Double Diffused Silicon Mesa 1956 Hermetic Seal Encapsulation

# Hermetic Seal: Closing the One Gap the Can From Step 17 Still Left Open

## 1. Why Placing a Can Over the Die Is Not the Same as Sealing It

Step 17 positioned a metal can over the mounted die, but a can that is merely placed over an assembly still has a seam where it meets the header — a path by which ambient moisture, airborne contaminants, or ordinary atmospheric gases can reach the die cavity, and this step closes that seam with a continuous weld or solder joint around the entire can-to-header interface, converting a mechanically protective cover into a hermetically sealed enclosure. This distinction matters specifically because of work done many steps earlier: the sidewall passivation from Step 11 reduced leakage from the mesa's exposed junction edges, but that passivation layer was never meant to be the device's only defense against moisture over its entire service life — it assumes a dry, stable cavity around it. A seal that leaks even slowly reintroduces the exact surface-state leakage mechanism Step 11 worked to suppress:

$$P_{\text{He}} = P_{\text{He},0}\, e^{-t/\tau}$$

a standard way this era's hermeticity is actually verified: pressurizing the sealed can with helium and measuring its leak rate $\tau$ afterward, since helium's small atomic size makes it a sensitive probe for exactly the kind of fine seam leak that would otherwise go undetected until the device failed in service.

A Seam Left Open Undoes Work Done Seven Steps Earlier Step 11's passivation assumed a dry cavity; this step is what keeps that assumption true CAN PLACED, SEAM OPEN moisture path, around the full perimeter CAN SEALED, THIS STEP continuous weld/solder seam, no gap P₄ᵉ = P₄ᵉ,₀ · e⁻ᵧᴰₕ — leak rate τ is measured, not assumed, before this die is trusted

## 2. Real Diagram: A Test That Checks the Seal, Not the Transistor

Hermeticity testing is the first test this entire process applies to the finished package rather than to the junction structure inside it — it verifies that the can-to-header weld is complete, independent of whether the transistor underneath it is electrically good, bad, or has not yet been tested at all.

A Mechanical Test, Not Yet an Electrical One this step's pass/fail criterion has nothing to do with transistor performance helium chamber, pressurized atmosphere leak rate measured, compared against spec a sealed-but-electrically-dead die and a leaking-but-electrically-good die are both failures, for different reasons, caught by different later steps

## 3. Why Hermetic Sealing Protects a Different Vulnerability Than the 1954 Diffused-Base Process Faced

The 1954 diffused-base germanium process this project has already documented also required a sealed package to protect its finished dies from the environment, so hermetic sealing itself is not a new concept introduced by this process. What differs is what, specifically, this seal is protecting from re-exposure: germanium's narrower bandgap already made it more leakage-prone at a given temperature regardless of moisture ingress, so a seal failure there compounded an existing weakness. Here, silicon's wider bandgap means the bulk junction leakage this seal protects is already comparatively low — the dominant leakage risk this seal actually guards against is specifically the exposed mesa sidewall from Steps 9 through 11, meaning a seal failure in this process reopens a narrower, more specific vulnerability than it would have in the earlier germanium lineage, but one this process is nonetheless just as exposed to if this step is skipped or done poorly.

Step 18 does not add anything a later electrical test could directly measure in a healthy device; it protects the conditions every later electrical test will assume are still true.

Take double diffused silicon mesa 1956 hermetic seal encapsulation 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