Special considerations for Bruker data
There are some peculiarities associated with Bruker data. Topics are
sequential data acquisition
byte swapping
directory structure
digitally filtered data
data transferred via Bruknet
2D Aspect data
See also: 2D processing, baseline correcting digitally filtered data
True quadruature (complex) detection is accomplished by splitting the signal from the probe into 2 channels, related by a 90-degree phase shift of the reference frequency. Bruker data is often acquired using "sequential" data acquisition, rather than true complex, or "simultaneous" data acquisition. To accomplish quadrature detection using this scheme, the data is digitized at twice the rate and points are placed into the 2 channels alternately. In other words, the nth point placed into each channel are not acquired the same time. This requires that a special type of FT be performed (BT). Nuts should correctly identify sequential data, which is identified by setting the Domain parameter to TPPI, and executing an FT should automatically perform the appropriate type of FT. If a complex FT is performed on sequential data, artifacts are created which resemble severe quadrature images - every peak has an out-of-phase mirror image. See example.
Sequential detection can cause a different artifact with similar appearance. If the 2 channels are switched, the data points become out of order in time, and following FT, each peak will have a large, out-of-phase mirror image. See example. If the data is complex, switching the channels simply reverses the spectrum. We have been unable to identify any header information to detect which channel is real and which is imaginary. In addition, we have some information that it depends on the method used to transfer the data from the Aspect to the PC. Executing an RI command (switching Real and Imaginary points) before the FT will fix the problem. An entry has been added to the NUTS.INI file which allows the user to automatically perform an RI on Bruker Aspect data at the time it is imported into NUTS, so that the user does not have to do this explicitly. The relevant entry in the NUTS.INI file is:
BRUKER_ASPECT_RI = TRUEIf the line is not present in the INI file then add it, left justified as typed above. There is also an entry in the nuts.ini file which is:
RI_ON_IMPORT = FALSE
This will apply an RI operation on all imported data, so will solve the problem only in cases where all data being imported into NUTS comes from an instrument that has this problem.
Some file transfer processes from Bruker Unix systems to the PC incorporate a "byte swap" in the file transfer process. The NUTS importing process wants the file to be an exact image of the Bruker Unix system fid and will get these files wrong on importing. This results in a corrupted file which does not resemble a normal FID. If the parameters of these files are imported correctly but the fid is wrong, set this parameter in the NUTS.INI file to TRUE.
BRUKER_UNIX_BS = TRUENote that changes made in the NUTS.INI file are not implemented until the next time the NUTS program is started.
The directory structure for Bruker Unix files must be exactly reproduced on the PC (or Mac) for NUTS to import the data successfully. The file name supplied by the user when the data are collected becomes the name of a directory. Within that is a subdirectory whose name is a number (1, 2, 3, etc.). The files associated with the data are in this subdirectory. This directory structure must be created on the PC (or Mac) and files placed in the correct directory. File names must be unchanged in moving data from the spectrometer, or NUTS will not be able to find the required information. When NUTS translates the data, it creates a file whose name is the name of the parent directory with a $ appended to the front.
Bruker data from Avance series spectrometers (DnX models) require special processing. The data have been digitally filtered and "decimated" prior to saving the FID. The FID, when first imported into NUTS, appears distorted because the initial points of the FID are zero. Before an FT can be performed, an RD (Rotate Data) operation must be performed. This is a circular left shift of a specified number of points. NUTS calculates the appropriate number of points for the shift, or number of rotations, from the Decimation Number found in the Bruker header. Any apodization or zero-filling must be performed before the RD. Show me how to process digitally filtered data.
Digitally filtered data have distorted baselines at the ends of the spectra, which can complicate baseline correction. See details.
The FT command was recently modified to perform the RD operation automatically, using the values of DECIM and DSPFVS found in the Bruker header. If an RD is done manually, the FT will not perform the RD again. This is not fool-proof, but works in most cases.
Data transferred from the spectrometer to PC via Bruknet
This transfer process splits the data file into 2 parts. For an original file called file.001, the 2 resulting files are file.001 and p_file.001. In this form, the data will not import correctly into NUTS. The solution is to re-combine the 2 files into a single file that NUTS will recognize and import. This is done on a PC with a simple batch file which is
copy p_%1 #%1
type %1 >> #%1
This batch file is placed into the directory containing the data files. Then execute the file with a single argument which is the name of the data file (e.g., file.001). The 2 files are combined and given the name #file.001, which can then be imported into NUTS.
2D data acquired on an Aspect-based spectrometer
The Aspect computer predates development of routine 2D spectroscopy. As a result, the Aspect data file format contains no provision for essential parameters for the indirect dimension. The missing parameters are essential to displaying a correct axis in the indirect dimension.
It is simple to enter values for the missing SF, SW and offset parameters, provided you know the correct values. It is suggested that SW and chemical shift information be placed into the Bruker TITLE when data are acquired, so they stay with the data.
Last updated 6/1/03.