1

Content
• Getting data organized
• Good data processing
• Plotting

2
 Documentation: Software
• Manuals available in TopSpin

Software manuals are accessible from within TopSpin under the icon. A good
place to start is with the Beginners Guide, which gives a brief introduction to the
NMR experiment and NMR hardware and then instructions about how to run some
frequently used experiments. Also useful are the 1D and 2D Step by Step – Basic
manual and the Processing reference manual. Also, if you type the name of a
command followed by a question mark into TopSpin (e.g. efp?), then you will be sent
to a help page explaining the command.

3
 How data is stored on disk by TopSpin
• <directory>\<dataset name>\<expno>\pdata\<procno>
• Example: c:\users\rstein21\data\RSS_01234\1\pdata\1
or
• <directory>\data\identifier\nmr\<dataset
name>\<expno>\pdata\<procno>
• Example: c:\data\AV400\nmr\RSS_5683b\3\pdata\1

• Every FID (raw data) has a different expno

• Each spectrum (processed data) has a different procno
• In IconNMR, you choose the dataset name
• You may also choose the expno, but you may also accept the default

• On the spectrometer PCs, the directory is c:\data\<your username>\nmr

Bruker is somewhat strict about how data is stored on disk. This is because its data
browser can only display data stored in certain locations.

Specifically, each piece of raw data has its own expno (experiment number), which is
a directory; each expno directory contains a directory called pdata which can holds
one or multiple pieces of processed data in different processing numbers (procno’s).

Every expno must be contained within a dataset name. Thus, for the dataset called
“RSS_01234”, for example, there could be a 1H experiment in expno 1, an HSQC in
expno 2, and an HMBC in expno 3. Within RSS_01234/3/pdata/1 would be the
processed HMBC, and RSS_01234/3/pdata/99 might contain a projection from the 2D
HMBC dataset.

Above the expno is the directory. The directory appears in the data browser.
However, if the directory has the form <directory>\data\<identifier>\nmr, then only
the part of the directory before \data appears in the data browser at the top level.

4
 View data: Data browser
• Once you have found the data in the data
browser, open it by:
• double-clicking
• clicking and dragging
• right-clicking and selecting
• display
• display in new window

• Close and open the databrowser by

typing Ctrl-D (this is especially useful
with small screens)

• The data browser can be made into a

separate window in the browser
preferences (right-click in the data
browser). After activating this
preference, quit TopSpin and restart it to
make the data browser float separately in
its own window.

After copying data to another computer with access to TopSpin, use the data browser
to open it.
The data browser lists all the datasets. A directory containing data can either be of
the form <directory>/<name>/<expno> or
<directory>/data/<user>/nmr/<name>/<expno>. In the first case, opening up a
directory (by clicking on the + sign next to the directory name) gives a list of dataset
names, beneath whch are the experiment numbers,and within those the processed
data numbers.

In the second case, opening up a directory gives a list of user names, beneath which
are the dataset names, and then the experiment numbers come below these:
 Opening Bruker data in TopSpin
• Drag and drop zip file from local Outlook email or Desktop to TopSpin:

If you use Outlook 2013 or similar on your computer (that is, Outlook not on the
Internet), you can simply drag and drop the zip file into the main TopSpin window. If
you aren’t using Outlook, you can save the zip file to your Desktop (or anywhere else
on your computer), and drag and drop it to the main TopSpin window. A window will
pop up that asks for the directory (DIR) to save the data to.

I recommend choosing c:\data\. (case A) If you prefer to sort your data by instrument
or project name, use c:\data\<project or instrument name>\nmr (case B).

Then, right-click in the data browser (to the left of the main part of the TopSpin
window), and select Add new data dir… Under DIR, if you previously chose a
directory whose path did not follow the form <directory>\data\<project or
instrument name>\nmr (case A), enter the full directory name here (for example,
c:\data)). However, if you previously chose a directory whose path did have the form
c:\data\<instrument or project name>\nmr, (case B), then enter only the part of the
directory before “data” (for this example, enter c:\). Then your data will be displayed
along the left side of the window. Left-click on the plus icons (+) to open the data
browser tree and show the data within it. If you right-click in the data browser and
choose Scroll to Active Dataset, the currently active dataset will be highlighted in the

6
tree.

6
 Getting Varian data onto your own computer
• Getting data from computer onto disk
• You will receive a .zip file in your email. Save this file to somewhere on your disk
where you will remember the location
• Browse to that location, right-click on it, and unzip it (note: Macs often unzip it
for you)
• You will end up with a directory whose name ends in .fid with four files in it
• Getting data into TopSpin
• In TopSpin, click on the icon or type vconv
• If needed, choose Open NMR data stored in special formats and file type = VNMR
• Browse to the directory ending in .fid, then choose an EXPNO (1) and a directory
(for example, c:\data\VNMRS500\nmr, or simply c:\data)

After you open the data, right-click in the data browser (to the left of the main part of
the TopSpin window), and select Add new data dir… Under DIR, enter the directory
you just chose (for example, c:\data\unknown). Then your data will be displayed
along the left side of the window. Left-click on the plus icons (+) to open the data
browser tree and show the data within it. If you right-click in the data browser and
choose Scroll to Active Dataset, the currently active dataset will be highlighted in the
tree.

