Proposal for dose and fluence testing for pixels

Proposal for dose and fluence testing for pixels

After extensive EMail exchanges with Tony Weidberg, and some assistance from Alfredo Ferrari, I would like to make a revised proposal for the lifetime fluences and doses for the pixel detector. Note there are two issues of interest in damage of Silicon (or GaAs) based devices: ionization damage, expressed here in kGy deposited in Si, and displacement damage, expressed in ATLAS in 1 MeV n equivalent (NIEL) fluences. Tony has also calculated the equivlent quantities for the GaAs VCSELs used in the optolink, including the proper irradiation protocols for different machines (LBL or CERN PS), and they will be finalized shortly in the optolinks requirements document he is writing. Before finalizing this document, he needs the agreement of the pixel community on the numbers below.

There are several conventions implied here which I wanted to make explicit:

1) B-layer: nominal lifetime goal is 2.3 years at 10³⁴. Note this includes

a safety factor of sorts already because the 10³⁴ year assumes that the average luminosity is 10³⁴, rather than the peak. At the Tevatron, the difference between peak and average can be as much as a factor 2... Tony and I propose to take no additional safety factors into account. This is in part out of concern that it is too late to significantly increase our dose/fluence goals for this layer.

2) Outer pixel layers: nominal lifetime goal is 7.3 years at 10³⁴. Of course

this includes the ”safety factor” described above. However, a convention established already at the time of the ID TDR was to increase this further by a factor of 1.5 due to the various uncertainties in the estimate (not least of which is the total cross section at 14 TeV).

With these goals in mind, the values that are extracted from the calculations of Ferrari at the time of the ID TDR provide the following:

1) B-layer lifetime definition = 2.3 years at 10³⁴, R=4.1 cm in present

layout, unchanged from the ID TDR. No safety factor has been included.

fluence = 1.0x10¹⁵

dose = 500 kGy (I rounded up the 433 value from Ferrari’s tables)

2) Layer 1 lifetime defintion = 7.3 years x safety factor of 1.5 since this

layer cannot be changed. Further note: the radius at the time of the ID TDR was given as 11.6cm in Ferrari’s calculation, whereas our present layout uses a value of 9.7cm. Naively, using R² scaling, I get a factor of 1.43 increase in fluence for this (I’m sure this is naive, but since the fluence is dominated by the charged particles from the vertex, it is a defensible guess...)

fluence = 7x10¹⁵ * 1.43 = 1.0x10¹⁵

dose = 343 kGy * 1.43 = 500 kGy

I propose that these values (10¹⁵NIEL and 500 kGy = 50 MRad, both for Si) should be our requirements for all of our components. Of course, the 10¹⁵ has been our requirement for several years already. The dose was usually given as 25-30 MRad in earlier discussions.

Achieving these in the two places we plan to irradiate in the near future requires the following exposures:

For the LBL machine (55 MeV protons), this corresponds roughly to 5x10¹⁴protons to give 10¹⁵ NIEL (hardness factor is 1.79 based on averaging existing measurements). This gives a dose (assuming the dE/dx for 55 MeV is 5.7*MIP) of about 75 MRad.

For the PS (24 GeV protons), this corresponds to 2.0x10¹⁵ protons to give 10¹⁵ NIEL (hardness factor is 0.51 based on measurements by ROSE collaboration). This gives a dose of about 55 MRad, ignoring the small relativistic rise in dE/dx in Silicon.

Because of the very significant increase of these doses over any real testing performed up to this time, I believe it is important to do the irradiations in three steps of 1/4, 1/2, and full dose. The 1/4 step approximately matches the full dose irradiations of the SCT community (10MRad and 2x10¹⁴ NIEL).

Feel free to ask about any details which are not clear...

Regards,

Kevin