag and Zn compared to those from Nambonga. Pd and Pt were below the detection limit in chalcopyrites for both deposits. The results for the Golpu pyrites indicated high concentrations of Pt, Au, Se, Mo, Sb, and Te and lower concentrations of Cu, Zn, As, Ag, Pb, Pd and Bi compared to those from Nambonga North. Au concentrations in the pyrites from both the porphyry deposits were higher compared to chalcopyrites, which mean that pyrite is the Au-bearing sulphide responsible for the higher Au content. In contrast, Cu values in pyrites from Nambonga North are higher than those from Golpu. Overall, it is envisaged that the ore ﬂuids were exsolved at different times during the evolution of both porphyry deposits, although these porphyry centres may be related in space and time.


Introduction
Papua New Guinea's unique geological heritage and its striking natural resources result from its location in the Southwest Pacific along the famous 'Pacific Ring of Fire'. Porphyry Cu-Au and epithermal Au deposits are common [1][2][3] along the major magmatic arcs of the northward-moving Indo-Australian plate and the westward-moving Pacific plate [4][5][6].
The northwest trending collisional zone in the interior of mainland PNG defines the New Guinea Orogenic Belt and hosts several porphyry and epithermal deposits such as Ok Tedi, Porgera, Mt Kare, Frieda River-Nena, Wafi-Golpu, Hidden Valley and Tolukuma ( Figure 1).
The metallogenesis of the porphyry and epithermal deposits is not fully understood because of the complex tectonic history and the lack of robust geological datasets over the region [6]. New trace element chemistry data of hydrothermal minerals associated with ore deposition in the region, particularly the New Guinea Orogenic Belt that hosts the Wafi-Golpu Mineral District (WGMD), can add more insights to the understanding of the metallogenesis.
The WGMD hosts the Golpu and Nambonga North porphyry Cu-Au deposits. Golpu has multi-phase, calc-alkaline, dioritic porphyries with associated epithermal and carbonatebase metal-Au deposits, whereas Nambonga North is a smaller porphyry system situated 2.5 km to the northwest. The Golpu deposit has a total resource of 830 Mt with 0.71 g/t Au, 1.1% Cu, 1.3 g/t Ag (measured and indicated grades) with an overall contained metal resource of 18.8 Moz Au, 8.7 Mt Cu and 32.6 Moz Ag [7]. The Nambonga North porphyry contains an inferred resource of 48 Mt, with a grading of 0.2% Cu and 0.69 g/t Au and with a contained metal of 1.1 Moz Au and 0.094 Mt Cu [8].
This paper focuses only on the pyrites and chalcopyrites which are the two most common Au-bearing sulphides of the Golpu and Nambonga North porphyry deposits. Although mineralisation also occurs in other mineral phases such as enargite, covellite, tennantite, pyrite and chalcopyrite are quite distinct, making them easier to work on in this study.
Using LA-ICPMS to determine their trace element compositions, we unravel the metal residence and concentrations and their elemental substitution and replacement, including closely associated mineral phases and their complexing agents. This information is also important for the mill and metallurgical processes that are required for optimal metal recovery.

Geological Setting
The geological evolution of Papua New Guinea is intricate with a mix of terranes comprising of sialic continental crusts, volcanic island arcs and oceanic micro-plates that have undergone fold-and-thrust belt deformation, metamorphism, magmatism, ophiolite obduction, uplift and sedimentation [4,5,[9][10][11][12][13][14]. Papua New Guinea has experienced major tectonic reconstructions since the start of the break-up of Gondwana in the Triassic period and continues to do so because of its location at the centre of the northerly moving continental Indo-Australian plate and the west-north-westerly moving oceanic Pacific plate [14][15][16]. How these large and exquisite Cu-and Au-mineralised porphyries formed on the island of New Guinea during the geological evolution has been a topic of debate for many decades. As new scientific data becomes available, our knowledge and understanding of porphyry deposits continues to improve [17]. Situated in the New Guinea Orogenic Belt, the geology of the WGMD has been reported in many company reports and publications e.g., [18][19][20][21] and is generally summarised in this paper.

