Green steel: Hot shortness from inherited Copper contaminated scrap in steel recycling

The global steel sector, a cornerstone of modern infrastructure, is simultaneously a primary contributor to global CO₂ emissions, accounting for approximately 8% of the total. The most expedient route to its decarbonization lies in maximizing the share of steel produced via secondary synthesis in Electric Arc Furnaces (EAFs), which can reduce carbon emissions by up to 85% compared to the conventional blast furnace route, provided the electrical grid is decarbonized. This transition to a circular economy, however, is threatened by a chemical contamination problem inherent to the scrap feedstock.

Post-consumer steel scrap is a heterogeneous mixture, leading to the intrusion and gradual accumulation of residual "tramp" elements that are not removed during standard steelmaking operations. Among these, copper (Cu) is the most problematic. Its sources are ubiquitous in modern goods, particularly in electrical components, motors, and electronic waste. The ongoing electrification of transport exacerbates this issue; a typical battery electric vehicle contains over 80 kg of Cu, compared to roughly 20 kg in an internal combustion engine vehicle. Projections indicate that the average Cu concentration in the global scrap supply will surpass the maximum tolerable limit for most high-grade steel products by mid-century.

The primary detriment of elevated Cu (typically >0.1 wt%) is its propensity to induce hot shortness, a phenomenon that renders steel susceptible to intergranular cracking during hot deformation processes such as rolling and forging. This defect severely limits the use of contaminated scrap in the production of high-performance sheet steels, which require exceptional formability and mechanical properties for applications in automotive and construction sectors. This paper provides a comprehensive analysis of the hot shortness mechanism, critiques existing mitigation strategies, and elaborates on a nascent thermodynamic approach that enables the active removal of Cu, thereby safeguarding the path toward a sustainable, circular steel economy.

The global steel industry's necessary transition toward a circular economy (CE), predicated on increased utilization of scrap in the Electric Arc Furnace (EAF) route, is severely constrained by the accumulation of residual elements, primarily copper (Cu). Copper, being metallurgically insoluble and non-removable by conventional secondary refining, leads to hot shortness—a catastrophic loss of ductility during thermomechanical processing. This paper delineates the physical metallurgy and chemical kinetics governing hot shortness, specifically the preferential oxidation of iron and subsequent Cu-enrichment at the steel-scale interface, culminating in liquid Cu film penetration into austenite grain boundaries. Traditional mitigation strategies, such as Ni-alloying and microstructure engineering, are reviewed alongside emerging, resource-efficient solutions. Crucially, the concept of metallurgical Cu removal via hydrogen plasma-based smelting reduction at a critical oxygen potential is presented as a paradigm shift, enabling simultaneous decarbonization and high-grade scrap upcycling, which is essential for realizing a sustainable, high-performance steel future.

Large-scale recycling of steel scrap is confronted by a key metallurgical challenge: the accumulation of copper (Cu). Introduced via mixed post-consumer scrap, Cu cannot be removed by conventional oxidative refining and leads to a severe processing defect known as hot shortness—a dramatic loss of ductility at elevated temperatures (>1000 °C) that manifests as surface cracking during hot rolling. This paper delineates the mechanistic underpinnings of hot shortness, rooted in the selective oxidation of iron and the subsequent enrichment and liquefaction of Cu at austenite grain boundaries. Current mitigation strategies, including dilution with primary iron and nickel alloying, are assessed and found to be either unsustainable or cost-prohibitive. A transformative alternative is presented, based on recent experimental findings: the efficient evaporation of Cu from molten Fe–Cu–O systems at a critical oxygen concentration of approximately 22 wt%. This principle enables a novel process route wherein Cu-contaminated scrap is smelted with iron ore under a plasma atmosphere, actively removing Cu and subsequently reducing the purified melt to iron using hydrogen. This approach represents a paradigm shift from impurity dilution to targeted extraction, offering a viable pathway for upcycling contaminated scrap into high-performance, low-carbon steel.

