Three ideas:

1. You probably put solid PET in the DSC crucible, right? If the thermal connection between PET and DSC crucicle changes after first melting, i.e. if the thermal resistance between PET and the bottom of the crucible changes, the data of the first cycle differs from succeeding cycles. Typically, the melting curve of the first run is shifted to slightly higher temperatures. This effect can explain the difference between the first and the second cycle.

2. The molecular structure of your PET sample can change from one cycle to another. Depending on the cooling speed after the first melting, the crystallinity prior to the second heating run will be different compared with the initial state.

3. However, a decrease of the melting temperature can also be the consequence of a thermal degradation of your PET sample (e.g. if you apply temperatures above the decomposition temperature of PET).

The latter effect should cause a more significant change in the melting behaviour than the first and second one.

Three ideas:

1. You probably put solid PET in the DSC crucible, right? If the thermal connection between PET and DSC crucicle changes after first melting, i.e. if the thermal resistance between PET and the bottom of the crucible changes, the data of the first cycle differs from succeeding cycles. Typically, the melting curve of the first run is shifted to slightly higher temperatures. This effect can explain the difference between the first and the second cycle.

2. The molecular structure of your PET sample can change from one cycle to another. Depending on the cooling speed after the first melting, the crystallinity prior to the second heating run will be different compared with the initial state.

3. However, a decrease of the melting temperature can also be the consequence of a thermal degradation of your PET sample (e.g. if you apply temperatures above the decomposition temperature of PET).

The latter effect should cause a more significant change in the melting behaviour than the first and second one.

In the first run if some impurities are there that melted and in second scan the melting is for pure polymer. It is normal for most of the synthesized polymers, first melting peak is higher and the consecutive runs showed lower values.  

Thank you all , nice comments

do u think , high or lower crystallinity has a high melting points

thanksa again

This may happen only if the crystal imperfection occurs. The crystallization is also associated with the secondary forces which may be differently organized in the second case because of rate of heating, leading  to poor orientation. Thus M.P. may come lower.