Introduction

About the tutorial
This tutorial is intended to explain the common steps of the structure solution and refinement of a 'small molecule' with less than 100 non-hydrogen atoms, using the SHEXTL program suite.
Optionally all steps explained within these pages for the structure example can be performed in practice by the user. Therefore you have to change to directory tutorial/user (look in the Linux basics if needed) and follow the green instructions in the bright blue coloured boxes (like this one).
All text chapters for steps 1-5 of the tutorial have been designed as follows: Paragraphs refering to user action (commands etc.) have a bright blue background. Other text explains the figures as well as important crystallographic concepts. Additional information and background knowledge has been left out on purpose. If you are further interested in crystallography you should consult the links to the glossary where many keywords are explained (still on beginners' level).
 

Should you get stuck on a certain stage of the project, look at the files in directory tutorial/save, where the (preliminary and final) results of each chapter are stored. If all works well, your files in tutorial/user should look the same in the end.

The examplary structure
Typical small molecules are either metallo-organic or 'purely' organic. The example in this tutorial is of the second type, it is a derivative of ascorbic acid. The lacton ring is substituted with a long alkyl(oxy) chain:
The 3D model has been created with the final atomic coordinates of the structure determination; thus it already shows the result of the tutorial steps. Of course, real life is more exciting, as the structure may turn out to be unexpected ...

Programs and files
During the whole structure determination process using the SHELXTL programs (as far as covered here) you will work exclusively with text files, that are edited sequentially with different programs. Regardless of the 'regular' tutorial steps, all those files can be displayed any time with several text editors. We recommend using nedit or kedit (see Linux tutorial or later chapters). The text files contain either experimental x-ray reflection data (HKL files, possibly also FCF files later) or atomic coordinates as well as some commands for the programs (INS and RES files):
The scheme shown here is an overview illustrating the sequence of programs and the information flow (i.e. file transfer) between the programs.
 

The tutorial starts at a stage where the x-ray diffraction experiment on the crystal has been finished and all reflection data has been collected. Data processing - integration, scaling and some other corrections - have also been made already. The structure determination so far resulted in a non-merged HKL file, called momo-new-unmerged.hkl in our case. (see step 2 - XPREP tutorial). Besides the cell parameters (and knowlegde about the expected molecule) you only have that file at the beginning of the tutorial. With the help of XPREP, the data set has to be merged after having confirmed/determined the cell geometry and space group. The merging reduces the number of data while enhancing the quality. During the following structure determination steps the merged HKL file momo-new.hkl will not be further modified. It will always be needed by the respective programs for structure solution and refinement. 
Like XPREP, also SHELXS - the program calculating phases for the corresponding reflections, thus solving the phase problem and enabling the calculation of electron density (step 3) - is used only once in normal cases. The first instruction file momo-new.ins contains some operational commands for SHELXS as well as information about the expected atom types and their estimated number (i.e. a rough sum formula of the structure). The structure solution results in the first atom coordinate file momo-new.res, that can be displayed and modified graphically with the program XP. At that point, the chemical interpretation of the still incomplete and/or inaccurate molecular model has to be done by the user (step 4).
The alternate use of SHELXL and XP (or a text editor, if no graphical display is needed) in step 5 is in principle a continuous update of the INS and RES file with a respectively better structural model. Within this process, the part of SHELXL is the computational parameter refinement for the model (creating a new RES file), whereas the user modifies the content of the model (i.e. type and number of the atoms, using XP or a text editor and creating a new INS file).

Hopefully, this short introduction has given you a first outlook on the tutorial. Next comes the first tutorial step explaining the use of XPREP. If you are not familiar with Linux, the crash course included in these pages is strongly recommended.