Mohammad Reza Hosseinzadeh

A state-of-the-art, robust classification model using artificial intelligence (AI) algorithms was developed in this study to estimate the gold ore classes based on trace elements measured by inductively coupled plasma (ICP). This approach involves the utilisation of stand-alone machine learning (ML) algorithms integrated within a committee machine (CM) framework. To this end, 19 trace elements including arsenic, bismuth, cadmium, cobalt, chromium, copper, iron, mercury, manganese, molybdenum, nickel, lead, antimony, selenium, tin, strontium, thorium, uranium and zinc acquired from eight drill holes in the Sari-Gunay gold-polymetallic mine (SGGM), Iran, were used as the inputs and three gold grade classes (ore, low-grade ore, and waste) were specified as outputs classes. The classes were defined based on their gold grades and mineable zones, using threshold values of > 1 g/t, 0.5-1 g/t, and < 0.5 g/t for ore, low-grade ore and waste, respectively. Eight stand-alone ML-based classifiers, including the back-propagation neural network (BP-NN), Takagi and Sugeno fuzzy inference system (TS-FIS), adaptive boosting (AdaBoost), GentleBoost, LogitBoost, random under-sampling boosting (RUSBoost), extreme gradient boosting (XGBoost) and adaptive neuro-fuzzy inference system (ANFIS) were used to perform the initial classification. The accuracy, precision and recall metrics were employed to evaluate the performance of the algorithms by comparing predicted results with reference data derived from ICP analysis and geological studies, providing a consistent reference benchmark. The parameters within each algorithm were tuned using the parameter-tuning method, aiming to identify the optimal model. The AdaBoost algorithm outperformed the other stand-alone algorithms. To further improve the stand-alone-derived classification outcomes and integrate the algorithms into a unified algorithm, genetic (GA) and simulated annealing (SA) optimisation algorithms were configured in the framework of the CM. Parameter-tuning of both optimisation algorithms produced multiple distinct models to determine the optimal weights and the best-performing model. The CM with SA (CMSA) outperformed the GA optimiser. Employing CMSA resulted in a 7.28% improvement in overall accuracy compared with the average performance of the stand-alone algorithms.

Chalcedony forms across a wide area in eastern Iran, particularly in the Sarbisheh district near Birjand city, with notable occurrences in the Qazdez-Bahamarz region. Here, cryptocrystalline quartz is found in hydrothermal veins and veinlets within volcanic rocks hosted by carbonate and intermediate igneous formations. Chalcedony samples of various colours—black, purple, green, blue, lavender, grey, lemon yellow and white—were analysed using diverse techniques. These chalcedony samples display fibrous and granular textures, comprising microcrystalline quartz, cryptocrystalline moganite, and opal-CT and opal-C interlayers. Elements that affect the colouration of chalcedony include iron (producing red and yellow tones), chromium (green), manganese (black, blue and grey patterns), nickel (purple) and copper (also purple). Altered carbonate host rocks are enriched in Fe, Cu, Zn and Cd, with Al and Mn depletion. Stable isotope analyses show δ ¹⁸ O values in agates range from +14.9‰ to +25.5‰, whereas δ ¹⁸ O and δ ¹³ C values in carbonate minerals and chalcedony range from +14.1‰ to +24.8‰ and –5.7‰ to +0.7‰, indicating agate formation from mantle-derived hydrothermal fluids mixed with meteoric waters. Raman spectroscopy detected moganite, α-quartz, goethite, aragonite and illite in agate interlayers. Analyses by field-emission scanning electron microscopy (FE-SEM) and energy dispersive X-ray spectroscopy (EDS) revealed minerals of Fe, Cr, Ni, Ti and Sn. X-ray diffraction confirmed chalcedony, moganite and opal-CT, whereas EPR spectroscopy showed strong magnetic backgrounds from goethite and silicon-vacancy centres formed by radioactive decay of U, Th and their byproducts.

