Planar Process 1959 Verify Junction Still Oxide Covered

# Verify the Junction Is Still Oxide-Covered Everywhere Outside the Contact: Testing This Series' Entire Premise

## 1. Why This Inspection Checks an Idea, Not Just a Dimension

This step inspects the field oxide across the finished wafer for pinholes, thinning, and any place the contact-window etch or the metal etch might have crept farther than intended, and it does so for a reason no earlier inspection in this project has had: every prior measurement step verified a dimension against a specification, while this one verifies whether the entire founding premise of this series — that the junction never has to be exposed — is still actually true on the wafer sitting in front of it. The 1957 series inspected field oxide for pinholes at its own Step 12, worried that a pinhole would admit an unwanted diffusion; this inspection reuses that exact technique, worried instead that a pinhole, or an unexpected bite from the metal etch, might reopen the one exposure this entire series exists to prevent. The margin this inspection measures is the accumulated outcome of every design decision this series has made so far, now checked as a single number:

where the first parenthesis is the covered region Step 2 and Step 3 established, the second is the contact window's final extent from Step 4, and $\delta_{\text{metal etch}}$ an additional bias this step must account for from Step 6's metal etch, which can occasionally attack the field oxide at a slow but nonzero rate. If $m_{\text{coverage}}$ is positive everywhere this inspection looks, the junction is genuinely still covered. If it is not, somewhere on this wafer the premise this series was built on has quietly failed.

## 2. Real Diagram: The One Failure This Entire Series Exists to Rule Out

If this inspection finds a site where the margin has gone negative, the cross-section at that site looks exactly like the thing this whole series was invented to avoid — a junction edge reachable from outside the device, indistinguishable from the exposed mesa sidewall the 1958 series spent nine steps fighting.

## 3. Why This Inspection Did Not Exist Before This Series Had an Idea to Protect

The 1957 series inspected oxide windows at Step 12 because a bad window threatened a diffusion. The 1958 series characterized surface leakage at Step 6 because an exposed junction was already an accepted, unavoidable fact of that process's geometry, and the measurement existed to quantify a cost already being paid. This inspection is different in kind from both: it exists to confirm that a cost this series was specifically designed not to pay has, in fact, not been paid. No measurement earlier in this project's history was framed around verifying the absence of a problem the process itself was built to structurally prevent — every earlier inspection accepted some risk as a given and measured around it. This step instead checks whether the risk ever had to be accepted at all.

Step 7 does not improve the wafer; it confirms, measurement by measurement, that everything this series set out to achieve starting at Step 1 actually happened the way it was supposed to.