PINN does not follow boundary condition

Hi everybody,
I am using the code GitHub - maziarraissi/PINNs: Physics Informed Deep Learning: Data-driven Solutions and Discovery of Nonlinear Partial Differential Equations as a base.
Instead of the burgers equation I am using the fick’s law of diffusion to calculate the concentration.

I don’t have any measured data so my PINN only uses the initial condition, the boundary conditions and the fick’s law of diffusion as physical losses.
However my result of the concentration does not follow the change over time, only the change of the radial coordinate.

These are my initial and boundary conditions. To make it easier, I used normalized values (from 0 to 1).

My training data is based on the initial and boundary conditions, my testing data are random values.

What could be the problem why my model does not follow the maximum boundary condition? What is the reason the concentration does not reach the value 1?

Thank you in advance