Alloy Junction 1952 Heat the Assembled Device

# Heat the Assembled Device: Two Disks That Don't Start From the Same Place

This is the step where step eleven's peak temperature finally gets reached for real, and where the dissolution budget that step fourteen's brief tack firing barely touched actually starts being spent in earnest — but the two disks heating up together are not starting from equivalent states. The collector already carries a shallow tack bond from step fourteen, a small head start toward the liquid state; the emitter, placed cold at step sixteen and never fired at all, has to reach melting from scratch. Bringing "both indium disks into the liquid state" sounds like one simultaneous event, but it's actually two separate melting processes racing to the same finish line from different starting points.

## 1. Heat Reaching the Far Face Lags Behind the Near Face by a Diffusion Time

$$\Delta t_{\text{lag}} \sim \frac{t_{\text{slice}}^2}{\alpha}$$

If the furnace supplies heat asymmetrically — more strongly from one side than the other — the face farther from the heat source reaches melting temperature later than the near face, delayed by a lag time that scales with the square of the wafer's own thickness, $t_{\text{slice}}$, divided by the material's thermal diffusivity $\alpha$. This is the same diffusion-time scaling that governed how deep step fourteen's brief firing could penetrate, now applied in the other direction: instead of asking how far heat gets into the wafer in a given time, this asks how long heat takes to cross all the way through it, from one disk's face to the other's.

## 2. Real Diagram: A Head Start on One Side, a Cold Start on the Other

Two Disks, Two Different Starting Conditions both have to reach the same liquid state by the time real dissolution begins germanium wafer collector tack-bonded already small head start toward liquid emitter placed cold, never fired starts from nothing a furnace heating only one side widens this gap instead of closing it

## 3. The Lag Has to Stay Small Against the Firing's Own Duration

$$|\Delta t_{\text{lag}}| \ll t_{\text{fire}}^{\text{main}}$$

For the two junctions to begin their growth clocks at nearly the same moment, the front-to-back thermal lag has to stay far smaller than the main firing's total scheduled duration, which is why this process favors furnace designs that heat both faces as symmetrically as practical rather than relying on conduction alone to carry heat from a one-sided source. Left uncorrected, an asymmetric lag doesn't just delay one side's melting — it skews step six's governing relationship directly, because a junction that starts dissolving later ends up shallower than its nominal schedule intended even if both sides are held at the same final temperature for the same total time, simply because one of them got less of that time to actually work with.

Temperature at Each Face Over Time the lag between the two curves is time one junction never gets to dissolve in time → face temperature indium melting point collector face, head start emitter face, cold start Δt_lag keeping this gap narrow is what symmetric heating actually buys

## Heat the Assembled Device's Place in the Process Lineage

Heating the assembled device is step twenty of RCA's forty-two-step alloy-junction manufacturing sequence — immediately after the furnace atmosphere was established, and before the germanium faces are actually wetted at depth. It is the step that finally reaches step eleven's peak temperature for the main dissolution cycle, and it has to manage a genuine asymmetry between the collector's tack-bonded head start and the emitter's cold start so that both junctions begin growing on nearly the same clock. Step twenty-one, wetting the germanium faces, is the step that follows only once both disks have actually reached the liquid state this step is responsible for delivering.

Take alloy junction 1952 heat the assembled device further

Ask the copilot about this term, or have our engineers assess it against your process.