In 1965, Gordon Moore predicted that the number of transistors on a chip would double every two years. For decades, engineers delivered on that promise by making transistors smaller and smaller. But around the early 2000s, they hit a wall—literally. The insulating barriers inside transistors had shrunk to just a few atoms thick, and electrons started doing something bizarre: they began passing straight through solid barriers that should have stopped them completely.

This wasn't a malfunction. It was quantum tunneling, and it threatened to end the decades-long march of Moore's Law.

The Barrier That Isn't

A transistor is essentially a switch. In the metal-oxide-semiconductor field-effect transistor (MOSFET)—the workhorse of modern electronics—a gate controls whether current flows from the source to the drain through a channel. Between the gate and the channel sits a thin insulating layer, typically silicon dioxide. This oxide barrier is what makes the switching possible.

The traditional MOSFET works by lowering energy barriers. Apply voltage to the gate, and you reduce the energy electrons need to flow through the channel. Remove that voltage, and the barrier rises again, blocking current flow. It's like adjusting the height of a dam to control water flow.

But there's a catch. As transistors shrink, that oxide layer gets thinner. By the 2010s, some gate oxides measured barely a nanometer—about ten silicon atoms stacked together. At these dimensions, classical physics breaks down. Electrons stop behaving like tiny billiard balls and start acting like waves. And waves can do something particles cannot: they can penetrate barriers.

This is quantum tunneling. An electron encountering a thin barrier has a calculable probability of appearing on the other side, even though it lacks the energy to "climb over" in the classical sense. The thinner the barrier, the higher that probability becomes.

When Your Switch Won't Stay Off

For transistor engineers, tunneling became a crisis. When a transistor is supposed to be "off," you want zero current flowing. But with barriers just a few atoms thick, electrons tunnel through even when the gate voltage says stop. This leakage current means billions of transistors on a chip constantly draw power even when idle, turning processors into pocket-sized space heaters.

The problem intensified as chip makers pushed toward the 5-nanometer node and beyond. At these scales, quantum tunnel current becomes substantial. Industry projections in the early 2010s suggested that by 2026, gate lengths would reach 5.9 nanometers—a regime where tunneling dominates and traditional MOSFETs might simply stop working as reliable switches.

Engineers tried several workarounds. They experimented with new insulating materials with higher dielectric constants, allowing thicker physical barriers while maintaining the same electrical effect. They redesigned transistor geometries, creating FinFETs and gate-all-around structures that better control the channel. But these were holding actions, not solutions.

The Judo Move

Then some researchers asked a different question: What if we stopped fighting tunneling and started using it instead?

Enter the Tunnel FET. Instead of trying to eliminate quantum tunneling, the TFET makes it the primary switching mechanism. The design flips the logic of conventional transistors. Rather than lowering an energy barrier to let current flow, a TFET maintains a high barrier and controls the tunneling probability through it.

The key is band-to-band tunneling. Semiconductors have two energy bands that matter for electronics: the valence band, where electrons are bound to atoms, and the conduction band, where they can move freely and create current. A gap separates these bands—an energy cliff that electrons normally can't cross without a significant energy boost.

In a TFET, the gate voltage aligns these bands at the source-channel junction. When aligned, electrons can quantum tunnel directly from the valence band to the conduction band, generating current. When misaligned, tunneling probability drops precipitously, and the transistor switches off.

The Low-Voltage Advantage

TFETs offer something conventional MOSFETs can't: sub-thermal switching. Traditional transistors need a minimum voltage to overcome the thermal energy distribution of electrons—about 60 millivolts per decade of current change at room temperature. This thermal limit has stubbornly resisted improvement regardless of materials or design.

Tunneling doesn't care about thermal energy. An electron either has the right quantum state to tunnel or it doesn't. This means TFETs can theoretically switch with much lower voltages, potentially reducing power consumption by orders of magnitude. For mobile devices and data centers where energy efficiency determines battery life and operating costs, this matters enormously.

The temperature dependence differs too. Conventional transistor current drops as temperature falls because fewer electrons have enough thermal energy to cross the barrier. Tunnel current shows weaker temperature dependence. At liquid nitrogen temperatures (77 Kelvin), tunneling can dominate completely, which opens possibilities for cryogenic computing where quantum effects enhance rather than hinder performance.

The Engineering Reality Check

TFETs aren't perfect switches yet. The same high barrier that enables steep switching also limits maximum current. When you need a transistor to drive high-speed signals, TFETs struggle. Their on-current remains lower than optimized MOSFETs, which means circuits run slower or need more space.

Material choice becomes critical. Silicon, the semiconductor industry's foundation, doesn't tunnel particularly well. Its indirect bandgap means electrons need to change momentum as well as energy to tunnel, reducing probability. Compound semiconductors like indium arsenide or gallium antimonide tunnel more readily, but integrating these exotic materials with existing silicon manufacturing poses immense challenges.

Device modeling gets complicated too. Engineers can't simply apply classical drift-diffusion equations. They need quantum mechanical models—Poisson-Schrödinger solvers that self-consistently calculate electron wave functions and energy distributions, or at minimum, quantum-corrected drift-diffusion models that approximate these effects. The computational demands increase substantially, and intuition developed over decades of classical device physics doesn't always apply.

Where Tunneling Takes Us

Quantum tunneling in transistors demonstrates a broader pattern in technology: barriers that seem absolute often aren't. When oxide layers reached atomic thickness and tunneling appeared to doom further miniaturization, the path forward required accepting rather than fighting quantum mechanics.

Today's processors already rely on quantum effects whether designers acknowledge it or not. Electrons in nanoscale devices occupy quantized energy levels. Their wave nature influences everything from threshold voltage to carrier mobility. The billions of transistors in a modern smartphone are fundamentally quantum mechanical devices.

TFETs represent an explicit embrace of that reality. Whether they become the next generation of mainstream transistors or remain specialized devices for ultra-low-power applications, they've already changed how engineers think about switching. The "wall" Moore's Law hit turned out to be a quantum tunnel with an exit on the other side.