NMR in biological Objects and Magic Angle Spinning

Proton NMR in Biological Objects Subjected to Magic
Angle Spinning

R A Wind and J Z Hu

, Pacific Northwest National

Laboratory, Richland, WA, USA

Introduction

Magnetic resonance imaging (MRI) of the spatial
distribution of water in biological objects has
developed as one of the major tools to diagnose le-
sions and diseases and follow the therapy response in
animals and patients in a minimally invasive way. In
addition, in vitro and in vivo localized magnetic reso-
nance spectroscopy (MRS) and spectroscopic ima-
ging or chemical shift imaging (CSI) are increasingly
used in biochemical and biomedical studies in cells,
tissues, animals, and humans. With these techniques
the resonance lines of several relatively small molec-
ular weight chemical compounds such as amino ac-
ids, lipids, and other key mobile metabolites are
measured, and their presence and intensities have
been linked to tumor phenotype, tumor formation,
tumor size, increased cell proliferation, and cell death
pathways. However, a major problem associated
with nuclear magnetic resonance (NMR) spectro-
scopy in intact biological tissues is that relatively
large resonance line widths are observed, often one
to two orders of magnitude larger than the widths
measured in liquids using established NMR tech-
niques. This is especially a problem for

H NMR,

which is the most widely used nucleus because of its
relatively large NMR sensitivity, but has a relatively
small chemical shift range of

B10 parts per million

(ppm). As usually many metabolites contribute to the
NMR signal, the result is a spectrum with severely
overlapping spectral lines, which seriously hampers a
quantitative analysis of the spectra, and sometimes
even makes it impossible to assign the spectral lines
unambiguously.

In biological samples, the main mechanisms for

this broadening are the local magnetic ﬁeld gradients
arising from variations in the isotropic bulk magnetic
susceptibility near boundaries of intra- and extracel-
lular structures, such as the various intracellular
compartments, air–tissue interfaces near the lungs
and sinuses, and bone–tissue interfaces. Then, chemi-
cally equivalent nuclei experience different local
magnetic ﬁelds, depending on their spatial locali-
zation, giving rise to line broadening. Using cell
extracts can eliminate this broadening, but this

procedure is time consuming, introduces spectral
artifacts, and cannot be applied to study intact tis-
sues and organs. Another method of improving the
spectral resolution is to increase the external magne-
tic ﬁeld B

, as the separation between the lines in an

NMR spectrum increases linearly proportional with
B

. However, the susceptibility-induced broadening

is linearly proportional to B

as well, and in many

cases zero or only marginal improvements in the
spectral resolution have been reported when higher
ﬁeld strengths were employed.

In principle, the susceptibility broadening can be

averaged to zero by the technique of magic angle
spinning (MAS), where the sample is rotated about
an axis making an angle of 54

144

relative to the

external magnetic ﬁeld. In a standard MAS experi-
ment, where the NMR signal is observed after a
single 90

1 rado frequency (RF) pulse (to be called

single-pulse MAS or SP-MAS hereafter), the spinning
frequency must be larger than the broadening in ord-
er to avoid the occurrence of spectral spinning side-
bands (SSBs) surrounding the various resonance
peaks, which can overlap with other resonance lines,
rendering the interpretation of the spectra difﬁcult
again. In fact, in practice often the spinning fre-
quency is chosen larger than the spectral width, i.e.,
a kilohertz or more, in order to avoid SSBs arising
from the water signal, which occurs in biological
objects in a concentration of

B30 mol l

, and

which is often much stronger than the metabolite
signals, arising from compounds with two to three
orders of magnitude smaller concentrations, even
when water suppression is applied. Figure 1 shows

H water-suppressed NMR metabolite spectra of ex-

cised rat liver, obtained in a 7 T external ﬁeld (i.e.,
300 MHz proton frequency) on a stationary sample
and with SP-MAS at different spinning speeds, illu-
strating the appearance of SSBs at low spinning
speeds. In fact, at 1 Hz spinning the side bands are so
dense that the spectrum is virtually the same as the
static spectrum. In these experiments, water suppres-
sion was achieved by preceding the 90

1 pulse by

a DANTE (delays alternating with nutations for
tailored excitation) water suppression sequence, con-
sisting of a train of equally spaced small-tip-angle
hard pulses.

