вход по аккаунту


Imaging Techniques Boost NMR Spectroscopy How To Remove Zero-Quantum Artifacts.

код для вставкиСкачать
Imaging in NMR Spectroscopy
Imaging Techniques Boost NMR Spectroscopy: How To
Remove Zero-Quantum Artifacts
Thomas Schulte-Herbrggen*
imaging techniques · NMR spectroscopy
Dedicated to Richard R. Ernst
on the occasion of his 70th birthday
Fourier imaging … can be considered as a
typical 2D spectroscopy technique.
R. R. Ernst et al.[1]
The Problem with Zero-Quantum
Magnetic resonance imaging[15] inherited its fundamental Fourier methods
largely from NMR spectroscopy. In turn,
spatial separation is just about to conquer spectroscopic domains by setting
landmarks of progress. After allowing a
multidimensional NMR experiment to
be performed in a single scan,[2] imaging
methods now solve the long-standing
problem of purging NMR spectra of
zero-quantum artifacts, as recently
shown by Keeler and Thrippleton in
this journal.[3]
Ironically, in the history of mathematics, the zero is widely recognized as
one of the most helpful ingenious concepts introduced in antiquity, whereas
zero-quantum coherences have all too
often shown up as the devil's signature
in a wide range of everyday NMR
spectra. It is a safe bet that nearly every
NMR spectroscopist has frowned over
ugly dispersive trails and misleading J
correlations, for example, in NOESY
spectra. What a relief if a protein
structure does not hinge on peaks interdispersed with (if not identical to)
“ghost peaks”!
[*] Dr. T. Schulte-Herbr ggen
Institut f r Organische Chemie und
Biochemie II
Technische Universit)t M nchen
Lichtenbergstrasse 4
85 747 Garching (Germany)
Fax: (+ 49) 89-289-13210
Why are zero-quantum coherences
so difficult to eliminate? They simultaneously escape two of the most powerful
purgative tools of NMR spectroscopy:
phase cycling and pulsed-field gradients.
They are both equally ineffective when
it comes to separating zero quanta from
z polarization. It is therefore worth
looking back in a bit more detail.
Only one year after the breakthroughs to pioneer 2D NMR spectroscopy,[4] Wokaun and Ernst showed in a
seminal paper[5] that every component
in a density operator (which represents
the state of a spin ensemble) can be
labeled with a phase factor eipf according to its quantum order p when subjected to a z rotation or phase shift of
angle f. Moreover, they demonstrated
how to exploit these phase labels to
select the components of desired quantum order, while annihilating all others
by way of phase cycling, that is, by
Fourier analysis with respect to phase
shifts of angle f.[5] This method led to
quantum filtering[6] and soon became
routine.[1, 7] Most prominently, it turned
the crude look of early COSY spectra
into the “well-groomed” appearance of
double-quantum-filtered (DQF) COSY.[8] Without dispersive trails from the
diagonal, they provide valuable connectivity and bond-angle information,
which is complemented with NOESYderived distance constraints to form the
basis of structure elucidation by NMR
Why, then, does the same trick not
purge NOESY-type spectra of zeroquantum artifacts? This reason lies in
the fact that the NOE transfers polarization from one spin to the other
(Ikz !
7 Ilz), and z polarization (Ikz) as well
as longitudinal multispin order (e.g.
2 Ikz Ilz) both remain invariant under
2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/anie.200301689
z rotation. Hence they appear invariably
with the phase label associated with p =
0—just as zero-quantum coherences do.
Separation impossible!
Even worse, these three types of
terms also remain largely unaffected by
magnetic-field inhomogeneity.[5] The
reason is similar: by applying pulsedfield gradients, terms of different quantum order receive different degrees of
dephasing, again proportional to their
quantum order.[10] This method, to
which the group of Keeler has significantly contributed over the years,[11]
quickly made it into the standard toolbox to select desired quantum orders in
a single scan (mostly one each for the
echo and anti-echo parts), whereas
phase cycling usually requires the combination of more scans. However, since
the phase labeling is similar in the two
methods (phase cycling and gradients),
both share largely the same blind
spots—but not entirely so.
