Junction capacitance and calculating diode capacitance

we all know that forward bias in a diode minimizes the junction width and after the threshold voltage junction width vanishes. but while calculating the capacitances (as most of the book mentioned) for forward bias case we take diffusion capacitance and neglect the junction capacitance(saying it is very small or negligible), but it is not considered completely zero. since in forward bias case there is no junction, junction capacitance should be considered as zero rather than negligible.

Re: DIODE CAPACITANCES

you write:
"since in forward bias case there is no junction...."
and thats wrong in my opinion; in forward bias case generally there is junction;
you said it yourself, after threshold is only one part of the forward bias;
for example 0.01 V and thats forward bias too but you havent reached the built in voltage.
the other point is that when you approach voltages at the junction around the built in voltage you get additional phenomena; the resistance in the non junction region of the diode become important because of the high current; so if you put Vbi on a real diode you cant say vbi is at the junction;
then you get high injection, causing the majority carriers going up at the junction;
the width goes down, but you get more and more carriers and the capacitor which causes the capacitance not to drop so slow;
i dont know the quantitative relationship for the high injection case, but i dont know reason why to say its contribution is zero instead of negligable;
i think the diode will start to burn because of high currents before you can make the junction width go to zero.