The Limits of Silicon: Quantum Tunneling and GAAFET at 2nm

One of the biggest illusions of modern technology is the belief that chips can keep shrinking forever. For decades, the billions of transistors inside our smartphones and AI servers have halved in size with every generation, carried along by Moore's Law. Today, however, semiconductor engineering has run into a dark frontier where classical physics ends and the rules of quantum mechanics take over: the 2-nanometer threshold.

At this scale, a transistor gate is only a few dozen silicon atoms thick. And once you get down to that thickness, electrons stop behaving like well-mannered particles that a classical switch can simply turn off. They begin to "teleport" straight through barriers to the other side. We call this phenomenon quantum tunneling.

1. The Crisis: Electrons Leaking Through a Closed Gate (Subthreshold Leakage)

A transistor's job is simple: it either passes or blocks current to represent 1 and 0. But once the gate barrier shrinks to a few nanometers, Heisenberg's Uncertainty Principle comes into play. An electron's position can no longer be pinned down precisely; its wave function spills beyond the potential barrier.

In engineering terms, this means one concrete thing: the transistor leaks current even when it is switched off. We call this leakage current. A processor whose billions of transistors are leaking electricity at every moment drains the battery in seconds even while idle, overheats, and eventually burns out the silicon through thermal runaway.

"Below 3 nanometers, the problem isn't getting electrons to move; it's getting them to stop when you want them to."

2. The Collapse of FinFET: Why It's No Longer Enough

For the past 10 years, the technology that kept the industry going was the FinFET (Fin Field-Effect Transistor) architecture. Unlike planar transistors, a FinFET raises the conduction channel into a three-dimensional "fin" and wraps the gate electrode around three sides of it.

Wrapping the channel on three sides gave excellent electrostatic control over the electrons. At the 3nm node, however, the base of the fin (the exposed fourth side) turned into a leakage path where electrostatic control weakens. Even with the gate voltage pulled to zero, the current leaking through the base could no longer be stopped.

3. The Solution: The GAAFET and Nanosheet Revolution

The solution is called Gate-All-Around FET (GAAFET). Instead of keeping the fin vertical, engineers split it into ultra-thin silicon ribbons (nanosheets) stacked horizontally on top of each other.

This is where the most critical engineering victory lies: the gate material passes not only over and beside these horizontal ribbons but underneath them as well, surrounding the conduction channel a full 360 degrees. In a channel enclosed by the gate on all four sides, there are no blind spots left for electrons to escape through. Despite the quantum tunneling limit, the electrostatic field gains full command over the channel.

Engineering Verdict: What Comes After Silicon?

The GAAFET architecture has bought us breathing room for a few generations at the 2nm and 1.4nm levels (Intel 18A / TSMC N2). But this is silicon's last stronghold. Once we reach the limits of atomic diameters, silicon will have to make way for two-dimensional carbon nanotubes (CNTs) or 2D monolayer crystalline materials such as molybdenum disulfide (MoS₂).

The road to 2 nanometers is not just a race to shrink; it is the greatest materials engineering battle humanity has ever fought against the laws of quantum physics, at the atomic level!