hartnonicity. Note that this is a conservative statetnent, because our energy difference between the bridged and the Y-shaped structures is somewhat smaller than the best theoretical estimate (6).T h e qualitative features of the distribution at 3000 K ( Fig. 2F) resetnble those obtained for the ground state (Fig. 2D). However, the magnitude of the outof-plane trans distortion has increase4 considerably at 3000 K to hy * 1-0.23 A and A0 -23". This is consistent with the qualitative change of the El-E2 projection with increasing temperature: T h e maximum at (S, E l , E,) --(0, 0, a) IS shifted to about (0, 0, 1.33) at 3000 K (Fig. 2E). This corresponds neither to the bridged nor to the Y-shaped structure, but rather to a distorted one with (€I1, 02, 0,) * (90°, 20°, 160").Ignoring the important differences between our real space and the velocity space of CEI, which is smeared by multiple scattering, this is very close t o the values (90°, 30°, 150") inferred from CEI (2). In addition, the features of Fig. 2E, with the six peaks connected by saddle points at about half the peak height and the (S, El, E,) peaks shifted away frotn (0, 0, ?fi), resemble those obtained from CEI (2). Thus, we observe only at quite high temperatures the behavior attributed to the ground state of C2H3+ by CEI (2). Again, note that for a quantitative cornparison of our real space distributions with CEI data reported only in velocity space, a sitnulation (1 1 , 13) of the Coulotnb scattering process itself has to be carried out. W e sumlnarize our main points pictorially by superimposing in Fig. 3 representative configurations obtained from the trajectolies of C2H3+. T h e resulting smearing is induced either by thermal or by quantum fluctuations of the tnolecular structure. As expected, the classical distribution at 5 K ( Fig. 3 A ) is nearly indistinguishable from the planar bridged minimum structure. In the quantum ground state (Fig. 3B) the protons are mainly localized around the bridged configuration in Fig. 3 A and exchange sites through tunneling. Thus, the quantutn ground-state structure of C2H3+ is best described as a quasi-planar bridged structure broadened significantly by anisotropic delocalization of the protons due to zero point motion. Rovibrational excitations at higher temperatures lead only to a broad distribution of nonplanar structures, including bridged and Y-shaped ones (Fig. 3 C ) .How does the scenario predicted by ab initio simulations cotnpare with the CEI data? As demonstrated, the quantutn ground-state structure is clearly different from what was deduced from CEI (2). W e obtain, however, good agreement with the CEI data at high temperatures. This strongly suggests that CEI (2) did not probe the quantum ground state of the floppy tnole-cule c , H~+ , bLlt an ensemble of ro-vibra-14, R, Car and MM. Parrinello. Phys. Rev. Lett. 55. 2471 (1 985). tionally excited molecules. This aspect calls 15, For reviews, see R. O. ones and 0. Gunnarsson, for further clarifications in view of the ex-Rev. Mod. Phvs. 61,689 (1 98...
