Can transistors replace capacitors and resistors?
It is done on IC very currently. The resistance is easy with the fets and the capacitor really cannot be totally replaced, it can be multiplied thus making the small capacity bigger and useful. Current sources can and are mainly used to replace resistance.
First understand the function of resistors and capacitors.
The resistors are used to limit the current. They resist or oppose the flow of current. Now, can we replace them? The is yes. Often, in the design of integrated circuits where it is difficult to manufacture resistors, we use transistors. This is called an active load.
A jfet (which is also a transistor) can be used as a variable voltage resistor.
Capacitors store charge. Well, you cannot replace capacitors in essential circuits such as filters, clamps or oscillators, but we use mosfets in computer memories.
The inter-junction capacity of the fets is used. However, complete replacement of the capacitors is not possible in a circuit.
There are many types of transistors, some of which have been discussed here. Basically, the fets are ideal for replacing resistors, because they can be configured so that the bias of the gate controls the current in the channel.
This makes it possible to use fixed resistors with fixed polarization and variable controllable resistors with variable polarization. It works with a lot of FET process.
Capacitors are more interesting: each semiconductor junction has a certain capacity that can be increased or decreased by reverse or direct polarization.
Fm modulators and demodulators and variable frequency oscillators were constructed using varicaps or varactors. By definition, not transistors, but rather diodes.
Because of the need to be able to provide a bias voltage to “adjust” the distances between the plates, they are used for AC signals rather than for charge storage.
Bjt have an intrinsic capacity in each junction (because of the miller effect). I do not know if this effect has already been considered as a useful / desirable capacitor substitution.
In integrated circuits, it is not necessary to make a pn junction, as long as you have a deposit of metal and silicon, or you can diffuse a sufficient isolation / dielectric between two layers of layers p or n doped.
Similarly, the intrinsic silicon between two layers of metal deposit would make a reasonable capacity, however small.