Recent University Research, Implementation and Education in Support of Pyrometallurgical Lead Processing

The increasing chemical complexity of lead process streams encountered in industrial high temperature processing operations, as the result of declining primary resources, increased metal recycling and increased overall range of metals in modern devices has highlighted the urgent need for new predictive tools, fundamental phase equilibria and thermodynamic information and thermodynamic models to characterise the chemical behaviour of these systems. The paper examines recent progress in experimental and thermodynamic modelling research on process fundamentals, the availability of advanced, predictive computer-based tools and the implementation of the research outcomes into industrial practice. A wide range of chemical systems and phase assemblages have been studied. Some examples are taken from the current research program at PYROSEARCH, which involves the characterisation of multi-component, multi-phase gas-slag-matte-speiss-metal-solids systems with the PbO-ZnO-“Cu2O”-FeO-Fe2O3-CaO-Al2O3-MgO-SiO2-S as major and As-Sn-Sb-Bi-Ag-Au-Ni-Co-Cr-Na as minor elements with focus on systems directly relevant to lead primary and recycling pyrometallurgical processes. Examples of the application of advanced analytical techniques to fundamental and applied industrial research are also given. The implementation of new research outcomes into industrial practice depends critically on commitments by research staff as well as industry management and the availability of well-trained metallurgical engineers. We examine the current status of research implementation, university research, metallurgical engineering education and the availability of suitable educational pathways and initiatives that can be taken to increase undergraduate enrolments. Active engagement and support by industry is critical in ensuring the continuation of academic programs and advanced technical skills required by the industry.


Introduction
As a society we face critical global challenges associated with sustainable energy and technological developments.We are using more metals, more different metals and more combinations of metals than ever before in the production of the new advanced materials needed to address these challenges.As a result, the chemical compositions of the process streams in primary and secondary metal production, refining and recycling processes, which are necessary to support these technological developments, are becoming more and more complex and with increased variability.These are significant and increasing technical and societal problems with major implications for industrial and economic sustainability, and environment impact.

Stages of process optimisation
For existing pyrometallurgical operations, improvements in the form of increased efficiency, technical and economic productivity can be achieved in several ways, for example, by increased process stability, process optimisation, adoption of feed-forward process control, increased recoveries and forward planning.Generally, three stages of process optimisation must be done (see Figure 1)i) measurement of all factors influencing the process output, ii) control of those factors and finally iii) optimisation of the process to achieve and maintain optimal process output.
Figure 1.Some of the factors influencing process uncertainty and the key steps in achieving process optimisationmeasurement, control and optimisation

