Published on: Mar 4, 2016
Transcripts - PoplettMESmithSSNMR1998v11p211FieldSweepNMR
Ž .Solid State Nuclear Magnetic Resonance 11 1998 211–214
Field sweep broadline NMR spectroscopy
I.J.F. Poplett, M.E. Smith )
School of Physical Sciences, UniÕersity of Kent, Canterbury, Kent, CT2 7NR, UK
Received 20 October 1997; accepted 2 December 1997
A novel NMR spectrometer is described that is uniquely versatile in its ability to accurately record broad lines by
sweeping the superconducting magnetic field and to perform standard high resolution solid state NMR experiments.
Broadline observation is illustrated by 27
Al spectra from static samples. Such an instrument opens up many nuclei for serious
study by NMR in the solid state for the first time. q 1998 Elsevier Science B.V. All rights reserved.
Keywords: Broadline NMR; Aluminium-27; Magnetic field sweep; NMR spectrometer
Today, NMR spectroscopy largely means a pulsed
Ž .Fourier transform FT approach using high field
w xsuperconducting magnets 1 . The relatively broad
frequency range simultaneously excited by a short
Ž .radiofrequency rf pulse brought great improve-
ments in time efficiency compared to the original
Ž .continuous wave CW approach of sweeping the
w xfield or frequency 1 . The pulsed FT approach has
been extremely successful for high resolution solu-
w xtion state experiments 2 and in high resolution
experiments of solids, such as magic angle spinning
w x 13
especially for spin-1r2 nuclei, e.g., Si 3 and C
w x4 . However, pulsed NMR has a weakness in that
the bandwidths of pulse excitation, probe and the
w xspectrometer itself 1 makes accurate recording of
broad lines very difficult, so that NMR of broad lines
Corresponding author. Fax: q44-01227-827558; e-mail:
firstname.lastname@example.org. now at Department of Physics, University
of Warwick, Coventry, CV4 7AL, UK.
has become a largely neglected field. There are also
effects of deadtime as the system recovers from the
rf pulse, even in the best engineered systems that
lead to observable distortions in broad NMR lines,
although these can be partially alleviated by using
w xecho techniques 1 . Recording broad lines has many
potential advantages for ascertaining NMR interac-
tions, such as first-order quadrupole and chemical
The neglect of broad line NMR spectroscopy has
meant that determining, e.g., the quadrupole interac-
tion for some nuclei can be generally difficult, and
those that experience significant quadrupole broaden-
Ž 25 45 47,49 59 63,65 67
ing e.g., Mg, Sc, Ti, Co, Cu, Zn,
91 93 135,137 139
.Zr, Nb, Ba and La are only considered
for NMR of solids in special cases. There is no doubt
that broadline information can be useful, with exam-
ples of relevance to materials science including 91
w x 139
of ZrO polymorphs 5 and La in lanthanum-con-2
0926-2040r98r$19.00 q 1998 Elsevier Science B.V. All rights reserved.
Ž .PII S0926-2040 97 00110-0
( )I.J.F. Poplett, M.E. SmithrSolid State Nuclear Magnetic Resonance 11 1998 211–214212
w xtaining ceramics 6 . ZrO polymorphs could be2
readily distinguished in static NMR spectra and mix-
tures of these polymorphs could be quantified by the
w xspectra 5 . The data from these materials were accu-
mulated by stepping the frequency but such experi-
ments are somewhat tedious, time-consuming and
require accurate retuning of the electronics at each
step, and very few examples of automation of such
an approach exist.
Alternatively, to record such spectra, the magnetic
field could be swept and high field superconducting
magnets have been used in this mode to examine
w xinternal fields of magnetic materials 7 and quan-
w xtum-tunnelling level-crossing effects 8 . Sweeping
of superconducting magnetic fields is also widely
used in physics to examine, e.g., de Haas–van
Alphen, magnetoresistance and magneto–optical ef-
w xfects 9 . Often, the superconducting magnets used in
Ž . 27
Fig. 1. a Field sweep Al NMR spectrum of a-Al O at an observation frequency of 79.1355 MHz with 10 G steps between successive2 3
Ž . Ž .spin-echoes which have been converted to a frequency, b the one pulse spectrum using a 0.5 ms pulse, and c a spin echo using an
Ž .interpulse spacing of 100 ms. The singularities of the quadrupolar pattern are marked Al 1 .1,2,3
( )I.J.F. Poplett, M.E. SmithrSolid State Nuclear Magnetic Resonance 11 1998 211–214 213
such experiments are very far removed from those
used in chemical NMR spectroscopy having low
Ž 3.magnetic field homogeneity e.g., 0.1% over 1 cm
Ž .and often cold i.e., -10 K access temperatures.
