The Dolomite Problem Is Solved. It Only Took Two Centuries.

For two hundred years, no one could figure out how it forms.

Not "how does this exotic quantum phenomenon work." Not "what happened in the first microsecond after the Big Bang." How does a rock form. A rock we've had samples of since the 1780s, that we've named mountains after, that we've analyzed across nine generations of scientists. The mechanism by which it comes into existence was genuinely unknown until now.

This is the dolomite problem, and it is one of geology's longest-running embarrassments.

The issue is crystal defects. Dolomite is calcium magnesium carbonate — its crystal structure alternates rows of calcium and magnesium ions in a precise sequence. In water, those ions land on a growing crystal somewhat randomly, and when they lodge in the wrong position they create structural defects that block further growth. The defects accumulate faster than the crystal can correct for them. Without intervention, forming a single ordered layer of dolomite would take approximately ten million years.

Ten million years per layer. Mountains at that rate are a very long project.

Researchers at the University of Michigan and Hokkaido University cracked it by figuring out what nature actually does, which turns out to be: cheat. Not slowly and carefully — cyclically. When the solution bathing a growing crystal cycles between supersaturated and mildly undersaturated — as it does constantly in nature through rain, tides, and seasonal change — the disordered defect regions dissolve away. Then, when conditions become supersaturated again, new crystal layers deposit more cleanly. Rinse. Repeat. Over geological time, mountains accumulate at a reasonable pace.

The researchers confirmed this with an experiment: they pulsed an electron beam on dolomite crystals in solution 4,000 times over two hours, inducing the same supersaturation-undersaturation cycling that nature does gradually. They grew approximately 300 layers. Previous experiments without the cycling had managed five.

The practical implication is almost funnier than the mystery itself. The received wisdom for growing defect-free crystals was: go slowly and carefully. Avoid disruption. Let the structure form without interference. That was wrong. The correct approach is to periodically dissolve your defects — make a mess, wash it away, make a cleaner mess. This has immediate applications for semiconductors, solar panels, and batteries, all of which are plagued by the same class of crystal defects and have been approaching the problem with the wrong strategy.

Two hundred years of the dolomite problem, and the answer was that the defects can leave if you give them somewhere to go. Which is, as geological insights go, suspiciously applicable to situations that have nothing to do with rocks.

The mountains were patient. The scientists, eventually, figured it out.

i · sources

• 200-year-old geology mystery resolved — University of Michigan News
• A fluctuating solution to the dolomite problem — Science, 2023
• After 200 years scientists finally crack the dolomite problem — ScienceDaily, 2026-04-20