Planar Process 1959 Oxide Already Covered Junction

# Recognize That the Masking Oxide Already Covered the Junction the Whole Time: The Insight, Not the Material

## 1. Why the Lateral Diffusion Everyone Already Knew About Was the Whole Answer

This step is not a physical operation on the wafer; it is the recognition, by Hoerni, of a geometric fact this project's own 1957 series had already measured and never thought to exploit — that a diffusion under an oxide window does not stop at the window's edge, but spreads laterally beneath the oxide by roughly eighty percent of its own junction depth, which means the place where the junction actually meets the silicon surface sits not inside the open window at all, but underneath the oxide that borders it. Every diffusion this project has documented since 1954 produced this lateral spread as an incidental, mildly inconvenient fact — the 1957 series measured it at Step 13 purely to budget for dimensional error in a second masking level. Nobody before Hoerni had asked what that same lateral spread meant for the one point on the wafer surface where a junction's own electric field terminates:

$$y_{\text{surface junction}} = y_{\text{lateral}} \approx 0.8\,x_j, \qquad y_{\text{surface junction}} > 0 \implies \text{already under oxide}$$

where $y_{\text{lateral}}$ is the lateral diffusion distance this project already quantified in 1957, and $x_j$ the junction depth. Because that lateral spread is strictly positive, the point on the surface where the junction meets it is never exactly at the drawn window edge — it is always some nonzero distance back underneath the oxide, which has been sitting there, uncelebrated, on every wafer this project has ever diffused.

The Junction's Surface Point Was Already Under Cover the same lateral spread 1957 measured, read for a different purpose CROSS-SECTION AT ONE WINDOW EDGE lateral spread, 0.8xj junction meets surface here, under oxide ysurface junction ≈ 0.8xj — always positive, always already under the oxide this fact was true in every wafer the 1954, 1956, and 1957 series ever diffused

## 2. Real Diagram: Why the Mesa Etch Was Cutting Through Protection That Was Already There

A mesa etch, by cutting a vertical sidewall straight down through the silicon, slices directly through the point this step identifies — the covered surface junction — and exposes exactly the location that was, before the etch, already protected. The insight is not that a new barrier needed to be invented; it is that an existing barrier was being deliberately destroyed by the very next step every earlier series performed.

Where the Mesa Etch Cut, and What Was Already There the exact point the old etch destroyed was the point this step just proved was already covered covered junction edge covered junction edge where a mesa etch would have cut, directly through this the old process was removing protection; it was never adding a new vulnerability by accident

## 3. Why This Is a Realization, Not a New Fabrication Technique

Every step so far in this project's seven series has introduced a new physical operation — a new furnace cycle, a new mask, a new etch. This step introduces none of that. The oxide this step relies on was already grown in Step 1 of this very series, using chemistry this project documented in 1957; the lateral diffusion this step relies on was already measured in that same 1957 series, at its own Step 13. What this step contributes to the project's history is purely conceptual: the realization that two facts this project had separately documented — oxide masks dopant, and dopant spreads laterally under the mask — combine, when neither mesa-etched away, into a junction that protects itself. No earlier series in this project reached this realization because every earlier series kept performing the one operation, the mesa etch, that made the realization pointless to notice.

Step 2 does not change the wafer at all; it changes what this project understands about what was already sitting on every wafer it has ever built, which is the only thing that needed to change for the rest of this series to exist.

Take planar process 1959 oxide already covered junction 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