Stratigraphy
The oldest rocks are Jurassic to Cretaceous in age and belong to the Owen Stanley Metamorphic Complex (OSMC) units. The units consist of interbedded siltstone, sandstone and conglomerate and shale; shale is carbonaceous in parts. These rocks have been regionally metamorphosed to greenschist facies, are weakly to moderately foliated with a northerly strike and dip 40-60 • to the east-southeast.
Middle Miocene Langimar Beds unconformably overlie the basement in the west. The Beds are comprised of volcanolithic pebble to cobble conglomerate with a tuffaceous matrix, silty sandstone, interbedded marl, mudstone, and lenses of detrital limestone, including lavas and pyroclastic units of basaltic to andesitic composition.
The Pliocene Babwaf Formation lies unconformably on the basement in the west. The Formation includes poorly sorted, weakly consolidated polymictic conglomerates, sandstones and siltstones.
Holocene and Quaternary Deposits fill depressions and cover topographic highs throughout the district. The Wafi Conglomerate occurs in the east and constitutes a sequence of unconsolidated conglomerate, gravel, sand, siltstone, clay and organic material.

Structures
The two major structural trends in the region surrounding WGMD are the northwest to north-northwest striking, arc-parallel (accretionary) structures and the northeast to northnortheast trending, arc-normal (transfer) structures [1,2,10]. These structures developed in response to the oblique plate convergence of the Indo-Australian plate and the Pacific plate, which created isoclinal folding, thrust faulting and sinistral transverse faulting parallel to the thrust direction [2].
The north-northwest striking structures are extensional faults related to the Wandumi and Watut Faults in the southern Wau Basin, a region that hosts the Morobe Goldfields ( Figure 2). The arc-normal structures developed due to the east-southeast-directed compression of the northwest-moving Pacific plate against the Indo-Australian plate, which also generated isoclinal folding and thrust faulting. The Golpu and Nambonga North porphyries are controlled by the northeast to north-northeast striking faults. Rafferty's Fault and Dokaton Fault bound the Golpu porphyry deposit and its associated intrusive phases to the east and west, respectively ( Figure 3).

Intrusions of the Golpu and Nambonga North Porphyry Centres
The Golpu intrusive phases are bounded by Rafferty's Fault in the east and Dokaton Fault in the west. At the time of the study, the Golpu intrusions were classified as the Main Golpu porphyry, Golpu West porphyry and Golpu North porphyry based on the textural and mineralogical characteristics (C. Muller, personal communication; 9 April 2010). The Golpu composite porphyry suite extends over about 800 m north-south by 500 m west-east, and was drill tested to a depth of more than 2000 m [8]. The Main Golpu porphyry is rich in quartz and feldspar. The Golpu West porphyry is directly west of and is more feldspar-rich than the main Golpu porphyry. The Golpu North porphyry is located at the northern margin of the main Golpu porphyry and is rich in hornblende.  Nambonga North is a quartz and feldspar-rich diorite porphyry similar to the Main Golpu porphyry located 2.5 km northwest of the Golpu intrusions. It intrudes both the OSMC units and Babwaf Formation and covers an area of about 200 × 200 m with a vertical extent of at least 800 m [8]. The diorite porphyry hosts the porphyry-related alteration-mineralisation with a minor high sulphidation overprint and argillic alteration assemblages [8].

