Jacques Dubochet, at the European Molecular Biology Laboratory, develops in the mid-1980s a technique for freezing biological samples so rapidly — cooling them within milliseconds — that the surrounding water has no time to form ice crystals, but instead solidifies into an amorphous, glass-like state that preserves the structure of biological molecules practically intact and in their natural aqueous environment, solving a fundamental problem that had severely limited the use of the electron microscope in biology: the vacuum and radiation of the conventional electron microscope destroyed biological samples before useful images could be obtained. Richard Henderson, at the MRC Laboratory of Molecular Biology, develops techniques for imaging proteins by electron microscopy with sufficient resolution to distinguish structural detail at near-atomic level. Joachim Frank develops mathematical algorithms capable of combining thousands of individual two-dimensional images of the same molecule, captured in different random orientations, to computationally reconstruct a complete, high-resolution three-dimensional model. The combination of these three innovations, refined and automated over subsequent decades, transforms cryo-electron microscopy into a complementary alternative to X-ray crystallography capable of resolving the structures of proteins impossible to crystallize, and proves decisive during the COVID-19 pandemic for rapidly determining the structure of the SARS-CoV-2 spike protein, accelerating vaccine development.