7
 Adding new directories to data browser
• Right-click in data browser to add new directories

As shown previously, the directory being added can be part of the path with form
<directory>/<name>/<expno> or <directory>/data/<user>/nmr/<expno>. The alias
chosen for the directory shows up in the data browser. (By the way, TopSpin 1 and 2
only allowed this second style of data storage.)

Directories can be removed from the list by right-clicking in the data browser and
selecting “Remove Selected Data Dirs.” This does not remove the data; it only
shortens the list displayed.

To delete data, one method is to choose the data (name, expno, or procno), right-
click, and choose “Delete…”

8
 Recently viewed datasets/groups

The “Last50” tab of the data browser shows the 50 most recent datasets used;
thumbnails showing the spectra can be displayed, if desired. To display a dataset
from the last 50, use the same techniques as for displaying data from the main data
browser (double-clicking, right-clicking, or clicking and dragging). 2D datasets are
indicated in blue, 3D in pink.

Groups of datasets can be defined under the “Groups” tab. Open a series of datasets
that you would like to have access to easily (for example, a 1H, a 13C, an HSQC, and
other spectra of a molecule that you study frequently). Then right click and select
“Add new dataset group…”. This just creates a group, but it has no datasets
associated with it. To add datasets to it, right-click and select “All open data
windows.” Now all those datasets can be recalled more easily.

9
 User-defined buttons
• To create a shortcut to a command that you use frequently, right-click
next to the buttons or between the buttons to add a user-defined button

The three fields in the User-defined Tool Buttons window are the text or icon that
appear on the button, the command that pushing the button executes (in this case,
icona starts IconNMR in automation), and the description (the tool tip) that appears
when the mouse cursor hovers over the button.

10
 Working with spectra
Axis units
Last 2D/3D
Axis scaling gridlines

1D

Intensity scaling Vert/horiz Distance

shift Overview
Multi display

2D

contours maintain scale 2D viewing options

when changing
spectra

A tool tip giving information about the function of a button appears when you hover
over it.

To change the look of icons to those of previous versions of TopSpin, right-click to the
right of the icons and choose “Toggle Icon Style.”

If you have a small screen, you can remove the space around the border using
“Toggle Icon Border.” After changing the icon style and removing the border, the
buttons look like:

11
 Setting preferences
• Type setres or go to Manage / Preferences

Auto-open last used dataset when restarting TopSpin: this makes it easier to pick up
your work where you left off.

12
 How TopSpin works: Parameters and
Commands
• Commands tell TopSpin to do something (for example, load data, process
data, or acquire data)
• Parameters describe how actions will be performed
• Status parameters show the parameters used to acquire or process the
spectrum as it currently exists

TopSpin commands can be input either by typing them on the command line or by
using the Flowbars to select the command. For example, use File / Open to open a
dataset, or type efp on the command line to process a 1D dataset.

Most commands that are executed depend on the values of certain parameters. For
example, the command efp applies exponential line-broadening according tot the
parameter lb in Hz. If lb is small, little line-broadening will be applied; if it is large, a
lot of line-broadening will be applied. To set a command, either type its name on the
command line or go to the ProcPars tab for processing parameters (such as lb for
line-broadening) or to the AcquPars tab for acquisition parameters (such as ns for
the number of scans to acquire).

TopSpin also keeps track of the parameters that were actually used by a command.
The parameters that were actually applied are called status parameters. After a
command is executed, the values of the parameters it used are updated as status
parameters. So, to see what line broadening you have already applied, type s lb
(the s refers to status ). Or, to see how many scans were actually acquired for an
experiment, type s ns.

13
The plot layout is a parameter – one way to set it is to type layout.

13
 Process flowbar / Basic automated processing
• The process flowbar contains the basic processing commands

• Clicking on Proc. Spectrum starts a basic automated processing

sequence, which can be customised by clicking on the arrow on the
right of the button

Basic 1D processing includes several steps, which can be performed by clicking on