Predictive tools
Critical for the metallurgical process optimisation of complex metallurgical processes is the availability of not only accurate data on the process performance and process conditions but also accurate predictive tools.Improved instrumentation and data storage on computer systems can provide the ability to gather key information about the process performance and changes in real time.Appropriate models and predictive tools can now be developed with the significantly increased computer speed and capacity.One approach is to analyse these data to determine trends, patterns, and correlations between measurements, currently referred to as data mining.These methods are useful to understand and quantify behaviour within a defined range of conditions, that is, within defined compositional and process conditions.However, these empirical correlations cannot be used to confidently predict behaviour outside these current conditions.Typically, chemical and physical properties in metallurgical systems do not vary in continuous or linear relations with key process variables.These non-linear relationships arise from the formation of new phases within the system, which change the compositions and proportions of all other phases present.For the reasons cited above, process development involving the use of novel chemical compositions or process conditions is preferably undertaken with the support of accurate fundamentally based models or tools.
In the case of pyrometallurgical processing, many factors can influence the outcomes of pyrometallurgical processes.Some of these factors are summarized in Figure 1.The major controlling input parameters for pyrometallurgical processes are the amounts and compositions of the materials introduced into the reactor.These input compositions control three key output process characteristics (see Table 1) -product compositions (chemical targets), heat balances and phase equilibria (liquidus temperatures or proportion of solids).All these three factors are directly determined by the thermodynamic and phase equilibria properties of the system (see examples given in Table 2).
Fundamental research can deliver fundamental scientific information on the heat capacities of elements and compounds, and enthalpies of reaction, to enable the calculation and prediction of mass and heat balances in pyrometallurgical reaction systems.It is now possible to better understand the relative stabilities of elements and compounds, and the conditions for chemical equilibrium, through the development of thermodynamic theories, such as the concepts of Gibbs free energies and Gibbs energy minimisation.The application of these fundamental scientific concepts would not be possible without quantitative descriptions of these systems, for example, thermodynamic data on enthalpies and entropies of all components, and potential reactants and products, in these systems.The development of phase diagrams has enabled detailed thermodynamic information on the relative stabilities of phases and phase changes under different process conditions to be summarised and translated into useful graphical and mathematical forms.The use of phase diagrams has provided industry professionals with reliable methods of predicting flux requirements to achieve smooth and efficient operation of smelting and refining systems.The increased use of complex ores, end-of-life materials and process wastes has led to the increased chemical complexity of metallurgical process streams.Phase diagrams of binary and ternary systems are no longer sufficient to design and select the optimum compositions and process conditions in these industrial process streams.Virtual reactors for reactor and process design Advanced computer-based predictive models are now needed in order to provide both improved accuracy and to undertake the mathematical operations necessary to determine chemical equilibria in multicomponent, multiphase systems.These thermodynamic predictive tools can be used to develop process predictive tools (virtual reactors) and be incorporated into systems similar to the GPS systems used to define directions of travel (see Figure 2).Input to the computer program interface is in the form of feed/phase compositions, temperature, pressure and solid phase characteristics.The outputs can be presented in a wide variety of formats providing quantitative data on the amounts of individual phases and their compositions, and the overall reaction enthalpies or heat changes associated with the chemical reactions.These advanced tools can be used to not only determine process outcomes form individual reactors but also can be applied to all process streams in an operation, making it possible to optimize multi-step processes and metal recycle streams within the operations.These then are powerful tools that have the potential to significantly enhance overall technical and economic performance.

Thermodynamic databases
Since the thermodynamic databases that are at the heart of these predictive tools describe the interactions of individual elements and the relative stabilities of a wide range of compounds and phases, they are more reliable predictors of reactions outside the current industrial practice.These thermodynamically based process tools also have the advantage that the predictions are independent of the technologies being used and therefore can be used to define and explore all combinations of process variables within the range of systems that have been optimised in the database.
Significant gaps in knowledge on slag properties still exist -these are due to difficulties associated with high temperature research.The demand for the accurate fundamental information on phase equilibria, thermodynamic and physicochemical properties of the complex multi-component systems from metallurgical and recycling industries is growing due to a) stronger economic competition, b) stricter environmental regulations, and c) better equipment and options in process control.Many other factors also contribute, including increasing complexity of ores, the number of available smelting technology options.The supply of the needed fundamental data on the chemistry of the processes is becoming possible since a) new experimental techniques are becoming available due to the developments of modern advanced analytical techniques, dramatic improvement of their capabilities and availability, and b) new theoretical modelling approaches due to the significantly increased computer power.This demand / supply combination is the basis for the renaissance in research on the high temperature chemistry of metallurgical systems.
The increased number of different elements present in the system results in an increase in the number of chemical components and species interacting simultaneously.Many of these chemical interactions, and the chemical subsystems defining these interactions, have not previously been studied or accurately characterised.To be able to accurately describe reactions in these systems it is necessary to undertake further experimental measurements and incorporate this information into suitable mathematical forms using thermodynamic models and appropriate model parameters.The unambiguous determination of the unique set of binary and ternary model parameters accurately describing a whole range of conditions from low to high order systems is critical for underpinning the predictive power of the tool, and requires experimental data for all binary and ternary subsystems, and for selected quaternary subsystems.Accurate predictive models cannot be developed if accurate, fundamental, experimentally determined information is not available.