This contrasts markedly with the necessary high
Ž y5 3.homogeneity e.g., -10 % over 1 cm and room
temperature access magnets of high field NMR spec-
Field-swept pulsed NMR spectra has been re-
w x w xported, e.g., from titanium metal 10,11 , NbSe 122
w x w xZn 13 and most recently alumina 14 . All these
measurements were carried out in magnetic fields
with resolution far below that required for normal
Ž .spectroscopy and mostly with cold e.g., -10 K
access. Hence they are essentially magnets dedicated
to this experiment and they often also have high
Ž .helium consumptions e.g., 30 lrday . The need for
a more flexible NMR spectrometer was highlighted
nearly a decade ago by the comment of Turner et al.
w x15 ‘‘ . . . a high field continuous wave approach
would appear most attractive for studies of extremely
broad lines in ceramic materials, but unfortunately
such an approach is not commercially available.’’
This approach has always been essentially available
but only on noncommercial, low resolution instru-
ments, as described above. For commercial exploita-
tion there have been two main perceived drawbacks:
Ž .1 the prohibitively expensive helium costs because
of increased consumption on sweeping the field, and
Ž .2 the need to have a specialised, dedicated instru-
ment since such a system would not have sufficient
performance for conventional spectroscopy. Neither
of these perceptions are necessarily correct.
A novel 7 T magnet from Magnex Scientific
delivered to the University of Kent has, in addition
to the main superconducting solenoid, a supercon-
ducting sweep coil that can alter the magnetic field
Ž .by 1 T. The boil-off rate is slightly increased ;10%
whilst sweeping and the sweep operation in no way
detracts from the performance of the system in high
resolution mode when it is not being swept. The
magnet has excellent resolution, being entirely
equivalent in all tests, due to a dedicated high resolu-
tion solid state NMR spectrometers, producing a 13
MAS decoupled linewidth on admanatane of 2.5 Hz
before and after a field sweep run using the same
room temperature shim file. Hence, this extremely
versatile NMR spectrometer provides both broad line
field sweep and conventional high resolution opera-
tion in a single instrument.
The novel wideline field sweep operation of this
system is illustrated by 27
Al for which the system has
a nominal resonance frequency of 79 MHz. Spectra
were accumulated using spin-echoes on a Chemag-
netics CMX 300 Infinity spectrometer with low
power 908 pulses of 45 ms duration. The additional
field is controlled by the current from a Lakeshore
610 that is set directly from the Spinsight software
via an RS-232 link. The field can be varied by "0.5
T with a settability of 0.0001 T and exhibits a highly
Fig. 2. 27
Al field sweep NMR spectrum of yttrium aluminum garnet using a frequency of 78.457 MHz with 4.5 G steps between successive
Ž .spin-echoes with 80 transients averaged at each point using a recycle time of 5 s. The singularities from the two sites are marked Al 1 1,2
Ž . Ž .and Al 2 . Note that the frequency axis is broken for convenience .1,2
( )I.J.F. Poplett, M.E. SmithrSolid State Nuclear Magnetic Resonance 11 1998 211–214214
linear current-field relationship. The current is set to
Ž .a particular value, allowed to settle ;1 min and a
series of spin-echoes accumulated using extended
w xphase cycling 16 to give sufficient signal-to-noise
before stepping to the next field setting and repeating
the experiment, eventually stepping through the com-
plete lineshape. All of this is done automatically
from within the pulse program and the results from
a-Al O are shown in Fig. 1a. A classical first-order2 3
quadrupolar powder pattern is seen with three singu-
Ž .larities from which the quadrupole coupling C ofQ
Ž .2.3 MHz and asymmetry parameter h of 0 can be
deduced, in good agreement with single crystal work
w x17 . The field sweep spectrum can be compared to a
Ž .conventional one pulse spectrum Fig. 1b where the
central transition is strongly recorded with only the
first two singularities of the broad outer transitions
Ž .showing up as weak features. An echo Fig. 1c still
records only the first two singularities, but better
records the overall outer transition intensity com-
pared to a single pulse. Both single pulse and echo
spectra do not have sufficient bandwidth to record
the complete lineshape and distort the intensity of
the outer transitions that is recorded.
