问题1：动力学计算和量化计算泛函选择不同的问题，详情如下：
One weak point of the paper is that the Gaussian calculations are performed using a different functional and temperature. The temperature is not the real problem, since the Gaussian calculations only take T to make the typical corrections in the rigid rotor/harmonic oscillator approximation. The main issue is the choice of the functional. M062X is a hybrid functional, while PBE (used in the MD simulations)is a pure GGA functional.
问题2：跑AIMD模拟过程中自旋状态的如何处理的问题，详情如下：
Another technical point I am not sure is the way in which the ab initio MD simulations performed using CP2K deal with spin states. I guess that the authors start their simulations with a closed shell system S=0. However, along the simulation radical species appear. I am not sure how CP2K deals with this situation. In localized basis sets codes this would mean that the system would evolve as an open shell singlet state (meaning that the overall spin state is kept equal to 0). However, DFT may fail in describing the electronic structure of those complex systems. I am not an expert in CP2K type simulations, so I am not sure about how this issue should be handled. If spin states occur in the process an adiabatic simulation, such as the ones performed in the paper, are completeley inadequate.