The Solution of Keeler and
Keeler and Thrippleton[3] have now
provided an elegant solution to the
notorious conundrum of how to suppress zero-quantum coherences efficiently, while retaining polarizations or
longitudinal multispin orders. Their experimental results are remarkably superior to any of the previous attempts at
tackling the problem. This is made
possible by state-of-the-art spectrometers that allow the combination of
pulsed field gradients and selective
pulses with adiabatic frequency sweeps,
features which the pioneers could hardly
have dreamt to be compatible with the
Angew. Chem. Int. Ed. 2003, 42, 5270 –5272
stability demands in high-resolution
spectroscopy some day.
The basic trick derives from imaging
and rests on spatial separation of different shift evolutions that sum to zero
when averaged over the entire sample.
The underlying refocusing module can
readily be understood on a pictorial
level in more detail. One needs to
remember that a z gradient makes the
resonance frequency in each slice of the
sample depend on the height (z coordinate) within the NMR tube (Figure 1 a).
In other words, every layer has its
unique Larmor frequency w. In the
presence of such a z gradient, a selective
1808 pulse is swept during a time interval
tf such that it will touch the resonance
frequencies of different disks at different times, reaching from say t = 0 for the
top slice to t = tf for the bottom one.
Thus, the chemical-shift evolution will
be inverted (in the sense of Hahn's spin
echo) first at the top (t = 0), and at the
bottom only from t = tf on (Figure 1 b).
In particular, this also holds for any
evolution under a nonvanishing zeroquantum frequency WZQ = Wk Wl of
Ik Iþl . If the spread of zero-quantum shift
evolution between the slices is large
enough, the average taken over the
entire sample will be zero—even if
another time interval (such as a mixing
time) follows (Figure 1 c).
This cancellation naturally results
after a single scan and is robust to
random processes such as diffusion
between the slices as well as to convection driven by temperature gradients
within the sample. However, in NOESY
experiments the resonance-frequency
bandwidth over the sample has to be
much larger than the spectral
width, otherwise the crosspeak buildup between two
spins will be attenuated depending on their shift difference and thus hinder a quantitative interpretation. Yet, owing to its robustness and general applicability, for example,
to isotropic TOCSY[12] and
other experiments with mixing
processes starting from z magnetization, the method will
surely find broad usage.
Multidimensional NMR
Spectroscopy in One Shot
Figure 2. Excitation scheme for a proposed singe-scan recording of
a 2D spectrum.[2] The z gradient ensures that every slice exhibits a
distinct resonance frequency. A selective 908 pulse then excites every slice at a different time, so that the spins evolve slicewise in t1.
Only mixing can proceed uniformly. Echo-planar imaging[14] methods finally allow the signals from each disk to be separated upon
Experimental robustness is
certainly not the key point in a
proposal for performing multidimensional NMR experiments in single
scans—speed (short measuring time) is.
Actually the paper by Frydman et al.[2]
might well have served to inspire Keeler
and Thrippleton. The principle of a onescan 2D NMR experiment resides in
parallel evolution of several t1 increments, each in a spatially separated slice
that also has to be detected distinctly. To
fulfill these experimentally tricky demands, the authors exploit ideas of
Fourier imaging.[13] However, in this
case the phase labels of the slice position
are not used for spatial resolution but
are taken as t1 evolutions (Figure 2). To
this end, a selective excitation pulse (of
flip angle 908) has to be swept over
discrete frequencies at discrete times in
the presence of a position-encoding
gradient. Thereby each slice is excited
Figure 1. Method to average out zero-quantum coherences while retaining polarization and
longitudinal multispin order.[3] See text for details.
Angew. Chem. Int. Ed. 2003, 42, 5270 –5272
with a unique Larmor frequency and
thus evolves with a unique t1 increment.
Only the subsequent mixing may proceed uniformly, whereas the detection
process is slice-selective by making use
of gradient echoes from echo-planar
imaging,[14] a trick “borrowed” from
medicinal NMR imaging. Confining
each t1 increment to a single spatially
separated slice and thereby to a very
small sample fraction requires high
concentrations; this is feasible in the
case of small molecules for which reduced resolution by diffusion and t1
artifacts by convection are not detrimental. Therefore, particularly in routine analytical NMR spectroscopy, the
new approach may well prove to be a
valuable time saver, provided it can be
reliably implemented and automated.
