[Nexus] Polarisation

[Mark Koennecke]

Hi,

On 04/22/2013 04:17 PM, yayahjb wrote:
> For what it is worth, for CBF we have (so far) gone through two 
> 2-parameter polarization definitions.  The second set (from Harry 
> Powell) has proven more useful than the first, and is a possibility to 
> consider for the current 2-parameter NXbeam definitions.  However, I 
> think Tobias is right that it would make a lot of sense to support the 
> four Stokes parameters as well (see 
> http://en.wikipedia.org/wiki/Stokes_parameters)
> To avoid confusion, I would suggest deprecating
>
> incident_polarization and final_polarization or defauting them to the 
> popular Denzo definitions
>
> and having 4 new alternate fields
>
> incident_polarization_Denzo:NX_FLOAT[i,2]  (in the order norm, then 
> ratio)
> final_polarization_Denzo:NX_FLOAT[i,2]  (in the order norm, then ratio)
> incident_polarization_Stokes:NX_FLOAT[i,4] (in the order I, Q, U, V)
> final_polarization_Stokes:NX_FLOAT[i,4] (in the order I, Q, U, V)
>
> The first pair two will make the most sense to a crystallographer and 
> the second pair will
> make the most sense to a physicist.
>
> I think Tobias is right about transposing the arrays so that the slow 
> index is beamline component
>
> The 2-parameter polarizations in CBF are:
>
> The first pair we started with was:
>
> _diffrn_radiation.polarisn_norm
>     The angle in degrees, as viewed from the specimen, between the 
> perpendicular component of the polarization and the diffraction
> _diffrn_radiation_polarisn_ratio.
>     Polarization ratio of the diffraction beam incident on the 
> crystal. This is the ratio of the perpendicularly polarized to the 
> parallel polarized component of the radiation. The perpendicular 
> component forms an angle of  _diffrn_radiation.polarisn_norm to the 
> normal to the diffraction plane of the sample (i.e. the plane 
> containing the incident and reflected beams).
>
>
> The second pair we got from Harry Powell (who works on mosflm, but who 
> was also thinking
> of ease of support for all crystallographic data processing) was:
>
> _diffrn_radiation.polarizn_source_norm
>     The angle in degrees, as viewed from the specimen, between the 
> normal to the polarization plane and the laboratory Y axis as defined 
> in the AXIS category.  Note that this is the angle of polarization of 
> the source  photons, either directly from a synchrotron beamline or 
> from a monochromater.  This differs from the value of 
> _diffrn_radiation.polarisn_norm in that 
> _diffrn_radiation.polarisn_norm refers to polarization relative to the 
> diffraction plane rather than to the laboratory axis system. In the 
> case of an unpolarized beam, or a beam with true circular 
> polarization, in which no single plane of polarization can be 
> determined, the plane should be taken as the XZ plane and the angle as 0.
>
> _diffrn_radiation.polarizn_source_ratio.
>     (Ip-In)/(Ip+In), where Ip is the intensity amplitude squared) of 
> the electric vector in the plane of polarization and In is the 
> intensity (amplitude squared) of the electric vector in the plane of 
> the normal to the plane of polarization.  In the case of an 
> unpolarized beam, or a beam with true circular polarization, in which 
> no single plane of polarization can be determined, the plane is to be 
> taken as the XZ plane and the normal is parallel to the Y axis.   
> Thus, if there was complete polarization in the plane of polarization, 
> the value of _diffrn_radiation.polarizn_source_ratio would be 1, and 
> for an unpolarized beam _diffrn_radiation.polarizn_source_ratio would 
> have a value of 0.
>     If the X axis has been chosen to lie in the plane of polarization, 
> this definition will agree with the definition of 'MONOCHROMATOR' in 
> the Denzo glossary, and values of near 1 should be expected for a 
> bending-magnet source.  However, if the X axis were perpendicular to 
> the polarization plane (not a common choice), then the Denzo value 
> would be the negative of _diffrn_radiation.polarizn_source_ratio.
> See http://www.hkl-xray.com for information on Denzo and Otwinowski & 
> Minor (1997).
>     This differs both in the choice of ratio and choice of orientation 
> from _diffrn_radiation.polarisn_ratio, which,  unlike 
> _diffrn_radiation.polarizn_source_ratio, is unbounded.
>      Reference: Otwinowski, Z. & Minor, W. (1997). 'Processing of 
> X-ray diffraction data collected in oscillation mode.' Methods 
> Enzymol. 276, 307-326.
>
>
I am in no position to discuss x-ray polarisation. I think polarisation 
in NXbeam and NXpolarisation has always been badly
defined. If we now have the expertise available, I would propose to fix 
this. For neutrons, the polarisation direction is
important. As well as:

- the polarisation_efficiency = (spinup - spindown)/(spinup+spindown)
- the transmission of the filter
- the relaxation_time, for He3 polarisers. The polarisation of those 
degrades with time.

Funny enough, the polarisation_norm, i.e. the measure of the cross 
component, is not so important for neutrons as
it is annihilated in the analyzer. That is what I was told....


Best Regards,

                Mar

>
> On 4/22/13 9:01 AM, Tobias.Richter at diamond.ac.uk wrote:
>> Hi all,
>>
>> I am meeting some unexpected problems recording the polarisation of 
>> the incoming beam.
>> All I could find is in NXbeam:
>>
>>        final_polarization:NX_FLOAT[2,j]
>>        incident_polarization:NX_FLOAT[2,j]
>>
>> There is no further clue about either the 2 or what j is. The values 
>> are float and units are NX_ANY. Hmm?
>>
>> In any case I guess I would struggle to record circular polarisation 
>> in this scheme. Or is someone up for the challenge to explain that to 
>> me.
>>
>> The most natural feeling thing for someone with a vague optical 
>> background (reda: me) would be to use the 4 Stokes parameters.
>>
>> So I assume I propose to change that 2 to 4 (optionally) in the 
>> existing NXbeam.
>> I'd prefer a [j,4] (transposed) array for these values.
>>
>> Leaves the question what j is. There is an i index sprinkled into 
>> that class as well, but I am not sure how they are supposed to line 
>> up. Any one?
>>
>> Regards,
>>
>> Tobias
>>
>
> _______________________________________________
> NeXus mailing list
> NeXus at nexusformat.org
> http://lists.nexusformat.org/mailman/listinfo/nexus