The Smoky Diagonal

[great_bear]

So I cleaned - or more correctly - de-dusted my 2” quartz dielectric diagonal today (I’ve not touched it for ten years) and I was struck by its pale smoky-blue colour. 

I remember when I bought (from new) I was a bit freaked out by this smoky appearance, but folks here on SGL reassured me that dielectric mirrors do look smoky-blue under casual inspection, but in proper use at 45 degrees no such colouration occurs (this certainly seems to be the case). 
 

What I don’t remember however, is any explanation as to why dielectric diagonals look smoky-blue under casual inspection. 
 

Can anyone enlighten me?

Edited June 24, 2022 by great_bear

[Sunshine]

Coatings? just a guess but that could be the case, I'm sure someone will have a definite answer.

[great_bear]

Well it will be coatings - of course. 
 

My question was why - and howcome it doesn’t affect normal operation?

Edited June 25, 2022 by great_bear

[Don Pensack]

The dielectric-coated mirror has no metal coating on the glass, just up to 75 (or more) layers of various oxides and fluorides that each reflect at different wavelengths.

These coatings interfere both constructively and destructively with each other depending on angle, so the stack of coating layers can result in, say, a 98% reflection at 550nm,

but only at one angle.  At other angles, the interference between layers changes because the layers are thicker or thinner than they are at the design angle.

Hence, the interference can change the spectrum of reflection depending on angle.

Properly designed, at the design angle, the reflectivity will be high (say, 95% or more) across the entire 400-750nm visible range (though 425-600nm is OK for nighttime use)

and it really doesn't matter what happens outside the visible range.

But, at different angles, you might see an enhancement of the blue, or yellow or even red depending on the materials in the stack and the angle of incidence.

 

In a star diagonal, the only thing to be aware of is reflection at non-design angles, so it is critical the interior of the diagonal be as black as possible so no off axis light can bounce around inside the diagonal.

Usually, an angle change of +/- 1° won't have any effect on the spectrum except perhaps a very small shift of a few nm, which won't matter in visual use.

[great_bear]

That’s great, thank-you!

[Louis D]

@Don Pensack Can really steep light cones in super fast designs cause issues with dielectric diagonals or dielectric coated secondary/tertiary mirrors?  I've read that dielectric narrow band filters have to be reformulated for super fast telescopes.

[Mr Spock]

For a f4 system the angle of incidence on axis will be a little over 7°. I suspect that won't make much difference to the coating. Not sure if that could have any effect on prism diagonals though.

[Don Pensack]

14 hours ago, Louis D said:

@Don Pensack Can really steep light cones in super fast designs cause issues with dielectric diagonals or dielectric coated secondary/tertiary mirrors?  I've read that dielectric narrow band filters have to be reformulated for super fast telescopes.

Yes it can.

See:

https://www.cloudynights.com/topic/527199-spectroscopic-analysis-comparison-of-nebula-filters/?p=7114958

The same is true of dielectric coatings on a mirror.  Note that at 7°, the bandwidth shift is minor, the transmission change is minor.

The same is true of dielectric mirror coatings.