The Sar-e-Yazd Cu (Mo) deposit is located in the Naein-Jiroft magmatic belt or Kerman belt. It is in the southern part of the Cenozoic Saveh-Nain-Jiroft magmatic belt of Iran. This deposit is spatially associated with Oligocene shallow porphyritic porphyries intruded into dacite porphyry to porphyritic quartzdiorite and middle Eocene volcanoclastic rocks. Hypogene ore minerals include chalcopyrite, magnetite, pyrrhotite, molybdenite, bornite and pyrite occurring as veinlets and disseminations mostly in the both intrusive rocks. Hydrothermal alteration from inside to outside is marked by potassic, carbonate alteration, an extensive phyllic assemblage and irregular zones of propylitic (actinolite, epidote and chlorite sub-zone) and clay assemblages with no consistent zoning patterns. The study of faults in the area reveals that, in general, they align with the pre-existing joints in the structure of the Sani Abad syncline. These faults are attributed to ancient fractures within the right-lateral shear system between the Anar and Dehshir right-striking faults. Three sets of faults are recognized in the eastern Mehriz area (Sar-e-Yazd region): 1– longitudinal direction joints (or tension faults), along with early intrusions and the first group of dykes with WNW–ESE extension, are attributed to old T2 fractures in the shear system. However, the second group of later dykes, often intruded in a N–S direction, is attributed to tensile fractures in the T1 shear system, 2 – cross direction joints (or expansion faults) and 3 – diagonal direction joints (or shear faults; the late intrusive, branches of the epithermal Cu-Fe and the Pb-Zn apophyse intruding aligned with these faults direction. The veins shows two main trends: a- tension split is related to T1 and tensile fractures (epithermal Cu-Fe veins) and b- the pennant vein pattern is related to R/R’ diagonal fractures. Fractures with the NW–SE trend (R), which are mostly filled by iron oxides, are younger and don’t have an effective role in mineralization and it seems that the fractures with the R’ shear system have been effective in the placement of intrusive. The total average of the highest metallic mineralization is enriched in supergene zone at each boreholes (BH): BH01, BH03, BH04 and BH23. Cu: 3881 ppm, Mo: 43.7 ppm, Pb: 13390 ppm, Zn: 18250 ppm and Ag: 6.7 ppm in the 25 to 58 m depth. Fluid inclusion studies indicate a wide range of homogenization temperatures in the veinlets and phyllic porphyritic quartzdiorite alteration, ranging from 136°C to 542°C, with salinities of 4–62 wt.% NaCl equivalent. The coexistence of vapour-rich inclusions (F3 type) with high salinity and liquid-rich inclusions (F2 type) with similar Th values provides strong evidence for boiling. Parts of the copper content in the solution could be deposited as a result of boiling processes as chalcopyrite which was the major mineral at the Sar-e-Yazd deposit (similar to porphyry copper deposits). The analysis of borehole samples indicates a moderate correlation between Cu and Mo, suggesting porphyry Cu (Mo) deposits in Sar-e-Yazd. Cu also shows a strong correlation with Pb, Zn, and Ag, which are key indicator elements of these deposits. Finally based on the distribution and zoning of the main metal ores, the porphyry mineralization of Sar-e-Yazd shows a significant resemblance to the Cu-Mo mineralization in the Mineral Park deposit in Arizona.