A serious problem associated with fast SP-MAS is

the large centrifugal force, F

, induced in the sample

by the spinning, which destroys tissue structures and
even individual cells at high spinning rates. F

NMR SPECTROSCOPY APPLICATIONS

/ Proton NMR in Biological Objects Subjected to Magic Angle Spinning

333

given by F

¼ mo

r, where m is the mass, o

¼ 2pF, F

being the spinning frequency, and r the distance from
the rotation axis to the point of interest. For exam-
ple, when F

¼ 2 kHz and r ¼ 1 cm, F

¼ 1.6 10

times the gravitational force F

. Therefore, the

SP-MAS method is not viable for MRS or CSI in
large intact biological samples or in vivo studies, and
methods are needed that yield high-resolution, SSB-
free spectra at reduced MAS frequencies. In the re-
mainder of this article, two such methods and some
applications will be discussed, following short back-
ground introductions about magnetic susceptibility
and MAS.

Magnetic Susceptibility

The magnetic susceptibility factor w arises from the
magnetization M induced in a material exposed to
an external magnetic ﬁeld H

, corresponding to a

magnetic induction B

¼ m

, in the absence of the

material, where m

¼ 4p 10

(T m A

) is the

magnetic permeability in vacuum. When the magneti-
zation is oriented in the same direction as B

, w is a

scalar and the susceptibility is called isotropic. w is
called the volumetric susceptibility and is a unitless
quantity deﬁned as w

¼ M/H, where M is the magne-

tic dipole moment per unit volume and H is the total
magnetic ﬁeld strength. w is equal to 0 in free space
and is practically 0 for air. In biological systems, the

additional magnetic ﬁeld strength induced by M is
much less than H

so that w

EM/H

. With this defi-

nition, in the material the net magnetic ﬁeld induc-
tion B, which is the fundamental magnetic ﬁeld
responsible for NMR, is given by B

¼ m

þ m

þ w)B

¼ m

, where m

is the relative magnetic

permeability. M can arise from several sources: (1)
The nuclear magnetization. Although this magnet-
ization is responsible for the MR signal, its contri-
bution to M can usually be neglected. (2) The
magnetization associated with slight changes in the
angular velocities of paired electrons in their orbitals,
induced by B

. According to Lenz’s law this will

cause a magnetic ﬁeld at the center of the orbital that
opposes B

, resulting in the well-known chemical

shift. Outside the orbital at a distance large com-
pared with the dimensions of the orbital, the effect of
B

on an electron orbital can be approximated by a

local magnetic ﬁeld arising from a magnetic dipole
associated with a ring current in the orbital. This
results in a negative value of w, the diamagnetism. (3)
Paramagnetic susceptibility arises in materials con-
taining unpaired electrons, resulting in a positive
value of w. (4) Ferromagnetism occurs in materials
possessing permanent magnetic moments aligned in
Weiss domains.

Most biological objects are diamagnetic. Water

and organic compounds possess susceptibility fac-
tors w of the order of

4p 10

. In liquids, this

(A)

(B)

(C)

(D)

ppm

9 8

7 6

4 3

Figure 1

300 MHz

H NMR spectra of freshly excised rat liver samples obtained with different methods: (A) static sample; (B) 1 Hz

SP-MAS; (C) 40 Hz SP-MAS; (D) 4 kHz SP-MAS. The external ﬁeld was 7 T. Line assignments: 1 (

B0.88 ppm), triglycerides CH

terminal, or neutral amino acid methyl, valine, leucine, isoleucine methyl; 2 (

B1.28 ppm), triglycerides –(CH

)

, lactate methyl,

threonine methyl; 3 (

B2.04 ppm), triglycerides CHQCH2CH

2CH

; 4 (

B2.24 ppm), triglycerides CH

–CH

–CO; 5 (

B2.8 ppm),

triglycerides CH

QCH2CH

2CHQCH; 6 (B3.2 ppm), choline methyl, phosphocholine methyl, b-glucose, trimethylamine-N-oxide

methyl; 7 (3.4 ppm) 8 (3.6 ppm), glucose, glycogen; and 9 (3.8–4.0 ppm), glucose, glycogen, amino acids. (Reproduced with per-
mission from Wind RA and Hu JZ (2003) Magnetic susceptibility effects in nuclear magnetic resonance spectroscopy of biological
objects. In: Recent Research Developments in Magnetism and Magnetic Materials, vol. 1, pp. 147–169.)

