Dear Rajesh,

with respect to x-ray attenuation Al is one of the worst materials you can ever use; especially when you operate in the high photon energy range.

Its linear attenuation coefficient µAl(10Mev) is about ~0,06/cm.

Heavy metals such as Lead or Tungsten are much more effective due to their large atomic number.

 For Lead we have   µPb(10MeV) ~0,5/cm.

You may pick up these data by yourself at the XCOM/NIST data base (see link below).

 According to the Lead attenuation coefficient you have to take at least 2cm for attenuating your beam down to 1/e ~ 37%.

So my advice: please take a stack of 1 to 2 cm thick diaphragms (of e.g. Lead) having a bore hole, whose cross-section is equal to the beam cross-section you would like to have and mount these plates one above another; but having a distance of about 1 to 2 cm in between each plate. So you will decrease the net total scatter form the inner surfaces of your diaphragms (compared to a complete pipe). Furthermore please round the edges/rims of your bore holes (upper and lower ones); this will also decrease the scatter.

You may also vary the cross sections of your bore holes in such a way that they match the beam geometry (please see case b of the attached drawing).

For your system  case c should be the right option.

Good luck. 

http://physics.nist.gov/PhysRefData/Xcom/html/xcom1-t.html

Hi Gerhard,

Thank you for your answer.

Please note that I meant to use Aluminum as an absorber while measuring depth-dose profile, NOT as main collimator material. I am thinking to use Aluminum sheet in collimator but in the following way.

I was planning to use a stack of lead sheets with no air gap in between. In order to stop the characteristic x-rays emitted from Lead, I was planning to use Tin next to Lead. Similarly to stop any characteristic x-rays from Tin, I was planning to use Copper and finally, to stop the characteristic x-rays from Copper, a thin sheet of Aluminum. With this design, the collimator should look like in Figure (attached). Note that this diagram is not in scale and the thickness of lead is to be determined using the minimum linear attenuation coefficient for the photons of range somewhere from 10 keV to maximum energy in the beam (such as 10 MeV for a 10 MV beam). And the thickness of each of other three materials to be determined using their linear attenuation coefficient for characteristic photons energies of preceding material. This way, I think we can absorb all the photons that incident on Lead and allow the photons to pass through the hole in the center. I will consider your idea of putting some air gap in between the lead blocks too.

With my design, the main problem is to avoid the scattered photons and electrons from the edges of the inner surface of the diaphragms in each layer. Here I meant the electrons and photons from Compton scattering and also 511 keV photons after pair-production (see objective iii in the question).

Can we really solve this by making round edges/rims of the bore holes like in your diagram (b)? If you can kindly provide any book or paper that explain this issue that would be great.

Thank you again for your help.

Dear Rajesh,

there is no need to stop the Pb characteristic lines  via sucessive Tin layer and the stopping the Tin lines via Copper etc.. The attenuation of lead is strong enough to stop its own charcteristic lines; µPb(75keV) ~ 30/cm ; thats a far large value. For your 10MeV photons I think you have to use about  20 to 30 cm of Lead. So there is really no necessity for stacking the material such as in your proposal.

In the attached file it is demonstrated how the air gaps, which are introduced in between the Lead layers will partially reduce the scatter (case a)  compared to the case where no air gaps show up (case b).  I think the drawing needs no further explanation.

Round rims cannot solve the general problem of scatter radiation; but they help to reduce the amount of scatter. Avoid sharp edges or rims; they are sources of excessive scatter.

By the way; there is no need for complete air layers such as they are displayed in case a. A saw tooth like inner surface of the collimator like in case c is sufficient and will do the same.

However Lead is very heavy, so you can save a lot of weight (case a), which you have to support in your experimental set up.

With respect to literature for such cases; sorry, but I havn't got any.

Maybe it could be useful for you to learn about Chandra X-ray telescope. I have read about it in the book by Paul R. Yoder "Mounting optics in optical instruments", SPIE Press, 2002, ISBN 0-8194-4332-8, Chapter 10.3.4, pages 364-369. I think it could be good idea how to built collimator for X-rays.

Not  in my expertise.  R.G.Rastogi.

Thank you, Mikhail, for your suggestion. I read the chapter 11.4.2 on The Chandra X-Ray Telescope. I am planning to use this collimator in megavolt x-rays where the intensity of photons is very high. With this collimator, I should minimize the scattering of photons both in the air and the collimator itself. And I should keep the properties of the primary beam (energy/angular distribution) that passes through the collimator aperture unchanged to the best possible way. In such case, I guess getting the reflected x-rays from mirrors is not a good idea. And I also have to fit the collimator within a space of maximum 0.6 to 0.8 m.

Dear Gerhard,

Thank you so much for your help. I am going to build one collimator a very similar to one that you suggested above (MeV-Collimator_2nd_GM.pdf). I have a couple of questions for you. The first one is a little general. 

Choosing the total thickness of collimator:

I took the attenuation coefficients of lead for the photons with energies ranging from 1 keV to 20 MeV. And I found that the minimum attenuation coefficient is 0.4757 cm^-1, which is the value for photons with energies from about 3.60 MeV to 3.75 MeV. Using this, I found the hundredth value layer to be 9.68 cm. Will this be enough if my collimator contains about 10 cm thick lead with an appropriate aperture at its center? And this is intended to be used in 4 MV, 6 MV, 10 MV, 15 MV and 18 MV x-rays beams.

