In this laboratory we use the SHELXTL software package most often. There are two other commercially available program for structure solution and refinement, Sir97 and . Often times program PLATON written by Dr. A.L.Spek is used to deal with disordered solvents and to perform other calcualtions. The correctness of the solved structures is normally verified by the CIFCHECK. Bruker software designed to deal with TWINNING is an excellent tool.

Molecular Structure Laboratory Software Links

Home | Facility | Personnel | Software | Policies | Data

 

 

  SHELXTL-97 Manual

  PLATON

  Sir97

  CIFCHECK

  Cambridge Structural   Database

  Bruker Software Site

  CHEM_613_2005

  

 

 

 

 

 

 

 

 

 

 

TWINNING

The Bruker-AXS twinning software was written by Robert A. Sparks. These programs are designed to help the user sort out data sets that were collected on samples exhibiting twinning of a non-merohedral nature. In this type of twinning some or all of the points of the different reciprocal lattices do not overlap. Merohedral twinning occurs when the two (or more) reciprocal lattices exactly overlap. The following programs may be run individually from a command prompt or may be run from the Gemini GUI in the Bruker Software Package. The Gemini interface is recommended because of program improvements and because of a nice outline of steps needed in a twin analysis. There are four programs in the twinning suite:

  • TWINDX produces one or more possible indexing solutions from a *.p4p file.
  • TWUTIL is a utility routine used to generate various *.p4p files.
  • TWHKL generates a combined *.hkl (type 5) file from two or more *.hkl files.
  • TWROT determines the rotation needed to superimpose one lattice onto another.

The programs may be called from the command line or may be accessed from the SAINTPLUS group of programs. There are no command line qualifiers for these commands. The programs do usually have reasonable values for default parameters. The displayed (default) value for a parameter is accepted by typing the slash "/" character followed by <enetr>. The usual course of action is as follows: Threshhold many reflections in SMART. Fill the reflection array if possible. Run TWINDX using the *.p4p file created in SMART. TWINDX creates two files, a *.twx and a *.sum file. The programs will use the *.twx file; people will find easier to read the *.sum file. These two files contain essentially the same information about the different indexing solutions. Look through the *.sum output to select the best possible indexing solution(s). Run TWUTIL to calculate the "complement" of the selected solution. Be sure to save the new *.p4p file with a different root name. Run TWINDX with the new *.p4p file saving the *.twx and *.sum with still a new root name. Look through the second *.sum output to select the solution that has similar cell parameters to the original solution. Once two separate solutions with similar cell parameters but different orientation matrix components have been found proceed with the next step. If no similar solution can be found, then select another solution from the original *.sum file and start at step 4. Run TWUTIL to generate a *.p4p file from the solution in step 3 but not the solution from step 6. Run TWUTIL a second time to generate a *.p4p file from the solution from step 6. Use these two *.p4p files in separate runs of SAINT to integrate the data in the two reciprocal lattices. Correct the two separate *.raw data sets for absorption generating two different *.hkl files. Run the TWHKL program to merge the two *.hkl files into one new *.hkl file with "HKLF 5" format. Run the TWROT program to determine the nature of the twin rotation. Data sets that may look twinned on the data frames (unreasonably close sets of spots in different parts of reciprocal space) may actually be split rather than twinned. Split crystals usually give one solution in step 2 above that uses nearly all of the spots; however, integration in SAINT using this solution produces only a very poor data set. When the three cell lengths of split crystals are relatively short, it is sometimes possible to integrate the data using a wide spot shape description.