334

NMR SPECTROSCOPY APPLICATIONS

/ Proton NMR in Biological Objects Subjected to Magic Angle Spinning

susceptibility gives rise to an induced additional
homogeneous magnetic ﬁeld in the sample, resulting
in a shift of the resonance lines rather than a line
broadening. However, in biological samples, which
contain many intra- and intercellular structures, the
various compounds are distributed heterogeneously,
and often close to a boundary of a material with a
different susceptibility factor. Then, the susceptibility
differences between the various compartments in-
duce magnetic ﬁeld gradients in the sample. As a re-
sult the resonance shifts become space dependent,
resulting in a line broadening. Often, line widths of
the order of 0.5 ppm are observed, at least one order
of magnitude larger than their intrinsic line widths.
These values are in accordance with estimations
based on the susceptibility differences one can en-
counter at interfaces in biological objects.

Magic Angle Spinning

In order to illustrate the effect of MAS we consider
the magnetic moment m arising from a sphere of a
stationary diamagnetic material (see Figure 2A). At a
distance R equal to or larger than the radius a, the
magnetic ﬁeld generated by this moment has two
components B

and B

, oriented in the direction and

perpendicular to R, respectively. These components
are given by

m cos g

and

m sin g

where g is the angle between R and the direction of
the magnetic moment, deﬁned as the z direction. It
follows that the dipolar ﬁeld component along the
z-axis, B

, is given by

¼ B

þ B

4pR

ð3 cos

1Þ

½1

Figure 2B shows the case that the sample is rotated
with an angular frequency o

about an axis making

an angle b with the external ﬁeld. Then, the angle g
and the azimuth angle f become time dependent, and
cos g(t) is given by

cos g

ðtÞ ¼ cos a cos b þ sin a sin b cos½fðtÞ

¼ cos a cos b þ sin a sin b cosðo

þ f

Þ ½2

is the angle between the rotation axis and R.

Equation [1] becomes

8pR

fð3 cos

1Þð3 cos

1Þ

þ 3 sin 2a sin 2b cos ðo

þ f

þ 3 sin

sin

cos

½2ðo

þ f

Þg

½3

Hence, B

contains a static term and two time-

dependent terms. Then, in a standard SP-MAS ex-
periment after Fourier transformation the NMR
spectrum arising from the interactions between the
nuclear spins and these local ﬁelds consists of a
center-band line, located at the frequency determined
by the static term, and SSBs, located at frequency

(t)

(A)

(B)

Figure 2

Magnetic dipole ﬁeld outside a sphere of a diamagnetic material with dipolar moment m induced by the external ﬁeld B

(A) static sample; (B) rotating sample. (Reproduced with permission from Wind RA and Hu JZ (2003) Magnetic susceptibility effects in
nuclear magnetic resonance spectroscopy of biological objects. In: Recent Research Developments in Magnetism and Magnetic
Materials, vol. 1, pp. 147–169.)

NMR SPECTROSCOPY APPLICATIONS

/ Proton NMR in Biological Objects Subjected to Magic Angle Spinning

335

distances

7no

, n

¼ 1, 2, 3,y, from the center-band

line and with amplitudes depending on the static line
width and o

. Hence, by choosing the angle

¼ cos

ﬃﬃﬃ

¼ 54:741, the effect of the static term

becomes zero, irrespective of the angle a, and for
spinning speeds much larger than the spectral line
widths the SSBs have very small amplitudes and can
often be neglected. As a result, the susceptibility shift
in a homogeneous liquid and the susceptibility line
broadening in heterogeneous samples are eliminated.
Moreover, MAS eliminates other line broadenings as
well, arising from other interactions with a similar
angular dependence as eqn [1]. Table 1 summa-
rizes the various interactions that can occur in a
material and the effectiveness of MAS in averaging
out these interactions. As already mentioned above,

in biological materials the (dia)magnetic susceptibi-
lity is the main source of the line broadening
observed in stationary samples.

Slow-MAS Techniques

In solid-state NMR, several methods have been
developed where slow MAS is combined with spe-
cial RF pulse sequences to suppress the spinning
sidebands or to separate them from the isotropic
spectrum so that a sideband-free high-resolution
isotropic spectrum is obtained. Examples of such
methodologies are total suppression of spinning side-
bands (TOSS), phase-adjusted spinning sidebands
(PASS), and phase-corrected magic angle turning
(PHORMAT). While TOSS cannot be used at low
spinning speeds because of serious spectral distor-
tions, it was found that PASS and PHORMAT can be
modiﬁed successfully for studies of biological sam-
ples at low speeds, an order of magnitude or more
lower than the speeds typically used in solid-state
NMR experiments. The basic RF pulse sequences
used in SP-MAS, PASS, and PHORMAT are shown
in Figure 3. In the following paragraphs PASS and
PHORMAT will be described briefly (SP-MAS is
self-explanatory).