Other Intrusive Suites
The Wafi Diatreme is a phreato-magmatic breccia or maar-related breccia pipe with steeply dipping, northeast-oriented margins that are vented to the surface west of the Golpu porphyries. The diameter of the diatreme is about 2 km from the north-northeast to southsouthwest direction. The diatreme is comprised of intrusive, sedimentary, volcaniclastic and tuffaceous rocks, which are poorly sorted, matrix-supported and heterolithic with clasts sizes not greater than 10 mm in diameter. The rock fragments include sub-rounded shale clasts, weakly bedded siltstone, undifferentiated metasedimentary rocks and crystals of quartz and feldspar in a matrix of fine-grained hydrothermally altered clays and fragments of quartz vein material.
The Wafi Dacite porphyry intrudes the Wafi Diatreme and Golpu porphyries in the east as small dikes. This porphyry is fine-to medium-grained and is comprised of quartz and feldspar phenocrysts. It also hosts discrete quartz and pyrite veins and can be termed as a late-mineral dike.
The Hekeng Andesite porphyry is situated south-southwest and south-southeast of the Golpu porphyries and the Wafi Diatreme, respectively. Similar to the Golpu porphyries, it is bounded by the Rafferty's and Dokaton Faults, and cross-cuts the Wafi Diatreme and Wafi Dacite porphyry. The andesite porphyry is fine-to medium-grained, is partly brecciated and denotes a late-to post-mineral intrusive phase. Kaolinite alteration is predominant.
The Wafi high sulphidation epithermal alteration-mineralisation consists of A and C Zones and is spatially, temporally and genetically related to the Golpu porphyry deposit ( Figure 3). The earlier formed hydrothermal alteration is zoned from silica-alunite (advanced argillic) to kaolinite-dickite-pyrophyllite (intermediate argillic) to illite-smectite (argillic) mineral assemblages. The late-stage mineralisation is characterised by the presence of enargite-covellite-tennantite-pyrite-arsenopyrite mineral assemblage and is comprised predominantly of refractory Au in arsenopyrite and arsean pyrite, including some pyritic phases.
The low sulphidation epithermal alteration-mineralisation is comprised of B and Link Zones and is located at the southeastern margin of the Wafi Diatreme. The zones consist of quartz veins with illite-smectite-chlorite-carbonate alteration selvages and pyrite ± marcasite mineralisation. The veins are cross-cut by more than one generation of pyrite-sphaleritegalena-carbonate mineral assemblages.
Several carbonate-base metal ± Au zones (i.e., Western, Malaria, Nambonga and Hesson Creek) occur at the marginal areas of the two porphyry Cu-Au deposits. These zones are considered to have formed through wall-rock interaction and/or the mixing of the magmatic-hydrothermal fluids derived from the source porphyry intrusion(s).
Supergene fluids have formed an oxide-hosted cap at the top, which is underlain by a weakly developed enrichment blanket consisting of secondary chalcocite and covellite. Iron oxides, quartz and alunite are the predominant minerals in the oxidised horizon.
Calibration was completed using external standards Po724 B2 SRM (Sulphide Standard, Memorial University Newfoundland) for Pd, Pt and Au and Mass-1 (Polymetal Sulphide Standard, United States Geological Survey) for Cu, Zn, As, Se, Mo, Ag, Sb, Te, Pb and Bi, respectively. A single spot with a 25 µm crater-diameter was analysed by a time-resolved analysis method with the measurements taken at the maximum peak and at a 15 Hz repetitive rate. A few smaller grains were also analysed using a 15 to 20 µm crater size. The analysis time was 20 s after a 20 s instrumental background analysis (carrier gas + LA-ICPMS). The carrier gases of the Agilent 7500i were He and Ar and the Plasma power was 1310 W.

