Summary of Findings
Every parameter on an electrode paste Certificate of Analysis (COA) was fixed at a specific process step, and the steps are not interchangeable. Ash is decided at raw-material selection. The conductivity floor of the aggregate is decided at calcination — anthracite shows a sharp rise in electrical conductivity at a heat-treatment temperature close to 700 °C, with the jump occurring across roughly 600–800 °C, and conductivity tending toward approximately 100 S·cm⁻¹ once calcination is complete (Celzard et al., Carbon, 2000). Binder distribution is decided at mixing and kneading. Block geometry is decided at forming.
One step is decided nowhere in the supply chain: the final baking. For self-baking paste it happens inside the buyer’s own furnace, at approximately 450–500 °C (Larsen, Platinum 2012), under a thermal regime the supplier does not control and no method standard prescribes. Understanding which step owns which parameter is what makes a specification enforceable and a quality discussion productive.
Key Findings
- Calcination sets the aggregate’s conductivity floor. The conductivity of anthracite rises sharply near 700 °C as insulating organic matter is removed and conducting domains come into contact — a percolation mechanism, with the conducting phase reaching a critical volume fraction of approximately 75 vol% (Celzard et al., 2000).
- Calcination also concentrates ash. Mineral matter survives while carbon-bearing volatiles leave, so the calcined aggregate carries a higher ash percentage than the raw material it came from.
- Kneading decides whether the binder actually coats the aggregate. Binder softening points across suppliers span 53–75 °C, and a mismatch between binder behaviour and the thermal conditions it meets produces layering and poor electrode formation (Wilkinson et al., Infacon IX).
- The last step belongs to the buyer. Baking occurs in the furnace, not at the supplier, which is why paste is specified on pre-baking properties and why a baking schedule has to be agreed before any baked-specimen parameter can be compared between two laboratories.
- Not every step is performed in-house, and the distinction should be stated. At Jiyang Carbon, materials screening and formulation are under direct control; mixing, kneading, forming and inspection are in-house; calcination, baking and graphitization are completed through qualified processing partners under documented specifications, process records, sampling and final inspection.
Contents
- What is the actual process chain, and who controls each step?
- What does calcination decide?
- What do mixing and kneading decide?
- What does forming decide, and what does it not?
- Why is the final step performed by the buyer?
- Which parameter is decided at which step?
- Frequently asked questions
- Related reading
1. What Is the Actual Process Chain, and Who Controls Each Step?
Electrode paste production runs from raw-material selection through calcined aggregate, formulation, mixing, kneading and forming to inspection and dispatch. Describing that chain as a single controlled process is common in supplier literature and is rarely accurate; the steps sit with different parties, and a buyer assessing a supply chain needs to know which.
At Jiyang Carbon the division is stated rather than implied:
| Step | Control |
|---|---|
| Raw materials screening | Direct control |
| Formulation development | Direct control |
| Calcination | Qualified processing partner |
| Mixing | In-house |
| Kneading | In-house |
| Forming | In-house |
| Inspection | In-house laboratory (JY QC) |
| Baking | The buyer’s furnace |
Table 1. The steps that set the properties on a Certificate of Analysis are distributed across three parties, and the last one — baking — is outside the supply chain altogether.
Where a step is completed by a qualified processing partner, it is performed under documented specifications, with process records, sampling and final inspection retained and reviewed. That is a different control model from an in-house step, and the reason for stating it is practical rather than rhetorical: a buyer auditing a supply chain needs to know which records exist where, and a claim of end-to-end in-house production that does not survive an audit question costs more credibility than the claim was worth.
2. What Does Calcination Decide?
Calcination is heat treatment of the carbon aggregate before it ever meets a binder, and it decides two things that no later step can revisit.