Proc. Spectrum or by typing commands:
• applying line broadening to the raw data (FID) to smooth away noise at the end of
the decayed data. The larger the amount of line broadening, the less noise there
is, but also the less resolution. Splitting from J coupling is harder to see when there
is more line broadening. Usually, between 0.1 and 1.0 Hz of linebroadening is
applied; 0.3 Hz is normally the default setting.
• performing a Fourier transform, to convert the raw data (FID) to processed data (a
spectrum).
• phasing the spectrum, so that peaks appear positive.
• referencing the spectrum, to put TMS at exactly 0 ppm (if there is no TMS in the
sample, an error message is generated, which can be ignored).
• correcting the baseline to make it perfectly flat.
• integrating the spectrum.
• plotting the spectrum.
 efp: Basic 1D Fourier transform command
• Manually performing the steps of the Proc. Spectrum tab show what
data processing actually does

efp

The command efp performs three processing commands sequentially: multiplication

by an exponential linebroadening function followed by a Fourier transform, followed
by an application of the phase parameters present in the dataset.
Multiplication by an exponential linebroadening function improves the
signal-to-noise ratio of the spectrum by multiplying the FID (the raw data) by a
decaying function. This means that the noise at the end of the FID is multiplied by a
very small number so as to minimize the noise in the spectrum. To change the
amount of linebroadening, type lb and enter a number (usually between 0.1 and 1.0
Hz), before typing efp.
𝑖−1 ∙𝐿𝐵∙𝜋
𝑦 = 𝑒 − 2∙𝑆𝑊𝐻
The Fourier transform is the mathematical transformation that
changes the FID into the spectrum.
Phasing is an adjustment of the appearance of the spectrum to give
positive lines with a good lineshape. This comes from adjusting the real and
imaginary parts of the spectrum so that the absorptive part of the spectrum (the part
with a good lineshape) is displayed as the real part.

15
 Other kinds of linebroadening
no linebroadening gm, -1 Hz; gb=0.3

em, 0.3 Hz

The basic linebroadening method is exponential multiplication, which tends to

improve the S:N ratio in spectra. On the other hand, to increase resolution, the
simplest thing to do is just omit the exponential multiplication by processing with the
command fp. For further resolution enhancement, the Lorentz-Gauss transform is
used to change the shape of the lines and make them sharper. This involves
multiplying the data by an increasing exponential, to avoid smoothing away the
resolution, while at the same time multiplying by a Gaussian function to make the FID
go to zero at the end. This can help to resolve small couplings.
𝑖−1 𝐿𝐵∙𝜋 𝑖−1 2
−𝜋∙𝐿𝐵∙ − ( )
𝑦=𝑒 𝑆𝑊𝐻 2∙𝐺𝐵∙𝐴𝑄 𝑆𝑊𝐻
The parameters that control the line broadening commands are lb,
which must be set to a negative value for the Lorentz-Gauss transform (to
approximately the width of the uncoupled lines), and gb, which controls the position
of the maximum of the Gaussian function and must be between 0 (meaning
maximum at the start of FID) and 1 (maximum at the end of the FID).
To do this, set the parameter lb to the negative of the peak width
measured following the command fp, and set gb to ~0.3. Then process with the
command gfp, instead of efp. In the above example, the peak width is about 1 Hz, so
lb has been set to ‒1 Hz.

16
 The cost of this resolution enhancement is sensitivity: the middle of
the FID is enhanced at the expense of the beginning. The spectrum with no LB has an
S:N ratio of 3900:1, that with 0.3 Hz LB has 11600:1, and that processed with gfp has
1300:1.

16
 Examining effects of linebroadening:
.winf

17

The .winf command allows you to interactively examine the effect of window
functions on a spectrum. Note that the default settings will be not always be those
expected, so if you have been applying Gaussian multiplication with gfp you may find
that the spectrum looks odd, since by default it is processed in the interactive
window multiplication window with em, despite the negative lb. It may be necessary
to set the phase mode, PH_mod, to pk so that the correct phasing appears.

17
 apk: Automatic phase correction

apk

Phasing creates an entirely absorptive lineshape in the spectrum. Two parameters

are adjusted to create this: phc0, which adjusts the phase of all parts of the spectrum
evenly, and phc1, which affects the spectrum linearly across it. These two
parameters are adjusted to maximize the positive signal in the spectrum
automatically by typing apk.
For many spectra, the command apk0 is more appropriate than apk, because only
zero-order phasing (the parameter phc0) should need to be adjusted to give a flat
baseline and absorptive lineshape with a modern spectrometer.
The various automatic phasing options can also be accessed from the Process
flowbar:

18
 Manual phasing

Manual phasing is used to correct the phasing when automatic phasing does not do a
satisfactory job. Click on the Adjust Phase button to enter manual phasing.
To phase manually, start by using 0th order phasing to make the peak with the pivot
point (red line) perfectly phased. Do this by clicking and dragging on the button
marked . To change which peak is the pivot point, right-click on the peak and
choose Set Pivot Point. Sometimes it helps to use a peak at the edge of the spectrum
rather than the largest peak, which TopSpin chooses by default. The sensitivity of the
mouse while clicking and dragging can be increased or decreased by clicking the
triangle buttons. After 0th order phasing, use 1st order phasing to bring the other
peaks into phase: do this by clicking and dragging on the button marked .

