Comment on ''Semiquantum-key distribution using less than four quantum states''

For several decades it was believed that information-secure key distribution requires both the sender and receiver to have the ability to generate and/or manipulate quantum states. Earlier, we showed that quantum key distribution in which one party is classical is possible [Boyer, Kenigsberg, and Mor, Phys. Rev. Lett. 99, 140501 (2007)]. A surprising and very nice extension of that result was suggested by Zou, Qiu, Li, Wu, and Li [Phys. Rev. A 79, 052312 (2009)]. Their paper suggests that it is sufficient for the originator of the states (the person holding the quantum technology) to generate just one state. The resulting semiquantum key distribution, which we call here 'quantum key distribution with classical Alice' is indeed completely robust against eavesdropping. However, their proof (that no eavesdropper can get information without being possibly detected) is faulty. We provide here a fully detailed and direct proof of their very important result.