Silicon is dying.
Not today. Not tomorrow. But the physics are running out of room. We’ve packed billions of transistors into modern processors, squeezing the element until it refuses to shrink any further. If we want faster computers, we can’t just keep making smaller versions of what we already have.
Scientists are looking at molybdenum disulumide (MoS2).
It’s part of the transition metal dichalcogenides (TMD) family. Atomically thin. Just three atoms high: molybdenum sandwiched between sulfur layers. Promising? Absolutely. The problem isn’t the material itself. It’s getting the stuff onto the chip without wrecking it.
The surgery is too precise
Manufacturers need to strip off the top sulfur layer. Leave the molybdenum alone. It’s delicate work.
Currently, they use plasma. High-energy particles. Same state of matter as the sun, a field the DOE’s Princeton Plasma Physics Laboratory (PPPL) has studied for 75 years. When those plasma ions hit the TMD surface, they knock sulfur atoms loose.
But physics is messy.
There’s almost no margin for error between “sulfur falls off” and “molybdenum gets destroyed.” Hit it too hard? Ruined wafer. Too soft? Nothing happens. Because the energy thresholds overlap, consistent manufacturing has been a nightmare.
Yury Polyachenko and his team found a cheat code.
They coated the material with oxygen or fluorine before hitting it with plasma. Computer simulations, published in The Journal of Physical Chemistry Letters, show the energy required to remove sulfur drops drastically. From roughly 30 electron volts down to 14 eV with oxygen. Even lower—about 10 eV —with fluorine.
Why does that matter?
Plasma isn’t uniform. Ions carry varying energy levels. On untreated material, that variation means inevitable collateral damage. By lowering the removal threshold, the gap widens. You can strip the top layer gently. The rest stays put.
“We are not directly breaking the bonds… We are forming some intermediate products, such sulfur dioxide. This intermediate product much easier to break off.”
Letting chemistry help out saves the hardware. When ions strike the oxygen-coated surface, oxygen grabs a sulfur atom. They bond to form sulfur dioxide —a stable gas that floats away on its own. Fluorine creates similar volatile compounds. It turns a blunt physical strike into a targeted chemical reaction.
Is it scalable?
Polyachenko, the study’s lead author and a Princeton grad student, is cautious about declaring victory too soon.
Right now, we know if the process causes damage. Next up: measuring how much damage occurs under different conditions. The real test will come when they swap out elements. Replace molybdenum with tungsten. Sulfur with selenium. See if this oxygen/fluorine trick works across the TMD family.
The research got support from the Department of Energy and various federal centers. Simulations ran at the National Energy Research Scientific Computing Center (NERSC) and Princeton University clusters.
If this holds up, the silicon barrier falls.
Maybe not instantly. Chips are complicated beasts. But for a moment, the future looks less like a dead end and more like a narrow doorway. One atomic layer thick.
Do you think TMDs will actually replace silicon?
Probably.
But not before we break a few more wafers first.
