Sulfation roasting has emerged as a promising alternative for lithium-first recovery from spent lithium-ion batteries, while reducing hydrometallurgical complexity. This study investigated the sulfation roasting of nickel–cobalt–manganese oxide (NCM) black mass to maximize lithium recovery while minimizing sulfuric acid consumption and thermal treatment time. The effects of preliminary calcination, sulfuric acid dosage, and roasting time were systematically investigated at 550 °C. Lithium and coextracted metals were quantified by inductively coupled plasma-optical emission spectrometry (ICP-OES), while phase transformations were analyzed using X-ray diffraction (XRD) and X-ray fluorescence (XRF). Statistical evaluation identified sulfuric acid dosage as the only factor significantly affecting lithium extraction. An integrated success index was developed to assess lithium recovery, contaminant dissolution, reagent consumption, and thermal treatment time. Under the optimum conditions (nLi:nH2SO4 = 1:0.95, without preliminary calcination, 45 min), lithium recovery reached 90.8
The rapid adoption of electric vehicles and energy-storage systems has generated a surge of spent lithium-ion batteries (LIBs), making the recovery of valuable metals (Li, Ni, Co, Mn) from cathode materials an urgent environmental and economic priority. A structured literature search covering 2000–2026 was used to identify peer-reviewed process data and industrial evidence. This review critically evaluates recent advances in resource recovery and direct regeneration of spent LIB cathodes, covering cascade utilization, pretreatment, pyrometallurgy, hydrometallurgy, emerging green technologies (deep eutectic solvents, ultrasonic intensification, reduction roasting), and direct-regeneration strategies (solid-state relithiation, dissolution–reprecipitation, sol–gel, electrochemical, and hydrothermal methods). The mechanisms, applicable cathode chemistries, process windows, advantages, and limitations of each approach are systematically compared. Conventional pyrometallurgical and hydrometallurgical routes are robust for mixed feedstocks but suffer from high energy demand, reagent consumption, wastewater generation, or poor lithium recovery. In contrast, direct regeneration and green coupled processes preserve cathode structure, reduce the carbon footprint, and enable closed-loop material cycles. The analysis also distinguishes the low-value lithium iron phosphate (LFP) case, for which value-preserving regeneration is especially important, from Ni/Co-rich nickel cobalt manganese (NCM) and lithium cobalt oxide (LCO) streams that can support profitable hydrometallurgical recovery. Direct regeneration remains predominantly at laboratory or pilot scale; industrial deployment requires accurate sorting, degradation diagnosis, impurity control, continuous-process design, and product-consistency evaluation. This review provides a comprehensive reference for researchers and engineers in metal recycling and battery material regeneration. This review critically examines technologies for recovering valuable metals (Li, Ni, Co, Mn) from spent lithium-ion battery cathodes, including pyrometallurgy, hydrometallurgy, emerging green routes (deep eutectic solvents, ultrasonication), and direct-regeneration methods (solid-state relithiation, hydrothermal repair). Direct regeneration offers the highest potential for value retention and carbon footprint reduction. Future recycling requires integrated sorting, diagnosis, and standardization for closed-loop sustainability.
Coke is one of the key raw materials used in blast furnaces (BF), serving as both a fuel and a reductant, with a particle size greater than 25 mm, which is essential for the smooth operation of the BF. As a result, the effective utilization of smaller coke particles (< 25 mm), referred to as nut coke, has become an important area of focus. In this study, the impact of nut coke particle size and its addition amount on the reduction behavior of pellets was investigated. The goal is to provide theoretical support for the use of nut coke in blast furnaces, thereby optimizing the utilization of coke resources. The experimental results show that the optimal mixture ratio of pellet–nut coke is 6
Gas atomization (GA), which is widely used for producing powders for laser powder bed fusion (LPBF), inherently generates a broad particle size distribution that includes fine particles outside the optimal size range for printing. Although particles smaller than 20 µm are unsuitable for LPBF due to limited flowability, they retain the chemical integrity imparted by the GA process and are typically discarded. This study evaluates ultrasonic atomization (UA) as a complementary recycling route for converting these fine gas-atomized 316L stainless steel powders into LPBF-compatible feedstock. Fine GA particles (< 20 µm) were consolidated into dense preforms using a press-and-sinter powder metallurgy route and subsequently re-atomized by UA, after which the powder was sieved to isolate a 20–75 µm fraction suitable for LPBF. LPBF parameters optimized using a GA reference powder were directly applied to UA powder to isolate the influence of powder characteristics. Parts produced from UA powder achieved density of 7.92 ± 0.01 g cm−3, comparable to that obtained using GA powder (7.90 ± 0.02 g cm−3), while exhibiting improved transverse rupture strength and hardness, attributed to enhanced powder morphology, flowability, and packing behavior. These results demonstrate that UA enables the effective upcycling of fine GA powder fractions into LPBF-suitable feedstock, improving powder utilization efficiency and reducing material losses associated with particle size classification in additive manufacturing.
Accurate prediction of the burn-through point (BTP) is essential for optimizing iron ore sintering processes; however, strong nonstationarity in industrial data limits the performance of conventional models. To address this issue, a Reversible Instance Normalization-enhanced Long Short-Term Memory Physics-Informed (RevIn–LSTM–PI) model is proposed. Key process variables are selected via Pearson correlation analysis, while RevIn alleviates distribution shifts through instance-wise normalization and reversible denormalization. Temporal dependencies are captured using an LSTM network, and physics-informed constraints derived from gas–solid thermal equilibrium equations are embedded into the loss function to ensure mechanistic consistency. Experimental results on industrial sintering data demonstrate that the proposed model achieves superior performance, with an root mean squared error (RMSE) of 0.0024, mean absolute error (MAE) of 0.0019, and R2 of 0.97. Importantly, an operational validation shows that improved BTP prediction significantly enhances energy efficiency by reducing excess fuel consumption from 0.121 to 0.004 kg/t-sinter, corresponding to a 96.7
