We present a detailed theoretical analysis of experimental rates for ddμ molecular formation and dμ hyperfine transitions at temperatures 25.5–150 K, which were reported by Zmeskal et al. [Phys. Rev. A 42, 1165 (1990)]. Theoretical effective ddμ formation rates are fitted to the observed rates by adjusting the ddμ binding energy ${\mathrm{\varepsilon }}_{11}$, the effective dd fusion rate λ${\mathrm{̃}}_{\mathit{f}}$, and the nonresonant ddμ formation rate ${\lambda }_{\mathrm{nr}}$. The value of ${\mathrm{\varepsilon }}_{11}$=-1966.1±0.2 meV is determined with extreme accuracy and agrees with the theoretical prediction within 0.1 meV. Experimental findings for ${\lambda }_{\mathrm{nr}}$ are compatible with theory. Since the value of λ${\mathrm{̃}}_{\mathit{f}}$ extracted from observed formation rates depends on the calculated value of ddμ formation matrix elements ‖${\mathit{V}}_{\mathit{i}\mathit{f}\mathrm{\Vert }}$, we present the region of pairs (λ${\mathrm{̃}}_{\mathit{f}}$,‖${\mathit{V}}_{\mathit{i}\mathit{f}\mathrm{\Vert })}$ allowed by experiment. The theoretical values of λ${\mathrm{̃}}_{\mathit{f}}$ and ‖${\mathit{V}}_{\mathit{i}\mathit{f}\mathrm{\Vert }}$ lie outside this region. A significant discrepancy remains for the dμ hyperfine transitions, where the theoretical rates, which consist of scattering and back-decay contributions, exceed experimental rates by ∼40%. Fits of the experimental data indicate that mostly the scattering contribution is smaller than calculated. The extrapolation of our fit to higher temperatures is in good agreement with other experiments on ddμ formation.

• Received 22 February 1993

DOI:https://doi.org/10.1103/PhysRevA.47.4691