A slow relaxing species for molecular spin devices: EPR characterization of static and dynamic magnetic properties of a nitronyl nitroxide radical

Article abstract of DOI:10.1039/c2jm35076a

Nitronyl nitroxides (NitR) are a family of persistent radicals widely used in molecular magnetism and recently suggested as potential candidates for spintronic applications. In this paper we characterize by X- and W- band Electron Paramagnetic Resonance (EPR) spectroscopy the new radical S-4-(nitronyl nitroxide) benzyl ethanethioate (NitSAc) designed for assembling on Au surfaces. We determined the radical magnetic tensors and studied by X-band pulse EPR its spin relaxation behaviour in fluid and glassy solutions of toluene. A comparison with the well known nitroxide 3-carbamoyl-2,2,5,5-tetramethyl-3- pyrrolin-1-oxyl (CTPO) is afforded. The advantages of using NitSAc in technological applications are discussed on the basis of the slow spin relaxation demonstrated by this study.

Full text of DOI:10.1039/c2jm35076a

View Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links <
C
Journal of
Materials Chemistry
Cite this: J. Mater. Chem., 2012, 22, 22272
www.rsc.org/materials
PAPER
A slow relaxing species for molecular spin devices: EPR characterization of
static and dynamic magnetic properties of a nitronyl nitroxide radical†
ac
b
A. Collauto,a M. Mannini,b L. Sorace,b A. Barbon,ac M. Brustolon
and D. Gatteschi
*
*
Received 30th July 2012, Accepted 4th September 2012
DOI: 10.1039/c2jm35076a
Nitronyl nitroxides (NitR) are a family of persistent radicals widely used in molecular magnetism and
recently suggested as potential candidates for spintronic applications. In this paper we characterize by
X- and W- band Electron Paramagnetic Resonance (EPR) spectroscopy the new radical S-4-(nitronyl
nitroxide) benzyl ethanethioate (NitSAc) designed for assembling on Au surfaces. We determined the
radical magnetic tensors and studied by X-band pulse EPR its spin relaxation behaviour in fluid and
glassy solutions of toluene. A comparison with the well known nitroxide 3-carbamoyl-
2,2,5,5-tetramethyl-3-pyrrolin-1-oxyl (CTPO) is afforded. The advantages of using NitSAc in
technological applications are discussed on the basis of the slow spin relaxation demonstrated by
this study.
approach makes them addressable using appropriate tech-
niques.20–22 On the other hand, they suffer from the same limi-
tations, for use in QIP, as any other potential electron spin qubit,
namely a stronger coupling with the environment which makes
decoherence related issues much more difficult to solve than in
e.g. photon based QCs.13 The absolute need for a deep charac-
terization of the relaxation dynamics of these systems in different
environments, trying to understand which factors are crucial in
determining their spin relaxation behaviour, is then clear. This
would then allow identifying the degrees of freedom that affect
more the relaxation processes, and on this basis to fine tune the
properties of the system by appropriate chemical modification.
This approach has been recently undertaken for chromium ring
based molecules,23 among the most relevant candidates for QC
applications of molecular magnets.24,25 It is worth noting that
such a deep characterization of the relaxation dynamics of
organic radicals may promote the routine use of NitR radicals in
more traditional applications, e.g. as spin probes or spin labels in
advanced Electron Paramagnetic Resonance (EPR) techniques.26
Among the vast family of NitR organic radicals we have
recently undertaken a full characterization by cw-EPR of p-
methyl-thio-phenyl-nitronyl nitroxide (MTPNN), both in
isotropic27 and ordered nematic phases,28 by exploiting an
innovative modeling strategy alternative to the usual spectral
simulation approach.29 This study was intended to open the way
to the characterization of EPR spectra of Self-Assembled
Monolayers (SAMs) of the same radical, in particular under the
aspect of the ordering and hindered mobility associated with the
bidimensional quasi-regular structure of the monolayer
anchored to a solid substrate. In this perspective, the research
activity of some of us in the assembling of NitR on surfaces led us
to tailor their chemical structure by appropriately selecting the
linking group which promotes the selective interaction of the
Introduction
Nitronyl nitroxides (NitR) are well-known stable radicals in
which the unpaired electron is essentially delocalized over two
equivalent NO groups.1,2 They have interesting properties, since
their ability as bridging ligands towards metal ions makes them
suitable as building blocks in designing molecular based
magnetic materials,3,4 and the strong p spin polarization favours
a direct ferromagnetic exchange between neighbouring molecules
in the solid state. A molecular crystal of a nitronyl nitroxide was
the first reported example of a purely organic ferromagnet.5
These radicals are also widely employed to build families of both
single chain magnets and single molecule magnets.6–11 More
recently, their use has been suggested as elementary molecular
two level systems to be used in molecular spin based Quantum
Information Processing (QIP).12 Among the main advantages of
these systems compared to other potential implementation of
Quantum Computers (QC), one has to consider that qubit scal-
ability can be achieved by appropriate synthetic strategy.13 The
use of organic radicals has further been suggested to present
some advantages compared to the use of transition metal based
clusters,14–19 as they do not contain metal centers which may
represent an intrinsic limitation for qubits encoding due to their
single ion anisotropy. More interestingly, organic radicals often
show a simple and flexible chemistry, which can be used to
promote grafting and organization in nanostructures; this
aDipartimento di Scienze Chimiche, Universitaꢀ degli Studi di Padova, Via
Marzolo, I-35131 Padova, Italy. E-mail: marinarosa.brustolon@unipd.it
bDipartimento di Chimica ‘‘Ugo Schiff’’, Universitaꢀ degli Studi di Firenze &
INSTM RU Firenze, Via della Lastruccia 3, 50019 Sesto Fiorentino, FI,
Italy. E-mail: dante.gatteschi@unifi.it
cINSTM RU Padova, Via Marzolo, I-35131 Padova, Italy
† Paper dedicated to the memory of Prof. Giacomo Martini.
22272 | J. Mater. Chem., 2012, 22, 22272–22281
This journal is ª The Royal Society of Chemistry 2012