The drive for rapid decarbonization has elevated steel scrap recycling as the most immediate and impactful pathway to reducing CO2​ emissions in the ferrous sector. However, the use of mixed scrap introduces tramp elements—those for which no viable removal route exists in conventional steelmaking—that accumulate over successive recycling loops. Copper (Cu) is the most notorious of these contaminants. Due to its thermodynamic nobility relative to iron (Fe) and its unlimited solubility in liquid iron, virtually all Cu entering the EAF remains in the final steel product, leading to concentrations exceeding 0.3 wt% in certain scrap-based products.

This accumulation presents a critical metallurgical barrier known as hot shortness, defined as the dramatic reduction in hot ductility and ultimate failure of the steel during hot rolling, forging, or other thermomechanical processes. The phenomenon limits the quality of recycled steel, necessitating the dilution of contaminated scrap with high-purity primary iron, a practice that defeats the purpose of the CE transition and perpetuates CO2​ emissions. A fundamental understanding of the Cu-induced damage mechanism and the development of cost-effective, high-efficiency removal or tolerance strategies are therefore paramount to the viability of green steel manufacturing.

Hot shortness is an interfacial failure phenomenon occurring at temperatures between 1000°C and 1200°C. Its genesis lies in a sequence of oxidation, segregation, and liquefaction events, in which the underlying key effect is the partitioning of Copper out of the surface oxide scale into the sub-surface bulk steels, forming a low melting Cu-rich phase. Thius means that hot shortness is an interfacial phenomenon driven by the fundamental difference in the Fe-O and Cu-O thermodynamic systems at high temperatures. The mechanism unfolds in three distinct, kinetic-driven stages during the reheating of the steel slab prior to hot deformation. The Preferential Oxidation and Interfacial Enrichment is here a key: When the steel containing residual Cu is heated in an oxidizing atmosphere (typically air in steel production mills) to temperatures required for hot working (∼1000∘C to 1300∘C), Iron (Fe) is oxidized preferentially due to its significantly lower thermodynamic stability compared to Cu: Fe(steel)+21​O2​(gas)→FeO(scale)(dominant reaction). The oxide scale, primarily wüstite (FeO) and magnetite (Fe3​O4​), forms and grows outward. Copper, being chemically  much more noble, is not oxidized and it is rejected (we also say it partitions) from the advancing steel/scale interface into the bulk. This process, governed by the mass transport kinetics of Fe2+ through the scale and the diffusion of Cu atoms in the steel matrix, results in the concentration of metallic Cu (or a Cu-enriched Fe-Cu solid solution) in a thin layer immediately beneath the scale. 

The Cu-enriched layer forms the basis for the embrittlement. The melting point of pure Cu is 1085∘C (∼1358 K). At typical hot-working temperatures, which often exceed this value, the Cu-rich phase liquefies.

The formation of this liquid Cu-rich phase is the immediate precursor to hot shortness. This liquid phase penetrates rapidly along the austenite (γ-Fe) grain boundaries due to a strong driving force dictated by the minimization of interfacial energy. The liquid Cu wetting the γ-Fe grain boundaries dramatically lowers the grain boundary cohesion energy (γgb​) of the steel. This energy criterion is often quantified by the ratio of the solid/liquid interfacial energy (γsl​) to the solid/solid grain boundary energy (γgb​), where penetration occurs when γsl​<21​γgb​.

Upon application of mechanical stress during hot deformation (e.g., rolling or forging), the steel's surface region, where the liquid Cu films reside, is subjected to tensile strains. Because the liquid films provide no resistance to shear and act as severe stress concentrators, the γ-Fe grain boundaries separate readily, leading to intergranular fracture and the formation of deep surface cracks.