PASS

PASS is a one-rotor-period (T

), constant evolution

time 2D experiment, during which ﬁve p pulses are
applied, with time intervals tm

–tm

(Figure 3B). In

PASS, the center-band spectrum and the SSB spectra
are separated by order. This is achieved by acquiring

acq

t m

(A)

(B)

(C)

Figure 3

Three RF pulse sequences used in combination with MAS. The 90

1 pulses are black while the 1801 pulses are gray: (A) SP-

MAS; (B) PASS; (C) the prototype PHORMAT. The various timing parameters are explained in the text. (Reprinted with permission
from Hu JZ and Wind RA (2002) The evaluation of different MAS techniques at low spinning rates in aqueous samples and in the
presence of magnetic susceptibility gradients. Journal of Magnetic Resonance 159: 92–100;

& Elsevier)

Table 1

The impact of magic angle spinning on the various spin

interactions playing a role in NMR

Interaction

Impact of MAS

Indirect spin–spin or J-coupling

Static spin–spin dipolar coupling

Isotropic chemical shift

Static anisotropic chemical shift

First-order quadrupolar coupling

Second-order quadrupolar coupling

Isotropic susceptibility

Anisotropic susceptibility

Spin–lattice relaxation time T

Intrinsic spin–spin relaxation time T

Y: MAS averages the interaction to zero, N: MAS does not affect

the interaction, P: MAS partially averages the interaction.

If determined by interactions rendered time-dependent by

molecular motions.

336

NMR SPECTROSCOPY APPLICATIONS

/ Proton NMR in Biological Objects Subjected to Magic Angle Spinning

the signal after a series of PASS experiments with
different values of the time intervals tm

–tm

. Each

combination of time intervals has been chosen in
such a way that the contribution of the signal in the
observed free induction decay (FID) in the acquisi-
tion dimension (t

), arising from the center band and

SSBs, is proportional to a phase factor given by
exp(

ikY), where k denotes the sideband order and

is a variable called ‘pitch’. Then after 2D Fourier

transform with respect to t

and y a series of spectra

is obtained that separates the contributions for each
k value, i.e., it separates the center-band and side-
band spectra. In practice, it sufﬁces to use n discrete
values of y, varying from 0 and 2p in steps of 2p/n,
where n denotes the total number of center-band and
sideband spectra that have to be resolved. Figure 4A
shows the stacked

H PASS water-suppressed spectra

obtained on excised rat liver in a 7 T ﬁeld using a
spinning speed of 40 Hz. Water suppression was ac-
hieved by preceding the PASS sequence by a DANTE
sequence. In this PASS experiment, 16 different com-
binations of delay times tm

–tm

are used, which

makes it possible to separate the center-band and 15
sideband spectra without spectral aliasing. Figure 4B
shows the center-band spectrum separately. It follows
that the spectral resolution is at least the same as that
obtained in the standard fast SP-MAS experiment
shown in Figure 1D, and similar results have been
obtained in other excised tissues and organs.

PHORMAT

PHORMAT is a regular 2D experiment, with a
variable evolution time and a (ﬁxed) acquisition time
(Figure 3C). The PHORMAT methodology is based
on the so-called magic angle hopping (MAH) experi-
ment, where the sample is hopped over angles of
120

1 about an axis at the magic angle. In PHOR-

MAT, the sample is spun slowly and continuously
instead of hopped. The prototype PHORMAT se-
quence is shown in Figure 3C. The parameter T

de-

notes the rotation period of the sample, t

is the

variable evolution time, and t

is the acquisition

time. The 90

1 pulses labeled r

, r

, and r

are

synchronized to 1/3 of the rotor period, and rotate
the magnetization into the transverse plane. Then
during the evolution periods, t

/3, the magnetization

precesses through angles F

, F

, and F

, respectively.

