Surface Barrier Transistor 1953 Philco Jet Etch

# The Surface-Barrier Transistor: Jet-Etching a Base Thin Enough to Reach Tens of Megahertz

The surface-barrier transistor, introduced by Philco in 1953 (with General Electric fielding a related rate-grown variant), solved the base-width problem every earlier germanium transistor had run into by abandoning mechanical thinning altogether. Grown-junction and alloy-junction devices set base width by how precisely a crystal could be grown or a disk could be aligned and fired — processes bottlenecked by lapping tolerance and firing-time control. Philco's answer was to stop cutting the base down and start dissolving it: two opposing jets of electrolyte, etching from both faces of a thin germanium wafer at once, hollow out a pit on each side until the undissolved germanium between them is thinner than any mechanical process of the era could reliably produce — then metal is electroplated directly into those etched pits, forming the emitter and collector exactly where the base is thinnest.

The frequency gain is not incidental; it is the entire point of making the base this thin. A transistor's upper frequency limit is set largely by how long an injected carrier takes to cross the base — transit time falls off faster than linearly as the base narrows, which is why a base a few micrometers thick, reachable only through electrochemical etching, pushed surface-barrier devices into the tens-of-megahertz range that lapped or diffused bases of the era could not reach. Every earlier germanium transistor in this project's own process histories — the point-contact device's spring-loaded whiskers, the grown-junction's pulled crystal, the alloy-junction's fired indium disks — set base width as a byproduct of a process built for something else. Jet-etching is the first process built to set base width directly, as the one variable the whole technique exists to control.

Fragility and low power are not separate defects alongside the frequency gain; they are the same thinness showing up in two other places at once. A base etched down to a few micrometers has almost no bulk material left to absorb mechanical shock, which is why surface-barrier units were notoriously delicate in handling and assembly compared with the thicker, mechanically robust grown-junction and alloy-junction devices built in this project's own process histories. That same thin cross-section also has little volume through which current can flow or heat can escape, which caps how much power the device can carry before localized heating damages the very region the whole process exists to make thin. A designer could not ask for the frequency response without also accepting both costs — they are three readings of one number, the base thickness, not three independent design choices.

Surface-barrier technology is the clearest demonstration in this project's own process history that a performance axis and a cost axis can be the same physical dimension measured twice. Every earlier device here reached its base width as a side effect of a process optimized for something else — crystal growth, disk alignment, firing time — and every one of them paid no particular fragility tax for it, because none of them pushed base width anywhere near a membrane. Philco's jet-etch process is the first to target base thickness directly, and it is precisely because that target was pursued without compromise that the resulting device became both the fastest and the most delicate germanium transistor of its era — not two separate outcomes, but one outcome described from two directions.