Dear Sir, How do we know that crystal is single and perfect?

zaw

Date:22-6-2005
Dear Sir,
How do we know that crystal is single and perfect? Why do we refine the
data after x-ray diffraction method? I hope your reply.

sincerely yours,

ZAW

Uncle Al

zaw wrote:
>
> Date:22-6-2005
> Dear Sir,
> How do we know that crystal is single and perfect? Why do we refine the
> data after x-ray diffraction method? I hope your reply.

You don't want a perfect crystal for a diffraction structure. You
want slight mosaicity to prevent destructive intereference. When the
primary beam makes the proper glancing angle "theta" with the planes
of a stack to produce a reflection, the reflected ray also also makes
this angle with the plane and is therefore reflected back parallel to
the primary beam. Each reflection changes the phase of the wave by
(pi)/2, so the twice-reflected ray is out of phase by "pi" with the
primary wave. They subtract amplitudes, and it gets worse the deeper
into the crystal you go. Primary extinction is proportional to

tanh(sq)/sq

where "s" is the number of planes in perfect registration and "q" is
the reflection efficiency of the plane. Primary extinction eats you
for reflections having large structure factors (large q) and large
spacings d(hkl), and is appreciable for paths greater than 10^(-4) cm.

A single crystal is often obvious from its diffraction pattern -
though twinning can be tricky. Quartz has at least seven different
ways to twin.

Library, Google, your local crystallographer. Show some initiative.


--
Uncle Al
http://www.mazepath.com/uncleal/
(Toxic URL! Unsafe for children and most mammals)
http://www.mazepath.com/uncleal/qz.pdf

Arnold Neumaier

zaw wrote:
> Date:22-6-2005
> Dear Sir,
> How do we know that crystal is single and perfect? Why do we refine the
> data after x-ray diffraction method? I hope your reply.

No crystal is single and perfect. This is just a convenient idealization.

Raw measurement data must be refined to extract the information of interest.


Arnold Neumaier

Torsten Becker

zaw wrote:

> Why do we refine the data after x-ray diffraction method?

During data refinement you use additional knowledge about your system
to increase the quality of the final structure.
If you have for example a crystal that scatters at 1.5 \AA resolution,
which is already pretty good for large molecules like proteins,
naively speaking you can't say much about atom positions below 1.5
\AA (which is pretty bad if you want to understand how an enzyme
works).

The solution is 'data refinement'. If you know for example that two
atoms are bonded covalently, it takes an enormous amount of energy to
stretch such a bond. It's most likely that these atoms will be close
to their minimum bond length (which are known quite often).
That way, by starting with a blueprint of your molecule, you can come
up with an energetically reasonable structure containing more
information than is inherent in the raw electron density.
(As a rule of thumb, with the above resolution one can trace back
structural changes in relative positions of atoms down to 0.3 \AA.
But i'm not sure about that number. You have to ask a crystallographer
for more details)

Hope it helps,
Torsten


--
*******************************************************
Torsten Becker Structural Biology/Bioinformatics
University of Bayreuth
+49-921-55-3547 (office) Universitaetsstr. 30, BGI
+49-921-7273406 (home) D-95447 Bayreuth; Germany

torsten.becker@uni-bayreuth.de
********************************************************

zaw

Date:
25-6-2005
Dear Mr Torsten Becker,
I just received your reply answer today. Thanks you for your reply.
One thing I confused is that 1.5\AA and 0.3\AA. Could you explain it
to me? I will be looking forward for your reply.

sincerely yours,

ZAW WIN
Research Student,
Nagoya Institute of Technology,
Nagoya, Japan
466-0063


Torsten Becker wrote:
> zaw wrote:
>
> > Why do we refine the data after x-ray diffraction method?
>
> During data refinement you use additional knowledge about your system
> to increase the quality of the final structure.
> If you have for example a crystal that scatters at 1.5 \AA resolution,
> which is already pretty good for large molecules like proteins,
> naively speaking you can't say much about atom positions below 1.5
> \AA (which is pretty bad if you want to understand how an enzyme
> works).
>
> The solution is 'data refinement'. If you know for example that two
> atoms are bonded covalently, it takes an enormous amount of energy to
> stretch such a bond. It's most likely that these atoms will be close
> to their minimum bond length (which are known quite often).
> That way, by starting with a blueprint of your molecule, you can come
> up with an energetically reasonable structure containing more
> information than is inherent in the raw electron density.
> (As a rule of thumb, with the above resolution one can trace back
> structural changes in relative positions of atoms down to 0.3 \AA.
> But i'm not sure about that number. You have to ask a crystallographer
> for more details)
>
> Hope it helps,
> Torsten
>
>
> --
> *******************************************************
> Torsten Becker Structural Biology/Bioinformatics
> University of Bayreuth
> +49-921-55-3547 (office) Universitaetsstr. 30, BGI
> +49-921-7273406 (home) D-95447 Bayreuth; Germany
>
> torsten.becker@uni-bayreuth.de
> ********************************************************