The 90

1 pulses labeled p

, p

, and p

are the storage

pulses, which project a component of the precessing
magnetization after the corresponding t

/3 period to

the z-axis, where it remains during the storage pe-
riods labeled L. A FID is acquired following the last
90

1 pulse (r

). For local ﬁelds arising from second-

rank interactions, such as interactions with the

magnetic susceptibility ﬁelds, the sum of the preces-
sion angles F

, F

, and F

averages to the isotropic

values of the interactions, i.e., F

þ F

iso

. With a proper phase cycling of the projection

pulses, p

, p

and the receiver, the FID can be

expressed as

FID

ðt

; t

Þ ¼ expðio

iso

ÞFIDðt

½4

After Fourier transformation as a function of t

and

then as a function of t

a pure absorption-mode 2D

spectrum is obtained. Figure 4C shows an example of
water-suppressed

H PHORMAT metabolite spectra

of excised rat liver tissue, obtained at a MAS spin-
ning frequency of 1 Hz and in a 7 T ﬁeld. Figure 4C
displays the 2D plot together with the projections
along the isotropic F

) and anisotropic F

) di-

mensions. By making slices parallel to the F

axis, the

anisotropic line shapes of each isotropic peak can be
determined separately, nine of which are plotted in
Figure 4D. In this way the susceptibility gradients
surrounding individual metabolites can be deter-
mined, which could be of diagnostic value. The spec-
tral line widths in the isotropic projection are
comparable to, albeit slightly larger than, the widths
obtained with 4 kHz SP-MAS (Figure 1D) and 40 Hz
PASS (Figure 4B). As will be discussed in the next
section, this slight increase (a few hertz) is attributed
to the diffusion of the metabolites in the susceptibi-
lity gradients.

Limitations of SP-MAS, PASS,
and PHORMAT

The question arises to what values the MAS fre-
quency can be reduced with the various MAS meth-
odologies, as this will determine the size and type of
biological object that can be studied with these tech-
niques. The answer depends on the MAS methodo-
logy used and the NMR properties of the magnetic
nuclei under investigation. Moreover, the NMR sen-
sitivity per unit measuring time and the total
measuring time itself are different for the various
techniques. Figure 5 shows 300 MHz

H SP-MAS,

PASS center-band, and PHORMAT isotropic spectra
of a mixture of water and spherical glass beads with
diameters of

B230 mm as a function of the spinning

frequency F. The susceptibility difference between the
beads and water, broadened the static water line to
3.7 kHz (12.5 ppm), at least a factor 20 larger than
the line widths observed in biological systems. For
SP-MAS it follows that F has to be at least of the
order of the static line width in order to reduce the
SSBs. In contrast, it follows from Figure 5 that PASS
produces a nearly sideband-free isotropic chemical

NMR SPECTROSCOPY APPLICATIONS

/ Proton NMR in Biological Objects Subjected to Magic Angle Spinning

337

shift spectrum at a spinning frequency as low
as 50 Hz. (The peaks marked by the symbol ‘

’ in

Figure 5 are aliased sidebands, arising because only
32 evolution increments were used to acquire the

PASS data, whereas at this spinning rate the spectrum
contained

B90 visible sidebands, requiring at least

90 increments.) Moreover, the isotropic line width
observed at 50 Hz is similar to that measured at

ppm

(ppm)

k = +4

k = 0

k =

−4

k =

−8

(A)

(B)

(C)

(D)

Figure 4

Top: The water suppressed

H 2D-PASS spectra of freshly excised rat liver acquired at a spinning rate of 40 Hz. The

external ﬁeld was 7 T. (A) The stacked 2D plot. The parameter k denotes the kth sideband; k

¼ 0 corresponds to the center band; (B)

The center-band spectrum. Bottom: The water suppressed

H PHORMAT spectra of a freshly excised rat liver sample, acquired at a

sample spinning rate of 1 Hz. The external ﬁeld was 7 T; (C) The contour plot of the 2D PHORMAT spectrum along with its anisotropic
(F

) and isotropic (F

) projections. The F

projection was obtained by summing only the data inside the dotted box; (D) The stacked plot

of the anisotropic line shapes corresponding to nine isotropic peaks, which were obtained by taking a slice parallel to the F

axis at the

center of each isotropic peak (F

). ((A) and (B) are reproduced with permission from Wind RA and Hu JZ (2003) Magnetic susceptibility

effects in nuclear magnetic resonance spectroscopy of biological objects. In: Recent Research Developments in Magnetism and
Magnetic Materials, vol. 1, pp. 147–169. (C) and (D) are reproduced with permission from Hu JZ, Rommereim DN, and Wind RA
(2002) High-resolution

H NMR spectroscopy in rat liver using magic angle turning at a 1 Hz spinning rate. Magnetic Resonance in

Medicine 47: 829–836.)