Analytical Results
The trace element results of the pyrites and chalcopyrites are shown in Appendices A and B, respectively. Logarithmic plots of the trace element concentrations of the pyrites are shown in Figure 5. The Main Golpu porphyry had 39 pyrite analyses where 13 of them were below the detection limit (<11-13 ppb Au) and the highest Au content was 23,700 ppb. The Golpu West porphyry had 40 Au pyrite analyses; 25 of these analyses were below the detection limit (<9-18 ppb Au) and the highest Au content was 4910 ppb. The Golpu North porphyry had 50 Au analyses of pyrite where 21 of them were below the detection limit (<8-24 ppb Au) and the highest Au content was 5580 ppb. The Nambonga North porphyry had 36 Au analyses of pyrite; seven of these analyses were below the detection limit (<1-12 ppb Au) and the highest Au content was 143,000 ppb.
Similarly, Figure 6 shows the logarithmic plots of the trace element concentrations of the chalcopyrites. The Main Golpu porphyry had 29 Au analyses of chalcopyrite; 17 of the analyses were below the detection limit (<20-35 ppb Au); the Au content was up to 530 ppb. The Golpu West porphyry had 28 Au analyses and 16 of them were below the detection limit (<19-43 ppb Au); the Au content was up to 160 ppb. The Golpu North porphyry had 19 Au analyses and eight of the analyses were below the detection limits (<20-34 ppb Au); the Au content was up to 1340 ppb. The Nambonga North porphyry had 38 Au analyses and 26 of them were below the detection limit (13 to 28 ppb Au); the highest Au content was 2300 ppb.
The Au in pyrite and chalcopyrite, together with the bulk-rock Au content, was evaluated using cumulative frequency plots and a geometric mean that was derived by taking into account those data below the analytical detection limit (Figure 7). The data revealed an approximate log-normal distribution of the Au concentration in the Main Golpu porphyry, Golpu West porphyry, Golpu North porphyry and Nambonga North porphyry.   The Main Golpu porphyry had a mean of 300 ppb Au (27-3000 ppb; 1 σ range variation) for pyrite (n = 26) and a mean of 48 ppb Au (34-90 ppb) for chalcopyrite (n = 12). The bulk-rock Au concentration for the Main Golpu porphyry (n = 7) had a mean of 550 ppb Au (200-1600 ppb; 1 σ range variation). The Golpu West porphyry had a mean of 200 ppb Au (28-1600 ppb; 1 σ range variation) for pyrite (n = 15) and 60 ppb Au (44-140 ppb; 1 σ range variation) for chalcopyrite (n = 12). The bulk-rock Au concentration for the Golpu West porphyry (n = 19) had a mean of 550 ppb Au (170-1800 ppb; 1 σ range variation). The Golpu North porphyry had a mean of 82 ppb Au (17-450 ppb; 1 σ range variation) for pyrite (n = 29) and 140 ppb Au (32-550 ppb) for chalcopyrite (n = 11). The bulk-rock Au concentration for the Golpu North porphyry (n = 13) had a mean of 160 ppb Au (60-350 ppb; 1 σ range variation). The Nambonga North porphyry had a mean of 1700 ppb Au (75-30,000 ppb; 1 σ range variation) for pyrite (n = 29) and 50 ppb Au (10-180 ppb) for chalcopyrite (n = 12). Its bulk-rock (n = 6) Au concentration had a mean of 700 ppb Au (200-2500 ppb; 1 σ range variation). The distribution shows that Au in the pyrites from the Nambonga North porphyry is higher than those from the Main Golpu porphyry, Golpu West porphyry and Golpu North porphyry.
The approximate log-normal mean Au concentrations in pyrite and the approximate log-normal mean Au concentration in the bulk-rock for each of the porphyries are plotted in Figure 8A. Note that these means represent the geometric means. The mean Au concentration in both the pyrite and bulk rock is the highest for the Nambonga North porphyry whereas the Golpu North porphyry has the least mean Au contents. The Main Golpu porphyry has slightly higher mean Au in the pyrite than Golpu West porphyry while the mean Au in the bulk-rock of both porphyries is the same. The approximate log-normal mean Au concentration in chalcopyrite and the approximate log-normal mean Au concentration in the bulk-rock for each of the porphyries are plotted in Figure 8B. It is apparent that the Golpu North porphyry has the highest mean Au in chalcopyrite compared to the other porphyries, whereas the Nambonga North porphyry has the highest mean Au in the pyrite. A close inspection of the pyrites at the Nambonga North porphyry deposit reveals two distinct generations. The two pyrite generations have similar vein mineralogy but discrete morphological textures which also correspond to the differences in their Au contents.
The first pyrite type, found in Sample N1, is from the diamond drill hole WR272 and is found at a depth of 396.00 m. The sample is a diorite that is pervasively altered by K-feldspar-sericite-chlorite mineral assemblage cross-cut by sheeted quartz veins and quartz stock-work, both of which are superimposed by magnetite-chalcopyrite-pyrite as veins, dissemination and fracture-fill. Hence, the first pyrite type is present as infill in a 2 cm-wide quartz ± magnetite ± pyrite ± chalcopyrite vein ( Figure 9). It forms granoblastic to idiomorphic textures which precipitated from the magmatic-hydrothermal fluids.
The second pyrite type in Sample N10 is from the same diamond drill hole but was collected at a depth of 294.30 m, a shallower depth than N1. The intrusive type, alteration minerals and sequence of alteration-mineralisation are the same as those of Sample N1. The second type of pyrite also occurs in 1 to 2 mm-wide quartz-magnetite ± chalcopyrite ± pyrite veins ( Figure 10). It is platy, tabular and lamellar in shape and is formed as pseudomorphic replacement of hematite, i.e., by a sulphidation reaction that is similar to the replacement of magnetite by pyrite. The pyrites in Sample N10 are variably replaced by chalcopyrite along their grain boundaries.
The coexistence of both N1 and N10 pyrites suggests several stages of deposition, and the pseudomorphic pyrite (N10) seems to be the most favourable for Au fixation. Figure 11 shows the logarithmic plots of the trace element concentrations of the pyrites from the Nambonga North porphyry; bulk-rock Au is used as an arbitrary reference. A total of 31 pyrites analysed from five samples from the drill hole WR272 were assessed.  For a quantitative evaluation of the Au contribution in pyrite to the bulk-rock Au content in each porphyry system, the maximum possible amount of pyrite in each system was calculated using the sulphur (S) concentration in the bulk-rock analyses multiplied by the stoichiometric molar ratio in pyrite of FeS 2 /S 2 . This calculation gives only an upper limit, as there are other but less abundant S-bearing minerals in the system (anhydrite, chalcopyrite, marcasite, sphalerite, molybdenite and galena). Overall, there seems to be no correlation between the calculated Au concentration in the pyrite and the Au content in the bulk-rock, as shown in Figure 12.  The pyrite from the Main Golpu porphyry provides about 16 ppb Au to the bulk-rock (5.34 wt% pyrite × 550 ppb Au in pyrite) which is about 3% of the total Au in the bulk-rock. The pyrite from the Golpu West porphyry supplies about 6.7 ppb Au (3.35 wt% pyrite × 550 ppb Au in pyrite) which is about 1% of the total Au in the bulk-rock. The pyrite from the Golpu North porphyry contributes about 6.5 ppb Au (7.93 wt% pyrite × 160 ppb Au in pyrite) which is about 4% of the total Au in the bulk-rock.
The pyrite from Nambonga North porphyry contributes about 63 ppb Au to the bulkrock (3.72 wt% pyrite × 1700 ppb Au in pyrite) with a mean of 700 ppb Au. This is about 9% of the total Au in the bulk rock.
This mass balance shows that more than 90% of the Au in each of the porphyries comes from elsewhere, apart from pyrite. Considering this and taking into account the low Au concentration in the chalcopyrite, there must be a third Au-carrier in the porphyries. The third Au-carrier is most likely sub-microscopic native Au observed as aggregates and inclusions.