The most severe hot shortness is observed in the temperature range ∼1000∘C to 1150∘C. This range corresponds to the optimal kinetic balance: high enough temperature for significant Cu enrichment and liquefaction, yet low enough that the rapid growth of a protective, solid oxide scale (which might mechanically contain the Cu layer) has not yet occurred. Tin (Sn), another tramp element, co-segregates with Cu, lowering the melting point of the liquid phase and exacerbating the penetration kinetics.

Below I illuminate a few detailed steps:

2.1. Selective Oxidation and Sub-Scale Enrichment
When a Cu-containing steel slab is reheated in an oxidizing atmosphere (e.g., air), an oxide scale composed predominantly of wüstite (FeO), magnetite (Fe₃O₄), and hematite (Fe₂O₃) forms on the surface. Iron, possessing a higher affinity for oxygen than copper, is selectively oxidized. The nobler Cu is rejected from the growing scale and accumulates in the metallic layer immediately beneath it. This process can elevate the local Cu concentration from a bulk level of, for instance, 0.3 wt% to levels significantly exceeding its solubility limit in austenite, which is approximately 8-9 wt% at 1100°C.

2.2. Liquid Phase Formation and Intergranular Penetration
Once the localized Cu concentration surpasses its solubility in the austenitic matrix and exceeds the melting point of the Cu-rich phase (1084°C for pure Cu), a liquid phase forms. This liquid does not remain dispersed but exhibits a high dihedral angle, favoring wetting and penetration of austenite grain boundaries. Under the tensile stresses applied during hot rolling, this liquid film is drawn deep into the material along the grain boundaries via capillary action. The presence of other tramp elements like Sn or Sb further depresses the liquidus temperature and intensifies segregation, aggravating the problem.

2.3. Embrittlement and Crack Initiation
The continuous liquid film at the grain boundaries drastically reduces intergranular cohesion. It acts as a preferential site for crack nucleation and provides an easy path for crack propagation. The result is a network of fine, intergranular surface cracks that can propagate into the material, compromising its structural integrity and rendering the final product unacceptable for stringent applications. Microstructural analysis of affected samples often reveals remnants of the Cu-rich phase along the crack paths and grain boundaries.

Traditional strategies have focused on mitigating the hot shortness effect in situ by altering the phase thermodynamics and kinetics of the Cu-enriched layer.

Dilution with Primary Iron
The most prevalent method involves diluting the Cu-rich scrap charge with primary iron, such as Direct Reduced Iron (DRI) or blast furnace hot metal, to lower the overall Cu concentration below a critical threshold (often <0.15 wt%). While effective for specific product grades, this approach is fundamentally a delaying tactic. It does not remove Cu from the material cycle, relies on carbon-intensive primary production, and constitutes a form of downcycling, as the resulting steel is often unsuitable for the most demanding applications.

Nickel Alloying: The Solubilization Approach

The most established metallurgical solution involves the alloying of Nickel (Ni). Ni exhibits complete solubility with Cu at all temperatures, including the hot-working range. When added to the steel, Ni partitions preferentially to the Cu-enriched liquid layer at the steel/scale interface, forming a Ni-Cu solid solution.

The incorporation of Ni raises the solidus temperature of the Cu-rich phase above the hot-working temperature, thereby preventing the formation of a liquid film. Moreover, Ni also alters the surface energy ratio, hindering the penetration of any semi-liquid phase into the grain boundaries. The requisite Ni content is typically governed by a stoichiometric rule of thumb: the Ni content must be approximately equal to or half of the Cu content (wt% Ni≥0.5×wt% Cu) to achieve effective suppression of surface cracking. However, this approach is severely limited by the high cost of Ni and the added complexity to overall alloy design.

Microstructure and Interfacial Control

Alternative strategies center on modifying the morphology and location of the Cu-enriched phase using less expensive alloying elements:

• Silicon (Si): Small additions of Si modify the oxide scale. Si promotes the formation of a stable SiO2​-rich layer and aids in the formation of globular Cu-enriched particles instead of the planar, thin films that facilitate grain boundary wetting. Furthermore, Si and B have been shown to help occlude the Cu phase into the growing oxide scale, pulling it away from the sensitive steel/scale interface.