Integration of experimental and thermodynamic modelling research
At the Pyrometallurgy Innovation Centre an integrated approach has been developed that involves the parallel and complementary characterisation of the thermodynamic properties of the systems using experimental and thermodynamic model and database development.
Experimental and thermodynamic modelling studies are fully integrated in the present research program.Thermodynamic assessment is used to identify discrepancies and to plan experiments.Experimental results are then used to fix model parameters to improve model accuracy.If there are no experimental data -there can be no reliable thermodynamic parameters, and consequently no reliable model.The thermodynamic parameters in the present study therefore are determined by targeted experiments specifically planned and undertaken to determine the parameters rather -this is different from studies fixing thermodynamic parameters only from the data available from literature.A brief summary of the experimental and modelling techniques, and their application is provided in the following paragraphs.

Experimental methodology
The accurate measurements of fundamental physical and chemical properties of lead/zinc containing systems pose particular experimental difficulties due to the high vapour pressures of the metal species and the highly reactive/corrosive properties of liquid phases in these systems.Conventional experimental techniques, such as thermal analysis can be used to determine liquidus temperatures and phase diagrams at low temperatures and in relatively simple low or silica-free systems containing eutectic reactions and stoichiometric compounds.However, these approaches cannot provide the accuracy required for thermodynamic database development at high temperatures, particularly for systems containing peritectic reactions, solid solutions and multicomponent systems.
The equilibration/quench/microanalysis methodology developed by PYROSEARCH has overcome all of these limitations and it is now the preferred approach to phase equilibria determination in these complex systems.The technique has been first discussed in 1995 [48] (Figure 3a).Essentially the methodology involves equilibration of small masses (typically less than 0.5g of material) under controlled process conditions at temperature.The sample is then rapidly quenched to room temperature thus retaining the phase assemblage and phase compositions present at the equilibration temperature.The phases and the phase compositions present in the samples are determined through the use of advanced microanalytical techniques including: i) electron probe X-ray microanalysis (EPMA), and ii) laser ablation inductively coupled plasma mass spectrometry (LA-ICPMS).The compositions and process conditions are deliberately selected so that multi-phase materials are formed in the equilibrated products.In these conditions' changes, so long as the phases present remain the same, any change to the bulk composition of the sample only changes the proportions of the phases, not their compositions.The methodology then has the important advantage that the results do not depend on small bulk composition changes that may take place during equilibration.Recent developments in experimental techniques mean that phase equilibrium measurements and elemental distributions between phases can now be obtained in slags/mattes/metal alloys/speisses/solid solutions in closed and open systems.Contamination from crucible materials is avoided by using synthetic substrates of the primary phase material.A range of metal substrates (Pt, Pd, Ir, Re, Rh/Pt, Au, Mo, W, Fe, Co, Ni, Cu) have also been used for specific systems and process conditions.Successful equilibrium experiments have been carried at temperatures up to 1740°C.In all cases, proof of equilibrium is established through the "4-point" test protocol, which requires that the results are independent of 1. reaction time and 2. the starting conditions of the system.3. the outcomes should be not restricted by the reaction mechanism and reaction pathway and 4. the compositions of the phases should be homogeneous and not vary with position.Overviews of the latest advances in the integrated experimental and modelling research approach to the thermodynamic database development for these pyrometallurgical systems has been provided in [49][50][51].The use of these microanalysis techniques has provided breakthrough capabilities in phase equilibrium studies for pyrometallurgical applications, greatly extending the range of elements, bulk compositions and conditions that can be characterised.Figure 3. a) Illustration of the PYROSEARCH approach to high temperature phase equilbria determination, enabling accurate measurement of phase compositions in multi-phase systems under defined process conditions, b) An example of a complex multiphase, multicomponent system at equilibrium.