For very crowded spectra where all peaks overlap, remember that the baseline
should be one continuous flat (or smoothly curved) line. Looking at the baseline
regions outside the spectral peaks (say less than 0 and greater than 10 ppm) can
therefore also be a guide when phasing.

19
 abs: Automatic baseline correction and
integration

abs

The command abs does two things to the spectrum: it corrects the baseline, and it
does its best to integrate the spectrum. Use abs n to correct the baseline without
integrating the spectrum.
The baseline correction is done by fitting a polynomial function y = a + bx + cx2 + dx3 +
… to the parts of the spectrum without peaks in them.

20
 Baseline correction: abs command

If the baseline is not at 0, use the command abs to adjust it automatically. This helps
make integrals more correct. The command abs also integrates the spectrum; the
command abs n adjusts the baseline without also integrating the spectrum.

21
 Manual baseline correction

In manual baseline correction mode, you can adjust the parameters of polynomial
subtraction manually, and you can choose points for spline baseline correction.
The most useful manual baseline correction option is probably the spline baseline
correction. This allows you to define a set of points that lie on the baseline which are
then fit to a cubic spline curve through these points. This can be used to subtract
background signals. To do this, select the “setup spline file baslpnts” option, left-click
on the points on the baseline, and save. If this produces an odd result, it may be that
points you have selected fall at the top and bottom of the noise. To get around this,
average over regions of points centered on the selected points. Do this by setting the
parameter nzp to 20 or more, and reapply the correction by redoing efp and then
sab.

22
 Manual integration

Divide
Draw region integral

To get into manual integration mode, just click on the Integrate button in the Process
flowbar. You can use the manual integration mode to define integrals from scratch,
or to modify those generated by abs. In this mode, right-clicking on an integral gives
options for calibrating the integral, normalising the sum of all integrals, and
deconvoluting integrals.

23
 sref: Automatic referencing of spectrum
to TMS/TSP (if present)

sref

Automatic referencing of TMS is performed using the sref command. If TMS is not
present, the command sref sets the calibration parameter sr to 0. This puts the
solvent peak in the correct location.

Automatic referencing can also be obtained by clicking on the arrow next to Calib.
Axis in the Process toolbar:
 Referencing

When spectra are acquired with locking (which is the standard when a deuterated
solvent is used), then the chemical shifts should be almost exactly right, even without
further referencing. However, if there is a compound in the sample whose chemical
shift is known, its peak can be used for more precise referencing. TMS (or TSP/DSS in
water samples) will be automatically found by sref, but if there is no TMS we could
use the solvent peak. Calibrate it by left-clicking on the peak to reference to, then
enter its exact chemical shift.

25
 ppf: Peak picking

ppf

The command ppf automatically picks peaks on the entire spectrum. The command
pps may also be used; it picks peaks only the displayed spectrum region.
Peaks are found according to the parameters pc, mi, and maxi. Pc is the peak picking
constant: a point is only considered a peak if it is a local maximum and bigger than pc
* a noise factor – i.e. if pc is small, all peaks will be found, but so might spikes. Mi is
the minimum intensity or threshhold, and maxi is the maximum intensity of peaks.
Usually, typing mi and increasing it to 0.1 or greater before running ppf or pps helps
with automatic peak picking.
Peak picking can be done through the Process toolbar:

26
 Manual peak picking

Define peaks manually

Drag to define region

In manual peak picking mode, peaks in a region can be picked by clicking and
dragging a box; any peak whose top is inside that box will be picked. Peaks can
also be defined manually, either by picking the peak top ( ) or by selecting
the peak to the right of the cursor ( ). Manual mode can also be used to
modify the peaks picked automatically by pps/ppf. For example, a peak can be
deleted by right- clicking on a peak and choosing “Delete peak under cursor.”

27
 Multiple display mode
• Use the multiple display icon to display multiple spectra simultaneously

Offset spectra Sum spectra

Scale individual
and save sum
spectra

Select
individual
spectra

Multiple display mode offers a way to view multiple spectra simultaneously. Once in
the mode, any command used to display a new spectrum (eg using the re command
to read another spectrum or dragging and dropping a spectrum into the main part of
the window) will display it on top of the spectrum already present.
Individual spectra can be selected in the lower left or on the right and scaled and
shifted individually. Sums and differences of spectra can be taken. These can be
saved with the floppy disk icon, which then prompts for a procno to save the new
sum or difference spectrum in.
The spectra can be displayed on top of each other (the default, shown in the picture
here) or offset from each other.
When a 2D spectrum is present, multiple display mode gives the option of viewing
and extracting individual rows and columns or of viewing 1D spectrum superimposed
on the 2D spectrum.

