New Chlorinator-Melt Study Makes Stage Losses the Scandium Recovery Story

A peer-reviewed study published on 2026-08-27 follows scandium through a complete recovery train built around spent titanium chlorinator melt. Its most useful result is not the highest percentage in the process. It is the gap between strong intermediate recoveries and the amount that reached a purified final oxide (research record).

The distinction matters whenever an upstream titanium project presents a critical-mineral co-product. A sorption percentage describes one boundary. A saleable-product case must account for every boundary after it.

A chloride-rich feed makes selectivity the first problem

The source describes spent titanium chlorinator melt as a potential technogenic scandium source. It also contains high levels of iron, aluminium and alkali and alkaline-earth chlorides. Recovery therefore begins with an impurity problem, not a clean scandium solution.

In the scaled-up test, 4 kg of spent melt was leached with 5% HCl. The process produced 25.5 L of filtrate containing 475.30 mg Sc. That set of numbers establishes the first useful denominator: batch mass, liquid volume and scandium inventory before selective recovery.

The route then used Fe(III) reduction with ascorbic acid, sorption on Lewatit MonoPlus SP112H, desorption with acidified ammonium sulfate, oxalate precipitation, calcination and an additional column purification. Each operation changes the material state and creates a new place for scandium or impurities to leave the intended stream.

High intermediate recovery can coexist with a lower final yield

Two-stage sorption recovered 90.9% of the scandium. Overall desorption reached 94.8%. Read alone, those values could support an impression that nearly all scandium will reach product.

It did not. Precipitation and calcination yielded 0.60 g of intermediate oxide containing 65.0 wt% Sc. Additional purification produced 0.45 g of final oxide. X-ray diffraction identified cubic Sc2O3 as the predominant crystalline phase, while elemental analysis put purity close to 98 wt%. Final recovery was 61.92% of the scandium in the filtrate.

There is no contradiction. The percentages use different stage boundaries, and later purification sacrifices mass while rejecting impurities. The study identifies iron co-sorption and scandium loss during the additional purification as the principal limitations.

Purity and recovery must share one ledger

A high recovery with a dirty product and a high-purity product with low overall yield describe different businesses. Feed variability, reagent demand, waste volumes, repeatability and product specification determine whether a laboratory route can become a stable co-product operation.

This study does not establish commercial economics, a production-scale recovery rate or any change in downstream titanium quality or price. That boundary is important. Processed-titanium buyers should not infer a material benefit from a scandium-recovery paper. The direct audience is the upstream operator, technology reviewer and supply-chain analyst evaluating whether a residue can become a controlled secondary resource.

The distinction also separates this topic from our earlier recycled-titanium charge analysis. Scrap pedigree asks whether titanium-bearing material is fit to re-enter a melt. The chlorinator case asks how a trace co-product survives sequential chemical separations.

A six-stage feed-to-product loss map

The map prevents denominator switching. Every percentage should identify the amount entering that step, the amount leaving in the desired stream, the impurities that followed, and the cumulative recovery from the original filtrate.

Scale-up begins where the abstract ends

The kinetic results were better described by a non-linear pseudo-second-order model with R2 = 0.890–0.938. That helps characterize sorption, but a model fit is not a plant guarantee. Commercial evaluation still needs multiple representative batches, reagent and water consumption, resin life, impurity variability, effluent treatment, product-specification consistency and a full cost and mass balance.

This analysis uses only claims in the publisher-deposited abstract because the MDPI article page returned HTTP 403 during this review. No unreported reagent condition, equipment detail or scale-up result is inferred.

The paper establishes a credible recovery path and a near-98 wt% final oxide in the reported test. Its stronger industry lesson is more restrained: the most attractive intermediate percentage is not the recovery story. The story is how much scandium crosses every stage and arrives as a specified final product.