OU Crystallography Lab

Department of Chemistry & Biochemistry
Chemical Crystallography Laboratory

| Home | Site Map | Contact

A paper that describes the results of experiments to optimize data collection conditions for area detector instruments, and in particular ccd instruments, was published by S. Rühl and M. Bolte. ("Strategies for data collection on a ccd-diffractometer," Z. Kristallogr., 2000, 215, 499-509.)

CCD Instrument Instructions

Data Collection

These instructions pertain primarily to using the SMART APEX CCD area detector with the Bruker Platform diffractometer at the University of Oklahoma Crystallography Lab.

  1. Log into the computer next to the instrument using the xtal account on the msgxray domain.

  2. Start the SMART program on the PC (Start, Programs, Bruker-AXS Programs, SMART).

  3. Optically align the sample using the Goniometer/Optical option. Start the VIDEO program (Start, Programs, Bruker-AXS Programs, Video) to view the crystal as it rotates. Be sure that the crystal stays centered when rotated to all positions using all four buttons A, B, C, and D of the manual control box.

  4. In the Video program, use the Tools/Vector Cursor routine to measure the dimensions of the crystal, by clicking the mouse when it is at one edge of the crystal and then moving the mouse to the other edge of the crystal. The crystal may be rotated using the Phi, Fast, Up/Down buttons on the manual control box. When done, exit the Video program.

  5. Check the status of the instrument by typing the letter "u" on the keyboard from the Goniometer/Optical routine. In particular check the tube power, which should be 50 kV and 30 mA, and the detector temperature, which should be -41 to -43 °C. If the Generator power reads 0 kV and 0 mA then restart the generator using the instructions in Instrument Operating Instructions. If the power level of the generator is low (20 kV, 5 mA) then use Goniometer/Generator to raise the power to 50kV and 30 mA. If the temperature of the detector is warmer than -41 °C then wait for 5 to 15 minutes for the detector to cool. If you have troubles with either the generator or the detector, please contact the lab director.

  6. Set up the project information for this sample in Crystal/New Project. Both the working and data directories should point to the same subdirectory under "c:\frames\". Be sure to include the chemical formula, crystal description, and data collection temperature.

  7. Take a rotation frame image of the crystal using the Acquire/Rotation option. If the rotation frame shows no spots or if it shows powder diffraction rings, mount and align another crystal of the sample.

    Typical rotation image 
         showing two mirrors of symmetry This image is a typical rotation image. Note that the image must show two mirror planes of symmetry. These mirrors are related to how the image was taken and have nothing to do with the symmetry of the sample. If these two mirrors are not present, then either the instrument is out of alignment or the crystal is not centered on the instrument.

  8. Determine initial cell parameters with the Crystal/Unit Cell command. Consider the following notes if problems arise during the cell determination step.

    • If there are very few "usable" spots in the cell parameter determination, when more were anticipated from the rotation frame, then repeat the Crystal/Unit Cell step with longer count times and a greater number of frames per run.

    • During the indexing step, you should see one or more lines of "1"s in the output. If the crystal indexes only with an implausibly large unit cell, try another crystal. If the program has trouble indexing the spots or refining the cell parameters and you feel this is an otherwise good crystal, collect triclinic data and try the twinning programs.

  9. Improve the cell parameters with the following steps.

    • Use Crystal/Modify to set the "c" flag for all located spots.

    • Initially set the Max RLV (Reciprocal Lattice Vector) Error to 0.05. In subsequent refinements, use smaller values for the RLV deviations in order to remove any outliers with non integer hkls. For the last refinement cycle, run Crystal/---LS with -1 for crystal system constraints.

  10. Test the validity of the unit cell.

    • Check the fit of the spots to the unit cell with Analyze/Display. Be sure to turn on the calculation of hkl locations. Examine several frames to confirm that the cell does properly fit the peak postions.

    • Estimate the volume of the chemical formula unit (18 × # of nonhydrogen atoms). The cell volume should be an integer multiple of the estimated volume of the chemical formula. This integer is usually related to the presumed symmetry of the cell, ie, related to the number of presumed symmetry operations for the crystal lattice symmetry.

    • Test to see if the structure has been previously determined by checking the Crystal Data File Database. This check is performed by the Crystal/CDF Search routine.

  11. Estimate a reasonable time (10 to 120 sec) to count each frame. Do not exceed 120 seconds per frame count time without discussing your data set with the lab director.

  12. Collect a new dark current correction frame using the Detector/Dark Current option. On the next menu enter the count time in seconds to be used for data collection, and change the name of the data file so that the last three characters of the root file name show the count time in seconds (eg. enter the time in the "###" field of "1071L###"). The program will ask you to confirm that the detector is operating at a speed of 400 MHz.

  13. Optimize the scan runs using the ASTRO program or use either of following sets of runs. Note these are already set up in the Acquire/EditHemi, Acquire/EditMultiRun, or Acquire/EditQuadrant. Change the count time in each run to the value you have selected above. Note that the following data collection conditions assume that the detector is set at a distance of 5 cm.