Thermodynamic models and databases
PYROSEARCH researchers use the FactSage computer platform and the associated thermodynamic models to develop the new thermodynamic database.For every portion of the database, the methodology used involves first constructing preliminary models using experimental thermodynamic and phase equilibria data available in the literature data, the identification of key areas of disagreement or thermodynamic inconsistency and the selection of critical experiments required to resolve these inconsistencies or omissions.The database is developed so as to achieve a self-consistent set of parameters for the thermodynamic models, accurately describing phase equilibria and all other thermodynamic properties for all phases over the complete range of conditions in terms of temperature, composition and atmospheric conditions.
In addition, there have been major revisions to the database in the form of implementation of an improved methodology to accurately describe the heat capacities of elements and compounds, corrections to the heat capacity and enthalpy of melting of SiO2 [20] and a number of other pure end members, and the introduction of parameters that accurately describe the intermediate silica concentrations with maximum ordering (near orthosilicate composition) as well as high silica slags.All of these improvements have been critical in obtaining improved thermodynamic description of high and low silica systems, the accurate prediction of liquid immiscibilities and the complete revision of the all subsystems in this 20-component system, including the FeO-Fe2O3-CaO-SiO2 subsystem, which is now being prepared for publication.
A major issue emerging in the development of multicomponent databases is the exponential increase in chemical interactions to be described with the increasing number of components.Adding a single component, or even incorporating new experimental data into the database, requires the iterative re-optimisation of all low-order and corresponding high-order systems; a significant task when handling a 20 component system that has over 1000 binary and ternary sub-systems.
To tackle this problem and to be able to efficiently update the databases, a new optimisation methodology has been implemented the PYROSEARCH team.The key points in this new methodology are as follows.The sets of target experimental points to be described and corresponding weights are selected based on the experimental information available.A matrix of first derivatives showing the sensitivity of each target point to each possible model parameter is calculated using traditional Gibbs energy minimization calculations, which is a relatively slow step.The initial slow Gibbs energy minimisation calculations are then replaced by a fast analytical approach with linear extrapolation of the existing values through matrix multiplications of the form: yx  =   A , where A is the matrix of first-order derivatives (n target values by k model parameters), .This non-iterative analytical (rather than numerical) optimization approach is orders of magnitude faster than the combination of the thermodynamic calculations using Gibbs energy minimisation and numerical non-linear minimisation.
The first-derivative-based linear extrapolation approach also enables a) the real-time graphical presentation of predicted and target points as well as b) the real-time systematic tabular presentation of the statistical analysis of agreement between predictions and target values, which makes the optimisation process truly interactive.Also, the new formalised and semi-automated methodology makes it possible to increase the efficiency and flexibility of collaborative work between researchers by organising parallel simultaneous optimisations by several researchers, thus distributing the database development efforts between the research team: once the procedure to optimise the model parameters is formalised and semi-automated, the group of researchers each will contribute to the thermodynamic parameters optimisation by planning and undertaking new experiments, adding corresponding target points and correcting weights, rather than "manually" optimising model parameters.Thus the discrepancies and conflicts within the system will be resolved by the formalised semi-automated system significantly more efficiently and with less oversight.Only periodically will the re-calculations using Gibbs energy minimisations be needed to update the matrix of derivatives.The new methodology enables researchers undertaking experimental work on a particular sub-system to personally contribute to the thermodynamic optimisation of that system and to select further experimental target points further increasing productivity.Optimisation is undertaken in iterative cycles, the major discrepancies are identified at each step, and new experiments conducted to specifically resolve discrepancies within time intervals from several days to several weeks rather than months and years using the more traditional approaches.

