Diffusion pump

The diffusion pump emerged because conventional vacuum pumps could not reach the pressures required for emerging technologies like radio tubes and particle physics—and Wolfgang Gaede recognized that a counterintuitive phenomenon could break through this barrier.

Textbooks taught that mercury pumps could not achieve vacuum lower than mercury's own vapor pressure. Gaede's experiments showed the opposite. If system dimensions were correct, mercury vapor flowing in one direction could produce a pressure drop the other way. Gas molecules could not diffuse against the vapor stream but would be carried along to the exhaust.

In 1915, working at the University of Freiburg, Gaede built a prototype where mercury was boiled and vapor flowed through narrow slits to condense beyond. The mercury vapor diffusion pump was the first vacuum pump with no moving parts—a radical simplification that enabled extreme pressures.

Gaede named it a diffusion pump because gas could not diffuse against the vapor stream. The design achieved ultimate vacuums around 10^-8 Torr (about one ten-billionth of atmospheric pressure), opening realms of vacuum previously inaccessible.

Independently and nearly simultaneously, Irving Langmuir at General Electric developed a similar device in 1915-1916. This convergent invention confirmed that the underlying physics had become accessible to multiple researchers. Langmuir's design featured larger nozzles with higher pumping speed, quickly adopted for industrial use.

Leybold commercialized Gaede's design rapidly. In 1928, Burch introduced oil diffusion pumps using low-vapor-pressure oils instead of mercury—safer and achieving even lower ultimate pressures. This became the dominant industrial vacuum technology for decades.

Gaede was the founder of high-vacuum technology, having invented the mercury rotary pump (1905), the molecular pump (1912), and the diffusion pump (1915). These devices made possible semiconductor technology, nuclear physics, electron microscopy, and vacuum deposition—industries that required pressures millions of times lower than any previous technology could achieve.

The diffusion pump's elimination of moving parts proved transformative: maintenance dropped dramatically while reliability soared. Gaede worked with Leybold from 1906 until his death in 1945, watching vacuum technology evolve from laboratory curiosity to industrial foundation.