    Table 1. MultiRun Collection Conditions with Mo Kα Radiation
    # 2Theta Omega Phi Chi Axis Range # Time
    1 -28 -28 0 54.8 2 -0.3 606 15
    2 -28 -28 90 54.8 2 -0.3 606 15
    3 -28 -28 180 54.8 2 -0.3 606 15
    3 -28 -28 270 54.8 2 -0.3 606 15
    • Start data collection using the Acquire/MultiRun command. Be sure to set the Job Name to the name being used for the project. Do not change any other options except the Job Name or Title.

    • Enter data collection information in the Log Book.

    • When data collection is finished improve the cell parameters by the following steps:

    • Clear the reflection array with the Crystal/---Clear command.

    • Threshold data frames in sets of 100 to 200 in different regions of reciprocal space using the Crystal/---Threshold command. Continue this step until the reflection array is filled (999 peaks) or there are no more frames to threshold. Do not threshold the same set of frames more than once.

    • Index these peaks with the Crystal/---Index command.

    • Refine the cell parameters using Crystal/---LS command using a Max RLV Error of 0.02. The program may warn that no spatial correction is being used. Since the APEX detector does not have an optical taper, there is no spatial distortion in the frames and hence no spatial correction is needed. Repeat this step using smaller values for the Max RLV Error until a value of 0.004 is reached. During least squares cell refinement it is not unusual for the program to reject a few peaks because they do not fit the model within the specified maximum reciprocal lattice vector error; however, if a large percentage of the peaks are being rejected, then cancel the refinement and begin again with a larger Max RLV Error.

    • Reduce the cell using the Crystal/---Bravais command.

    • Refine the cell parameters again using a Max RLV Error of 0.02 and Constraint of -1 Triclinic. The -1 allows a more sophisticated non-linear least squares refinement of the cell and instrument parameters.

  14. If an analytical absorption correction is needed, index the faces of the crystal using the instructions in the SMART manual (platform goniometer with Video camera).

The instructions above serve as a general guide for nearly all data collection experiments with Mo Kα radiation. It may occasionally be necessary to use some of the more sophisticated commands found in either the online help or the manuals of the Bruker SMART manual.

If you have any problems using the instrument or the SMART program, please contact the lab director either in person or by telephone. His home telephone number is located on the outside of the door to the lab.

If you feel there is a safety problem with the instrument, please close the front doors to the instrument, contact both the campus radiation safety officer (325-0820) and the lab director as soon as possible, and place a note on the front of the instrument stating "INSTRUMENT PROBLEM" and include the date, your name, and a telephone number where you can be reached.

Troubleshooting the Data Collection

Rotation Image troubles

If the shadow of the beam stop is not visible on the rotation image, the power to the X-ray generator is either turned off, or is at low levels. Turn on the generator by beginning with instruction 5 at the CCD instrument operating procedures page. Raise the power of the generator using the Goniom/Generator command in the SMART program to 50 kV and 30 mA.

If the shadow of the beam stop is seen, but no spots appear on the rotation image, then try a single run of frames with # frames set to 10, count time set to 60, and the name of the run set to "temp". If no spots appear on these frames, then the sample is either not crystalline or it is too small in mass to contribute sufficient scattering for use on this instrument. Either grow bigger crystals or send the sample to someone with access to a synchrotron. If spots are see then try to collect the data with very long, 90-120 seconds/frame, count times.

A rotation image of a glass sample. This image is a rotation image of a glass sample.

A rotation image showing the 
   rings of a polycrystalline sample. This image is a rotation image of a polycrystalline sample showing the rings of diffraction.

Indexing troubles

Indexing troubles occur as three main types of problems. First, check the crystal to detector distance in the Edit/Config menu of SMART. This distance (in cm) is shown on the instrument in the channel of the V-block that supports the detector. The detector-sample distance is given (in mm) as the distance at the front of the detector support bracket.

A second cause of indexing problems is non-merohedral twinning as shown by difficulties with indexing or unexpectedly long cell lengths. Use the "cell_now" or Gemini progams to index the domains of the twins. If one twin domain is dominant over the other twin domain, then threshold the spots again using a larger I/σ low threshold, say 20.0 instead of the default 10.0. Also set the Fraction which must be fit to a smaller value, say 0.6, in the indexing routine. If the domains are not significantly overlapped, then using these tips can allow you to collect and process the data as if the sample were a single crystal.

A third indexing problem occurs for very strongly diffracting samples. Such samples will sometimes have what appears to be satellite peaks around the main peak. The satellite peaks are actually due to the shape of the crystal and the fact that narrow slices of reciprocal space are being collected at any given time. As with the twinning case, threshold the spots with a larger I/σ low threshold. This subset of stronger spots should index without further trouble.

Multiple Exposure Times

This information is based on the generous contribution of Dr. Ludger Häming, Bruker-AXS, to the Bruker CCD e-mail list. This example is intended solely to demonstrate the use of a script file to collect different runs at different exposure times. This method is needed when collecting data with Mo radiation and the detector needs to be moved more than 7 cm from the sample or when using Cu Kα radiation. The example below shows a setup for collecting intensity data on a sample with triclinic symmetry using Cu Kα radiation and a 4 cm sample-to-detector distance.

