Lead Attachment and Can
# Lead Attachment & Can: Turning a Correct Slice Into a Testable Device
Everything that makes a grown-junction transistor electrically correct was decided before this step ever starts — base width, both junction depths, collector and emitter doping are all permanent facts about the slice by the time it arrives here. What lead attachment and can sealing add is not electrical function but *access to* that function: three fine wires bonded to the emitter, base, and collector regions so an external circuit can actually reach the structure the crystal already grew, followed by a hermetic can that protects the exposed germanium surface and the fragile wire bonds from the atmosphere, mechanical damage, and contamination for the rest of the device's working life. This is the step where a correctly-grown, correctly-sliced piece of crystal either becomes a transistor someone can solder into a circuit, or stays a laboratory curiosity that only a probe station can read.
## 1. Why Germanium Needs the Can More Than the Wires Need It
Germanium's exposed surface is far more electrically sensitive to contamination and ambient humidity than its bulk — surface states and adsorbed moisture create leakage paths that scale with surface-trap density $n_{\text{surf}}$ and surface conductivity $\sigma_{\text{surf}}$, both of which drift upward quickly in open air. A perfectly-formed junction, with exactly the base width and doping the earlier five steps worked to guarantee, can still show unacceptable leakage current or an unstable gain if its surface is left exposed to atmosphere for any meaningful time after slicing. The hermetic can is not a mechanical convenience added on top of a finished device — it is the step that makes the electrical specification computed from steps one through five actually hold once the device leaves the lab, by sealing the surface against the exact failure mode most likely to undo that specification.
## 2. Real Diagram: Three Wires, One Can, No New Electrical Function
## 3. Parametric Test Confirms, Never Corrects
The parametric test at the end of this step — measuring gain, leakage, and breakdown voltage on the finished, sealed device — exists to confirm that the device behaves as the five earlier steps' doping and geometry predict, not to fix it if it doesn't. If a tested device fails, there is no adjustment available at this stage: the base width was fixed at the second pellet drop, the junction depths were fixed by the melt-dosing timing, and the surface was just sealed into a can. A failing parametric test at this point identifies a defect that happened somewhere upstream — a dip that went wrong, a pellet drop that landed late, a slicing cut that wasn't quite perpendicular — and the only available response is to discard that specific device and attribute the failure rate back to whichever earlier step's tolerance it violated. This is also why parametric test data, aggregated across a full ingot's worth of devices, is the real feedback signal for the whole process: it doesn't fix the device in hand, but it's the only measurement that tells the operator whether the upstream timing targets in steps one through five are actually being hit.
## Lead Attachment & Can's Place in the Process Lineage
Lead attachment and can sealing is step six — the final step — of the six-step grown-junction transistor device process flow, after crosswise slicing has produced an individual device wafer. It adds electrical access and environmental protection to a structure whose doping and geometry were permanently fixed as early as the second pellet drop, and it ends with the one measurement in the whole process that can only confirm what upstream already decided, never correct it.