Rate Grown Transistor 1954 Prepare Seed Crystal Compensated Melt

# Prepare the Seed Crystal: An Orientation Choice That Now Has to Survive a Melt That Never Simplifies

The seed's crystallographic orientation is fixed once, exactly as in both earlier grown-junction series, but here it has to hold up against a melt that is compensated for the entire run rather than starting simple and being complicated by a later pellet drop. In the 1951 germanium and 1954 silicon series, the seed touches down on a melt carrying only one dopant species — whatever complexity the junction eventually requires is still steps away. In this process, the moment the seed makes contact, it is already freezing against a liquid carrying both donor and acceptor simultaneously. That does not change what orientation to choose — ⟨111⟩ remains the stable, low-facet-risk axis for germanium in every series this project has documented — but it does raise the stakes on getting the choice right, because any instability in the freezing interface here will be read, later, as if it were a pull-rate-driven doping change rather than what it actually is: a seed-orientation defect.

## 1. The Same Orientation Equation, a Melt That Offers No Second Chance to Fix a Bad Choice

$$v(\theta) = v_{\langle hkl \rangle} \, f(\theta)$$

Growth rate anisotropy in germanium's diamond-cubic lattice is identical to what step three of the 1951 series already established: a ⟨111⟩-oriented seed grows a stable, axisymmetric interface, while an off-axis seed risks facet formation that disrupts diameter control. What's different is the consequence of getting it wrong. In the 1951 series, a facet-driven diameter wobble is a growth-quality problem, cleanly separable from the later pellet-drop composition change that forms the junction. Here, any growth-rate instability the seed introduces changes the local boundary-layer thickness $\delta$ that this project's pull-rate-dependent effective segregation coefficient depends on — meaning a seed-orientation defect can masquerade as an unintended, unscheduled doping shift, since both processes act through the same growth-interface physics.

## 2. Real Diagram: Why a Facet Defect Here Is Read as a Doping Defect Later

A Facet Defect and a Doping Shift Now Share One Root Cause in the 1951 series these two failure modes are unrelated; in a compensated melt they are not 1951 germanium, single-dopant melt bad seed orientation → facet defect a growth-quality problem only junction composition set later, independently, by pellet drop this series, compensated melt bad seed orientation → facet defect changes local boundary-layer thickness δ δ feeds directly into k_eff(v) reads as an unscheduled doping shift because this process reads composition from growth physics, growth defects are no longer composition-neutral

## 3. Mechanical and Chemical Preparation: Unchanged in Method, Tightened in Tolerance

$$\sigma_{\text{thermal}} \approx \frac{E \, \alpha \, \Delta T}{1 - \nu}$$

The seed is ground to a precise cross-section and etched to remove saw-damage micro-cracks that would otherwise seed thermal-stress fractures on contact with the melt, identically to the 1951 series's preparation and governed by the same thermal-stress relation. The germanium melt temperature is unchanged at 937°C, so the seed faces none of the radiative-flux severity that the 1954 silicon series's far hotter 1414°C melt introduces. What tightens here is dimensional tolerance on the seed tip itself: because the pull-rate schedule planned for later steps depends on reaching a known, repeatable initial growth velocity quickly after immersion, a seed tip whose geometry causes an extended, uneven thermal equilibration period delays the point at which $k_{\text{eff}}$ for both dopants settles into its intended steady-state value — stretching out exactly the transition region that later pull-rate modulation is supposed to control precisely.

Seed Tip Geometry Sets How Quickly k_eff Settles a longer thermal equilibration delays the point where the pull-rate schedule can be trusted time since immersion → k_eff, approaching steady state well-prepared tip: fast, clean settling poorly prepared tip: delayed, uneven settling later steps assume k_eff is already at steady state when the planned pull-rate schedule begins

## Prepare the Seed Crystal's Place in the Process Lineage

Preparing the seed crystal is step three of this project's rate-grown series — the same ⟨111⟩ orientation choice and the same mechanical and chemical preparation established in the 1951 germanium series, now carrying an added requirement that this project's earlier series never faced: the seed's geometry must produce a fast, clean thermal equilibration, because this process reads junction composition from growth-interface physics rather than from a separate, independently-timed pellet drop. Step four, loading the crucible, is where this prepared seed and the compensated melt from step two are brought together for the first time.

Take rate grown transistor 1954 prepare seed crystal compensated melt 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