338

NMR SPECTROSCOPY APPLICATIONS

/ Proton NMR in Biological Objects Subjected to Magic Angle Spinning

higher spinning frequencies. However, at this speed
the signal is seriously attenuated. This is caused by
the decay of the magnetization during the rotor pe-
riod T

prior to the acquisition (cf. Figure 3B).

During this period the magnetization dephases in
the transverse plane, and is partially refocused by the
180

1 pulses. Hence, the time constant governing the

decay is approximately given by the intrinsic spin–
spin relaxation time T

, which means that the mini-

mum spinning frequency should be larger than
(T

)

, i.e., the isotropic line width, in order to

avoid serious signal losses. This has as a consequence
that for biological samples, where the minimum

metabolite T

values are of the order of 30 ms, with

PASS spinning frequencies of

B40 Hz or larger

should be used.

The T

attenuation is avoided in a PHORMAT

experiment. Here, the magnetization is stored paral-
lel to the main ﬁeld direction with a maximum
duration of 2/3 times the rotor period, which means
that the spinning frequency has to be large compared
with the spin–lattice relaxation rate (T

)

of the

spins rather than (T

)

in order to avoid signal

attenuation. Hence, for the water/bead mixture,
where the water T

is several seconds, the spinning

speed can be made as low as 1 Hz without causing

SP-MAS

1 kHz

250 Hz

50 Hz

1 Hz

−10

ppm

−10

ppm

−10

ppm

PASS

PHORMAT

∗

Figure 5

H MAS spectra obtained at different spinning rates by SP-MAS, PASS, and PHORMAT on a mixture of H

O and spherical

glass beads with diameters of 230

720 mm. The PASS spectra are the center-band spectra, the PHORMAT spectra are the isotropic

projections. (Reprinted with permission from Hu JZ and Wind RA (2002) The evaluation of different MAS techniques at low spinning
rates in aqueous samples and in the presence of magnetic susceptibility gradients. Journal of Magnetic Resonance 159: 92–100;
& Elsevier.)

NMR SPECTROSCOPY APPLICATIONS

/ Proton NMR in Biological Objects Subjected to Magic Angle Spinning

339

serious signal attenuation, and the same is true for
biological samples, where the metabolite T

values

are at least an order of magnitude larger than the
T

values. However, it follows from Figure 5 that

at ultralow spinning speeds the isotropic line
width increases. This is due to the diffusion of the
water molecules in the susceptibility gradients. This
diffusion-induced broadening, which cannot be elimi-
nated with MAS techniques, is proportional to
G

ﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ

, where G is the susceptibility gradient, D

is the diffusion coefﬁcient, and F is the MAS fre-
quency. It follows that this broadening is much less
in biological samples, where both G and D are at
least an order of magnitude less than in the water/
bead mixture. Comparing the SP-MAS, PASS, and
PHORMAT experiments of the excised rat liver
shown above, it was estimated that the diffusion-
induced line broadening in a 1 Hz PHORMAT ex-
periment is

B2 Hz at 7 T, considerably less than the

intrinsic line width.

Finally, it is worth noting that compared to PASS,

PHORMAT has a considerably lower NMR sen-
sitivity and often requires a longer measuring time.
The sensitivity loss is mainly due to an intrinsic loss
of a factor of 4 resulting from the use of two storage
pulses p

and p

in the PHORMAT sequence, cf.

Figure 3C (the storage pulse p

is omitted in a regular

PHORMAT experiment). Also, PHORMAT often
requires a relatively large number of evolution steps,
resulting in a long measuring time, an hour or more.
Although the performance of PHORMAT can be
improved, e.g., by applying multiple-echo acquisi-
tion, PASS should be considered as the method of
choice if the sample can tolerate spinning speeds of
tens of hertz. PHORMAT should be reserved for re-
search in larger biological samples, including in vivo
applications.

