fxdung said:
I know that because quantum mechanics particle ''lies'' in wave packet,so it has multi-frequence due to Heizenberg relation.
The uncertainty relation in the case of electron wavepacket is between position and momentum. Localized electron wavefunction in free space is supposed to be realized by introducing uncertainty in momentum, I guess it is this momentum spread which you referred to as "multi-frequency".
fxdung said:
Why is coherence state when the number of photons tend to infinite?
I think your English skill obscures your intention about what you actually wanted to say. The state of a quantum system is something one can control, for the states of photon, it can be
number state, coherent state, squeezed state, etc. It's not like when the (expectation value of) number of photons increases, the state of light becomes more and more coherent. You can have number states with arbitrarily larrge number of photons, yet the electric field corresponding to such state is not coherent at all.
However, for the special case of coherent state, the observed electric field does exhibit certain dependency between the coherency, which may be defined as the uncertainty in phase at a given time, and the expected number of photons ##\langle \hat{n} \rangle##. For sufficiently large value of ##\langle \hat{n} \rangle##, the uncertainty in the phase of the electric field turns out to have the form ##\Delta \phi = 1/(2\sqrt{\langle \hat{n} \rangle})##. Therefore, if the number of photons is very large, the electric field becomes more and more coherent (i.e. it gets closer to being a classical field oscillation), in the sense that its phase becomes more definite.