Implications of the Trace Element Composition of Pyrite
The trace elements of pyrite from the Main Golpu, Golpu West and Golpu North porphyries display no significant correlation except where a high content of Au exists. High Au contents are associated with high Cu (e.g., Sample G22 from Main Golpu: WR331, 1214.20 m). This pyrite is associated with the potassic alteration. Locally, a high Au content is also associated with high As, Ag, Te and Bi contents (e.g., Sample G17 from Main Golpu: WR331, 1481.00 m). This element association may be pyrite that was deposited by high sulphidation fluids. At the Golpu West porphyry, high Au contents correlate positively with high Cu and Ag contents; Bi content is locally high too (e.g., Sample G6: WR327A, 756.40 m). Such a correlation may be indicative of pyrite deposited during phyllic alteration. There are instances where a high Au content correlates well with high Cu, As, Ag, Te and Bi contents (e.g., Sample GW15: WR321, 941.20 m). This relationship identifies pyrite deposited by high sulphidation fluids that remobilised and redeposited elements deposited during the porphyry stage. Sample GN23 (WR351, 957.00 m) from Golpu North reveals a positive correlation between high Au, As, Sb, Te and Bi contents with local high Cu contents, suggesting pyrite deposition by high sulphidation fluids. In all, fluids associated with potassic, phyllic and high sulphidation alteration types are differentiated from the local overprint of porphyry-related mineralisation by high sulphidation fluids.
Nambonga North pyrites show a positive relationship between Au, Cu, As and Ag contents. High Au contents correspond to high Cu, As and Ag contents (e.g., Sample N10: WR272, 294.00 m). In addition, high contents of Sb, Te and Bi are associated with the elements mentioned above (e.g., Sample N11: WR272, 376.30 m). This relationship is considered to be associated with pyrite deposited during phyllic alteration, implying that there is a high potential of remobilising and redepositing elements deposited during the Potassic-associated mineralisation. It can also be inferred that the Au and Cu deposited immediately after phyllic alteration were transported as bisulphide complexes.
The pyrite generations at Nambonga North show that high Au grades are associated with the phyllic alteration assemblage and a particularly negative correlation of the Au is displayed by As, Bi and Zn. Some studies show that Au is spatially associated with the enrichment of local As in pyrite grains [22][23][24]. Research on the As and Au intake of the arsenic-rich pyrite, marcasite and arseno-pyrite in sediment-hosted Au deposits at Carlin in Nevada (USA) revealed that Au is removed from the ore fluids by chemosorption at As-rich and Fe-deficient surface sites and is incorporated into the solids in a metastable solid solution [25]. However, this is not usually common since the Au hosted in the pyrite from Yangzhaiyu lode Au deposit in China is associated with Te and not with As [26]. As noted above, it is evident here that the As-deficient pyrite associated with the phyllic alteration at Nambonga North is enriched in Au. Hence, this is indicative of pyrite metal speciation and partitioning e.g., ref. [27] where in this case the granoblastic textured pyrite detected lower Au and deficient As whereas the platy, prismatic and lamellar textured pseudomorphic pyrite has a higher Au content and is rich in As.

