Alloy Junction 1952 Solidify the Residual Alloy Buttons

# Solidify the Residual Alloy Buttons: The Same Melt, Two Different Fates

Every step from 20 through 27 has followed the fraction of the indium-germanium liquid sitting directly against the original crystal — the fraction that recrystallized epitaxially and now carries the electrical junctions Step 27 just located. That was never the whole melt. Step 21 wet more of each germanium face than Step 22 ever dissolved into, and Step 13's disks were sized with real excess material to begin with — so once the regrowth interface at each face comes to rest, a visible excess of indium-rich liquid is still sitting on the *outside* of that interface, never touching the original lattice at all, with nothing left to regrow onto. Step 28 is what happens to that leftover liquid as the same furnace cooling that drove Sections 25 through 27 keeps running: it freezes too, but it freezes as a free-standing alloy button with none of the constraints Step 25's stability criterion placed on the junction regrowth, because it has no original crystal surface to continue.

## 1. No Template to Match, So No Instability to Avoid

Section 1 of Step 25 showed that the regrowth front *inside* the device is only stable if it advances slowly enough, relative to the local thermal gradient, to avoid breaking down into a defective cellular front — because a defective front there means dislocations and trapped indium pockets frozen directly into the electrically active base. The residual button has no such requirement, because there is no pre-existing lattice it needs to extend. Once its composition cools past the indium-germanium eutectic temperature $T_E$, the remaining liquid simply freezes as whatever two-phase mixture the phase diagram dictates at that composition — a fine-grained eutectic mixture of nearly pure indium and germanium-saturated indium, with no requirement that any of it continue the original crystal's orientation. The governing relationship is the same lever-rule mass balance used throughout this series' phase-diagram reasoning (echoing Step 3's Scheil treatment and Step 11's solidus-limited dose), applied here to the two phases coexisting below $T_E$ rather than to a single solid solution:

where $f_{\text{Ge-rich}}$ is the mass fraction of the button that freezes as the germanium-rich solid phase, $C_0$ is the bulk composition of the liquid remaining once regrowth stopped, and $C_{\text{In-rich}}$, $C_{\text{Ge-rich}}$ are the two phases' compositions at the eutectic. Whichever value $C_0$ actually lands on — a direct consequence of how much germanium Step 22's dissolution removed from the original melt composition Step 11 and Step 13 established — determines whether the finished button is mostly soft, nearly pure indium, or a harder, more germanium-rich mixture.

## 2. Real Diagram: Inside the Interface Versus Outside It

## 3. The Button Is Not Waste — It Is the Mechanical Anchor for Everything After It

It would be easy to treat the residual button as leftover material with no further role, now that the electrically meaningful work — Steps 25 through 27 — is finished. That reading misses what the button mechanically becomes: it is the only solid material bridging the finished junction region to the outside world at either face, and every later step in this process depends on it having solidified into something sound. A button riddled with shrinkage porosity (a real risk, since indium and germanium have different densities and the eutectic mixture shrinks on freezing like any cast alloy) or one that pulled away from the regrowth interface during cooling, rather than freezing in continuous contact with it, leaves a mechanically and electrically unreliable boundary exactly where Step 29's lead attachment will need a sound surface to bond to. A simple shrinkage-driven void-fraction estimate, following the same density-mismatch reasoning used for any binary casting, is:

where $\rho_{\text{liquid}}$ and $\rho_{\text{solid}}$ are the residual button's densities just above and just below freezing. A larger density jump across freezing means more shrinkage, more risk of an internal void or a pulled-away interface, and a weaker mechanical foundation for the finished device's electrical connections — which is exactly why this step, unglamorous as it sounds next to Steps 25 through 27's electrical content, still has a real pass/fail criterion of its own: a continuous, void-free, well-bonded button, not merely "the liquid that's left over has gone solid."

## Real Diagram: Where the Button's Composition Lands on the Eutectic

## Solidify the Residual Alloy Buttons's Place in the Process Lineage

Solidify the Residual Alloy Buttons follows Step 27, Complete Both Junctions, which located the electrical boundaries inside the device; it precedes Step 29, the first step of Phase 4, which will begin working with the finished assembly's exterior surfaces, starting with inspection of exactly the kind of structural and interfacial soundness this step's void-fraction criterion addresses. It is the eighth and final step of Phase 3, and the point where the single melt that Steps 19 through 24 heated, wetted, and dissolved finally splits into two permanently different materials: a single-crystal continuation of the original lattice carrying the junctions Step 27 found, and a free-standing eutectic button with no lattice to match at all — the same starting liquid, resolved into two entirely different microstructures purely by which side of the regrowth interface it happened to sit on.