Hello Afzal,
as far as I know the difference is not on the hardware but it is in the way you use it:
an E1 or T1 has a time based multiplexing framed structure and can carry some DS0 (64 Kbps each) and a dedicated signalling channel.
E1 is made of 32 64kbps timeslots with 30 DS0 one signalling channel and one DS0 used for framing
T1 is made of 24 timeslots with 23 user DS0, one signalling channel and an additional 8 kbps overhead for framing.
From the point of view of the router you can have installed an E1 or T1 controller.
When you configure the controller E1 for example you can decide how the DS0 channels will be used and the usage of the whole link.
Among the possible choices the controller can be instructed to remove the framing structure and to be unframed.
This kind of setup is good for data communication to have the equivalent of a single synchronuous serial line.
When connecting to PSTN and to ISDN the frame structure described above and the signalling channel are needed to interact with the ISDN switch in order to be able to setup calls or receive calls. This is the usage for voice.
Actually an ISDN PRI can also provide data services over ISDN using Dial on Demand Routing where usually remote routers call and can get 1 or 2 B channels (=DS0).
So there is not a single answer all depends from what usage you want to do of the link:
if it is for DATA and Voice over IP an unframed solution can be fine.
Some scenarios may require to allocate groups of timeslots (different channel groups) to setup fractionary E1 or T1 or to have two logical serial links on a single port
see
http://www.cisco.com/en/US/products/hw/routers/ps274/products_tech_note09186a00800b6e0f.shtml
the hardware modules are dependent on your router platform the link above is about a submodule that can be used on devices from C2600 to C3800.
Edit:
a good starting point about E1 and T1 is the protocol support page
http://www.cisco.com/en/US/tech/tk713/tk628/tsd_technology_support_protocol_home.html
Hope to help
Giuseppe
