Re: Digest for power-satellite-economics@googlegroups.com - 1 update in 1 topic

[Geoffrey Landis]

People have been talking about up-conversion and down-conversion for decades, but the efficiency of wavelength shifting is so poor that in real-world applications you always lose more than you gain.

Efficiency is only slightly awful if you only want to up-convert a single wavelength (eg., frequency doubling a laser) but that’s not good enough for solar cells, you need broadband.

—Geoff

Fusing photons for upconversion

[Keith Lofstrom]

On Wed, Sep 20, 2023 at 02:01:28PM -0500, Geoffrey Landis wrote:
> People have been talking about up-conversion and down-conversion for decades,
> but the efficiency of wavelength shifting is so poor that in real-world
> applications you always lose more than you gain.
> Efficiency is only slightly awful if you only want to up-convert a single
> wavelength (eg., frequency doubling a laser) but that’s not good enough for
> solar cells, you need broadband.

The article mentions the "decades of work", but also claims
some new discoveries. It is typical IEEE Spectrum "equation
bereft, but with diagrams and vague citations" engineering
journalism, interesting and suggestive of further delvings
into the peer-reviewed literature.

And if the article is indeed invalidated by RECENT WORK ...
please CONTACT SPECTRUM'S EDITORS ***PRONTO***, before
the online article goes to press. It appeared online
19 September, so there will be a day or three to fix this.

A few months ago, with the help of other subscribers on
this list, we successfully spiked what would have been
a Spectrum cover story about "Spinlaunch". Spectrum
pulled that article and ran a different cover story.

That will happen often; with 400,000 IEEE members and
MANY academics and persnickity geeks among us, Spectrum
demonstrates Linus Torvalds' first law of open source
software: "with enough eyes, all bugs are shallow."

-----

As an optics adept, Geoff might also appreciate the
July Spectrum lead article about advanced Extreme UV
semiconductor patterning systems from ASML in Netherlands
... and (parenthetically) for Intel in Hillsboro Oregon.

https://spectrum.ieee.org/high-na-euv

Hundred megabuck optical marvels the size of houses that
print fractional wavelength transistors the size of
molecules, using 13.5 nm photons and reflective optics.

Note that the destruction of the huge Antonov cargo
aircraft by the Russian invasion of Ukraine will limit
future ASML designs to what fits in a 747 cargo aircraft.
Probably multiple subsystems shipped from the Netherlands
for final assembly and calibration at an ASML factory in
Oregon (and dozens of other customer sites worldwide).

My Linux hacker buddy Jaap works as a field engineer for
ASML. 8 digit dollars buys a community, not just a
fantastic optical machine.

'Shipped" involves special flight controls on the jumbo
cargo jets to minimize vibration and gee loads. We cannot
ship these delicate monsters to hypothetical chip factories
in space using rockets punching through Earth's turbulent
atmosphere.

That said, semiconductor chips made on Earth are getting
smaller and lighter; there is "no Earthly reason" why we
would want to manufacture those tiny dabs of semiconductor
in space. Launching unpackaged silicon chips (nanograms
to grams) is 100 times cheaper than manufacturing them.

But with /THAT/ said, launching "naked" unpackaged silicon
chips, then packaging and wiring them together in orbit,
makes excellent sense. We can probably make the packages
with space-source material, that is mostly ceramics and
metal forming (and again, supercleanliness).

Contamination control won't be effortless, even in vacuum;
"shit floats", literally. We must develop new technologies
to do with space systems vacuum and microgravity what we've
already learned to do on Earth. If you are a technologist
allergic to rocket fuel and broadcast TV, there will be
many other space-sourced products worthy of your talents.


Keith

--
Keith Lofstrom kei...@keithl.com