MOSFET 1962 Define Source Drain Diffusion Windows

# Define Source and Drain by Diffusion Through Oxide Windows: Two Identical Regions, Not a Base and an Emitter

## 1. Why This Device's Two Diffused Regions Don't Need Different Names

This step opens two windows in the gate oxide Step 2 grew, and diffuses a heavily-doped region into each one using the identical planar diffusion technique the 1959 series established for base and emitter — but where 1959's two diffusions were deliberately different from each other, nested and asymmetric, with different doping levels performing different electrical roles, this step's two diffusions are identical: the same doping type, the same depth, the same diffusion schedule, separated by a gap rather than nested inside one another. Every bipolar device this project has documented needed a base and an emitter to be structurally distinguishable, because current flows through one into the other in a fixed direction defined by the structure itself. This device's two diffused regions have no such fixed roles: which one is called the source and which the drain depends entirely on which way the applied bias happens to push current, not on anything different about how the two regions were built. The structure itself does not know, or need to know, which terminal is which.

$$L_{\text{eff}} = L_{\text{mask}} - 2x_j$$

where $L_{\text{mask}}$ is the gap between the two oxide windows as drawn, $x_j$ the lateral diffusion extent spreading inward from each window's edge — governed by the same lateral-to-vertical diffusion ratio this project established back in 1959 for junction coverage — and $L_{\text{eff}}$ the actual electrically active channel length between the two diffused regions, always shorter than the mask drew it, because each diffusion spreads sideways under the oxide edge exactly as 1959 found a base or emitter diffusion does, only here that lateral spread eats directly into the one dimension this entire device's performance will depend on.

A Mask Gap, Then a Shorter Real Channel lateral diffusion spreads inward from both windows, symmetrically TOP VIEW, TWO SYMMETRIC DIFFUSIONS CLOSING IN diffused region A diffused region B Lmask, the drawn gap Leff, what's left xj xj Leff = Lmask − 2xj — the same lateral spread 1959 found, now eating the channel neither diffused region is distinguishable from the other by its own construction

## 2. Real Diagram: Two Identical Wells, a Channel Not Yet Controlled by Anything

The cross-section below shows the finished structure after this step: two symmetric, identically-doped diffused regions separated by an undisturbed gap of silicon still covered by the thin gate oxide — a region that will become an electrically active channel only once the next step adds a gate electrode above it.

Two Identical Regions, One Undisturbed Channel Between Them symmetric by construction — neither region is structurally the source or the drain yet gate oxide, grown in Step 2, thinned only over the channel diffused region diffused region channel — not yet gated substrate silicon beneath, opposite doping type to the diffused regions source versus drain is decided later, by which way the applied bias pushes current

## 3. The First Structurally Symmetric Device This Project Has Ever Built

Every bipolar device this project has documented since 1954 required its two outer diffused regions to be distinguishable from each other — a base and an emitter, or a collector and an emitter, built with different doping levels and nested in a fixed geometric relationship, because current in a bipolar device has a structurally fixed direction through a specific sequence of regions. The 1959 series' own planar diffusion technique, used to build those asymmetric regions, nonetheless taught this project the lateral-spread behavior this step now reuses directly: a diffusion driven through a mask window spreads sideways under the oxide edge by a fraction of its vertical depth, a fact 1959 needed to manage junction coverage and this step now needs to manage channel length instead. What is genuinely new here is not the diffusion technique but the geometry it is building: two regions of identical doping, identical depth, and identical role, separated by a gap rather than nested — a structure with no built-in asymmetry for current to follow, because this device's current direction is not fixed by construction at all.

Step 3 does not diffuse a base and an emitter under new names; it diffuses this project's first pair of structurally interchangeable terminals, leaving the one asymmetry this device actually needs — which terminal pushes current and which receives it — to be decided later, by bias, not by anything built into the silicon itself.

Take mosfet 1962 define source drain diffusion windows 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