Pulsed NMR

Concepts
 Nuclear magnetic resonance (NMR); pulsed NMR (PNMR), magnetic resonance imaging
(MRI).

Background Reading
 Magnetic Resonance Experiments: Melissinos 7.1-7.5 or equivalent
 Pulsed NMR Fundamentals: Teachspin manual

Special Equipment and Skills

LeCroy 9400 Digital Oscilloscope; High-field permanent magnet; radio frequency (rf) pulses, rf-
amplifier, mixer, etc.

Precautions
 The high-field magnet is delicate and dangerous. Most metal objects are strongly attracted to
the magnet, potentially causing injury. The rare-earth material of the magnet is brittle, and can
be damaged by metal objects that are pulled into the magnet.
 The sample can be traversed between the pole pieces of the magnet, but do NOT force the
sample probe past its limits of travel. The carriage should work smoothly, do NOT force it.
 The amplifier can easily overheat if it is operated without being attached to the sample probe,
or if the pulse duty cycle exceeds 1 % (time on/time off).

Background
In this lab, you will measure the dynamics of nuclei precessing in a static magnetic field,
and in oscillating fields that are applied for various lengths of time (pulse times). The precession
rate (resonance frequency) of protons (hydrogen nuclei) depends linearly on the magnetic field
according to the equation f0 (MHz) = 4.258 B0 (kilogauss). In the experiment, most of the field
comes from a permanent (rare-earth) magnet that has a field of about 3.6 kG. Small variations of
the field (chemical shifts) come from the local environment of each nucleus. A classical picture of
a nuclear spin is analogous to a spinning top, where the precession frequency is governed by the
force of gravity. Although local variations in gravity are rarely significant, vertical acceleration
(such as in an airplane) change the precession rate in an analogous manner.
You will focus on characterizing two relaxation times, T1 and T2. The spin-lattice relaxation
time (T1) is the typical time for nuclear spins to forget their initially-induced (non-equilibrium)
alignment, and return to their preferred alignment in the direction of B0. The analogy is when a
spinning top falls over due to insufficient conservation of angular momentum as it slows down, or
due to strongly fluctuating random forces on the top. The spin-spin relaxation time is the typical
time for nuclear spins to stop precessing synchronously. An apparent spin-spin relaxation time
(T2*) comes from differences in their precession frequency, causing changes in their precession
phase. However, the more-fundamental spin-spin relaxation time (T2) involves changes in their
precession frequency due to changes in the local magnetic field. The analogy involves changes in
the local gravity (or vertical acceleration) that would change the precession frequency of the top.
Additional details regarding the behavior of nuclear spins in static, and oscillating magnetic
fields can be found in the background reading.

F18 pulsed NMR.docx p1/5 8/18/2018

I. Free-Induction Decay (FID)
In the first set of experiments you will tune the spectrometer, and examine how the
amplitude of an induced spin alignment relaxes as the spins de-phase via T2 and/or relax towards
the external field via T1. Mineral oil has convenient values for both T1 and T2, of 10-30 ms. Obtain
(or prepare) a mineral oil sample that is roughly “spherical” (as tall as it is wide) in a vial, see Fig.
2.3 pg. 30. The O-ring near the top of the vial is used to position the sample in the spectrometer.
After you achieve an initial FID, you should explore the influence of the sample position by
carefully sliding the vial up and down and observing changes in the FID. Then you should ensure
that the sample is centered in the spectrometer to maximize the signal for the remainder of the lab.

900 pulse FID (pgs. 28-29 in Teachspin manual).

1. Connect equipment as shown in Fig. 1.3 (reproduced above). Use the recommended settings,
also reproduced above. If the intensity is too low, you may wish to increase the Gain up to 50%.
Typical oscilloscope settings: Time/Div 20 μs, Ch1 1V/cm, Ch2 2 V/cm, Trig: Ext, +2 V, POS
2. Adjust “Tuning” until the FID (detector out, channel 1 on scope) is maximized.
3. Use the “Frequency Adjust” until the beat oscillations (mixer out, channel 2) are minimized.
4. Adjust “A-width” until the FID (channel 1) is again maximized (roughly 900 pulse).
5. Repeat 3-5 to ensure minimum beat oscillations (channel 2) and maximum FID (channel 1).
6. Now change the “Frequency Adjust” to observe beat oscillations from channel 2. Sketch these
oscillations. Describe the mechanism involved; what causes the beat oscillations?
7. Return the “Frequency Adjust” to the ideal frequency (f0) that minimizes (eliminates) the beat
oscillations. Use this f0, and the resonance condition for protons (4.258 MHz/kGauss), to determine
the applied magnetic field on the sample from the pair of external rare-earth magnets.

