Planar Process 1959 Deposit Metal Planar Surface
# Deposit Metal Across the Entire Planar Surface: Why Flatness Is a Property Worth Fighting For
## 1. Why a Step Height Is an Enemy of an Evaporated Film Long Before It Is an Enemy of Anything Else
This step evaporates a thin aluminum film across the entire wafer in one operation, coating the field oxide, the contact window, and the thin step between them uniformly — and this step is only straightforward because this series never cut a mesa, which means the wafer this metal lands on is, to the resolution this process cares about, flat. Metal evaporated from a source in one direction arrives at a surface with a thickness that depends on the angle between that surface and the source, and a surface tilted steeply away from the source receives a film that thins toward nothing as the angle approaches grazing incidence:
where $t_0$ is the film thickness on a surface facing the source directly and $\theta$ the angle between the surface normal and the source direction. A flat field oxide and a flat contact-window floor both present $\theta \approx 0$ and receive full thickness; the gentle isotropic-etch sidewall at the window's edge, sloped rather than vertical, presents a modest $\theta$ and still receives a workable film. A mesa sidewall, cut nearly vertical by the 1956 and 1958 series' own etch, would have presented $\theta$ near ninety degrees at exactly the place a contact lead most needed to cross it — thinning the metal to the point of cracking or open-circuiting, a failure this series never has to confront because the surface it is evaporating onto never has that cliff to begin with.
## 2. Real Diagram: A Blanket Film That Needed No Topology to Fight
The cross-section after this step shows uniform metal everywhere — thick over the field oxide, thick down into the contact window, thick across the modest slope between them. Nowhere on this wafer does the metal cross anything resembling a cliff.
## 3. Why This Step Could Not Have Existed, as Written, on Either Earlier Series' Wafer
The 1956 and 1958 series both eventually needed a metal contact, and both had to form that contact on a wafer whose surface stepped sharply down at every mesa boundary — a topology that made any blanket metal deposition risky at precisely the sidewall a lead had to cross to reach the mesa's base or emitter contact. This series never produces that sidewall at all, which means the single biggest hazard those earlier metallization efforts had to manage by careful process control, this step simply does not encounter. The planar surface this series has built since Step 1 was not designed with metal deposition in mind — it was designed to keep the junction covered — but it happens to make this step almost trivial as a side effect, which is exactly the kind of compounding benefit a genuinely better geometry produces without anyone having asked for it.
Step 5 does not introduce a new metal or a new deposition technique; it is the first metallization step in this project's history that gets to assume the surface underneath it has no cliff worth worrying about.