28
 Processing 2D spectra
• 2D: automated processing with proc2d

Like 1D spectra, 2D spectra can be processed in a semi-automatic fashion in TopSpin

3. The processing can be customised using the proc2d window, accessible from the
menu next to the Proc. Spectrum tab.
Often, errors will appear regarding phasing and/or phase-sensitive data: these occur
when the option of “auto-phasing” is selected, but the data is not phase-sensitive. In
this case, it doesn’t make sense to phase the data. The error can be ignored.
HMBC 13C spectra must be processed with the command xf2m after the automatic
processing.

29
 Processing 2D spectra
• xfb: Basic 2D processing command

xfb

The basic 2D processing command is xfb. This command always performs a Fourier
transform of both dimensions. It applies window multiplication (as long as the
parameter wdw is not set to “no”) and it applies phasing according to the parameter
ph_mod. (A value of ph_mod of “no” means that no phasing will be applied, “pk”
means that phases given by phc0 and phc1 will be applied, “mc” means that the data
will be processed in magnitude (non-phase sensitive) mode, and “ps” means that it
will be processed in power mode. In magnitude mode, the intensity plotted at any
point (where a point is defined by a frequency in the F1 and a frequency in F2
dimensions) is the square root of the sum of the squares of the measured real and
imaginary components at that point. In power mode, the intensity plotted is the sum
of the squares of the measured real and imaginary components at that point.)
Usually, either pk or mc are used.
Note that the standard 13C HMBC implemented on the Bruker spectrometers is
phase-sensitive in F1 only. So, you must use two commands to process these spectra:
xfb, then xf2m.

30
 Linear prediction
• To keep experiment time down, 2D data is often restricted (truncated) in
f1. To improve the quality of the spectrum, linear prediction can be used
to extrapolate the data:

Zero filling

TD
LPBIN
SI

• In F1 column set: ME_mod = LPfc

NCOEF = about 40
LPBIN = size (0 means double TD)

The parameter ME_mod stands for “maximum entropy” mode, a different name for
linear prediction. The parameters to define linear prediction (the options for
ME_mod) are:
LPfc = Linear prediction, forward, complex (the “complex” does not matter as the
software automatically determines the type of data: LPfb will work just the same)
NCOEF = number of coefficients. Broadly speaking, this is the number of
“frequencies” to try and find in the signal for extrapolation; it should be greater than
the number of points expected. Increasing this number excessively will increase
computation time, and usually will not make the spectrum any better.
LPBIN = the total number of points in the extrapolated fid, so if TD > LPBIN no
extrapolation is done. The more data is extrapolated the more unreliable the answer
will become, and in general a doubling of TD is reliable. This is indicated by the
special case of LPBIN=0. However, if signal to noise is good, you may be able to get
good results using a bit more linear prediction (e.g. LPBIN = 4 * TD).
After any linear-prediction, the FID is then zero-filled to a total of SI points and is
ready for Fourier transformation.

31
 Linear prediction example
• Linear prediction – example

TD(1) = 128 TD(1) = 128

No linear prediction Linear prediction to 256

After linear prediction the peaks are better resolved in the vertical (f1) dimension.
As with zero filling in 1D, there is nothing to be gained by linear predicting too many
points – the errors in the calculations become more and more significant until the
peak shape deteriorates and no information is gained.

32
 Processing 2D spectra

2D truncation

• Truncation artifacts can be reduced by using window functions that force

end of fid to zero:

no adjustment
 FT

FT fid multiplied by
cosine function

Linear prediction is usually not sufficient to keep 2D data from being truncated or
adding enough points in the FID to define the points in the spectrum properly. So,
zero-filling is usually necessary. However the “step” that occurs between the end of
the fid and the added zeros give rise to so-called “sinc wiggles” in the spectrum. If
the FID is smoothly taken down to zero at the end, then this artifact is removed,
although this broadens the peaks slightly.

33
 2D window functions
• Use WDW = SINE (sine-type) or QSIN (sine squared)
• Parameter SSB determines shift of function
• SSB = 2 for maximum sensitivity (cosine function)
• SSB = 3, 4, 5, 6 … shifts function to increase sharpness
• SSB = 1 for maximum sharpness (sine function, emphasises centre of fid)

2 3 5 1

increasing sensitivity increasing resolution

Usually, this parameter (WDW) is set correctly. But, if you read Varian data in with
TopSpin, it may not be set correctly.
All the “sine” based functions explicitly end at zero, avoiding truncation.
By shifting the function from a cosine function (SSB = 2 and first point = 1) towards a
sine function (SSB = 0 or 1 and first point = 0), the function additionally provides
resolution enhancement (emphasising the middle of the fid) at the cost of worse
sensitivity (the beginning of the fid is suppressed). Note that setting SSB = 0 is the
same as 1 (i.e. maximum sharpness).
Other window functions, such as the traditional Gaussian or exponential
multiplication, can still be used with 2D datasets, but they are not as generally
applicable and usually require more fine tuning to get good results.
As a rule, use SSB = 2 with phase-sensitive experiments whose first points have non-
zero intensity and that give intrinsically sharp (Lorentzian) peaks. Use SSB = 0 with
non-phase-sensitive experiments whose first points are 0 and which give broader
peaks.
 Processing 2D spectra

