Scientists from the University of Michigan and Hokkaido University have successfully grown dolomite in the laboratory for the first time after more than 200 years of failed attempts.
How atomic simulations revealed the flaw in dolomite formation
The breakthrough came from understanding that dolomite’s alternating layers of calcium and magnesium often form with atoms attaching randomly instead of in an orderly sequence, creating structural defects that halt growth. Researchers used detailed atomic simulations to show these defects are unstable in water and dissolve when exposed to natural cycles like rainfall or tidal changes, effectively resetting the surface for proper layering.
Why this solves a long-standing geological puzzle
Dolomite is abundant in rock formations older than 100 million years but rare in modern settings, a discrepancy known as the “Dolomite Problem.” The study shows that natural wet-dry cycles act as a built-in mechanism to clear growth-inhibiting defects, allowing the mineral to accumulate over geological time despite its slow molecular growth rate.
What this means for materials science
Wenhao Sun, Dow Early Career Professor of Materials Science and Engineering at U-M and corresponding author of the study published in Science, said understanding dolomite’s growth could inform strategies to promote crystal formation in modern technological materials. The team’s software, developed at U-M’s PRISMS Center, simplifies atomic interaction modeling by leveraging crystal symmetry to predict energy states without exhaustive computation.
Why has dolomite been so demanding to grow in labs before?
Previous lab attempts failed because they did not account for the need to periodically remove atomic-scale defects through water exposure, which occurs naturally in environments with wet-dry cycles.
How long does it take to form a single layer of dolomite under natural conditions?
Without the defect-clearing mechanism, forming a single well-ordered layer could take up to 10 million years due to the high frequency of growth-blocking errors.