Edit the Multi-Run sets of frames to have parameters similar to:

Table 3. Multi Run Data Collection Conditions
# 2Theta Omega Phi Chi Axis Width # Time
1 -38 -38 0 54.8 2 -0.5 724 10
2 -38 -38 20 54.7 3 -0.5 362 10
3 -38 -38 92 54.7 3 -0.5 362 10
4 -38 -38 164 54.7 3 -0.5 362 10
5 -38 -38 92 54.7 3 -0.5 362 10
6 -38 -38 308 54.7 3 -0.5 362 10
7 -104 -104 0 54.7 2 -0.5 724 50
8 -104 -104 0 54.7 3 -0.5 362 50
9 -104 -104 60 54.7 3 -0.5 362 50
10 -104 -104 120 54.7 3 -0.5 362 50
11 -104 -104 180 54.7 3 -0.5 362 50
12 -104 -104 240 54.7 3 -0.5 362 50
13 -104 -104 300 54.7 3 -0.5 362 50

Obviously, the count times should be adjusted for the sample at hand. The high scattering angle frames are best collected with counting times that are 5 times counting times for the low scattering angle frames. This 1:5 ratio of counting times tends to produce even scaling between the sets of frames.

Next create a simple text file in SLAM format must be created. An example file is shown below. Assign this file a name such as "multi.slm", and place this file in the directory that will receive the data frames.


After the runs have been properly set up, start data collection by going to the Command line item in the Level menu and typing @multi. If data collection needs to be stopped, press the cntl-c characters, followed by enter. At this point you may return to the menu commands by typing menumode

Data Integration

  1. After data collection has completed, start the data integration program Saint+ (Start, Programs, Bruker-AXS Programs, SAINTPLUS).

  2. Select New under the Project menu to set up the project description. Enter the project name and indicate the location of the data frames by selecting the *.p4p file corresponding to the first set of data runs.

  3. Two additional steps are needed before starting SAINT. In the SAINT menu, first select "Initialize". Then under the SAINT menu select "Execute".

  4. Select the "Integrate" option. Change the cell parameter constraints during integration to "Unconstrained". Also, uncheck the box to constrain crystal translations during cell parameter refinement. Start integration by hitting the "Integrate/Merge" button at the lower left corner of the menu.

  5. When integration has completed perform the following tasks at a command prompt window that is in the proper \frames\*\work directory.
    ren newdirm.p4p newdir.p4p
    ren newdirm._ls newdir._ls

  6. Run the Sadabs program to correct for absorption and related systematic errors in the data. In general, most questions can be answered with the default response. Be sure to correctly specify the Laue symmetry.

  7. Copy necessary data files to a network hard disk for further analysis.

Editing the Active Pixel Mask

This information is based on the generous contribution of Dr. Ludger Häming, Bruker-AXS, to the Bruker CCD e-mail list.

There are times when parts of the data frames must be excluded from data analysis. This can be done by editing the active pixel map files to exclude these regions so that SAINT will exclude these regions. Note that there is a separate active pixel map (*._am) file for each run of the data set.

Preparing the *._am file:

  1. Run SAINT with no change for the active pixel mask.
  2. Find an old .\work\*._am file from the old integration run.
  3. Use the FILE/LOAD command in SMART to read this file and the corresponding background *._ib file. Set the Max. Display Counts to 1 in order to get a good contrast.
  4. Use the ANALYZE/BOX to select a part of the active mask to be excluded. The shape of the box can be changed with the escape button. Once the box is shaped and positioned, hit the backspace button to set a new value, 0, for this region.
  5. Use SAVE to export this edited mask to the hard disk. Copy this pixel mask to the pixel masks for the other runs of the data set.

Integrate the data again. This time, in the Advanced Integrate menu, select the box for "Use initial background and active pixel mask from last integration". Note that both the background and pixel map files must match the data runs. If the data set contains three runs be sure that the files name1._ib, name2._ib, and name3._ib; name1._am, name2._am, and name3._am are in the appropriate work directory.

Edit the frame headers

Occasionally frames are collected with incorrect information in their headers. For example, it is possible to set the detector at a particular distance from the sample, but forget to change the sample-to-detector distance in the configuration parameters before collecting the frames. Luckily, the folks at Bruker have written a program that can correct the headers of the files. This GUI-based program is called FrmUtility.exe and is in the saxi\sxtl directory.

To use the program, open the program and set the Project directory to the desired directory. Then enter one frame from the run that you wish to change into the Input filename field. Select the Import defaults and complete as many of the items on all submenus. Change the file extension in the Input filename field to the wildcard character*. From the Fixup Header tab select the appropriate group of items to be updated--in the example above, select Update detector distance, center, mounting fields. Then click the Update button. After a short time (sometimes a minute or two for hundreds of frames) a new screen will appear listing the files that have been updated.

Thank you for visiting our site.