We solve a long-standing experimental discrepancy of NH 3 BH 3 , which-as a molecule-has a threefold rotational axis, but in its crystallized form at room temperature shows a fourfold symmetry around the same axis, creating a geometric incompatibility. To explain this peculiar experimental result, we study the dynamics of this system with ab initio Car-Parrinello molecular dynamics and nudged-elastic band simulations. We find that rotations, rather than spatial static disorder, at angular velocities of 2 rev/ps-a time scale too small to be resolved by standard experimental techniques-are responsible for the fourfold symmetry. Ammonia borane NH 3 BH 3 has drawn significant interest in recent years because of its potential as a hydrogen storage material, with a gravimetric storage density of 19.6 mass% [1][2][3][4][5][6][7]. The structure of its solid phase has been explored previously [8][9][10][11][12][13], but the literature does not agree about the hydrogen behavior at room temperature. The molecule consists of a dative B-N bond and a trio of H atoms (henceforth referred to as a "halo") bonded to each of those two atoms, forming an hourglass shape, visible in Fig. 1. At low temperatures (0 ∼ 225 K), the solid exhibits an orthorhombic structure with space group Pmn2 1 . Heated above 225 K, it undergoes a phase transition to a body-centered tetragonal structure with space group I4mm. It is this room-temperature phase that exhibits unexpected experimental results: While the molecule itself has a threefold symmetry about the B-N axis, neutron [11,14] and x-ray [12,15,16] diffraction on the solid reveal a fourfold symmetry about the same axis, creating a geometric incompatibility within the structure. Investigating the dynamics of the system with ab initio methods, we find that the individual halos are rotating with angular velocity on the order of 0.7 • /fs ≈ 2 rev/ps, such that standard experiments can only probe the time-averaged positions, leading to the tetragonal host structure with fourfold symmetry.The precise behavior of these hydrogen halos has been the subject of several studies over three decades. In 1983, Reynhardt and Hoon [8] found threefold reorientations of the BH 3 and NH 3 groups with a tunneling frequency of 1.4 MHz in the orthorhombic phase. Penner et al. [9] found in 1999 that these groups reoriented independently. Deciphering the behavior in the tetragonal structure has been less straightforward. In the same 1983 study, Reynhardt and Hoon concluded that the BH 3 , and possibly the NH 3 groups, rotate freely. Brown et al. [11] found that they could describe the disorder entirely with threefold jump diffusion. Bowden et al. tried using a larger unit cell to model the same disorder as spatial variation rather than higher-order rotation; however, they found no evidence to support this model [12], leaving this disagreement unresolved in the literature. The present study aims to elucidate how the hydrogen halos behave in the solid, especially in the high-temperature, tetragonal structure. To this...