We consider minimum-energy routing in a wireless network. We assume the use of ideal rateless codes, so that a node can accumulate transmission rates from the transmission of previous nodes on the routing path. Mutual information accumulation has significant advantages when compared with classical cooperative schemes that use energy accumulation. However, the resource allocation problem becomes more complex, as it involves the determination of 1) Routing path, 2) Transmission duration of each node, and 3) Transmission power of each node. We formulated the problem as an optimization problem, where the objective function is the total energy expenditure and the constraints are minimum mutual information for each node and the maximum total transmission time. We make a slottedtime assumption, and given the routing path and transmission duration, power optimization problem becomes convex. The optimal routing path and transmissions durations are found using a Branch-and-Bound technique. A distributively implementable greedy algorithm is also found and performance are compared by numerical simulations.