Thank you for your valuable efforts Sir Constantinos D. Zeinalipour-Yazdi , Dilawar Singh Sisodiya & Bobby Solanki
The Answer is provided below:
As I was trying to optimize the structure using B2PLYPD functional. The problem is that the default resource is insufficient for an expensive calculation like this.
Note that double-hybrid functionals, like B2PLYPD, follows a similar procedure like MP2:
First, obtain the reference DFT wavefunction (while MP2 uses HF as the reference);
Second, performs a MP2-like calculation for the correlation energy (but with the DFT reference).
The final result is given by adding the (scaled) E2 contribution to the DFT reference energy. Therefore the cost is similar to MP2 instead of DFT.
To avoid this error we need to consider a different model chemistry, or switch to a 64-bit binary and increase the resources available to Gaussian to run it. Also, we can improve the initial geometry so it'll take less steps to converge.