The conductivity of the aggregate. Raw anthracite is effectively an insulator. As heat-treatment temperature rises, volatile organic matter is progressively removed and the conducting carbon domains come into contact with one another; conductivity rises sharply at a critical heat-treatment temperature close to 700 °C, with the jump observed across roughly 600–800 °C for the varieties studied, and tends toward approximately 100 S·cm⁻¹ once calcination is complete (Celzard et al., Carbon, 2000). The mechanism is percolation: the conducting phase has to reach a critical volume fraction — approximately 75 vol% in that work — before continuous conduction paths exist at all.
Two consequences follow for a buyer. First, an under-calcined aggregate cannot be corrected downstream; mixing and kneading redistribute material but do not remove residual volatile organic matter from within the particles. Second, resistivity on a COA is a composite property of the aggregate, the carbonised binder and the specimen preparation, and the aggregate’s calcination sets the floor beneath it.
The ash percentage of the aggregate. Mineral matter survives calcination while carbon-bearing volatiles leave. The calcined material therefore reports a higher ash percentage than the feed did, for the same absolute quantity of mineral matter. This is the same arithmetic that makes ash on baked material read higher than ash on the paste as supplied, and it is treated in the ash article.
Calcined petroleum coke sits at the other end of the aggregate range, carrying less than 0.3 wt% ash at greater than 97 wt% carbon (White et al., 2024). The choice between aggregate types is a formulation decision, made upstream of calcination and treated in the ash article.
3. What Do Mixing and Kneading Decide?
Mixing brings the graded aggregate fractions and the binder together in the specified proportions. Kneading is what makes the mixture a paste: under heat and mechanical work, the binder is distributed over and into the aggregate surfaces.
The parameter kneading governs is not on the COA. It is homogeneity — whether every part of the batch carries the binder proportion the formulation specifies. Inadequate kneading produces a batch whose average composition matches the formulation while its parts do not, and the resulting variation appears later as batch spread rather than as a failed parameter.
Temperature is the operative variable, and it has to be set against the binder rather than against habit. Binder softening points across suppliers span 53–75 °C, and the consequences of a mismatch run in both directions (Wilkinson et al., Infacon IX). At the furnace, a softening point too high for the operating current density produced low liquid paste levels, layering of paste from successive charges, and poor electrode formation — the same class of failure that inadequate kneading produces at the plant, arriving by a different route.
That symmetry is worth stating explicitly, because layering at the furnace is routinely reported as a paste manufacturing defect when its cause frequently lies in the thermal conditions the paste met at the paste level rather than in the mixing hall.
4. What Does Forming Decide, and What Does It Not?
Forming decides geometry: block shape, nominal dimensions, and the surface condition that determines how much material survives handling as blocks rather than as fines. It decides nothing about composition. The same formulation supplied as trapezoidal blocks and as pebble-form briquettes returns identical values on the COA and behaves differently between the charging platform and the melt.
The practical consequence is that a form change and a formulation change must never be made in the same consignment. Where both move, no subsequent observation at the furnace can be attributed to either. Form selection and its effects are treated in the block shape article.
5. Why Is the Final Step Performed by the Buyer?
Because a self-baking electrode bakes in service. That is the defining property of the Söderberg system, and it has a specific consequence for how paste can be specified and tested.
Every parameter that requires a baked specimen — compressive strength and electrical resistivity among them — has to be measured on material baked under some schedule, and no schedule is prescribed for electrode paste by the method standards, whose scope assumes an already-baked article. Two laboratories baking specimens under different schedules are measuring two different materials, and the numbers will not reconcile no matter how carefully each was determined. That mechanism, and the contract clauses that close it, are treated in the standards comparison article.
The corollary shapes what a supplier can and cannot be held to. A supplier is accountable for composition, homogeneity, form and the properties measurable on the paste as supplied. The baking regime — heating rate, current density, slipping cadence — is the buyer’s, and a large part of what the electrode finally becomes is decided there. Assigning responsibility after a field failure begins with that division, and the diagnostic guide sets out how the evidence separates the two.
