Double Diffused Silicon Mesa 1956 Emitter Drive in

# Emitter Drive-In: Fixing the Junction Depth Without Letting It Catch Up to the Base

## 1. Why This Drive-In Has the Opposite Goal From the Base's

This step anneals the phosphorus dose banked in the previous step, but where the base drive-in was run long and hot enough to stretch a dopant profile into a deliberately steep gradient, the emitter drive-in is run only as far as needed to set a stable, well-defined junction depth — and no further. The emitter does not need its own graded field; it only needs to arrive at a fixed, shallow depth $x_{jE}$, positioned inside the base's existing gradient at the location chosen back in Step 6. Pushing the emitter drive-in too far in either temperature or time risks letting the emitter-base junction migrate deeper into the base than intended, eating into the region the drift field needs in order to act over any meaningful distance:

$$W_B = x_{jC} - x_{jE}$$

where $W_B$ is the effective base width the drift field has to work across, $x_{jC}$ the base-collector junction depth fixed back in Steps 3–5, and $x_{jE}$ the emitter-base junction depth this step is responsible for fixing. Because $x_{jC}$ is already set and cannot be revisited without discarding the wafer, this step's only free variable is $x_{jE}$ — and every minute of extra drive-in time shrinks $W_B$ by eating into it from the wrong side.

The Base Width This Field Acts Over Is Set Here, Not Before xᵀᴲ was already fixed; this step's only job is to not let xᵀᵉ drift too close to it depth, surface → collector base gradient, already fixed by Steps 3–5 xᵀᵉ, set by THIS step xᵀᴲ, fixed earlier, cannot move Wᵇ = xᵀᴲ − xᵀᵉ — the only distance the field has left to act across every extra minute of this drive-in shrinks Wᵇ by pushing xᵀᵉ rightward, toward a boundary that cannot itself move

## 2. Real Diagram: A Tight Time Window Compared With the Base's Wide One

The furnace schedule for this step has to be controlled more tightly than the base drive-in's was, because the base step had a wide acceptable range — any reasonably steep gradient would do, as long as it was steep. The emitter step has a narrow acceptable window: too short, and the junction may not be well-defined or repeatable wafer to wafer; too long, and the base width it depends on for the drift field collapses.

A Narrow Process Window, Not a Wide One the base drive-in tolerated a range; this drive-in tolerates a window drive-in time, increasing too short: junction not well-defined target window: xᵀᵉ repeatable, Wᵇ stays wide enough for the field too long: xᵀᵉ migrates toward xᵀᴲ, Wᵇ collapses drift advantage erodes even though junction looks fine this is why emitter drive-in time is specified far more tightly than base drive-in time was

## 3. Why the Base-Collector Junction Being Already Fixed Changes This Step's Stakes

The 1954 diffused-base germanium process this project has already documented also ran an emitter drive-in after an emitter predeposition, and in that process a slightly long drive-in mostly cost base width margin the device could often still tolerate, because no built-in field depended on that margin staying wide. Here, the base-collector depth was fixed back in Steps 3 through 5 specifically to support a target field strength, so this step inherits a boundary it cannot adjust and a budget — the remaining base width — that only shrinks from this point forward. The emitter drive-in time chosen here is effectively the last chance to decide how much of the drift field this device gets to keep.

Step 7 does not add new electrical behavior to the device; it decides, by how precisely it holds the line on $x_{jE}$, how much of the base width this process spent five earlier steps engineering actually survives to do useful work.

Take double diffused silicon mesa 1956 emitter drive in 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