In conclusion, imaging techniques
allow the spectroscopist to separate the
sample into spatially divided layers with
different excitation, evolution, and detection. Depending on the transformation properties of the pertinent operator
terms, spatial editing can thus be combined with spectral editing. Beyond
purging or accelerating known techniques, this opens a perspective of creativity in designing novel experiments.
[1] R. R. Ernst, G. Bodenhausen, A. Wokaun, Principles of Nuclear Magnetic
Resonance in One and Two Dimensions,
Clarendon Press, Oxford, 1987; see also
Nobel Prize Lecture: R. R. Ernst, An 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
gew. Chem. 1992, 104, 817 – 836; Angew.
Chem. Int. Ed. 1992, 31, 805 – 823.
L. Frydman, T. Scherf, A. Lupulescu,
Proc. Natl. Acad. Sci. USA 2002, 99,
15 858 – 15 862.
M. J. Thrippleton, J. Keeler, Angew.
Chem. 2003, 115, 4068 – 4071; Angew.
Chem. Int. Ed. 2003, 42, 3938 – 3941.
W. Aue, E. Bartholdi, R. R. Ernst, J.
Chem. Phys. 1976, 64, 2229 – 2246; “The
Unpublished Baško Polje (1971) Lecture Notes”: J. Jeener in NMR and More
(Eds.: M. Goldman, A. Porneuf), Les
Nditions de Physique, Les Ulis, 1994,
pp. 266 – 272.
A. Wokaun, R. R. Ernst, Chem. Phys.
Lett. 1977, 52, 407 – 412.
G. Bodenhausen, H. Kogler, R. R.
Ernst, J. Magn. Reson. 1984, 58, 370 –
H. Kessler, M. Gehrke, C. Griesinger,
Angew. Chem. 1988, 100, 507 – 554;
Angew. Chem. Int. Ed. Engl. 1988, 27,
490 – 536.
[8] U. Piantini, O. W. Sørensen, R. R. Ernst,
J. Am. Chem. Soc. 1982, 104, 6800 –
[9] K. WPthrich, NMR of Proteins and
Nucleic Acids, Wiley, New York, 1988;
see also Nobel Prize Lecture: K. WPthrich, Angew. Chem. 2003, 115, 3462 –
3484; Angew. Chem. Int. Ed. 2003, 42,
3340 – 3363.
[10] C. J. R. Counsell, M. H. Levitt, R. R.
Ernst, J. Magn. Reson. 1985, 64, 470 –
[11] G. Kontaxis, J. Stonehouse, R. D. Laue,
J. Keeler, J. Magn. Reson. Ser. A 1994,
111, 70 – 76; A. L. Davis, J. Keeler, R. D.
Laue, D. Moskau, J. Magn. Reson. 1992,
98, 207 – 216; see also: Encyclopedia of
Nuclear Magnetic Resonance (Eds.:
D. M. Grant, R. K. Harris), Wiley, Chichester, 1996.
[12] S. J. Glaser, J. Quandt, Adv. Magn.
Reson. 1996, 19, 59 – 252.
[13] A. Kumar, D. Welti, R. R. Ernst, J.
Magn. Reson. 1975, 18, 69 – 83; W. A.
2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Edelstein, J. M. S. Hutchinson, G. Johnson, T. Redpath, Phys. Med. Biol. 1980,
25, 751 – 756; P. T. Callaghan, Principles
of Nuclear Magnetic Resonance Microscopy, Clarendon Press, Oxford, 1991.
[14] P. Mansfield, J. Phys. C 1977, 10, L55 –
L58; P. Mansfield, Magn. Reson. Med.
1984, 1, 370 – 386; see reference [15 b] as
well as: P. G. Morris, NMR Imaging in
Biology and Medicine, Oxford University Press, Oxford, 1986.
[15] During the revision of this Highlight, it
was announced that Lauterbur and
Mansfield had been awarded the Nobel
Prize in Medicine for 2003. For more
information on the techniques, see:
a) P. C. Lauterbur, Nature 1973, 242,
190 – 191; b) P. Mansfield, P. G. Morris,
NMR Imaging in Biomedicine, Academic Press, New York, 1982; c) see also
reference [14].
Angew. Chem. Int. Ed. 2003, 42, 5270 –5272
Без категории
Размер файла
997 Кб
spectroscopy, remove, nmr, quantum, zero, artifact, imagine, boos, techniques
Пожаловаться на содержимое документа