Re: [ccp4bb]: Problems with low mosaicity crystals

[Marc SCHILTZ <schiltz@globalphasing.com>]

***  For details on how to be removed from this list visit the  ***
***          CCP4 home page http://www.ccp4.ac.uk         ***


Jim Pflugrath wrote:
> Consider that each Bragg reflection has a 3D dimensional volume in
> reciprocal space.  For the sake of argument, let's assume it is
> ellipsoidal.  When the loop vibrates, a refln near the surface of the
> Ewald sphere (in other words near the Laue condition; in other words near
> the diffraction condition) will move in and out of the diffraction
> condition.

I have some doubts on whether this is an adequate explanation.
If the crystal undergoes a translation, this does NOT imply
a translation of the reciprocal lattice with respect to the
Ewald sphere. In the Ewald construction, the incident beam
vector always points to the origin of the reciprocal lattice.
So if you 'translate' the reciprocal lattice, the Ewald sphere
will undergo the same translation and a reflection that is in
the diffraction condition will remain there. (Although the
notion of a 'translation' of the reciprocal lattice is not
very meanigful). In other words, if a translational movement is
applied to the crystal (as is the case when the loop vibrates),
reflections will NOT move in and out of the diffraction
condition (provided, of course, that the crystal does not
move out of the incident beam...).
The diffraction condition is a function of the ANGULAR
position of a reciprocal lattice vector. Therefore, the only
movement which could make move a reflection away from the
diffraction condition is an angular vibration. (This would
correspond to a rotation of the reciprocal lattice around
its origin, with fixed Ewald sphere).

I do not think that turbulences in the cryo-stream can produce
such an angular vibration. This would only be possible if the
crystal has a highly anisotropic shape (plates for instance)
so that the turbulences can exert an angular moment to the
crystal.


Possible sources of the problem might be:

(1) If there is a translational movement of the crystal,
then different parts of the crystal will be illuminated
during the vibration (the crystal might even move out of
the beam for a minute fraction of time). Hence, the
diffracted intensity might change during such a movement,
according to the changes of the crystal volume that is bathed
by the X-ray beam at each instant. If the crystal is small
and if the vibrations are such that the crystal is always
fully bathed in the incident beam, then this problem should
not occur. So it really depends on the crystal and beam
sizes and on the amplitude of the vibrations.

(2) If the spots a very narrow, standard profile-fitting
algorithms might do a poor job. I have seen data frames
collected at an ondulator beamline at the ESRF where the
spots in the centre of the image did not extend over more
than 4 pixels ! It is uncertain to me whether profile-
fitting does a good job on such reflections.

(3) Further, for such narrow spots, each pixel will
receive a high dose of X-rays and the well-known problems
associated with the measurement of high-intensity
reflections (e.g. non-linearities, saturation
effects, count-rate limitations etc...) will show up.

(4) As Jim points out, shutter jitter and goniometer errors
will also have a deleterious effect if the reflection range
is small. This should be more pronounced on partially recorded
reflections. So if you see a clear difference in the data
quality of (summed) partials versus fully recorded reflections,
it might point to hardware problems.

Marcccp4bb@

-- 
================================================================
* Marc SCHILTZ                                                 *
*         Global Phasing Ltd.                                  *
*         Sheraton House                                       *
*         Castle Park                                          *
*         Cambridge CB3 0AX                                    *
*         UK                                                   *
* email:schiltz@GlobalPhasing.com                              *
* Tel: +44-(0)1223-353033   Fax: +44-(0)1223-366889            *
================================================================