The Sungun porphyry ore deposit is located in Eastern Azarbaijan province, Northwestern Iran. The oldest intrusive pulse in the region is a quartz-monzonite pluton, which hosts the porphyry copper-molybdenum mineralization. The Sungun Copper Mine includes the mineralized Sungun porphyry as well as six groups of cross-cutting and lithologically distinct post-mineralization dykes. The composition of these dykes ranges from quartz diorite, gabbro, diorite, dacite, lamprophyre, and microdiorite. Quartz diorite and dacite dykes are the oldest and youngest dykes, respectively. Based on their cross-cutting relationships, the composition of the dykes tend to become more primitive through time. The dykes strike Northwest–Southeast with Southwest dip, sub-parallel to the reverse faults within the deposit area. The lamprophyric dykes range from phonotephrite, to trachybasalt, tephrite, and basanite. The quartz-monzonite porphyry (SP) and the post-mineralization dykes (DK1-DK3) have clear and distinct negative anomalies of Ti, Zr, P, Pr, Ce, and Nb, as well as positive anomalies of Cs, U, K, Pb, and Nd with respect to primitive mantle. Microdioritic dykes (MDI) show depletion of Ti, Nb, P, Ta, Th, Yb, and Zr, and enrichment of Cs, Ba, U, Pb, Nd. The similarities in trace element abundances and patterns in the porphyry and post-mineralization calc-alkaline dykes implies a single source and fractional crystallization as the main mechanism controlling magmatic evolution in a collisional environment. Lamprophyric dykes have enrichment of LREE and LILE and depletion of HREE and HFSE such as Ti, Nb, and Ta. The parent magma of the lamprophyric dykes (LAM) was likely derived by low degrees of melting of a garnet lherzolite mantle peridotite. The ⁸⁷Sr/⁸⁶Sr and ¹⁴³Nd/¹⁴⁴Nd ratios range from 0.704617 to 0.706464 and from 0.512648 to 0.512773 for the dykes suggesting that the parental magmas came from a progressively more enriched mantle. Isotope ratios of ⁸⁷Sr/⁸⁶Sr and ¹⁴³Nd/¹⁴⁴Nd support a cogenetic relationship of porphyry and calc-alkaline dykes, except for the microdiorite ones. A common primary melt underwent gravity differentiation in a deep magmatic chamber to form a dioritic magma. This subsequently migrated to shallower levels to evolve further and feed individual dyke groups into the Sungun porphyry.

The Shah-Ali-Beiglou epithermal base metal-silver deposit is located in the Tarom-Hashjin metallogenic province (THMP) in northwestern Iran. This deposit is hosted by quartz monzonite dikes of Oligocene age and surrounded by andesite to trachyandesite volcanic and volcaniclastic rocks of Eocene age. The subvolcanic rocks in the study area vary in composition from quartz-monzonite to monzonite and have metaluminous, calc-alkaline to shoshonitic affinity. These rocks have I-type geochemical characteristic and are related to post-collisional tectonic setting. The mineralization occurs as NE-SW and E-W-trending brecciated veins controlled by strike-slip and normal faults, which are associated to the Late Oligocene compressional regime. The mineral paragenesis of the vein mineralization is subdivided into pre-ore stage, ore stage, post-ore stage, and supergene stage. Pre-ore stage is dominated by quartz, sericite, and subhedral to anhedral pyrite as disseminated form. Ore-stage is represented by quartz, sphalerite (from 0.1 mol % to 4 mol % FeS), galena, chalcopyrite, tetrahedrite-tennantite, minor seligmannite and enargite, as vein-veinlet, cement and clast breccias. Post-ore stage is defined by deposition of quartz and carbonate along with minor barite, and supergene stage is characterized by bornite, chalcocite, covellite, hematite, goethite, and jarosite. The ore mineralization is associated with the silicic alteration. The styles of alteration are silicic, carbonate, sericitic, chloritic, and propylitic. Fluid inclusions in sphalerite have a wide range of salinities between 0.35 wt % and 21.4 wt % NaCl equivalent and homogenization temperatures range from 123 to 320 °C. The isotopic values of sulfides vary from 2.8‰ to 6.7‰ suggesting a magmatic source for the sulfur. In the present study, based on geological setting, alteration style of the host and wall rocks, main textures, mineral assemblages, composition of ore minerals, and structural features, it is suggested that the mineralization in the Shah-Ali-Beiglou is similar to intermediate-sulfidation style of epithermal deposits.