SSB examples

ssb 2 5 1

pk

2 5 1

mc

The difference between the two examples above is the result of the intrinsic
lineshape difference between a phased spectrum (normal absorption Lorentzian
lineshape) and the much broader magnitude-mode lineshape (a mixture of the
absorption and dispersion lineshapes). If you want to illustrate this just type “mc” in
a normal 1D spectrum (normal 1D spectra are phase-sensitive; they can be phased).
Hence, for magnitude mode spectra, resolution enhancement is an important
component of processing.

Typically, for phase-sensitive experiments (HSQC, DQF-COSY, NOESY, ..), start with
SSB=2, and then increase it you need extra resolution and have enough signal.
For magnitude mode experiments (HMQC, COSY, …), start with SSB=1. If more
sensitivity is needed, try a very large number, and then decrease it.

35
 Manual 2D phasing
• Start phase mode:
• Then right click on a few peaks (choose peaks that span the width and
height of the spectrum to phase) and select “Add” to choose them for
phasing

• Click rows or columns to show traces

• Phase all rows first, using 0 and 1, then save
• Then phase all columns and save

Phasing a 2D experiment is very similar to a 1D, except that there are 2 dimensions:
both f1 (vertical) and f2 (horizontal) can be phase corrected.
Keep in mind how the spectrum should look after phasing: NOESY spectra should
have the diagonal negative and crosspeaks positive. DEPT-edited HSQC spectra
should have peaks either positive (CH3 and CH) or negative.
Rows and columns are extracted from the 2D for phasing, because it would be too
slow to interactively phase the whole 2D matrix. The aim is to extract 1D traces that,
taken together, contain peaks across the width of the spectrum (if you are phasing a
NOESY with many cross peaks, this may only require one trace, but for an HSQC with
single discrete peaks it may be necessary to use three or more traces to cover the
left, middle, and right of the spectrum).
It may occasionally be necessary to repeat the phasing after the first attempt for best
results. Remember that, as for 1D spectra, it can be helpful to try to get a flat
baseline as well as look at peak shapes.
If, after phasing, the data are reprocessed using the command xfb, the stored phase
parameter will be re-applied if the PH_mod parameters are set to “pk”.

36
 Exact 2D phasing
• In some cases phasing requires more attention!
• For experiments such as NOESY/ROESY, the large diagonal peaks and
small cross peaks make exact phasing important, but negative peaks can
make the lineshape less obvious
• In general, phase for a smooth baseline:

When phased correctly, the bulk of the baseline should usually be flat (even if not at
zero) or at least a smooth continuous curve. You might see deviations at the extreme
edges of the spectrum due to digital filters.

37
Referencing 2D spectra
• either sref (which puts TMS at 0 ppm, if found)
• or manually:
• then click on a peak and enter the correct chemical shift when prompted

38
 pp2d: peak picking in 2D spectra
• After setting the contour levels, automatic peak picking can give the
chemical shifts of all peaks:

pp2d

The pp2d dialogue box lets you set the threshold based on the contour levels. Set the
contours so you can see (only) the peaks of interest and save them before starting
pp2d. Then, under Minimum intensity / Set to, choose “Lowest contour level.”

39
 int2d: automatic integration of 2D spectra
• After peak-picking, peak integration is simple:

int2d

The command int2d allows for integration of the peaks chosen with the pp2d
command.

40
 Baseline correction of 2D spectra
• 2D spectra may have baseline offsets in one or other dimension
• abs1 corrects F1 direction, abs2 corrects F2

NOESY/ROESY are usually the worst spectra for needed baseline corrections,
especially in t1, because of t1 noise.

41
 Setting projections
• Right-click above or to the left of spectrum to set projection

Right-click on the projection above or to the left of the 2D dataset to set an actual
spectrum to be displayed in place of the projection. If you want to calculate a
projection (or row/column sum) and store it in another procno, the command proj
can be used. The projections should be defined in the standard spectrum viewing
window before plotting is done.

42
Another option is to right-click on a spectrum listed in the data browser and choose
Display as 2D Projection

42
 Using multiple display mode to view and
extract rows or columns

To view individual rows and/or columns, go into the multiple display mode by clicking
on the icon The scan rows/columns buttons within multiple display mode allow
you to look at individual rows/columns of a 2D spectrum. If you want to do more with
a row/column, like compare it in multiple display mode with another spectrum, you
can save it to its own procno by right-clicking while the desired row and/or column is
displayed.