II. Apparent spin-spin relaxation time (T2*), non-uniform field

(Discussion on pgs. 8-9).
1. Continue with the optimized settings for observing the FID, from the previous section.

F18 pulsed NMR.docx p2/5 8/18/2018

2. Adjust the oscilloscope to clearly see the full FID. Typical values: Channel 2 off; Channel 1: 50
mV/div; 50 μs/div.
3. Measurements of amplitude and time are most-accurately made using the cursors and their
digital readouts on the oscilloscope screen. To get started: Push the “Curser Measurement Time”
button (two cursors should appear on the screen); Adjust “Reference” until one cursor aligns with
the trigger time (upward arrow at the bottom of the screen, expand time scale for accuracy); Adjust
“Difference” until the other cursor it is at the maximum value of the FID (channel 1). To facilitate
reading the voltage numbers on the screen, increase the “Repetition Time” to 1 s and 100%.
4. Time dependence of the FID. Measure the amplitude of the FID (M) as a function of time, t.
Specifically: adjust the “Difference” cursor to a chosen time, then record several (6-12) values for
the amplitude of this cursor. Repeat for several (6-12) different times that cover the full range of
the relaxation. Obtain a table of FID amplitudes, with uncertainties (σM), as a function of time
5. Apparent spin-spin relaxation time. Fit the data from step 4 to an exponential relaxation,
M=M0exp(–t/T2*). This can be done using “Curve Fit” to the amplitudes, or “Linear Fit” to the
logarithm of the amplitudes. What is your value of the apparent spin-spin relaxation time, T2*?
You are likely to find a small value, T2*<0.1 ms. In fact, the actual spin-spin relaxation time (T2)
for mineral oil at room temperature is on the order of 10 ms. Why is T2>> T2*? How can the actual
T2 be measured? Hint: look at the title of this section, and the title of this lab.

III. Approximate spin-lattice relaxation time (T1’)

(Single pulse FID; section G, pg. 31)
1. Continue with the optimized settings for observing the FID, from the previous section. Make
sure that the “Repetition Time” is still long (1 s and 100%) to allow reading numbers on the screen.
2. Maximum amplitude. Move the “Difference” cursor back to the maximum amplitude, and
record several (8-12) values for the amplitude of this maximum.
3. Spin-lattice relaxation time. Shorten the “Repetition Time” until the readout fluctuates around
1/e times the maximum amplitude. (You will need to adjust the variable “Repetition Time” on the
10 ms scale). Accurately measure the variable “Repetition Time” by changing the time scale on
the oscilloscope until you can see two pulses, then determine the time between these pulses. This
time between pulses is an estimate for the spin-lattice relaxation time, T1’. What is the mechanism?

IV. Accurate spin-lattice relaxation time (T1)

(Two pulse - zero crossing, pg. 32)
1. Start with the optimized settings for observing the FID from the previous section. Temporarily
return the “Repetition Time” to 100 ms and 100% to avoid delays in observing the response.
2. Turn on the B pulse (B on, B pulses=1); start with a small delay between pulses (e.g. 0.2 ms).
3. Adjust the oscilloscope until both A and B pulses can be clearly observed on the screen.
4. Adjust the “A-width” until the amplitude following the first pulse is minimized (1800 pulse).
5. Adjust the “B-width” until the amplitude following the second pulse is maximized (900 pulse).
6. Switch “Sync” to B to focus on the maximum value following this second pulse.
7. Minimizing the amplitude. Find the “Delay Time” (τ) that minimizes the amplitude following
the second pulse. Because the actual minimum is difficult to discern directly from the scope,
measure several amplitudes (6-12) around the minimum value. Again, this is best done using the
cursors, with the “Reference” set to a fixed time preceding the second pulse and the “Difference”
set to a fixed time at (or near) the maximum that follows the second pulse, then do not change this

F18 pulsed NMR.docx p3/5 8/18/2018

cursor delay time. As before, increase the “Repetition Time” to 1 s, 100% to help in reading the
readout on the screen. Record several (6-12) amplitudes at each value of τ for accurate averages.
8. Spin-lattice relaxation time. T1 is given by the “Delay Time” that minimizes the amplitude
after the second pulse. Be sure you understand the mechanism. Use a quadratic function to fit the
amplitude as a function of τ around the minimum amplitude. Use the fit parameters to extract your
best experimental value for T1, and its uncertainty.

