Corrosion Mechanism of Molten Glass on Molybdenum Electrodes and Its Influencing Factors

After a year or two in the melt, the molybdenum electrodes of an all-electric furnace are often visibly thinner in diameter. The loss is not a simple dissolution of metal like salt but the result of several mechanisms acting together. Only by clarifying the corrosion chain can we explain why the same electrode has very different life in different glass compositions and furnace types, and identify the operating parameters that can be adjusted.

The three stages of corrosion

The first is chemical oxidation. Sulfates, nitrates and fining agents such as arsenic, antimony and selenium in the glass melt, together with dissolved oxygen, progressively oxidize the molybdenum electrode surface to higher oxides that enter the melt. The oxidation rate depends on temperature, oxygen potential and contact area.

The second is electrochemical action. When the electrode conducts as an anode, metal ions migrate into the melt under the electric field, and the dissolution rate is higher than in static immersion without current. The third is convective scouring. Circulating flow and fresh batch flow in the tank continually strip the surface reaction layer, exposing fresh metal to the corrosive medium. The three stages follow one another, and the loss accelerates itself. Their relative weight varies with duty: in a low-temperature, weakly oxidizing melt, chemical oxidation may be the main cause, while in high-current-density, strongly convective regions the electrochemical and scouring contributions rise. When analyzing a failed part, determine first which stage dominates so that the countermeasure is aimed correctly.

Temperature and redox state

Dissolution rate rises with temperature, and local overheating at the hot end noticeably shortens life. The electrode tip, the liquid level line and the vicinity of the orifice are locations where temperature tends to run high, and where loss usually appears first. Placing thermocouples or infrared measurement points in these regions is more meaningful than measuring only the center of the bath. The effect of temperature on corrosion is not linear; near the upper limit of the glass working temperature the reaction rate rises faster, and a difference of a few tens of degrees at the tip can show up in service life.

The redox state of the melt is equally critical. Melts fined with reducing components such as carbon powder or sulfur dioxide are usually less corrosive; melts rich in oxidizing components are more aggressive. If a batch is designed only for fining performance while the electric melting duty is ignored, electrode loss may only become apparent after production starts.

Current density and flow field

The higher the anodic current density, the larger the electrochemical dissolution component. Pursuing high power per electrode can make up heat in the short term, but over the long term it accelerates tip consumption and alters the local flow field. Keeping current density within the empirical range and sharing the load by adding electrodes is usually better for balanced life.

Current density also affects the structure of the reaction layer on the electrode surface. At low density the reaction layer is thicker and may be stripped off in whole patches by the flow; at high density it is thin and active, and loss is more continuous. The two forms call for different inspection priorities.

The influence of the flow field is often underestimated. Flow velocity is high near the batch inlet, the throat and the corners of the tank walls, where the reaction layer is easily carried away and local grooves often appear. Electrode layout should avoid strong scouring zones, or these zones should use a thicker section and a lower current density.

Surface condition and bubbles

The surface finish of the electrode affects the initial corrosion density. Polished and sintered surfaces have different reaction activity, and residual cutting fluid, abrasive and oxide tint on the surface can become starting points for local corrosion. Cleaning and inspection before installation are a low-cost line of defense. Chamfered and rounded tip transitions reduce local electric field concentration and help delay pitting.

Gas evolved at the electrode surface locally shields the current, and periodic bubble detachment causes point corrosion. Bubbles gather more easily at the tip and near the liquid level line, producing an uneven loss pattern. If the current of one electrode fluctuates abnormally in service, check whether bubbles are retained there or the reaction layer has thickened abnormally.

Countermeasures to slow corrosion

• Keep current density within the empirical range; do not pursue the power ceiling of a single electrode.
• Design the batch with the electric melting duty in mind and avoid excessive oxidizing fining agents.
• Keep the liquid level stable and reduce repeated exposure of the electrode in the gas-liquid alternation zone.
• Use several smaller-section electrodes to share current in highly corrosive tank sections, so that loss is slower and more uniform.
• Shorten the inspection interval in highly corrosive zones and schedule replacement based on measured diameter.

The amount of molybdenum corrosion product that migrates into the melt is small, and unlike some metal electrodes it does not cause obvious coloring; this is one of the conditions for its long use in container, photovoltaic and electronic glass furnaces. For particularly severely corroded tank sections, power stability can also be maintained by rotating the supply and replacing electrodes in batches.

Distinguishing corrosion from other failure modes

Electrode failure is not always corrosion. Thermal shock cracks, mechanical impact, overheating at connections and oxidation also appear as tip loss or fracture. Combining fracture morphology, crack direction and operating records is the only way to identify the main cause.

For example, intergranular cracking is more likely related to thermal stress or impurities, whereas surface grooves and uniform diameter reduction fit the corrosion signature. Mistaking mechanical damage for corrosion leads to ineffective changes in batch and power.

Evaluation and verification methods

To judge whether corrosion is abnormal, do not look only at operating hours. Record each electrode's commissioning date, current, voltage, cooling water temperature and the corresponding melt temperature; measure diameter and tip morphology at shutdown, and convert the loss into diameter change per unit time. Photograph and measure in several directions during shutdown inspection, because single-point data is easily misread. Only after several campaign records can normal consumption be distinguished from abnormal acceleration.

The effect of batch or power changes should also be verified by comparison. Change one variable at a time, observe diameter change and glass defects over a period, then decide whether to roll it out. With a clear mechanism, operating adjustments have a direction.