Choosing the tapering amount for air gap:

I guess this also depends on the size of the focal spot of the beam. But, if I choose the space for air gap as shown in the diagram which includes tapering at ~45 degrees in various depths along the aperture of the collimator, will this be okay?

Dear Rajesh,

first of all: merry Christmas and a happy new year from Germany.

Total thickness: Your derivation of the 10cm (Lead)  value for at least 1/100 transmission is ok. So if you want to go to less than 1/100 of photon flux outside your field of view  is sufficient to take effectively  10cm of Lead. You may go to higher Lead thicknesses; but I think with respect to your therapy application this 1/100 transmission value is a quite good number; furthermore this value only holds for about 1,5MeV to about 5MeV.  In the other spectral parts of the x-ray  spectrum the transmission is much smaller than 1/100.

Air gaps: from the 1/100 width of Lead (10cm) one can easily deduce the 1/10 width of Lead to  5cm(=10cm/2) and from that the half width to about 1,5cm (~ 5cm/3). So your your effective thicknesses of your lead layers should be at least about 1,5cm. Taking the air gap widths nearly equal to the Lead layer thicknesses you end up with about 1,5cm for the air gaps too; this is already noticed in your drawing. So any x-ray flux components impinging onto the air gap regions of your collimator walls will be attenuated to at least by 1/2 when scattered into the direction of your field of view. So I agree with your 1,5cm air gap proposal.

Total scatter: This series of Lead diaphragms should be arranged in the middle between your x-ray source and your field of view. Just in front (but not too close)  of your x-ray source and just in front (but not too close) of your field of view  I would propose only one of such Lead diaphragms respectively. In the attached file the distance dependence of the single scatter contribution of an object (such as your diaphragm walls) in between the x-ray source and field of view is presented. This curve is calculated for example for  a distance of source to field of view of 1000mm. It is just your SSD. Other distances will change only the absolute values but not the general shape of this distance dependence.

Dear Rajesh,  you can take a look at our project ELI-NP-GBS, where we designed and we are testing a collimator for gamma rays in the 1-20 MeV energy range, very precise and stable (microrads) and with minimum scattered photons. It is described in https://www.researchgate.net/publication/263930417_Technical_Design_Report_EuroGammaS_proposal_for_the_ELI-NP_Gamma_beam_System   and more specifically in https://www.researchgate.net/publication/273398448_Monte_Carlo_simulation_of_a_collimation_system_for_low-energy_beamline_of_ELI-NP_Gamma_Beam_System , hope they can be useful for you.

Article Technical Design Report EuroGammaS proposal for the ELI-NP G...

Article Monte Carlo simulation of a collimation system for low-energ...

Dear Friend

With respect

I have a question and I would be grateful if you answered my question

What is the difference between pre and post patient collimators in the CT scan? (Material, Thickness, and Dimensions)

Dear Bahare Bigham ,

please see for example:

https://www.google.de/url?sa=t&rct=j&q=&esrc=s&source=web&cd=11&ved=2ahUKEwjFga2jxernAhXPwqYKHVWRBCwQFjAKegQIARAB&url=http%3A%2F%2Fwww.staempfliverlag.com%2Fannot%2F564C42696D677C7C393738303831333831373437377C7C504446.%3Fsq%3D1%26title%3DVeterinary%2520Computed%2520Tomography&usg=AOvVaw3f2z4t5Gj2Ryo99t0TlV-4

https://www.facebook.com/pg/alwayzchakri/photos/?tab=album&album_id=493052737404545

https://books.google.de/books?id=TH_sMhXgzxAC&pg=PA43&lpg=PA43&dq=pre+and+post+collimator+CT&source=bl&ots=M7Vpcin-3-&sig=ACfU3U3trijlJ1EiGci_BRF1_TvU-XB9-Q&hl=de&sa=X&ved=2ahUKEwjFga2jxernAhXPwqYKHVWRBCwQ6AEwEXoECAkQAQ#v=onepage&q=pre%20and%20post%20collimator%20CT&f=false

https://books.google.de/books?id=kB_mCgAAQBAJ&pg=PA232&lpg=PA232&dq=pre+and+post+collimator+CT+philips&source=bl&ots=qlhHKS97ch&sig=ACfU3U05aicq4MqBGgM5A77QmlpkUu74hA&hl=de&sa=X&ved=2ahUKEwiDiv2Gx-rnAhWTh1wKHZWpABE4ChDoATAEegQIChAB#v=onepage&q=pre%20and%20post%20collimator%20CT%20philips&f=false

Dear all, I know that I am late, but there are gamma-ray spectrometers in operation worldwide - one of them is the GAMS-spectrometer at ILL / Grenoble, the website is

https://www.ill.eu/users/instruments/instruments-list/pn3-gams/description/instrument-layout/

I think that people there are confronted with the same problems, and it might be very helpful to contact them. Good luck - if I am not too late ..., Dirk

You can find a huge amount of collimator related papers via the CERN electronic library (Type CDS CERN in Google) or via the JACOW Page (do not use keyworrd search but rather full tect search. Or simply talk to Redaelli at CERN or Assmann at Desy..and /or have a look at the CERN collimator page