By looking at the Ebers-Moll equation for a BJT transistor:
Is this formula explicitly showing why Ic starting to decrease towards zero in saturation region. Which part of the equation shows that?
As has been pointed out, no part of Ebers-Moll will explain your effect. Ebers-Moll explicitly deals with linear operation, not saturation.
You should note that, for voltages above 798 mV, the transistor is not in saturation.
For instance, at V = ~824 mV, current is 1 mA. This means that the voltage across R1 is only one volt, which means 4 volts across C-E and R2. Since the base voltage is 824 mV, the maximum possible voltage across R2 must be less than one volt, and there must be at least 3 volts across the transistor. With Vce greater than Vbe, the transistor is, by definition, not in saturation.
What is going on in the simulator model - I have no idea. But you've clearly found a mode where Something Is Wrong. I suggest you rerun your simulation with voltage and current meters on every node.