43
Contour levels
• Right-click on the spectrum and choose Edit contour levels…

• Or try the command clev

44
 Tidying up the spectrum
• To further improve spectra:
• Baseline correction for samples in water: abs2.water. Also consider using
BC_mod=qfil in f2 to suppress the central noise stripe.
• Fine-tune contour levels (manually, with the mouse button and/or by setting
nlev to the number of levels, then running levcalc or clev, or with edlev)
• Symmetrization (for experiments such as COSY): sym or syma (for phase-
sensitive experiments like NOESY), but use this with caution
• Reducing t1-noise: t1away (an AU program available upon request)

When symmetrizing, watch out for generating false peaks where noise bands
correlate. Only symmetrize after applying noise-reducing techniques such as abs1/2
and t1-noise subtraction first, and after adjusting levels so that only real peaks are
displayed. Then perform sym (or syma for phase-sensitive data). Remember that this
loses or creates information, so it must only be applied with caution. An automatic t 1
noise subtraction is done by the AU program t1away (ask me for a coyp) which sets
the lowest intensity peaks in each column of the 2D spectrum to 0, allowing peaks to
be displayed.

45
 Pseudo-2D data
• 1D data: Processed data has two frequency axes (raw data has two time
axes)
• 2D data: Processed data has two frequency axes (raw data has two time
axes)
• Pseudo-2D data: Processed data has one frequency axis and one axis
that is not frequency
• The non-frequency axis may be time between experiments, relaxation delay,
diffusion coefficient, or some other variable
• Raw data has at least one time axis, in the direct dimension
• To process the time axis in the direct dimension only, use the command
xf2

Most NMR data is a function of frequency (chemical shift). However, some data, such
as relaxation measurements, diffusion spectroscopy measurements, or kinetics
measurements, have a second axis which is not a chemical shift axis. In this case, the
main or direct dimension has to be processed using the command xf2. Then the
other dimension can be processed afterwards.

The direct dimension can be phased using standard 2D phasing methods, using the
most intense row of the pseudo-2D dataset. Baseline correction is done using the
abs2 command to baseline correct the direct dimension. Traces can be extracted for
further analysis, if desired, using the multiple display mode.

46
 Dynamics Center

The Dynamics Center is Bruker’s approach to processing the non-frequency

dimension of non-frequency 2D datasets. After doing the xf2 transform to process
the frequency dimension of relaxation, diffusion, or similar datasets, the Dynamics
Center is used to process the second dimension.

After doing xf2, type expl to open a Windows Explorer browser. If this does not work,
the workaround is:
Use a text editor like Notepad to open the text file called
c:/users/<username>/.topspin-<pc>/prop/globals.prop
Add the line FILE_EXPLORER_PATH=c:/Windows/explorer.exe at the bottom.

Drag the file 2rr over the appropriate box in the Dynamics Center window and follow
through the prompts to finish processing the data and output a report.

47
 Varian 2D and pseudo-2D data in TopSpin
• Several processing parameters tend to be set wrong
• They are (generally) set correctly for non-phase sensitive spectra (eg COSY),
but are wrong for phase-sensitive spectra

• For 2D, phase sensitive datasets

• MC2 in F1: try echo-antiecho
• ph_mod in both dimensions: PK
• SSB in both dimensions: 2 (also try setting WSW to QSIN)
• Set ME_mod in F1 to Lpfc
• Then redo xfb and phase with apk2d and/or manual phasing
• For pseudo-2d datasets
• SI in F2: a large number (32k or 64k)
• WDW: EM
• LB: 0.3 Hz or so

These changes are needed to get the right lineshape for Varian 2D datasets. If you do
this frequently, the changes can be automated.

48
 Generating a printout quickly: prnt or the Plot
tab

There are several ways to create a printed or PDF output of a spectrum.

Typing prnt generates a printout of the active window, exactly as it appears, with text,
gridlines and all. A dialog box gives a choice of printers and all the usual printing
options.

Clicking on the Plot tab (or going to the Publish flowbar and choosing Plot Layout)
opens a plot editor which displays the portion of the spectrum shown in the
spectrum tab, according to currently chosen plot layout. The command prnt will print
the displayed layout, but much more customization is also possible. A PDF can be
generated by choosing Publish / PDF; other file formats (JPG, PNG, TIF) can be
generated under Publish / PDF / Other formats…

49
 Customizing the appearance of the plot

The plot tab can be used in two ways: to customize a view of the current spectrum, or
to create a more generally useful plot arrangement, known as a layout.

A layout is simply an arrangement of items on the page; the items include the
obvious things like spectra, text about the experiment, and titles, and they can even
be extended to things like shapes, pictures, or structures.

