Printing metal on an extrusion machine is usually a three-part job: print a part from a polymer-bound metal feedstock, get the polymer out, then sinter what remains into solid metal. The middle step, debinding, is where much of the difficulty sits. With aluminium it is especially awkward, because the metal oxidises easily and the oxide film works against good sintering. A paper published on October 15, 2026 in Progress in Additive Manufacturing gives one answer. Researchers led by Empa compared the usual solvent-plus-thermal debinding route for AA6061 aluminium feedstocks with a single-step wick debinding process in a powder bed. They concluded that the wick route is preferable. Their best parts reached 97.9% relative density.
The work was a collaboration between Empa and AGH Krakow in Poland and FHNW in Switzerland. 3D Printing Industry lists the authors as Montakhab, Linchard-Syta, Hadian and Clemens; the journal page also names Kopyściański. An earlier version appeared as a Research Square preprint on May 12, 2026, with the same core findings.
The setup
The team made feedstocks from gas-atomised AA6061 powder at a 60 vol% solids loading. AA6061 is the general-purpose structural alloy machinists will recognise. They mixed the powder with three commercial binder systems: EnCeram from Budenheim, Kcmix3.3 and EmbemouldCC. The third did not get far. According to 3D Printing Industry's account, EmbemouldCC separated during extrusion and was dropped. That left two binders, and each was carried through both debinding routes.
Test parts were 20 mm discs printed on a Tumaker Voladora NX+. A disc is not a demanding geometry. The study is about the chemistry and thermal processing after printing, not about printability, and you should read it that way.
Why aluminium makes debinding hard
Debinding is usually a staged process. A solvent step dissolves the soluble part of the binder and opens pore channels. A thermal step then burns out or decomposes the backbone polymer that held the part's shape. Aluminium leaves little margin for that second step.
The paper's abstract says the powder starts to oxidise at around 400 °C, so the team kept thermal debinding below that temperature. That narrows the process window considerably. The binder has to come out at temperatures where some polymers are still reluctant to leave, and every extra degree risks growing oxide on the particle surfaces.
The solvent bath was not oxidation-free either. The authors report that even high-purity acetone caused slight surface oxidation, particularly at the printed sample surfaces. For a metal whose sinterability depends heavily on surface condition, that is a real cost, and it is a central part of their argument against the solvent route.
Wick debinding, one step
The alternative was wick debinding. The printed part is buried in a bed of alumina powder and heated. The powder draws the liquid binder out of the part by capillary action. There is no solvent bath and no separate solvent stage before the thermal step. Everything happens in one heated bed.
Both routes worked in terms of the main measure. The abstract states that both removed at least 96% of the binder. The detailed numbers favour the wick process. According to 3D Printing Industry, 24 hours of wick debinding at 350 °C removed more than 97% of the binder. Going up to 400 °C made things worse: in a separate test the aluminium powder gained about 0.25% in mass over 24 hours at that temperature, which is consistent with the oxidation onset the authors identified, and parts debound at 400 °C sintered to lower density. In this system, running the debind hotter does not help. It moves the process into oxidation.
The density numbers
Binder removal only matters if the part sinters well afterwards. The team sintered in nitrogen at 635 °C. Here the gap between the two routes is clear.
With solvent debinding, the reported densities were about 75% for the EnCeram feedstock, which was soaked in water, and about 94% for Kcmix, which was treated in high-purity acetone. The report says the water turned the EnCeram powder surface into hydroxide, which became oxide on heating and left a porous outer layer.
With wick debinding, a 2-hour sinter gave 96.7% relative density for Kcmix and 93.6% for EnCeram. Extending the sinter to 10 hours raised those figures to 97.9% and 96.8%. Kcmix was the better binder on both routes. The wick process narrowed the gap between the two binders a great deal. EnCeram's result went from about 75% to nearly 97%.
| Route | EnCeram | Kcmix3.3 |
|---|---|---|
| Solvent + thermal debinding | ~75% | ~94% |
| Wick debinding, 2 h sinter at 635 °C | 93.6% | 96.7% |
| Wick debinding, 10 h sinter at 635 °C | 96.8% | 97.9% |
The cost is time. As 3D Printing Industry notes, the best result needs a 24-hour debind followed by a 10-hour sinter. This is not a quick turnaround process.
What It Means for Makers
If you have been following bound-metal extrusion, you know aluminium has been the material that is easy to promise and hard to deliver. Its oxide problem has kept it difficult. This paper does not solve that problem, but it shows a route that removes one step and one source of contamination.
Three points matter in practice.
The solvent step is not a harmless intermediate. The finding that even high-purity acetone slightly oxidised the powder is a useful result in its own right. Anyone trying to run reactive metal feedstocks through a hobby-scale solvent bath should take note. If laboratory-grade acetone causes measurable surface changes, a garage process is unlikely to be cleaner.
Temperature discipline matters more than speed. The jump from 350 °C to 400 °C brought oxidation mass gain and lower sintered density. For aluminium, the debinding furnace needs to stay accurately below the oxidation onset, and that is a requirement for furnace control, not just a recipe setting.
The binder choice drives the result. Out of three commercial binders, one failed at the extrusion stage. Another lagged badly on the solvent route and only caught up with wick debinding and a long sinter. A feedstock that prints cleanly is only the first part of the job.
Equipment is the obvious limit. A nitrogen-atmosphere furnace holding 635 °C for 10 hours, plus a powder bed for a 24-hour wick cycle, is not desktop hardware. The results are also from small discs. Nothing here tells us how larger or more complex parts would shrink, distort or densify, and the team measured only density and microstructure; mechanical properties still need evaluating. For now, this is a process-engineering result: it shows that, in AA6061, the simpler debinding route is also the better one.
Bottom Line
Replacing a two-stage solvent and thermal debind with a single wick step in an alumina bed gave the Empa-led team denser aluminium parts, up to 97.9% relative density. It also avoided the surface oxidation the solvent step introduced. The trade-off is long furnace time and a narrow temperature window. For extrusion-printed aluminium, the result is a clear point in favour of wick debinding.
Sources
- Material extrusion-based additive manufacturing of aluminum feedstocks with a focus on the debinding process (Progress in Additive Manufacturing, DOI 10.1007/s40964-026-01995-z)
- Swiss-Polish team skips the solvent bath, gets printed aluminium to 98% density (3D Printing Industry)
- FDM Printing of Aluminum Feedstocks with focus on the debinding process (Research Square preprint, May 2026)