Fundamental research outcomes
Examples of the experimental results on chemical systems relevant to lead smelting, reduction and refining obtained during recent studies are provided in the following figures.These demonstrate the number of different phases, both solid and liquid, that can form in these polymetallic systems in the presence of minor elements, as exemplified in the examples given in Figure 4.The predicted phase diagrams show the influence of bulk composition on the liquidus and primary phases formed, as illustrated in Figure 5.The application of the database to the prediction of phase equilibria in complex multicomponent systems is illustrated by the fluxing diagram reproduced in Figure 6, which clearly shows the effect of CaO/SiO2 ratio on slag liquidus and primary phase formation.Once the database has been established the program can be used to systematically quantify the effects of changing composition or process conditions.The current 20-component database under construction for the in system with "Cu2O"-PbO-ZnO-FeO-Fe2O3-CaO-SiO2-Al2O3-MgO-S major, Cr and Na slagging and As, Bi, Sn, Sb, Au, Ag, Ni, Co other minor elements contains oxide, matte, and speiss phases.In order to prepare an accurate description for all compositions and process conditions the numbers of subsystems to be characterised is illustrated in Figure 7.To date the PYROSEARCH team has completed ~20,000 experiments in 340 subsystems (investigated or in process).Data from 88 systems are available in the literature, 109 systems are impossible to experimentally investigate using the current methodologies.There are no measured equilibrium data on 109 major + 550 minor systems.The current thermodynamic database contains 26 major solutions, 79 small solutions, >380 stoichiometric solid phases, > 100 gaseous species, > 1148 endmembers + stoichiometric compounds, defining S298, ∆H298 and cP vs T for each.The slag solutions include data on 24 metal cations multiplied by [O 2-, S 2-] anions = 48 endmembers, 276 binary and 2024 ternaries systems, involving the determination of 990 excess parameters.

Implementation
The collaboration between the industry partners and PYROSEARCH has enabled the development of a set of powerful predictive tools, which can be used to analyse a broad range of industrial systems independent of the technologies used.The collaborative program does not stop there -the important next step is demonstrating the value of these advanced tools and their implementation in industrial practice.The tools can be used by companies to improve process efficiencies and productivities, optimise the utilisation of existing plant, predict changes to plant practice to adjust for changes in process feed compositions, design new processes or operations.It has been recognized and acknowledged by industry sponsors that the successful implementation of the tools into industrial practice requires adequate training in their use.This has been achieved through a variety of actions in collaboration with sponsor companies.
Upskilling of engineering professionals is achieved through providing • Dedicated courses in chemical thermodynamic theory, pyrometallurgy fundamentals and the application of FactSage tools to industrial problems, delivered by distance and in person.• Extended visits, placements and secondments of engineers with the research team.
• Undertaking one-on-one projects with the research team on industrial problems.
• Undertaking Research Masters and PhD studies at PYROSEARCH.The active collaborations between industry and the research team provide important opportunities to demonstrate the capabilities of the predictive tools.Testing the tools against industrial practice provides confidence in their potential for extended use in an industrial contexta critical factor in technology transfer.

Education
Universities are viewed traditionally as institutions where knowledge is stored to be passed on through academic debate and learning, in this way they serve society through providing continuity in knowledge and expertise.Through teaching they transfer advanced knowledge and skills to undergraduate and postgraduate students.Much less understood are other important roles played by University academics; sponsor-nominated Universities, • Minimise the adverse impacts of periodic economic cycles on the retention of knowledge and expertise for the metallurgical industry.Such is the importance of the discipline that every country should have a strong University academic with expertise and research capabilities in metallurgical engineering.The number of students selecting studies in metallurgical engineering and the number of university programs offering educational opportunities in metallurgical engineering have been in decline for many years -this is a worldwide trend.There appear to be a number of contributing factors, • Unfavourable public perceptions of the industry and lack of information on the role of metallurgy in our society.• The lack of industry engagement and support in promoting studies and careers in the field.
• The funding formulae used by Government to support Universities, which are in the main based on the number of enrolments rather than industry or societal needs.• University and Journal Ranking systems do not favour small specialist fields.
The lack of professionals in the field poses significant risk to the successful implementation of process improvements to existing operations and is an impediment to the major changes required for the future economies.The general public and students are now acutely aware of the impacts caused by climate change and the need to transform our industrial operations to meet these challenges.These changes in attitude represent an opportunity to change the narrative and perceptions about the industry from a negative to a positive view.The dramatic electrification of all aspects of our economies from renewable energy, transportation and electrical and computational devices has raised awareness of the need for critical metals to enable these transitions to take place.Added to this is the need for materials conservation and increased recycling of metals to minimize environmental impact and energy usage and achieve circularity in the use of valuable non-renewable resources.These are messages that the industry can use to advantage to explain the role of lead metallurgy in facilitating the recycling of critical, strategic and precious metals.
Increasing public awareness of the key role of metallurgy is necessary but is not necessarily sufficient to attract students and graduates.The industry needs to be proactive in improving workplace environments and practices so that they are attractive to prospective students and engineers.Students are looking for opportunities where they feel they can make a positive contribution to the environment and are able to use the latest technologies to achieve this.Industry can take positive steps in attracting undergraduates in their early years at university when they have an opportunity to redirect their careers into the discipline, through for example, providing on-site experiences in the form of vacation work, plant visits and internships, financial support scholarships, professional development and social activities.
The allocation of university funding is for the most part determined by class size rather than strategic needs of industry or the country.Small classes for specialist courses are more likely to be removed from the curriculum, small programs are closed in the name of efficient use of limited financial resources.This is the pattern around the world -the only way to change the current trend and provide educational resources is for industry to be more proactive.Industry leaders need to inform university management and governments of the importance of the discipline and actively engage with and provide support to those institutions providing programs in metallurgical engineering.
We strongly believe there are opportunities to increase the number of metallurgy graduates if Industry, Universities and Governments are proactive and work collaboratively together to solve this problem to their mutual benefits.