In Vivo PHORMAT

Two effects associated with spinning a live animal
can cause harm: the centrifugal forces induced in the
animal, and the effects induced by the external
magnetic ﬁeld, rendered partially time-dependent
from the rotating animal’s point of view when spin-
ning inside the magnet. This time-dependent ﬁeld
dB/dt can cause nerve and cardiac stimulations in a
similar way as pulsed ﬁeld gradients. However, both
effects can be small in a PHORMAT experiment. For
instance, when a mouse is placed in a cylinder with a
1 cm radius, and rotated around the magic angle in a
2T ﬁeld and with a spinning speed of 1 Hz, the maxi-
mum centrifugal force at the periphery of the animal
is 0.04 times the gravitational force F

, about two

orders of magnitude below the forces applied in

chronic centrifuging experiments, and dB/dt is
B10 Ts

, well below the threshold of

B90 Ts

for which nerve stimulations have been reported. In
fact, 15 mice were spun in a 2 T ﬁeld at frequencies
up to 8 Hz and for durations up to 60 min without
causing any apparent short-term or long-term health
effects. Hence, it is possible to use ultraslow PHOR-
MAT for in vivo studies on live animals. Figure 6
shows the ﬁrst result of such an experiment, obtained
on the middle section of the body of an (anesthetized)
female BALBc mouse between the dotted lines shown
in Figure 6A and the arrows shown in Figure 6B.
Figures 6C and 6D shows the

H 85 MHz water-

suppressed metabolite spectra obtained in a station-
ary mouse and a mouse subjected to 1.5 Hz MAS,
respectively. Even in this relatively low ﬁeld a signi-
ﬁcant increase in spectral resolution is obtained.

Future Perspectives

It can be concluded that it is possible to signiﬁcantly
increase the resolution in the proton NMR metabo-
lite spectra in intact biological samples by PASS and
PHORMAT. It was found that for small samples,
with sizes of a few millimeter or less, where spinning
speeds of 40 Hz or more can be tolerated, PASS is the
method of choice because of its superior sensitivity
and short measuring time. For larger biological sam-
ples, including animals, ultraslow-MAS PHORMAT,
allowing spinning speeds as low as 1 Hz, has to be
used. In a 7 T ﬁeld with PASS the spectral line widths
are reduced by an order of magnitude or more to
values determined by the intrinsic T

, originating

from the various spin–spin dipolar interactions, ren-
dered time-dependent by the molecular motions.
With 1 Hz PHORMAT slightly larger isotropic line
widths are observed, resulting from the molecular
diffusion in the susceptibility gradients. Increasing
the spinning speed, if allowed, will reduce this con-
tribution. In external ﬁelds larger than 7 T the line
width reductions are expected to become even larger,
as the residual isotropic line widths are essentially
ﬁeld-independent, whereas the susceptibility broad-
ening increases linearly proportional to the ﬁeld.
Hence, with PASS or PHORMAT the full beneﬁts of
high-ﬁeld NMR are obtained.

It is worth noting that the slow-MAS methodology

is a new area of research, and that several further
improvements are necessary to make PASS and
PHORMAT more viable methods for biochemical
and biomedical research. Particularly, specific com-
binations of RF and pulsed-ﬁeld-gradient sequences
need to be developed so that PASS or PHORMAT
spectra can be obtained of a select volume in the
sample or the animal rather than the whole sample or

340

NMR SPECTROSCOPY APPLICATIONS

/ Proton NMR in Biological Objects Subjected to Magic Angle Spinning

the main part of the body. If these implementations
are successful, it can be expected that PASS and
PHORMAT will signiﬁcantly increase the utility of
proton metabolite NMR spectroscopy for biochem-
ical and biomedical studies in cells, tissues, organs,
and animals.

Finally, the question arises whether the slow-MAS

approach will ultimately reach the hospital to
investigate patients. Obviously in this case PHOR-
MAT is the only candidate. Although spinning a pati-
ent at a frequency of (e.g., 1 Hz) may cause no phys-
ical harm, it is likely that it will induce unacceptable
stress in many patients. An alternative approach
might be to rotate the external magnetic ﬁeld instead
of the patient, either mechanically or electronically,
or by rotating both the magnetic ﬁeld and the patient
in opposite directions. In this way the spinning
speed of the patient can be reduced. The future will
tell whether this approach will become a viable
option.

See also: Nuclear Magnetic Resonance Spectroscopy:
Overview; Principles. Nuclear Magnetic Resonance Spec-
troscopy Applications: Pharmaceutical. Nuclear Magne-
tic Resonance Spectroscopy Techniques: Solid-State.