Conductors have positive temperature coefficient of resistance. So as temperature rises resistance also increases. What is the reason for this?
While in insulators and semiconductors, resistance decreases with temperature?
When temperature is rised , electrons should break free. Why that does not happen in insulators and semiconductors?

 

I'm not sure why, but I don't think it can be lumped such as you have "Conductors have positive temperature coefficient of resistance" & "insulators and semiconductors, resistance decreases with temperature"

not all semiconductors have a negative temperature coefficient; IGBTs do but MOSFETs don't.
as far as conductors, I can't think of one that has a negative temperature coefficient, but I know I've heard of one.

 

Read this
http://www.learnabout-electronics.org/resistors_01a.php

 

and watch this.

http://ocw.mit.edu/courses/physics/.../lecture-9-currents-resistivity-and-ohms-law/

The reference given by Jony130 is very good at explaining the different behavior of insulators and conductors. The video I'm referencing is very good at deriving the details of conductivity of conductive metals (Cu is given as the example).

In a nutshell, the reason why resistance of a conductor goes up when temperature goes up is that the conductor has many free charges. These free charges have thermal motion and the average speed goes up with temperature as expected intuitively. The higher speed means that charges colide more often and with a shorter average time. The shorter time between colisions means that the electric field has less time to accellerate the electron; hence the drift velocity is lower when temperature goes up. This is non-intuitive, but if you think about it, it makes sense. Clearly lower drift velocity is going to translate to higher resistance.

 

I'm not sure why, but I don't think it can be lumped such as you have "Conductors have positive temperature coefficient of resistance" & "insulators and semiconductors, resistance decreases with temperature"

not all semiconductors have a negative temperature coefficient; IGBTs do but MOSFETs don't.
as far as conductors, I can't think of one that has a negative temperature coefficient, but I know I've heard of one.

Let us be careful not to confuse the OP. In very broad-brush terms, his statements are in fact correct.

What is key here is that in materials we generally describe as conductor (metals, mostly), free electrons are plentiful. Their resistivity generally increases with temperature, because the increasing thermal vibrations impede the movement of electrons.

Materials classed as insulators generally show an opposite behaviour with resistivity reducing with temperature, because they normally contain very few free electrons, but their numbers tend to increase with temperature as the extra thermal energy increases the probability of electrons breaking free. This generally more than offsets any reduction in their freedom to move.

Semiconductors are an intermediate case: the temperature coefficiient of the material depends on doping levels changing the availability of carriers, among other things. I am no expert about this but clearly positive and negative coefficients are possible, as witness the different types of thermistors which can be made.

Finally, the apparent temperature coefficient of a device is not always just a question of material resistivity. The RDS(ON) of a MOSFET probably is largely due to this, but the turn-on characteristics of a BJT are affected by other things, including very prominently junction saturation current, which is very dependent on heat energy freeing electrons.