Indium Pellet Placement
# Indium Pellet Placement: Metering Volume Before the Furnace Ever Sees It
By the time a wafer enters the furnace for alloying, the single most important number in the whole process — how much indium is actually available to dissolve germanium — has already been fixed, twice, by a room-temperature stamping operation that never touches heat at all. The alloy-junction process's saturation-limited strategy depends on oversizing each indium pellet so that it runs out of dissolving capacity before furnace-timing variation can matter; but that strategy only works if "how much indium" is itself a precisely metered, repeatable quantity. Indium pellet placement is the step that meters it — not with a furnace, but with a punch, a calibrated foil, and an alignment fixture that has to get both faces of the wafer right at once.
---
## 1. Volume, Not Just Diameter, Is the Controlled Quantity
The alloy-junction process article's saturation-limited strategy rests on a specific inequality — the total germanium a pellet can dissolve is capped by how much indium volume is physically present, $V_{\text{Ge dissolved}} \le V_{\text{In}} \cdot C_s(T)$ — but that inequality only does useful manufacturing work if $V_{\text{In}}$ itself is a quantity someone actually controls, rather than whatever happens to result from however a pellet was formed. Indium pellet placement answers that by treating pellet volume as the product of two independently settable manufacturing parameters, fixed before the furnace is ever loaded:
Foil thickness is calibrated once, upstream, by whatever rolling or electroplating process produces the raw indium stock; punch or die-cutting diameter is then set per pellet type at room temperature. Because indium is exceptionally soft — among the softest metals used anywhere in semiconductor processing, with a Mohs hardness barely above talc — die-cutting it to a precise diameter requires no special heating or exotic tooling, which is a large part of why the alloy-junction process could meter pellet volume this precisely using 1950s-era mechanical stamping rather than anything resembling a furnace-controlled deposition step.
---
## 2. Why Indium Cold-Welds to a Clean Surface on Contact
Indium's extreme softness produces a second practical consequence beyond easy stamping: pressed gently against a clean metal or semiconductor surface at room temperature, indium cold-welds — plastically deforming just enough at the contact interface to form an actual metallurgical bond without any melting at all. This matters directly for pellet placement, because a pellet that is merely resting against the wafer surface, with microscopic air gaps at the interface, would alloy non-uniformly once heated: the furnace's thermal energy has to reach an intimate, void-free contact for dissolution to begin evenly across the pellet's full footprint, and indium's cold-welding tendency is what lets a simple mechanical placement step — light pressure, no heat, no adhesive — reliably deliver that intimate contact before the wafer ever enters the furnace.
---
## 3. Double-Sided Registration: Aligning a Pellet Through an Opaque Wafer
Because the emitter pellet sits on one face of the wafer and the collector pellet sits on the opposite face, placing them requires aligning two components the operator cannot see simultaneously through an opaque germanium wafer. Production fixtures solved this with mechanical registration — jigs holding the wafer at a fixed position with alignment pins or edge references common to both the top and bottom placement steps, so that each pellet's center is set relative to the same fixed reference rather than relative to the other pellet directly. Registration error between the two faces does not change either pellet's own diameter or volume, but it does break the device's rotational symmetry: current crowds toward whichever side of the perimeter the two regrowth regions happen to overlap most closely, producing a non-uniform effective base width around the junction's circumference rather than the single, well-defined $W_B$ the rest of this process flow is built around measuring and controlling.
---
## Indium Pellet Placement's Place in the Alloy-Junction Process Flow
| What Placement Fixes | What It Feeds Downstream |
|---|---|
| Foil thickness × punch diameter | Total $V_{\text{In}}$ — the saturation-limiting dissolving capacity the soak phase depends on |
| Area ratio (collector dot larger) | Collector-to-emitter area matching for carrier collection |
| Intimate, void-free contact | Uniform dissolution onset across the pellet footprint once heated |
| Through-wafer registration | Rotational symmetry of the finished base width around the junction |
Read indium pellet placement through a *metering-before-melting* lens rather than a *loading step* lens: every parameter the furnace-alloying step later depends on for saturation-limited, timing-insensitive control — how much indium is available, how evenly it contacts the wafer, how concentric the two faces are — was already decided here, at room temperature, by a stamping operation that never gets a second chance once the furnace door closes.