6. Which Parameter Is Decided at Which Step?
| Parameter | Decided at | Correctable later? |
|---|---|---|
| Ash | Materials screening, formulation | No |
| Sulfur | Materials screening | No |
| Volatile matter | Formulation, binder selection | No |
| Softening point | Binder selection | No |
| Moisture | Forming, packing, storage | Partly — by storage and handling |
| Resistivity | Calcination (floor), kneading, and the buyer’s baking regime | Not after dispatch |
| Compressive strength | Kneading, forming, and the buyer’s baking regime | Not after dispatch |
| Homogeneity / batch spread | Mixing, kneading | No |
| Block form and fines | Forming, packing, handling | Partly — by handling practice |
Table 2. Only moisture and the handling-related properties can be influenced after dispatch. Everything else is fixed before the consignment leaves, which is why a specification that omits a parameter cannot be recovered by inspection at the receiving plant.
7. Frequently Asked Questions
Q1. Does a supplier that outsources calcination have less control over quality? It has a different control model, not necessarily a weaker one. What matters is whether the outsourced step is performed under documented specifications, whether process records and samples are retained, and whether the final inspection is independent of the partner. Those are auditable questions, and they are the ones to ask — rather than whether the step is in-house, which by itself establishes nothing.
Q2. Can under-calcined aggregate be identified from the COA? Not directly, and this is a limitation worth knowing. Resistivity on a baked specimen is a composite of aggregate, carbonised binder and specimen preparation, so an aggregate problem can be partly masked. Where an aggregate question arises, the useful request is the calcined aggregate’s own test data and the calcination records, not a repeat test on the finished paste.
Q3. Our resistivity results differ from the supplier’s by more than the specification tolerance. Where does that usually come from? Specimen preparation, in most cases. Resistivity cannot be measured on green paste, the baking schedule is not fixed by the method standards, and two schedules produce two materials. Compare the specimen preparation statements from both laboratories before comparing the numbers; where the schedules differ, the results are not comparable and a joint re-test on an agreed schedule is the route forward.
Q4. Which process records should a purchase specification require? Those that correspond to steps whose output cannot be verified on the finished product. Calcination records and formulation records fall into that category; forming records generally do not, since form is inspectable on arrival. Requiring records indiscriminately produces a document package nobody reads and does not improve control.
Q5. If baking happens in our furnace, what is the supplier actually responsible for? Composition, homogeneity, form, packing and the properties measurable on the paste as supplied — and for stating the method and basis for each. The buyer is responsible for the baking regime and the operating cadence. Most field failures involve both sides, which is why the diagnostic path collects the COA and the operating trace together rather than treating either alone as sufficient.
8. Related Reading
- Ash Content in Electrode Paste: Where It Comes From, and What a Grade Actually Buys (companion article) — the parameter fixed earliest in the chain
- Reading a Seven-Parameter Test Report for Internal Consistency — how the outputs of these steps are checked on the document
- Electrode Paste Raw Materials Explained — the aggregate and binder entering the chain
- Paste Block Shape and Size: How Geometry Affects Feeding and Sintering — what forming decides
- YB/T 5215 and International Test Standards: A Clause-Level Comparison — why an unagreed baking schedule makes two results incomparable
- Batch-to-Batch Variation: Judging Supplier Stability with Distributions — how homogeneity from mixing and kneading appears to the buyer
Author and Review
Written by Sterling Yang, Senior Vice President at Jiyang Carbon.
This article was reviewed by Mr. Yang, Technical Director at Jiyang Carbon, drawing on 40+ years of hands-on experience in carbon processing, weekly review of production data and personal approval of custom formulations.
Values attributed to published sources are listed in Appendix B. The division of steps in Table 1 is Jiyang Carbon’s stated process control arrangement: materials screening and formulation under direct control; mixing, kneading, forming and inspection in-house; calcination, baking and graphitization completed through qualified processing partners under documented specifications, process records, sampling and final inspection. Production area is 20,000 m² with annual capacity above 50,000 tonnes; the quality management system is certified to ISO 9001 and audited annually.
Technical Consultation
For assistance in defining which process records a purchase specification should require, or in assessing an aggregate or homogeneity question, please contact the technical team.
