Assignment 1
NMR Guided Reading
January 26, 2000

The first thing necessary to do is learn to control the spectrometer very well.  Having a few basic skills for optimally locking the instrument and shimming the field to an extremely good homogeneity will help us immensely later on.  For this first exercise, we are going to hone your present skills in these areas considerably.

Locking:  Please read in Derome the section on "The deuterium lock" on  pages 44-46 and pay particular attention to the following parameters/concepts:  lock power, lock gain, lock phase, lock field, and saturation of the lock signal.  These are also discussed in Braun, Kalinowski, Berger  in Section 1.3 pp. 6-7.

For an optimum lock signal:

Lock power:  We want this as high as possible without saturating our signal.
Lock gain:  We want this as low as possible without risking losing our lock signal.
Lock phase:  We want this as close to directly on phase as possible to maximize the response from the lock circuitry (this provides a steadier lock signal)
Lock field:  We need to adjust this correctly prior to trying to obtain the lock (otherwise the automatic lock routine will be fouled up)
Saturation is bad, bad, bad.  This occurs when we are pumping to much energy (via the lock power parameter) into the sample.  This results in the equalization of the spin states for the deuterium solvent and the lock signal disappears.  If the lock signal disappears, the instrument cannot lock.

After you understand those concepts in general, turn to page 120 of Derome and read section 5.3.4.  This gives a better picture of what the lock is trying to do.

Now we are ready to look at the effects of varying these parameters.  Place in the spectrometer our standard H2O/D2O sample with the white cap.  Spin the sample and lock the sample up as you would normally do.

Note the important parameters:

  Beginning Saturated Not Saturated Optimized
Lock power:        
Lock   gain:         
Lock phase:         
Lock   field:         

Let's start with lock power. 

  1. Increase the lock power until the signal starts to saturate.  This is indicated by instability in the lock signal (it starts waving up and down).  When this occurs, note the value of the lock power again. 
  2. Play with the lock gain to see if this makes any difference in the behavior, note your observations.  Reset the gain to its original value.
  3. Reduce the level of  the lock power until it no longer is saturating.  Note the value. 
  4. Now, watching very carefully drop the level of the lock power by 3 or 4 units and watch the lock signal.  It should appear to drop, and then slowly rise back up.  Repeat this until dropping the lock signal by 2 or so units gives a steady signal (no upward climb).  The reason that we see this unsteady behavior is that we are still near the saturation limit.  When you can change the signal and get a completely steady signal you are out of the saturation danger zone.  Note the value.
  5. Our lock power is now optimized and we should now play with the lock phase.  The lock phase is dependent upon the lock power (i.e. changing the lock power also changes the lock phase).  This is because on old spectrometers like ours, the amplifiers for power output are not linear in their phase outputs.  What that means for us, is that anytime we change the lock power, we need to readjust the lock phase.  Do that now by simply adjusting the lock phase to achieve a maximum signal.  Note the new value.
  6. It is possible that if the lock phase changes enough that we may start saturating our signal again.  So, if there is a big change in lock phase you need to check the lock power again for saturation.  If it changes then you have to adjust the lock phase again and so on (usually one iteration is enough).
  7. After you have adjusted the lock phase, turn off the lock for a little bit and look at the pseudo-fid of the lock.  Notice how both waves are more or less symmetrical and have the same phase.  Relock the instrument.
  8. After these two parameters are set, we need to adjust the lock gain.  The lock gain amplifies the deuterium lock signal once it leaves the probehead.  The value of the lock gain does not affect signal saturation or lock phase.  The only thing that matters here is decreasing the amount of noise in the signal while still maintaining a strong enough lock signal to maintain the lock.  Turn down the lock gain until the lock is lost (light starts blinking) and then turn it up 3 or 4.  Note the value.  You should notice that the lock signal is now a lot less noisy.  This is better for the lock circuitry to "see" small changes in field and correct for them.  If there is a lot of noise, bits of noise get counted as field changes resulting in small constant shifts in the lock frequency and hence the spectrometer frequency.  This leads to broad ugly lines.

Shimming:

All right, the next part of the exercise is shimming.  Shimming is the art of homogenizing the magnetic field such that all nuclei in the NMR tube experience exactly the same magnetic field (excepting of course, differences in chemical shift, coupling, and gyromagnetic ratio).  Read the sections in Derome on shimming, Sec. 3.3.4 on pages 42-49.  Don't worry about the bit on shimming on the fid yet. 

We will work on the spinning shims z, z2, and z3 first.  Our z4 shim doesn't seem to act quite right so don't touch it for now.

The shims and a reasonable shimming procedure are discussed in Braun, Kalinowski, Berger (BKB) in Section 1.4, pp. 7-12. 

  1. A good first step to the shimming process is to save what you've got right now so that you can always get back to it if you really mess things up.  You should do that now.  In the MacNMR software, pick out the little NMR console button and save the shims under some new filename.  If things get fouled up during the shimming exercise below, just read in the old values.
  2. You can now start shimming.  This can be done either from the MacNMR console or from the old Bruker console.  I find the latter easier, but maybe it is just cause I know it better.  It really doesn't matter, they are doing the same thing.  If you are going to use the Bruker then you need to turn off the console page and get back to the main program of MacNMR (otherwise the Mac doesn't release control of the Bruker to you).  Either way you do it, you need probably to set the lock gain to a higher value so that small changes in homogeneity result in large changes in the signal (you can set it back down to our optimized level later).
  3. We are going to assume that the shimming is not so bad and therefore skip the First Round the BKB describe.  We will go straight on to the second round steps 1 and 2.  Follow the procedure to maximize the lock signal as described.  When the signal is maximized, note the values of the z, z2, and z3 shims.
  4. Take a one-scan proton spectrum and expand around the signal.  Does it look pretty good?  If not shim some more, otherwise print out the expanded spectrum.
  5. When your shimming is good and you have a printed spectrum, deliberately change z pretty drastically.  Take a new spectrum and plot it out.  Your nice singlet for water should now be a doublet.  Reset z and repeat this procedure for z2 and z3 plotting as you go.  If you are interested, try changing z2 and z3 the other direction.  How does that affect the spectrum differently?
  6. Reset all of your shims to their "good" values and turn off the spinner.  The level of the lock will probably drop, sometimes it drops enough to lose your lock.  This is because a substantial amount of inhomogeneity in the xy plane is removed by averaging it by spinning the sample.  We can actually get reasonable homogeneity without spinning the sample if we adjust the non-spinning shims.  We should use steps 1-4 of the third round of shimming to fix the non-spinning shims.  Do so and then take another spectrum to compare.  Note that z, z2, and z3 will need to be fixed slightly if you are not spinning (the sample moves up and down slightly when the normal cushion of air from the spinner is absent).
  7. Turn spinning back on and then readjust z, z2, and z3 to optimize their values.  Take a spectrum.  Now deliberately mess up the "x" shim by A LOT.  Take another spectrum.  Notice how there are now absolutely huge spinning sidebands.  Measure the distance between those spinning sidebands in Hz and verify that the difference is a multiple of the spinning frequency.

That is the end of the first assignment.  The next assignment will involve learning all of the various dashboard parameters for MacNMR to allow complete control over the acquisition of spectra.