A second illustration is given by the 27
Ž .of yttrium aluminium garnet YAG that has octahe-
drally and tetrahedrally coordinated aluminium sites
whose C s differ by a factor of 10, and have beenQ
extensively characterised by NMR of single crystals
w x w x18 and powder MAS 19,20 . A highly detailed
distortion-free field sweep NMR spectrum is ob-
Ž . Ž .tained Fig. 2 and gives C s of 0.6 MHz AlOQ 6
Ž .and 6.0 MHz AlO both with hs0, in good4
agreement with previous work. In both cases, it
should be noted that the field sweep NMR spectra
are as they come from the spectrometer. There has
been no correction of the baseline nor need to phase
Al NMR broadline spectra obtained by sweeping
a superconducting solenoid allow much information
to be extracted. These experiments were performed
on a spectrometer that additionally has all the capa-
bilities of a conventional high resolution solid state
NMR spectrometer. This greatly enhances the range
of nuclei that can be meaningfully and routinely
studied. There are other new experiments that could
be envisaged such as one-dimensional stray field
imaging of solids without moving the sample. These
possibilities mean that such an NMR instrument
could be a standard commercial configuration and
have a widespread application in solid state chem-
istry, mineralogy, materials and biological sciences.
Ž .The authors thank Mr. R.S. Thomas Kent and
Ž .Dr. P. Jonsen Chemagnetics for their help in imple-
Žmenting the field sweep, Dr. T.J. Bastow CSIRO,
.Melbourne for his helpful comments on the
manuscript and the University of Kent for its support
w x1 E. Fukushima, S.B.W. Roeder, Experimental Pulse NMR,
Addison-Wesley, Reading, MA, 1981.
w x2 A.E. Derome, Modern NMR Techniques for Chemistry Re-
search, Pergamon, Oxford, 1987.
w x3 G. Engelhardt, D. Michel, High Resolution Solid-State NMR
of Silicates and Zeolites, Wiley, Chichester, 1987.
w x4 K. Schmidt-Rohr, H.W. Spiess, Multidimensional Solid-State
NMR and Polymers, Academic Press, London, 1994.
w x Ž .5 T.J. Bastow, M.E. Smith, Solid State NMR 1 1992 165.
w x Ž .6 T.J. Bastow, Solid State NMR 3 1994 17.
w x Ž .7 Y. Furukawa, S. Wada, J. Phys. Condensed Matter 6 1994
w x8 S. Clough, A.J. Horsewill, P.J. McDonald, F.O. Zelaya,
Ž .Phys. Rev. Lett. 55 1985 1794.
w x9 O. Madelung, Introduction to Solid-State Theory, Springer-
Verlag, Berlin, 1981.
w x Ž .10 A. Narath, Phys. Rev. 162 1967 320.
w x Ž .11 H. Ebert, J. Abart, J. Voitlander, J. Phys. F 16 1986 1287.
w x12 E. Ehrenfreund, A.C. Gossard, F.R. Gamble, T.H. Geballe, J.
Ž .Appl. Phys. 42 1971 1491.
w x13 J. Abart, E. Pelangie, W. Socher, J. Voitlander, J. Chem.
Ž .Phys. 78 1983 5468.
w x14 X. Wu, E.A. Juban, L.G. Butler, Chem. Phys. Lett. 221
Ž .1994 65.
w x15 G.L. Turner, R.J. Kirkpatrick, S.H. Risbud, E. Oldfield, Am.
Ž .Ceram. Soc. Bull. 66 1987 653.
w x16 A.C. Kunwar, G.L. Turner, E. Oldfield, J. Magn. Reson. 69
Ž .1986 124.
w x Ž .17 R.V. Pound, Phys. Rev. 79 1950 685.
w x Ž .18 K.C. Brog, W.H. Jones, C.M. Verker, Phys. Lett. 20 1966
w x19 R. Dupree, M.H. Lewis, M.E. Smith, J. Appl. Cryst. 21
Ž .1988 109.
w x20 D. Massiot, C. Bessada, J.P. Coutures, F. Taulelle, J. Magn.
Ž .Reson. 90 1990 231.