Implications of Trace Element Composition of Chalcopyrite
Main Golpu chalcopyrites have relatively low Au contents and elevated Ag and Te contents, including Bi found locally (e.g., Sample G21: WR331, 1214.20 m). It is most likely that these chalcopyrites were deposited during potassic alteration. Local High Au content correlates with high As, Ag, Sb, Te and Bi contents, indicating a possible deposition of the chalcopyrite during phyllic and/or chlorite-sericite alteration during boiling and metal precipitation. Golpu West and Golpu North have trace element compositions that are similar to Main Golpu. Therefore, the interpretation of Main Golpu can be applied to both Golpu West and Golpu North.
An analysis of Nambonga North chalcopyrites shows that there are no clear relations between the elements; although few elements appear to correlate positively, there is no consistency. Locally high Au content is positively correlated with high As, Ag, Te and Bi contents. This mineral association can be interpreted as being related to the deposition of chalcopyrite during phyllic and/or chlorite-sericite alteration types.
A quantitative evaluation of the Au contribution from the pyrite to the bulk-rock Au content in each porphyry system shows that 90% comes from elsewhere. This means that Au is in the other sulphides such as bornite and is present as native Au. Figure 13 shows the graph of the Au versus Cu contents of large porphyry deposits in the Circum-Pacific region. The Cu content in the Golpu porphyries plots below the Grasberg deposit in Indonesia and above the Bingham deposit in USA, Batu-Hijau in Indonesia and Panguna and Ok Tedi in PNG. The Au content of the Golpu porphyry deposit is also slightly higher than Ok Tedi and Batu-Hijau. However, the Nambonga North Au is higher than all three deposits including the Cadia Hill deposit in Australia. As alluded to earlier, the high Au content was deposited during the phyllic alteration assemblage and/or the chlorite-sericite alteration. Sulphur isotope studies confirmed that the high Cl contents corresponded to fluid inclusions with high salinity brines ranging between 40 and 74 wt%, with the NaCl equivalents trapped in quartz formed during potassic alteration at 423 to 600 • C [28]. While the presence of chlorine implies that the magmatic-hydrothermal fluids depositing secondary biotites and phlogopites during the potassic alteration were enriched in chloride ions and saline, the magmatic-hydrothermal fluids also transported Au as chloride complexes.
Similar to the study undertaken by [27] and as the hydrothermal pyrite suggests, Cu and Au occur as micro-to nano-sized particles of chalcopyrite and native Au (or Au tellurides) while As occurs as a structurally bonded element in pyrite. Hence, this clearly confirms that pyrite controls metal speciation and partitioning during porphyry Cu mineralisation [27].