• Carbon (C): Higher C content decreases the activity of Fe in the γ-phase, potentially retarding the preferential Fe oxidation kinetics.

• Process Timing (Strip Casting): A kinetic control approach involves rapid processing routes, such as strip casting combined with direct hot rolling. The extremely short time interval between reheating and deformation minimizes the time available for Fe oxidation and the necessary critical Cu accumulation to occur, allowing even high-Cu steels (up to 2.5 wt%) to be processed successfully under specific conditions.

The most visionary and fundamentally appealing solution to Cu contamination is its primary metallurgical removal from the molten charge, aligning the contaminant problem with the urgent need for decarbonization.

The Critical Oxygen Potential Concept for Cu Evaporation

A transformative approach is presented by Souza Filho et al. (2024), which leverages the thermodynamic landscape of the Fe-Cu-O system to achieve Cu volatilization. This method exploits the ongoing shift to green steelmaking, particularly the use of hydrogen-based reduction.

The study demonstrates that Cu can be effectively evaporated from Fe-Cu-O melts by controlling the oxygen concentration in the liquid. Key findings indicate:

1. Cu Activity Dependence: Cu removal via evaporation is intrinsically linked to the activity of Cu (aCu​) in the melt.

2. Critical Oxygen Concentration: The evaporation rate is maximized when the melt reaches a critical global oxygen (O) concentration of approximately 22 wt%. At this high O potential, the Cu activity is drastically reduced, promoting the volatilization of Cu (likely as a sub-oxide or through a complex plasma-gas-metal interaction) even without a strong reducing agent.

3. Hybrid Processing: The process uses Cu-contaminated scrap mixed with fresh iron ore (e.g., Fe2​O3​), which provides the high global oxygen content required for the initial purification phase. The Cu is effectively removed (from 1 wt% to <0.1 wt%) in this high-oxygen, plasma-driven melt. Subsequent treatment with a hydrogen plasma then reduces the resulting Cu-purified iron oxide melt back into high-purity metallic iron, completing the upcycling process.

This hybrid scrap upcycling strategy simultaneously addresses the two greatest challenges in the steel sector: decarbonization (via H2​-plasma reduction) and tramp element removal (via critical O potential-driven evaporation).

Vacuum Treatment and Flux-Based Solutions

Other removal techniques, though technologically demanding, rely on physical-chemical property differences:

• Vacuum Tapping: Utilizing the relatively higher vapor pressure of Cu compared to Fe under deep vacuum, Cu can be evaporated. However, the required vacuum levels and processing times are not yet economically feasible for high-throughput industrial processes.

• Sulfide/Chloride Fluxes: Earlier research explored the use of Na2​S- or Na2​SO4​-based fluxes or oxychlorination to form Cu compounds that partition into the slag phase. These methods are typically plagued by high operational costs, refractory attack, and environmental concerns related to byproducts.

Addressing the Cu challenge requires a multi-faceted strategy that combines prevention, tolerance, and removal:

1. Prevention at Source: Aggressive implementation of AI and sensor-based sorting technologies is essential to minimize Cu contamination in scrap yards. Policy measures promoting design for disassembly can further enhance sorting efficiency.

2. Enhanced Metallurgical Tolerance: Research must continue into "impurity-tolerant" alloy design, such as grain boundary engineering and the development of uni-alloy concepts, to create steels that are more resilient to residual tramp elements.

3. Adoption of Active Removal Processes: The oxygen-controlled Cu evaporation process represents a viable technological leap. Future work must focus on scaling up the technology, developing refractory materials resistant to highly oxidizing melts, and designing efficient gas-cleaning and Cu-recovery systems.

4. System-Wide Integration: A holistic view connecting metallurgy, product design, and policy is crucial. Creating closed-loop recycling streams for specific steel grades and fostering markets for recovered tramp elements are vital components of a circular economy.