Diffused Base 1954 Perform Base Predeposition

# Perform Base Predeposition: Metering a Dose This Process Will Spend the Rest of Its Schedule Protecting

This step actually runs the base dopant's predeposition — holding the wafer surface at the constant-source boundary condition Step 11 set up, for a time this step alone controls, to lay down a total dopant dose $Q_{\text{base}}$ that every later step in this sequence treats as fixed. The overview article covering this process's physics already describes the mathematics: a saturated surface concentration pinned at $C_s(T)$, held for a short time $t_1$, producing a dose that scales with $\sqrt{D t_1}$ regardless of how long predeposition actually runs beyond the point saturation is reached. What that overview does not dwell on is what this step specifically has to get right in practice — because this process's signature advantage, decoupling dose from depth, only holds if this step's dose is actually metered correctly before drive-in ever begins.

## 1. This Step's Only Real Output Is a Number, Not a Depth

$$Q_{\text{base}} = \int_0^\infty C(x,t_1)\,dx = \frac{2 C_s(T)}{\sqrt{\pi}} \sqrt{D t_1}$$

This step's entire deliverable is $Q_{\text{base}}$, the total dopant dose per unit area — a quantity that depends on predeposition time $t_1$ only through $\sqrt{D t_1}$, which means once the surface has actually saturated, running this step noticeably longer barely changes the dose at all. This is the opposite failure mode from an etch or lapping step, where running long directly spends a margin; here, running long mostly wastes furnace time without proportionally changing the one number this step is supposed to control; the real risk is running this step too briefly, before the surface has actually reached $C_s(T)$ everywhere Step 11's flux was supposed to deliver it.

## 2. Real Diagram: Dose Saturates Early; Depth Is Not Decided Here at All

Dose Plateaus; This Step Has One Job, Not Two junction depth is Step 14's responsibility entirely — this step only has to meter the dose correctly predeposition time, t₁ → dose laid down, Q_base dose plateau — this step's real target minimum correct stopping point too brief — surface never actually saturated once past the plateau, extra time here barely moves the dose, because this is not the depth-setting step

## 3. The Dose This Step Sets Becomes the Fixed Numerator in Every Later Depth Calculation

$$x_{j,\text{base}} \approx 2\sqrt{D_{\text{base}}\,t_2} \quad \text{requires } Q_{\text{base}} \text{ already fixed before } t_2 \text{ is chosen}$$

Step 14's drive-in depth calculation treats $Q_{\text{base}}$ as a known, fixed quantity already supplied by this step, which means any error in this step's actual delivered dose propagates directly into Step 14's depth calculation as a silent, unflagged input error rather than a visible failure of its own. This step therefore carries a verification obligation this process's sheet-resistance checkpoint, run immediately after this step completes and before drive-in begins, exists specifically to catch — confirming the dose this step actually laid down matches the target Step 2 and this step's own furnace recipe intended, before that number gets locked in as an assumption for every calculation downstream.

An Error Here Travels Silently Into Step 14's Calculation this step's own dose error looks, downstream, like a drive-in error instead this step dose, Q_base Step 14 depth, assumes Q_base correct without the sheet-resistance checkpoint between these two steps, a dose error is indistinguishable from a drive-in error

## Perform Base Predeposition's Place in the Process Lineage

Performing base predeposition is step twelve of Bell Labs' twenty-six-step diffused-base manufacturing sequence — immediately after the furnace atmosphere was established, and before any dose-verification checkpoint or drive-in anneal. It is the step that actually meters $Q_{\text{base}}$, the one quantity this process's entire double-diffusion advantage depends on being correctly fixed before depth ever enters the calculation, and an error here propagates invisibly into Step 14's depth calculation rather than announcing itself as its own distinct failure. Step thirteen, verifying this step's delivered dose before sealing the wafer for drive-in, exists specifically to catch that error while it can still be addressed.

Take diffused base 1954 perform base predeposition 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