Comment on Feasibility of an electron-based crystalline undulator

Tabrizi et al. [physics/0701342] discuss the feasibility of an electron-based crystal undulator (e-CU) by planar channeling of 50 GeV electrons through a periodically bent crystal. We show that their scheme is not feasible. First, their undulator parameter is K >> 1 always, which destroys photon interference. Second, they overestimate the electron dechanneling length in e-CU by an order of magnitude, which shortens the number N of e-CU periods from 5-15 (as they hope) to just 1-2. This kills their e-CU concept again. We made first simulation of electron channeling in undulated crystal and conclude that an electron-based crystal wiggler is feasible with wiggler strength K=10 and number of periods N=2.