Potassic Alteration
Potassic alteration is comprised of secondary biotite-quartz-magnetite ± K-feldspar ± anhydrite mineral assemblage. Secondary biotite is paragenetically the earliest mineral to be deposited, followed by quartz veinlets, magnetite ± quartz and anhydrite ± quartz. Quartz gradually forms dense stock-work or occurs as sheeted veins with K-feldspar selvages. Most quartz veins correspond to A-, B-and D-type veins e.g., ref. [29], which are characterised by secondary biotite-quartz-magnetite-anhydrite-chalcopyrite-pyrite, quartz-molybdenite-pyrite ± anhydrite and quartz-sericite-chalcopyrite-pyrite mineral assemblages, respectively. Some of the ear-liest quartz veins are pale gray to grayish white, sinuous, associated with abundant magnetite, local plagioclase halos, barren of infill minerals and are cross-cut by late mineralised quartz veins; the former quartz veins could be a variant of the A-type vein or correspond to the M-type vein e.g., ref. [30].

Propylitic Alteration
The propylitic alteration was formed at about the same time as the potassic alteration via heat transfer but in a peripheral area [2]. This alteration type is subdivided into inner propylitic and outer propylitic based on the predominance of actinolite-epidote and chlorite-carbonate (mostly calcite), respectively, as index minerals [31]. The mineral assemblage of porphyry-related propylitic alteration is similar to that of sub-greenschist to greenschist metamorphic facies of the OSMC, making differentiation very difficult.

Chlorite-Sericite Alteration
Paragenetic work suggests that chlorite-sericite alteration may occur or is transitional between potassic and phyllic alteration [32]. It is considered to have formed after potassic alteration but before phyllic alteration. Chlorite replaces mafic minerals and sericite replaces plagioclase. Both mineral species are disseminated and fracture controlled. The condition and sequence of the formation of chlorite-sericite alteration in the evolution of the porphyry Cu system is similar to that described in [33].

Phyllic Alteration
The volatiles released during retrograde boiling (induced by the fracturing of the carapace or the cupola of the cooling porphyry) subsequently condensed upon interaction with groundwater. The condensated fluid attained its acidity due to the presence of HCl ± SO 2 gas(es) and formed phyllic alteration as it migrated downwards, overprinting the potassic and inner propylitic alteration zones. The phyllic alteration is comprised of quartz, sericite and pyrite, and is accompanied by varying amounts of chlorite and illite. The occurrence of sericite indicates a pH of 4-5.

