You can try the following in the route section

# model chemistry opt(maxcyc=500,loose) Freq=noraman scf(tight,novaracc,noincfock) IOP(5/7=999)

Observe that in your last iteration you got the message:

Optimization stopped.

-- Number of steps exceeded, NStep= 151

Just increase the number of cycles as previously mentioned, I often use:

opt(maxcycles=400,maxstep=400)

Best, Pablo

Pablo Mtz is right. Apart from that you use the Cartesian coordinates to optimise the structure. I looked at the optimisation plot of you calculations (obtained with GaussView), attached as png file. You might observe that your optimisation got the resonance, so to say and it has no chance to converge.

This might be the result of using Cartesian coordinates together with opt=loose option. I personally see only one reason for a practical purpose - it the optimisation with internal coordinates yields the known Gaussian error of too small torsion angle.

I would advice you to use the default "option", it shall at the end of the day, go quicker.

Kindest regards

Juliusz