The article presents the study results for the processing of industrial waste from copper production at copper smelter Kazakhstan. The samples taken were analysed with the help of X-ray fluorescence and phase analyses which showed that the composition of the studied raw materials was determined as silicate, oxidized, and copper was in a mixed form with a total content of 0.481%, including some in the form of sulfates and sulfides. Mineralogical analysis showed the presence of magnetite, hematite and martite, while copper was present in various mineralogical formations - from magnetic fractions mainly with very fine dusty micron dissemination to native copper and copper minerals. Beneficiation studies performed included flotation and gravity methods. As a result, a concentrate with a copper content of 9.35% was obtained during gravity beneficiation, and a concentrate with a copper content of up to 46% was obtained during flotation. Copper was extracted from beneficiated raw materials with a sulfuric acid leaching method in agitation mode. The solid residue analysis conducted after (cake) leaching also showed the content of the noble metal - gold at the level of 0.47 g/t enabling us to consider its extraction in the future as an additional valuable component.

This paper provides an overview of hydrometallurgical copper extraction studies in which liquid extraction technology has been used with four copper deposits of different compositions. The sulfuric acid consumption rate and copper extraction efficiency, which are dependent on the initial content and forms of calcium compounds and other impurities in ore samples, were calculated, and the results are presented herein. It was established that during the leaching process, silicate compounds of alkaline earth metals, in addition to calcium and magnesium carbonate compounds, would affect the levels of sulfuric acid consumption, thereby actively lowering the acidity of the environment. Moreover, these compounds could partially sorb copper ions from sulfuric acid leaching solutions. Thus, the analysis of waste ore samples showed that residual copper is mainly contained in the form of complex silicate complexes. The presence of divalent iron compounds in the composition from one of the deposits also allowed us to perform a biochemical leaching experiment with preliminary oxidation using an Acidithiobacillus ferrooxidans bacterial culture adapted to the ore composition. The use of this biochemical method in the copper leaching process resulted in a significant reduction in sulfuric acid consumption, by 40%, and a copper recovery rate of 87.2%.

Sustainable copper extraction presents significant challenges due to waste generation and environmental impacts, requiring advanced predictive methodologies to optimize production processes. This study addresses a gap in applying deep learning to forecast hydrometallurgical copper production by comparing six recurrent neural network architectures: Vanilla LSTM, Stacked LSTM, Bidirectional LSTM, GRU, CNN-LSTM, and Attention LSTM. Using time-series data from a full-scale industrial operation, we implemented a data augmentation approach to overcome data scarcity limitations. The models were evaluated through rigorous metrics and multi-step forecasting tests. The results demonstrated remarkable performance from five architectures, with Bidirectional LSTM and Attention LSTM achieving the highest accuracy (RMSE < 0.004, R² > 0.999, MAPE < 1%). These models successfully captured and reproduced complex cyclical patterns in copper mass production for up to 500 time steps ahead. The findings validate our data augmentation strategy for enabling models to learn complex known cyclical patterns from limited initial data and establish a promising foundation for implementing AI-driven predictive systems that can enhance process control, reduce waste, and advance sustainability in hydrometallurgical operations. However, these performance metrics reflect the models’ ability to reproduce patterns inherent in the augmented dataset derived from a single operational cycle; validation on entirely independent operational data is crucial for assessing true generalization and is a critical next step.

This article presents the results of cyanide leaching of gold-containing concentrate using the trichlorocyanuric acid (TCCA) oxidizer. Gold-containing concentrate was obtained from a gold tailings sample from a gold recovery factory (GRF) in one of the deposits of Kazakhstan that have not previously been studied for concentrability. According to X-ray phase analysis and energy dispersive spectrometry (DSM) data, the main compounds in the tailings sample under study are pyrite FeS₂, quartz SiO₂, calcite CaCO₃, albite NaAlSi₃O₈, muscovite KAl₂Si₃AlO₁₀(OH)₈, dolomite CaMg(CO₃)₂, and oxidized iron compounds. Microscopic studies of the concentrate have established the presence of ultrafine gold with sizes from Au 0.9 to 10.2 μm in pyrite. Obtaining the gold-containing concentrate with a gold content of 15.95 g/t is possible according to the enrichment scheme, which includes centrifugal separation, classification according to the fineness class −0.05 mm, additional grinding of hydrocyclone sands to a fineness of 90.0–95.0% of the class finer than 0.050 mm, and control centrifugal separation. Since pyrite in technogenic raw materials is the main gold-containing mineral, this paper presents studies on the oxidizability of pyrite with the TCCA oxidizer. The results of studies on the oxidation of pyrite using the TCCA oxidizer show the products of its hydrolysis oxidize pyrite with the formation of various iron compounds on its surface. Pretreatment of gold-containing concentrate with oxidizer TCCA for 3 h before the cyanidation process (20 h) allows for an increase in the recovery of gold in the solution by 5.8%.

The article discusses the interaction in amino acid–gold cyanide systems using amino acids of different structures. The formation of complex compounds of gold cyanide with amino acids with the participation of carboxyl and amino groups is shown. A relationship has been established between the formation of gold complexes with amino acids and the degree of its extraction in the process of leaching from low-grade ore with sodium cyanide together with amino acids: the higher the degree of participation of the amino group in the formation of the complex, i.e., covalent bond, the more pronounced the effect of the amino acid on the degree of gold leaching. The contribution to the formation of the complex of the carboxyl group (ionic bond) and the amino group (covalent donor–acceptor) can be assessed by the intensity of the band at a wave number of 1419 cm⁻¹ of the IR spectra of the systems: gold cyanide–amino acid. This approach makes it possible to predict the effect of amino acid structure on gold recovery during cyanide leaching based on IR spectra.

Gold-containing ore processing technology has been developed by using chemical, X-ray phase, mineralogical, X-ray fluorescence analyses. The results of experiments on the gold recovery from a concentrate with a particle size of 10 and 4 µm are presented, including options for direct cyanidation without preliminary oxygen oxidation, as well as with the use of preliminary oxidation of Trichloroisocyanuric acid (TCCA) and Sulfanol surfactant (SAS). It was found that the test sample contains 24.74 g/t Au; 11.66% S; 18.51% Fe. The content of sulfide minerals is 36.1%, the main sulfides are arsenopyrite (25.4%) and pyrite (10.7%). Microscopic examination of the combined gravity-flotation concentrate revealed ultrafine, fine and visible gold dust particles both in free form and in aggregations with waste rock. The particle size of Au in the product varies from 0.5 to 18.8 μm. 77.12% of gold is recovered from concentrate of 10 μm by direct cyanidation. “Sulfanol” was used to recover 83.71% of gold. Pre-treatment with TCCA yielded 85.61% of gold. Recovery of gold from the 4 μm concentrate was 75.38% for direct cyanidation, 78.79% for cyanidation with Sulfanol surfactant, 76.52% for cyanidation with TCCA.