Relation of the Wafi High Sulphidation Au Deposits to the Golpu Porphyry Cu-Au System
Highly acidic, oxidised fluids generated after the formation of the potassic alteration produced a downward zonation from vuggy silica/quartz, to advanced argillic (vuggy silica-alunite ± pyrophyllite), through intermediate argillic (dickite-kaolinite ± sericite ± illite) and to argillic (smectite-illite-chlorite ± carbonate) zones at deeper depths [34]. This zonation indicates an increasing pH, a decreasing temperature and increasing wallrock dilution as the acidic fluids condensed from rising magmatic volatiles that migrated downwards via fractures, fracture systems or faults [2]. Figure 14 is a diagram showing the paragenesis of the porphyry-related alteration-mineralisation in the Golpu and Nambonga North porphyry Cu-Au deposits.
Deposition of Au occurred after the high sulphidation alteration. Most (refractory) Au resides in arsenic-rich pyrite and minor enargite-luzonite in the intermediate argillic zone e.g., ref. [34]. Such an occurrence is unusual because most of the Au is in the central vuggy silica of similar deposits such as Lepanto in the Philippines e.g., ref. [35], Nena in PNG e.g., ref. [36] and Nansatsu in southern Kyushu Island, Japan e.g., ref. [37]. Therefore, Au was transported by magmatic volatiles exsolved via boiling to permit deposition due to condensation and the chemical reaction with previously acid altered wallrocks in the intermediate argillic zone cf., [38].
Sulphur isotope signatures of chalcopyrite and chalcopyrite-pyrite at Golpu range from 0.23 to 0.60 δ 34 S; the sulphur isotope signatures of anhydrite and chalcopyrite-pyrite at Nambonga North are 10.06-11.74 δ 34 S and 2.16 δ 34 S, respectively [28]. The range of sulphur isotope signatures approaches zero, which implies that sulphur and mineralising fluids were derived from a magmatic source. It also means that metals such as Cu and Au are of magmatic origin too e.g., [39]. The anhydrite-chalcopyrite and anhydrite-pyrite mineral pairs from Nambonga North are considered to be in chemical equilibrium and estimate sulphur isotope fractionation temperatures of 551-634 ± 40 • C and 522-603 ± 40 • C, respectively [28], estimating the temperature of Cu-Au mineralisation at Nambonga North. Under such a high temperature range (i.e., >350 • C), Cu and Au at Nambonga North and by inference Golpu, would be mostly transported as chloride complexes such as AuCl 2− [39]. However, bisulphide complexes may be the main carriers of Au and Cu during sericite-chlorite and quartz-sericite-pyrite alteration events if the temperatures were less than 500-300 • C. This means the potassic-related Au and Cu would be remobilised, transported as a bisulphide complex (e.g., Au(HS) 2− , [39]) and redeposited.

Conclusions
The porphyry-related alteration and mineralisation are centred at the Golpu Porphyries and Nambonga North Porphyry, which are separate porphyry systems but cooling intrusions responsible for supplying magmatic-hydrothermal fluids may be sourced from the same, larger magma chamber at depth. Although the characteristics of the alteration-mineralisation are similar, there are some major differences. The Golpu porphyry deposit has a diatreme emplaced before the expulsion of the highly oxidised, magmatichydrothermal fluids and late high sulphidation epithermal system. In contrast, Nambonga North has no diatreme and high sulphidation epithermal alteration-mineralisation.
Copper-gold mineralisation at both the porphyry centres is characterised by pyritechalcopyrite-bornite-gold-molybdenite mineral assemblage that is hosted in sheeted quartz veins and fractures. Pyrite and chalcopyrite are the main ore minerals. Native Au also occurs as minute inclusions in chalcopyrite, pyrite and bornite. Most ore mineralisation is closely associated with and occurred after the potassic alteration. Some of the Au and Cu deposited after the potassic alteration were remobilised and re-deposited as components of the mineralisation associated with the sericite-chlorite and/or phyllic alterations.
Petrographic studies and age dating [32] including the sulphur isotope studies [28] confirm that the Wafi high sulphidation epithermal Au deposit overprints and is spatially, temporally and genetically related to the Golpu porphyry system. Similar observations have been noted by previous investigators such as [18,20,33,40].                        Figure A2. LA-ICPMS trace element results of chalcopyrite (all results are in ppm).