Selenium Xtal Structure?

Uncle Al UncleAl0 at
Fri Jun 21 12:05:29 EST 2002

"B. DeBoer" wrote:
> Nope, You can't do any better from the ICSD, at least up to the 2001/1
> release. From a rather fast look at the ICSD entries, your Phys. Rev.
> reference looks best for both Se & Te.  There was mention of crystals
> grown from the vapor in vacuum or with iodine, so there are alternatives
> to hot aniline. Unless your proposed Eotvos experiment requires paired
> crystals of the same substance but opposite chirality (sorry, didn't
> read it thoroughly), what amount a mass of stretched DNA fibers?

Tellurium data is in hand.  We are still worrying the selenium.

Phys. Rev. B16 4404 (1977)
J. Phys. C 8 L445 (1975)
J. Phys. Chem. Solids 35 1089 (1974)
Acta Cryst. 21 A46 (1966)
Donohue, Jerry, The Structures of the Elements (Wiley-Interscience:
NY, 1974), p. 370
Zingaro, Ralph A., Cooper, W. Charles, Eds., Selenium (Van Nostrand
Rheinhold Co.: NY, 1974), pp. 87, 137

They perversely present the x-axis Se fractional coordinate in a
non-standard setting, then don't list the y- and z- reset
coordinates.  A lot of that data is purely awful. One group did a
hugely complex neutron diffraction study on a melt-grown cm Se
crystal.  Said crystal was severely disordered (largely amorphous)  by
their own data - before manipulation to put lipstick on the pig. 

Even opposed chirality probably isn't good enough.  It must be opposed

and Eotvos balance test masses are typically 2 cm diameter and high
cylinders or 1-2 cm diameter spheres.  My experiment challenging the
Equivalence Principle requires single crystal test masses for maximum
structural density and perfection.  Biological materials are
horrible.  Even if both enantiomers/parity pairs were available as
large single crystals, C,N,O all have the same mass near enough and it
isn't large.  We've calculated alanine amide - crappy.  We can
guestimate the parity divergence of custom-synthesized DNA given the
geometry (inertial axis distribution) of its unit cell contents.  CHI
is approximately zero, where 1.000000 is perfect divergence.

I've talked with the ICSD and they offered to forward some CIF files
for Se, especially the Phys. Rev. B16 4404 (1977) at ambient
pressure.  That was two weeks ago.  They are very busy and will do it
as possible.

It turns out grey Se is incredibly fragile as a crystal.  Minor
surface mechanical damage causes deep lattice damage.  Tellurium can
be Czochralski grown under hydrogen then thermally annealed under
hydrogen to give semi-conductor grade material, including dislocation
density.  It can be cut and machined without deep lattice damage.  It
isn't quite as tightly wound as a helix, but as we are only interested
in nuclear positions (where all the mass resides) helix isolation in
space is the same as in selenium even though Te atoms are bigger.  (Te
is an incredible contact hazard.  It's a physicists' problem until
they get the fabricated shape gilded).

No other parity-pair space group crystal structure is anywhere nearly
as good as Te and Se.

There is no theory predicting how quantitative parity will fare as
larger volumes of crystal lattice are considered.  We have calculated
through 2921 Te atoms (99,256 A^3) and see a rapid ramp up (better
inertial axis equality of the sample) followed by a third decimal
place decline.  Artifact or trend?  That lattice chunk required 15 CPU
hrs in a RS6000/Power3.  The next step up will eat 150-300 hrs, which
isn't in the budget.  That is for inertially equiaxial hexagonal
cylinders of identical helices.  Te spheres are in the computer as you
read this.  Selenium will follow.

I'm going to take another crack at publishing in Phys. Rev. D when the
data is complete.  However onerous the real work, getting it into
REVTeX promises to be a vastly bigger pain.

Uncle Al
 (Toxic URL! Unsafe for children and most mammals)
"Quis custodiet ipsos custodes?"  The Net!

More information about the Xtal-log mailing list