V. Fundamental spin-spin relaxation time (T2)

(Two-pulse spin echo, pgs. 33-34).
1. This procedure involves adjusting A and B pulses, as in section IV, but now for opposite phases.
Again you should temporarily return the “Repetition Time” to 100 ms and 100% to avoid delays
in observing the response.
2. Start with a small “Delay Time” (τ) between pulses (e.g. 0.2 ms).
3. Temporarily switch “Sync” to A and adjust the oscilloscope until both A and B pulses can be
clearly observed on the screen.
4. Adjust the “A-width” until the amplitude following the first pulse is maximized (900 pulse).
5. Adjust the “B-width” until there is a spin echo that occurs at a later time τ after the second pulse
(1800 pulse).
6. Repeat steps 4 and 5 to maximize the amplitude of the spin echo. Note that a more-sensitive
way to adjust the A-width is to minimize the amplitude of the signal immediately following the B-
pulse (not the spin echo that occurs time τ after the second pulse, which should be maximized).
7. Switch “Sync” back to B to focus on the spin echo that occurs time τ after the second pulse.
8. Spin-spin relaxation time. Obtain several (6-12) values for the echo amplitude as a function of
total “Delay Time” (2τ). Note that this total delay time is twice the “Delay Time” setting, τ; be
sure you understand why. Use the cursor for Channel 1 to obtain accurate values of this amplitude,
similar to what was done in sections II 3 and III 2, but with the “Difference” cursor set to the
maximum amplitude of the echo. (As you change time scales on the scope to observe much later
spin echoes, it may be useful to accurately set the position of the peak amplitude by shifting the
delay time of the cursor by the shift in τ each time.) As before, switch the “Repetition Time” to 1
s and 100% to facilitate reading the digital values on the screen. Again record each value 6-12
times to obtain an accurate amplitude. Similar to section II 5, fit your data to an exponential decay,
which yields the fundamental spin-spin relaxation time, T2, and uncertainty.

F18 pulsed NMR.docx p4/5 8/18/2018

 Pulsed NMR: specific requirements

Figure 1, Apparent Spin-Spin Relaxation Time (T2*): (20 points) Semi-logarithmic graph of free-
induction decay amplitude as a function of time. Fit the decay to a simple exponential function (straight
line on the semi-log graph). Show the best fit, and parameters, on the graph.
In the caption: Briefly describe (1-3 sentences) the physical mechanism of how the pulse sequence rotates
the spins to yield the FID; what causes this FID and why is it only an apparent spin-spin relaxation time?
Give your experimental value, and its uncertainty, for this T2*. Also give your value for the magnetic field
B0 (with uncertainty) on the sample due to the pair of external rare-earth magnets.

Figure 2, Spin-Lattice Relaxation Time (T1): (30 points) Graph of the amplitude of the magnetization
immediately following the B-pulse as a function of delay time. Show the best fit to this amplitude using a
quadratic function, with the fit parameters and their uncertainties.
In the caption: Briefly describe (1-3 sentences) the physical mechanism of how the pulse sequences rotate
the spins to yield an amplitude after the second pulse, and what causes this amplitude to be minimized at
the spin-lattice relaxation time. Give your best experimental value for T1, and its uncertainty, deduced
from the fit parameters for the quadratic. Also give your value for the approximate spin-lattice relaxation
time T1’, and compare it to your value for T1’; are they consistent, why or why not?

Figure 3, Spin-Spin Relaxation Time (T2): (30 points) Semi-logarithmic graph of spin-echo amplitude
as a function of the total delay time (2τ). Fit the decay to a simple exponential function (straight line on
the semi-log graph). Show the best fit, and parameters, on the graph.
In the caption: Briefly describe (1-3 sentences) the physical mechanism of how the pulse sequences rotate
the spins to yield a spin echo after the second pulse, and what causes this amplitude to decrease with
increasing delay time. Give your best experimental value for T2, and its uncertainty, deduced from the
exponential fit parameters. Contrast your value for the apparent spin-spin relaxation time T2*; why is it so
different from your value for T2?

F18 pulsed NMR.docx p5/5 8/18/2018