D1 Reflectivity

Detlef Smilgies, CHESS

The ion chamber used for GISAXS line-up can also provide some useful reflectivity data for about 3 to 3.5 orders of magnitude. Particularly the reflectivity between the critical angles and just beyond can be studied with high resolution (up to 3000 A). In the following I describe how the reflectivity data can be used for fitting XR curves.

SPEC file

• for every sample name provided with the newfile or new command, a file of equal name and no extension is created in the data subdirectory
• all scan files used for sample line-up and taking images are appended to this ASCII file
• a particular scan is identified by the scan number, which is given in a line starting with #S followed by the scan number
• each image with a given number is associated with a tseries scan of the same number which contains all the monitor reading while this image was taken
• in order to extract files it is the easiest to use a simple text editor like gedit in LINUX or Wordpad in Windows XP, and use the find command to find the correct scan
• the file header contains further information, such as time and date of the scan, and all motor positions before the start of the scan
  
• the line starting with #L identifies the scanned motor and all the readout monitors and detectors
• the #L line is followed by the data in x y1 y2 y3 ... yn format
• it is best to copy this section to a separate file and then read it into Origin or Excel for further processing

Reflectivity macros

• the macro xrfine takes a reflectivity scan and also measures the electronic offset Idark of the V/f converter of the ion chamber electronics
• the offset scan is found as a tseries either right after (old version) or right before (new) the reflectivity scan
• the electronic offset should be subtracted from the reflectivity signal Idet as the first step; note that the offset measurement uses 10 sec per point while the reflectivity scan typically uses 1 sec - the offset has to be scaled accordingly
• if your reflectivity program can handle raw reflectivities, you can feed the offset-corrected data to it
• the incident beam intensity is contained in the initial line-up scans, in particular the samz scan - from this you obtain Idirect
  
• The beam in the synchrotron decays over time - typical CHESS lifetimes are around 10 hours. In order to monitor this decay the beamline is equipped with severeal monitor detectors, all contained in the scan file:
  
• Imon (incident beam after collimating slits)
• Itot (total monochromator output)
  
• CESR (positron beam current)
• In order to detect 1% changes, a beam monitor should read at least 100,000 count/s. The in-line Imon monitor in helium often does not get that high. In this case Itot may be a more reliable signal. CESR outout also works quite well.
  
• if you use the freeware program PARRATT32 from Hahn-Meitner-Institut, Berlin the data need to be processed further:
• reflectivities need to be normalized - the most complete way is R = {(Idet/Imon) - (Iback/Imon)}/(Idirect/Imon)
• the incident angle should be checked whether the ZERO was correctly determined; it is good to fit a straight line to the onset to refine the absolute zero of the incident angle
  
• the incident angle samth needs to be converted into q_z: q_z = 4*PI*sin(samth)/lambda
• PARRATT32 requires a third column with the counting error which can be set to zero; the monitopr ion chamber data could be also treated like a scintillation counter. However, keep in mind that the electronic really measures an anolog signal which is the converted into counts by the V/f converter, so the usual Poisson statistics error = sqrt(Idet) may not be the correct way and over- or underestimates the statistical error
• the energy resolution of the D1 multilayer optics is 1.5%; this can be included in a PARRATT32 fit under "velocity selector"
• often a better fit can be obtained by including a scale factor in the fit: R' = factor R. Unfortunately PARRATT32 does not allow to include a scale factor in the fit, it has to be adjusted by hand.
  
• other fine points of XR, such as the correction of the finite area effect, depend on your experience with reflectivity measurement and analysis
• happy fitting!