Summary
As a society we are faced with significant technical, societal and environmental challenges that must be addressed in a relatively short period of time.The urgency is such that we should use all means at our disposal to meet these important goals.
To respond to the changes in metals sources and the usage of new combinations of metallic elements changes must be made to metallurgical practice and the technologies employed.Process systems are becoming more chemically complex and process optimization is becoming increasingly more difficult.To address these changes we need to develop both the technologies and the workforce for the future.
Closer collaborations between industry and universities can assist in meeting these goals and can bring benefits to both parties by • Delivering the new scientific information and advanced predictive tools required by industry • Providing training on the use and implementation of these tools in industry • Increasing enrolments in metallurgical engineering at our tertiary institutions

Figure 2 .
Figure 2. Illustration of the key role of fundamentally based thermodynamic tools in predicting process outcomes and assisting in process optimisation.
is the difference between the final set of model parameters and their initial approximation, and 0 y y y  = − is the difference between the final model predictions and model predictions at 0 x .The optimum values of

Figure 4 .Figure 5 .
Figure 4. Example of experimental results: backscattered electron images of microstructures of multielement, multiphase equilibria and compositions of liquid phase for that equilibria plotted on the Pb-Cu-As triangle

Figure 6 .
Figure 6.Fluxing diagrams for the oxidizing smelting and reducing stages of Pb/Zn smelting examining the impact of CaO/SiO2 ratio on slag liquidus and primary phase formation predicted using PYROSEARCH database.Also shown are microstructures of the typical plant samples.
Major: Pb-Cu-S-Fe-As-Sb-Sn-Ni-O Minor: Zn-Bi-Co-Ag-Au(+Na) Major: Pb-Cu-Zn-Fe-Ca-Si-O Minor slagging: Al-Mg-Cr (+Na) Minor: As-Sn-Sb-Bi-Ag-Au-Ni-Co knowledge in the metallurgical engineering can actively support the metallurgical industry in a number of ways, by • Undertaking knowledge generation through fundamental and applied research, • Developing and maintaining collaborative research links with industry, • Attracting additional financial support for research from Government sources, • Developing research infrastructure within institutions, facilities that can be utilised to support industry related research, • Facilitating communication and support relationships between the metallurgy industry and senior university management, • Promoting the important role of the metallurgical industry to young people and the broader community.In addition, by insuring the sustainability of strong research teams with specialist knowledge and expertise, University academics can • Develop and maintain industry consortia for research and education of common interest to sponsors, • Actively assist in the optimisation of existing and development of new metallurgical processes, • Ensure the successful implementation of research outcomes into industrial practice • Ensure the availability of expert capabilities and advanced technological for application to individual confidential R&D support, • Facilitate active technological exchange and links between sponsors metallurgists and between

Table 1 .
Typical input, control and target parameters used in pyrometallurgical processing

Table 2 .
Examples of the application of fundamental thermodynamic research to industrial pyrometallurgical practice.