A bipolar junction transistor has two p-n junctions sharing a thin base region, arranged n-p-n or p-n-p. In its forward amplification regime, the emitter junction injects minority carriers; many cross the base before electron-hole recombination and reach the reverse-biased collector junction.[ref]
William Shockley's proposal used transport through the bulk of a semiconductor, rather than coupling between two surface points. The base must be thin enough for injected minority carriers to reach the collector. Two separately wired diodes do not supply that shared transport region. The proposal was a design principle; producing a working device required further control of material properties.[ref]