MOSFET 1962 Field Effect Surface State Finally Solved

# Why a Field Effect, Attempted Since 1947, Finally Works: The Surface-State Problem Bardeen Found While Looking for It

## 1. Why a Voltage on a Gate Electrode Used to Do Almost Nothing

A field-effect transistor is, in principle, the simplest active device imaginable — a voltage on a gate electrode should reach through an insulator and directly modulate how easily current flows beneath it — and Shockley tried to build exactly that device first, years before the point-contact transistor this project's own 1947 series documented, and it did not work. The applied gate field, instead of bending the semiconductor's own energy bands and controlling its surface conductivity, was absorbed almost entirely by a dense population of electronic states trapped at the semiconductor's own surface — states that could fill and empty as fast as the gate voltage changed, soaking up the field like a sponge before it ever reached the bulk material beneath. It was Bardeen's own investigation into exactly why this failed, searching for the physical cause of the surface states defeating Shockley's field-effect device, that led directly to Bardeen and Brattain's accidental discovery of point-contact transistor action instead. The field effect itself sat unsolved for over a decade afterward, waiting for a semiconductor-insulator interface clean enough to let a gate voltage actually reach the material beneath it.

$$\Delta\psi_s \approx \frac{C_{\text{ox}}}{C_{\text{ox}} + qD_{\text{it}}}\,\Delta V_G$$

where $\Delta\psi_s$ is the resulting change in surface potential for a given change in gate voltage $\Delta V_G$, $C_{\text{ox}}$ the oxide's own capacitance per unit area, and $D_{\text{it}}$ the density of electronic states trapped at the semiconductor surface — when $D_{\text{it}}$ is large, the denominator is dominated by the trap term, and $\Delta\psi_s$ collapses toward zero no matter how large $\Delta V_G$ is made, because the gate voltage is spent filling and emptying surface traps rather than bending the semiconductor's own bands; only a surface clean enough to make $qD_{\text{it}}$ small compared to $C_{\text{ox}}$ lets the gate voltage actually do its intended work.

A Gate Voltage, Spent on Traps Instead of the Channel the same applied field, two very different interfaces beneath it HIGH Dit, 1947-STYLE SURFACE dense trapped states absorb the field Δψs ≈ 0 — bands barely move this is the surface Bardeen was investigating LOW Dit, THERMALLY OXIDIZED SILICON few trapped states, the field passes through Δψs tracks ΔVG directly this is the interface Atalla and Kahng found Δψs ≈ Cox ÷ (Cox + qDit) · ΔVG the same equation, two wildly different outcomes depending on one surface property

## 2. Real Diagram: A Gate, an Insulator, a Semiconductor — and the Interface That Decides Everything

The simplest possible structure capable of demonstrating a working field effect is a metal gate separated from a semiconductor by a thin insulating layer — a structure this project will build in later steps of this series — with the single property that decides whether it works at all located entirely at the boundary where the insulator meets the semiconductor beneath it.

The Minimal Structure, and the One Interface That Matters gate, insulator, semiconductor — everything in this series depends on the boundary in the middle metal gate electrode thermally grown insulator, built later in this series the interface — Dit lives here silicon semiconductor the surface potential here is what the gate is trying to control every step after this one in the series depends on this one interface staying clean

## 3. The Discovery That Happened Because This Device Didn't Work Yet

This project's own 1947 series documented the point-contact transistor, discovered by Bardeen and Brattain at Bell Labs — but that discovery did not begin as a search for a point-contact device at all. It began with Shockley's own attempt to build a field-effect amplifier, a device whose entire operating principle is the one this article has just described, and Bardeen was assigned to understand theoretically why Shockley's device produced far less amplification than the physics seemed to promise. Bardeen's answer was the surface state: a population of electronic states trapped at the semiconductor's own surface, dense enough to absorb an applied field almost entirely before it could modulate the bulk material beneath. That same investigation, probing the semiconductor surface with a metal point under bias, led directly to the observation of point-contact transistor action — a genuine, useful amplifying device, but a structurally different one from the field-effect device the investigation had originally set out to build. The field effect itself remained unsolved for more than a decade afterward, not because the underlying physics was wrong, but because no one had yet found a semiconductor-insulator interface clean enough to make $D_{\text{it}}$ small enough to matter. A thermally grown oxide on silicon, refined through this project's own 1957 series for an entirely different purpose — masking diffusions, not gating a channel — turns out to be exactly that interface.

Step 1 does not build a working field-effect device yet; it establishes why every attempt before this series failed, and names the single surface property every later step in this series exists to control.

Take mosfet 1962 field effect surface state finally solved further

Ask the copilot about this term, or have our engineers assess it against your process.

Ask CFSGPT about this → Request a technical assessment → Open the CFS app