Answer: Pbnm Refinement, Whole Range
Refinement with the group Pbnm in the whole range of data
The match of the pattern calculated with the structure obtained in the previous
refinement (done with the lower angle area data only) with the full range of
experimental data shows
that the peakshape and halfwidth parameters taken from there are actually very bad:
Higher angle area
The picture shows that, before any conclusions about group Pbnm are drawn, a further refinement with the whole range of data with a stress on the peak shape parameters would be necessary:
| Step | Circles | Rwp | Scale (refined) | Other parameters refined (values after the refinement) |
| 1. | 2 | 20.280 | 2.53 | PEAK SHAPE PARAMETERS Gam0 = .409982 Gam1 = -.016668 Gam2 = 9.8788e-5 |
| 2. | 4 | 15.676 | 2.5131 | CELL PARAMETERS (a, b, c) = 6.928331 8.418787 5.370543 ZEROPOINT = .08996 U = .015437 V = .098346 W = .152900 |
| 3. | 6 | 15.095 | 2.5873 | CELL PARAMETERS (a, b, c) = 6.928464 8.418672 5.370291 ZEROPOINT = .08889 Gam0 = .185117 Gam1 = -.008665 Gam2 = .000072 |
| 4. | 6 | 14.747 | 2.6299 | CELL PARAMETERS (a, b, c) = 6.928179 8.418952 5.370339 ZEROPOINT = .08877 HALFWIDTH PARAMETERS U = .004823 V = .047713 W = .161651 PEAK SHAPE PARAMETERS Gam0 = .123027 Gam1 = -.003742 Gam2 = .000054 |
| 5. | 3 | 14.394 | 2.6296 | CELL PARAMETERS (a, b, c) = 6.928209 8.419052 5.370334 ZEROPOINT = .08892 PB (x, y) .16890 .18964 S (x, y) .18148 .43201 O1 (x, y) .09514 .59230 O2 (x, y) .03782 .30685 O3 (x, y, z) .30676 .41891 .97427 |
| 6. | 3 | 14.116 | 2.6754 | PB (B) .3600 S (B) .4294 O1 (B) .5585 O2 (B) .5989 O3 (B) .3755 |
| 7. | 4 | 14.061 | 2.7085 | PB (x, y, B) .16915 .18962 .3654 S (x, y, B) .18019 .43324 .3430 O1 (x, y, B) .09510 .59201 .6061 O2 (x, y, B) .03900 .30685 .6595 O3 (x, y, z, B) .30725 .41871 .97464 .3850 CELL PARAMETERS = 6.928194 8.419136 5.370313 ZEROPOINT = .08896 HALFWIDTH PARAMETERS U = -.000104 V = .044196 W = .162877 PEAK SHAPE PARAMETERS Gam0 = .255005 Gam1 = -.006164 Gam2 = .000066 |
A big drop of R-factor value in only one step (from 20.280 after the first step to 15.676 in Step 2) shows how bad peak shape parameters were determined from data from the lower angle area only, and how big hindrance they were for refinement of the other parameters.
After these 7 steps any further refinement attempt proved to be futile, and it seems that 14% is the best Rwp value that can be obtained with the orthorhombic group.
