Pressurized, tankless commodes are a relief in these days of low-flow toilets. The latter are so infamously ineffective as to be the topic of a recent and very humorous Dave Barry column.
The present case involved a pressure commode in a men's room at a shopping mall. The brass male fitting leading from the city water supply corroded through just where it was screwed into the female fitting from the commode. Major water damage resulted before the failure was discovered and the water turned off. I was hired by the insurers of a cleaning service, which was accused of causing the failure because they were the last ones in the lavatory prior to the flood. I was told that the cleaners could be held liable if they applied only the final nudge to a nearly failed fitting.
Simple metallography revealed the cause of the failure. The brass at the root of the first exposed thread of the male fitting had been replaced by a porous mass. The failure was a classic case of dezincification. The zinc had diffused to the outer surface, where it combined with oxygen. The remaining copper had degenerated to a porous composite of copper and copper oxide.
Corrosion occurred only at the first exposed thread, probably because a Cr-plated cover tube protected the rest. (I presume that the cover tube was for cosmetic reasons.) The tube probably even accelerated the corrosion, in that Cr is noble with respect to zinc.
What provided the corrosive medium? A lavatory has plenty of water being splashed around. In addition this was a men's room. Poorly aimed urine would provide both ammonia and salt to the electrolyte. Both substances should accelerate corrosion. This was the first case in which I dealt with a female trial lawyer. I explained male bathroom habits to her with some embarrassment.
So just how did the zinc get from the inside of the alloy to the surface? An earlier column noted that solid-state diffusion in alloys is perceptible only above about half the absolute melting point. Room temperature is way below half the melting point of brass so that the atoms should have been frozen in place. But the boundaries between grains are disordered regions where the atoms may diffuse much faster than through the matrix. Accelerated diffusion along grain boundaries has been observed in many alloy systems.
Several authors have suggested that during dezincification the zinc diffuses along these grain boundaries to reach the surface. The ability of small additions (&1 wt.%) of such unlikely alloying elements as As, Sb, and P to stop or greatly retard dezincification jibes with this proposal. Non-metallic atoms are dimensionally and electronically dissimilar to the matrix atoms and as such are ill suited to occupying the usual lattice sites. They quite naturally segregate to the grain boundaries to find a different and more hospitable environment. It is not unusual for a small matrix concentration of a non-metal to result in a monolayer at the grain boundary. Once in the grain boundary these atoms may block the paths of the Zn atoms seeking to diffuse to the surface.
I note finally that room temperature de-alloying has been observed in a variety of Ni, Fe, and Cu- based alloys.