[Solid_cc] WLS coated MAPMT performance question

[Sylvester J. Joosten]

Hi Brad,

To answer your first question:

When applied through vacuum deposition, the p-Terphenyl layer is barely visible - it mostly makes the PMT face look less glossy due to slight surface irregularities. Large degrees of cloudiness are indicative of crystallization, which *will* decrease the optical transparency.

We coated and tested 200 5-inch PMTs for the CLAS LTCC, and we performed precise gain measurements between 200 and 400nm with a monochromator. I included the results of such a measurement in the attachment. The measured gain is compared to the a theoretical calculation for a 110ug/cm^2 coating. Above 320 nm, the performance of uncoated and coated PMTs is identical, while below 320 nm the coated PMT performs significantly better. At no point did we measure any performance degradation.

Regarding the timing:

We measured a 2ns delay due to the WLS with a narrow time distribution. Light above 320nm does not interact with the WLS and is therefore not delayed. While I don’t think this is a big problem due to the 1/wavelength^2 dependence of the Cherenkov spectrum, I do agree we need to take this into account when designing our trigger (especially if we are thinking to use an ultra-narrow window).

Cheers,
Sylvester
-------------- next part --------------
A non-text attachment was scrubbed...
Name: comp.pdf
Type: application/pdf
Size: 16613 bytes
Desc: not available
URL: <https://mailman.jlab.org/pipermail/solid_cc/attachments/20160829/ecb51dab/attachment.pdf>
-------------- next part --------------

> On Aug 29, 2016, at 10:51 AM, Brad Sawatzky <brads at jlab.org> wrote:
> 
> 
> On Mon, 29 Aug 2016, Michael Paolone wrote:
> 
>>> 2. above 350nm, it shows no change comparing to uncoated PMT. If the
>>> main feature of WLS is shifting light to UV, why QE won't drop at
>>> high wavelength?
>> 
>> It works the other way, Higher energy (low wavelength) photons are
>> absorbed and reemitted as lower energy (high wavelength) photons.  So
>> why isn't there a bump at higher wavelengths, you might ask?  The QE
>> spectra has the WLS effects folded in.  So a photon at 250nm on the QE
>> plot has an average effective QE of the 250nm photons that were not
>> absorbed averaged with the QE of the reemitted photons at higher
>> wavelengths.  On the QE plot, photons at 350nm are not absorbed at
>> all, so their effective QE is not changed.
> 
> If the wavelength shifter material is fully transparent above 300-350
> then why does the material look cloudy when applied -- shouldn't it be
> clear?
> 
> 
> One other question that came to mind was whether the WLS florescence
> delay messes with the timing you want for background suppression.  If
> you're assuming coincidence gates of ~40ns, then it won't matter a bit.
> If you're thinking of something really tight, like 10ns, then the tail
> of the florescence decay time could matter.
> 
> -- Brad
> _______________________________________________
> Solid_cc mailing list
> Solid_cc at jlab.org
> https://mailman.jlab.org/mailman/listinfo/solid_cc