Hi Dean,

That's definitely interesting; unfortunately the surface in question is SiO2 -
and it's on top of a finished CMOS chip, so I can't simply clear away the SiO2
to expose Si.  This also means I have to keep an eye on thermal budget and
minimize high-temperature processes.  Does the grass formation depend on having
single-crystal Si, or could it be formed in, say, a low-temp CVD poly-Si film
deposited on top of my SiO2?

(And you're correct; spin coating is not a true requirement.  I'd mainly been
thinking along the lines of various spinnable polymers like we'd used in the
past, but something that could be deposited or fabricated and patterned with
other standard fab equipment would likely be suitable too, as long as it doesn't
involve high-temperature processing.  I'm mainly ruling out things like
laminated tapes, sprays, paints, and high-viscosity pastes which are used to
blacken macro-scale components.)

To Shay:
Thanks for the suggestion; I'll contact those people to see if it might be
possible to make something work.  However, this project has some ITAR
restrictions slapped on it, so sending my samples to Israel will involve
significant paperwork.  It's not impossible if they're truly a unique provider
of an essential service, but it's a lot of hassle on my end.

-Mike

On Jan 5, 2012, at 14:43, Dean Hopkins wrote:

> Hi Mike, would you consider "Black Silicon"?
>
> Anyone with a deep RIE system can treat your wafers to form a black silicon
'grass' on horizontal exposed silicon surfaces. The black silicon process is
frequently used  in tuning the process for best vertical sidewalls.
>
> * High broadband optical absorption.
> I don't know the spectral range of absorption, but it is really black, black,
black.
>
> * Spin coatable.
> No, but is that a performance requirement or an artifact of the existing
process sequence?
>
> *
> Electrical resistivity as high as possible.
> As etched the 'grass' is conductive, but a thin oxide or nitride layer can
passivate it.
>
> *
> Patternable.
> Simple. Only open the areas you want to blacken to the RIE.
>
> * Long life.
> As long lived as the silicon in the rest of the device.
>
> I think I might be able to intercede with a fab to get samples run, if you are
interested.
>
> Best,
>
> Dean Hopkins