LINACs are the best practice to accelerate electrons up to 25 MeV. Approximately 1m of subsequent resonance cavities can force electrons to accelerate in each cavity up to the final energy. The whole setup is like an arm where at the end proper magnets change the electron tracks so as to hit the target cells. The whole system can be move in a circular manner.

The whole technology is much better than cyclotron which is a stable set-up and needs to be adjusted for the change of electron mass at the end of their acceleration path.

As far as I know, Cobalt Units were the state of the art before LINACS.

--

Dr.Dimitrios Nikolopoulos

Environmental-Health Physics

Associate Professor

_________________________________________________________

Department of Electronic Computer Systems Engineering,

Technological Education Institute (TEI) of Piraeus,

Petrou Ralli & Thivon 250, GR122 44, Aigaleo, Greece

e-mail : [email protected]

[email protected]

Mobile : +0030-6977-208318

Home : +0030-213-0308432

Tel (TEI): +0030-210-5381560

Fax (TEI): +0030-210-5381436

Skype : dr.dimitrios.nikolopoulos

assos.prof.dimitriosnikolopoulos

Internet-1 : http://env-hum-comp-res.teipir.gr/

Internet-2 : http://env-hum-comp-res.daidalos.teipir.gr/

url-1 : http://scholar.google.gr/citations?user=BuZ2IJQAAAAJ&hl=el

url-2 : http://www.researchers.gr/el/node/71649

url-3 : https://www.researchgate.net/profile/

Dimitrios_Nikolopoulos6?ev=hdr_xprf

_________________________________________________________

Electrons have rest mass of 0.511 MeV c-2 and thus for energy above this electrons behave relativistic. 

This frequency of power source is in sync with frequency of particle being accelerated which in cyclotron governed by the equation:

f = q*B/(2*pi*m) 

As electron is accelerated above 0.511 MeV, the mass of electrons keeps increasing and from above formula the frequency of revolution (of electron) also changes. Thus it leads to mismatch in frequency of power source and that of particle. 

Soon electron doesn't remain in sync and may start decelerating. 

So we use modified form of cyclotron called synchrotron to avoid this ambiguity.

@Abhishek´s comment is 100% perfect. It´s just a matter of relativistic behaviour and the phase of electrons and HF. I attend  a chapter from one of my textbooks for private use. It´s in German but you will understand the formula and principles. Have a special look on pages 271-281.

An 'ordinary' cyclotron - i.e. one with a constant magnetic field within it, and where the accelerating AC voltage in the Dees has a constant frequency - requires the particles to be in the classical (non-relativistic) regime. Ernest Lawrence's key insight was that as a classically-moving particle gains energy in a cyclotron its radius and velocity increase together such that the revolution frequency stays constant; hence a constant AC frequency can be used in the Dees.

Electrons could in principle be accelerated in a cyclotron, but only up to energies where they become significantly relativistic. Practically, this would be only around 100 keV, so you might as well use a DC accelerator or a linear accelerator (linac) for this.

In the early days electrons were also accelerated in betatrons, which look superficially like cyclotrons but which work in quite a different way. A betatron works by induction - the magnetic field the electrons circulate within varies with time, and an induced emf gives rise to the electrons' acceleration. In the case of the betatron there is actually a mis-match between the radius and the acceleration, and so the magnetic field *inside* the electrons' path must be larger than the field the electrons themselves see. This is why the poles of a betatron stick inwards at the centre of the accelerator.

I mention betatrons as they have been used in the past to accelerate electrons for radiotherapy. Linacs are the main accelerator type used these days though, as they are more cost-effective and reliable for the typical electron energies that are needed, which is mostly between 6 MeV and 40 MeV.

Hywel Owen gives a very good discussion of the problem.

Thank you gave the answer to my question.