Laser-cut kitchen foil does the work of a terahertz polarizer that costs thousands

The material was never the expensive part of a wire-grid polarizer. What costs money is how the wires get made: lithography inside a cleanroom, or a precision machine that winds microscopic tungsten filament one strand at a time, which the TMOS team likens to threading an incandescent bulb ^[6]. Both are capital, and capital gets amortised into a per-unit price in the thousands ^[1]. Laser ablation moves the scarce input from a machine tool to a recipe, in this case a pulse-energy window that took months to locate: enough to cut the foil, not enough to buckle the wires it leaves behind ^[7]. Recipes travel in journals. Winding machines do not.

Fifteen seconds is the figure for the small part ^[3]. The large-area process the group also developed runs one to two minutes across a range of metals ^[5], so four to eight times the fast case ^[14]. Neither number is the bottleneck. Characterisation is.

That is where the phys.org account goes quiet: it carries no extinction ratio, no transmission figure, no wire pitch and no frequency band ^[15]. A lab that fabricates its own component can measure it on the bench it already owns and move on. A buyer paying thousands for a catalogue part is paying for the datasheet and the tolerances printed on it, and that document is the one still absent. Until it exists, this displaces polarizer purchases inside the group that invented it, not inside a procurement system.

The team is candid about the second gap. Doctoral student Oleg Kameshkov, who led the design, says the fabrication problems are solved and the trade-off between mechanical stability and optical properties is what remains ^[9]. Aluminum foil, on the group's own account, is not robust enough for a commercial product, which is why tungsten and copper are next ^[8]. The freestanding construction with no glass or plastic behind it ^[4] is part of why the thing is nearly free, and it is also the reason durability is an open question rather than a solved one. The cheap device and the shippable device may not turn out to be the same device.

None of this reads as an accident of curiosity. The idea came out of an internal TMOS pitch competition last year explicitly aimed at commercial potential ^[10], the fabrication was done by Vladlen Shvedov at ANU ^[11], and Kameshkov describes the foil work as the first experiment with the simplest material, already followed by industry-grade metals including tungsten ^[13]. The framing around beyond-6G is the long horizon ^[12]. The near-term customer is duller and more real: spectroscopy, imaging and materials groups that already buy these parts ^[12], and that break them.

Professor Ilya Shadrivov, a co-author of the study in Optics and Laser Technology, says the group looked at how the components were made and asked whether they could be made cheaper and faster ^[2]. On fabrication time and bill of materials, the answer in the paper is yes. On whether the result meets a spec someone else wrote, the answer has not been published.