The floppy disk icon can be used to save a modified layout with a new filename for
general use, or to save a layout for single use in a particular dataset.

With a small monitor, it can be difficult to see all parts of the left panel. To help, click
on the icon that looks like a monitor ( icon). Alternatively, turn off the Acquisition
Status Bar (right-click on the Acquisition Status Bar and choose Acquisition Bar
On/Off), reduce the TopSpin icons to one row ( icon), and/or hide the flowbars
(right-click on the flowbars and choose Set All Tabs Invisible). To bring things back,
click on the Spectrum tab.

50
 Modifying objects already on the plot
• Clicking on an item brings up options for modifying it (in the left panel)

In this window, the spectrum is selected, so the options in the left panel are those
appropriate for modifying the appearance of the integrals, the peaks, the axis, etc.
The meaning of the buttons is:
Undo Return to main panel (to add
Save items and adjust printer
options, etc.)

Do / do not display peak Format text and line style

marks or labels of peak marks and labels
Format peak labels
(decimal places, Hz / ppm,
Do / do not display integral position of label)
curve or labels Format text and line style
of integral curve and labels

Change units of chemical shift Format text and line style

axis of chemical shift axis

Move spectrum on the page

Change other parameters

51
 Plot range
• Usually, the spectrum range shown is whatever was in the spectrum tab
before entering the plot tab
• There are several ways to modify the range:

Reset to Automation Zoom in on a particular

limits area of the spectrum

Return to spectrum limits

on Spectrum tab

• Automation limits: set in the Automation Actions panel and allow x and y
ranges to be set differently depending on the spectrum

The Automation limits allow various options to be selected if the Reset icon is
chosen. This is particularly important for high-throughput or repeated work. For
example, if you always want to view the spectrum between 8 and 5 ppm, then set
parameter f1p to 8 and f2p to 5, then set the Automation Action for X to be “Values
of F1P/F2P”.

52
 Getting access to multiple spectra for plotting
• Use multiple display (the icon) to view all spectra for plotting

• Then, in the plot editor, load these spectra:

Use the multiple display mode in the Spectrum tab (the main view of the spectrum),
not the plot editor tab. Use the standard techniques for viewing spectra (dragging
and dropping from data browser or typing re <#>, for example) to load up many
spectra.

Then, in the plot tab, in the plot portfolio portion of the window (bottom left), use
the down arrow to get access to the option “Load collection from multiple display
mode.”

In future version of TopSpin, it will be possible to drag spectra from the data browser
directly in to the Plot Portfolio window.

With the stand-alone plot editor, the Data button allows you to add spectra to the
portfolio and then to choose which one is displayed where.

53
 Plotting multiple spectra
• Use “Click here to insert new elements” / “NMR” to add an element to
the plot window

• After selecting an element, click and drag

on the plot layout to add it
• Then drag and drop a spectrum from the
plot portfolio list onto the new element
to display it there (here, spectrum #2 is displayed)

The number displayed over an element when it is selected shows which number
spectrum in the plot portfolio list is being plotted. It can be difficult to see, so here a
grey box was plotted behind the spectrum so that the highlighted number appears
clearly.

54
 Stacked spectra
• Use “Click here to insert new elements” / “NMR” to insert a stacked plot
• Draw a box for the first spectrum
• Click on the spectrum, then use “Stacked spectra” to customize stacked
plot (including adding more spectra)

The green box only defines the size of the initial spectrum. The spectrum which
appears here (#1 in this example) is taken from the position in the Plot Portfolio; if
you wanted to plot spectra 2-5, for example, you would drag spectrum 2 into the
stacked spectra region and then plot 4 stacked spectra.

55
 Adding other objects to the plot
• Use “Click here to insert new elements” / “Standard” to add an element
to the plot window

• For shapes, click and drag to place them

In earlier versions of the plot editor, objects were added by selecting them, clicking
on the page to place them, and then editing them.

56
 Plotting 2D spectra
• Plotting

list of spectra (2D and projections), as set as projections, in

multiple display mode

Plotting is much the same as for 1D. By default, the spectra set as projections in the
spectrum view are used as projections in the plot tab. If the projections are not
correct, go back to the Spectrum tab and use Multiple Display mode to bring them in,
then in the plot editor, use the arrow icon in the plot portfolio to load the multiple
display mode datasets.

57
Saving layouts
• Save your layouts so can reuse them
• Remember: a layout is a set of objects on a page
• Reusing layouts saves you time and effort
• To reuse a layout:
• Set the parameter layout to your layout name
• Or, automate the process
• edau setlayout
STOREPAR(“layout”, “+/my1D_H.xwp”)
VIEWDATA_SAMEWIN
QUIT
• Then File/Save
• Then type setlayout from your dataset and the layout will change
• You can even make a button

58