Alloy Junction 1952 Cool and Inspect the Crystal
# Cool and Inspect the Crystal: Where Step Three's Tolerance Band Gets Measured, Not Assumed
Step three managed a drift it could only anticipate; this step is the first point in the entire process where that drift becomes a measured number instead of a prediction. But measurement is only half of what happens here — before anyone probes the crystal electrically, it first has to survive its own cooldown intact, because a crystal with no junctions in it yet can still fail catastrophically in a way no later electrical test will ever reveal: it can crack, or accumulate enough dislocation density from thermal stress alone to be structurally unusable, regardless of what resistivity it would have measured. Cooling and inspecting the crystal is a two-part gate — survive the cooldown, then find out what the pull actually delivered — and neither part can substitute for the other.
## 1. Thermal Gradients During Cooldown Generate Stress That Can Exceed the Lattice's Own Strength
As the rod cools, different cross-sections along its length and radius contract at different rates because they pass through the same temperature drop at different times — a steep local temperature gradient $dT/dx$ multiplied by the material's thermal expansion coefficient $\alpha_{\text{th}}$ and elastic modulus $E$ produces a real mechanical stress $\sigma_{\text{th}}$ purely from uneven cooling, with no electrical cause at all. If that stress exceeds the lattice's critical resolved shear stress, new dislocations nucleate or, in the worst case, the crystal cracks outright — a failure mode entirely independent of whether step three's resistivity drift stayed inside its tolerance band. This is why cooldown rate is itself a controlled parameter, not simply "let it cool": too fast a cooldown can destroy a crystal that was doped and pulled perfectly.
## 2. Real Diagram: Two Independent Failure Modes, Checked in Sequence
## 3. The Probe Sweep Turns Step Three's Predicted Band Into a Measured Curve
A four-point probe swept along the rod's length produces an actual resistivity-versus-position curve, which is the direct, measured answer to the question step three's pull schedule could only estimate in advance. Wherever $\rho_{\text{meas}}(g)$ falls outside the tolerance band carried over from step three, that section of the crystal is flagged as out-of-specification — not discarded yet, since a narrower cut from a borderline section can sometimes still be usable, but marked for the decision the very next step in this sequence actually makes. This step produces the data; it does not decide what to keep.
## Cool and Inspect the Crystal's Place in the Process Lineage
Cooling and inspecting the crystal is step four of RCA's forty-two-step alloy-junction manufacturing sequence — immediately after the single crystal has been grown, and before any section of it is selected for use. It is the step that first confirms the crystal survived its own cooldown intact, independent of any electrical consideration, and then converts step three's predicted resistivity drift into an actual measured curve. Step five, selecting usable crystal sections, is where that measured curve — and the structural pass/fail from this same step — actually gets acted on.