Chemistry Reference
In-Depth Information
precessions at frequencies
J
/2 relative to the rotating frame. Because this frame is
significantly tilted from the rotating frame, application of frequent
1
Hpulsesresultsin
incomplete cancelation of the cross term. Therefore, infrequent
1
H pulses are applied
to suppress DD/CSA correlation when the off-resonance effect is significant [
38
,
39
].
Other parameters that have to be specifically considered in the
R
1
r
experiment
are the spatial homogeneity of the
B
1
field strength and linearity of the power
amplifier. Although
B
1
homogeneity has been improved in recent NMR probes,
estimation of the inhomogeneity remains important to confirm the accuracy of the
obtained
R
1
r
. For this purpose, measurement of the
B
1
inhomogeneity using inverse
detection is useful [
40
]. Amplifier linearity has also improved. However, since
pulse power is switched for spin-lock, it is important to check that a phase shift
accompanies a power change, and make a correction should one be needed.
2.1.2 Carr-Purcell-Meiboom-Gill
R
2
Experiment
In CPMG
R
2
,
limitation of applicable B
1
field strength
is in general smaller than that
of the
R
1
r
experiment. Since CPMG 180
pulses are applied with interpulse delays
(2
than that of spin-lock can be
applied. However, in a protein, the RF field strength used in the
15
NCPMG
R
2
experiment must be carefully considered because the interpulse delay is set short
(typically, 2
t
CP
), CPMG pulses with much stronger
g
N
B
1
/2
p
t
CP
~1 ms) to suppress generation of antiphase terms, N
X
,
Y
H
Z
,causedby
1
H-
15
N J coupling. For example, if 6 kHz CPMGpulses are appliedwith 2
1ms,
the RF power delivered to the probe is nine times stronger than that of a 2 kHz spin
lock. Given that the duty cycle is ca. 10% in the CPMG experiment, about the same
amount of energy is deposited in the probe in both CPMG and
R
1
r
experiments.
In CPMG
R
2
,
off-resonance error
is negligible at low magnetic field strength but
significant at high magnetic field strength [
41
,
42
]. Although the stronger
t
CP
¼
for
each CPMG pulse inverts magnetization more uniformly than the spin-lock, CPMG
pulse train accumulates error caused by a combination of pulse imperfections and off-
resonance effects. The CPMG error of a signal located at off-resonance frequency
f
off
(
f
off
is the difference between the signal and carrier frequencies) is estimated
with a function of 2
g
N
B
1
/2
p
t
CP
and
B
1
. Importantly, the off-resonance error is maximized
at 2
t
CP
f
off
¼
n
(
n
is an integer): when 2
t
CP
¼
1 ms, the off-resonance error is signi-
ficant at
f
off
¼
1, 2, and 3 kHz [
41
]. Magnitude of the error depends on the
B
1
field
strength the 180
pulses. This relationship indicates that
R
2
canberecordedwithout
significant off-resonance error at 61 MHz for signals because the entire chemical
shift range for amide backbone
15
N signals spans approximately less than
15 ppm.
t
CP
is used at
15
N 91MHz resonance frequency, signals located
However, if the same 2
at
11 ppm off-resonance suffer from significant errors in measured
R
2
values.
It is a disadvantage of CPMG
R
2
experiment that there is no simple equation to
correct for CPMG
R
2
off-resonance effects. In practice, it is recommended to
discard
R
2
data obtained at
f
off
¼
t
CP
frequency, and record the data at two
different carrier frequencies. As an alternative, a phase cycle to average out the off-
resonance effect in CPMG
R
2
may be used [
43
,
44
]. Using this method,
R
2
is
n
/2