View Online
molecules with a gold surface. To avoid problems related to the
chemical instability of NitR in the presence of free thiols, a
methyl-sulphide functionalization was originally chosen. This
strategy allowed the optimization of the spacer group between
the surface and the radical in order to achieve the highest control
of the organization on the surface.20 However, the use of this
linker suffered from the ambiguity about the way alkyl-sulphides
anchor on gold, i.e. with or without a C–S cleavage.30–32 To avoid
this problem and to improve further the assembling of NitR on
gold, a possible solution33 is the use of a thio-acetyl function as a
linker group, by preparing the S-4-(nitronyl nitroxide) benzyl
ethanethioate radical (hereafter NitSAc).
In this paper we describe a complete characterization of the
EPR-related properties of NitSAc with a particular emphasis on
its spin relaxation behaviour, both in fluid and glassy solutions
of toluene. As previously stated, a thorough characterization of
the EPR-related properties of the radical, with a special
emphasis on the determination of the magnetic tensors, is
required in order to interpret the spectra obtained from SAMs.
We wish to show how our multifaceted approach can be a first
step to describe the magnetic properties of NitRs, both static
(magnetic tensors) and dynamic (relaxation rates), in view of
their potential applications either as spin probes/spin labels in
EPR spectroscopy or as fundamental units of QC exploiting
pulse EPR techniques. We will also characterize in the same
experimental conditions a nitroxide radical, CTPO (3-carba-
moyl-2,2,5,5-tetramethyl-3-pyrrolin-1-oxyl), to compare, in
particular, the relaxation rates of the two radicals; in this respect
it is worth noting that – among other nitroxides characterized by
similar size and magnetic tensors – CTPO is characterized by a
quite rigid structure, and it is expected to have rather slow
relaxation rates as the impact of internal motions on the electron
spin relaxation should be small.
Scheme 1 Structural formulae of the radicals S-4-(nitronyl nitroxide)
benzyl ethanethioate (NitSAc) and 3-carbamoyl-2,2,5,5-tetramethyl-3-
pyrrolin-1-oxyl (CTPO).
equipped with a Bruker ER4118X-MD5 dielectric ring resonator
inside a cryostat (CF935 Oxford Instruments) cooled by nitrogen
flow for variable temperature operation. CW-EPR spectra in the
fluid phase were recorded at a microwave power of about 20 mW;
the modulation frequency was set to 100 kHz with an amplitude
of 0.1 G. Higher resolution spectra were recorded setting the
modulation frequency to 3.5 kHz with an amplitude of 25 mG;
these parameters allowed us to avoid spectral line shape distor-
tion at the expense of the signal-to-noise ratio. In frozen solution,
much broader linewidths allowed us to use a modulation
amplitude of 1 G.
In pulse EPR experiments the primary echo was generated
with a standard p/2–s–p–s–echo sequence; the two-pulse echo
decay (or Hahn decay) traces were recorded by monitoring the
top of the echo on varying the delay s between the pulses,
whereas the echo-detected EPR (ED-EPR) spectra were obtained
by integrating the whole primary echo while sweeping the
magnetic field for a fixed delay between the pulses. For this last
experiment soft pulses (p/2 ¼ 48 ns) were used to improve
resolution, whereas for the echo decay experiment soft and hard
pulses (p/2 ¼ 8 ns) were used in frozen solution and in fluid
solution, respectively (see Results).
Experimental
All starting materials were purchased from Aldrich. High
purity solvents were used without additional purification
steps. The radical was obtained by following the procedure
previously reported by Gorini,34 which we briefly summarize in
the following: NitSAc (Scheme 1) was synthesized starting from
the coupling of S-4-acetylbenzyl ethanethioate with 2,3-di(hy-
droxyamino)-2,3-dimethylbutane exploiting
a TPAP-NMO
based strategy to avoid overoxidation problems.35 The obtained
product, purified by flash column chromatography, yielded a
blue microcrystalline powder of NitSAc (yield 56%). Anal.
calc. for C16H21N2O3S C 59.78, H 6.58, N 8.72. Found: C
61.12, H 6.07, N 9.52%. IR (KBr, cmꢀ1): 2964–2852 (nC–H, m),
1688 (nC]O, s), 1366 (nN–O, s), 1132–1100 (nS–C, m), 802
(nbCe–nHding, m).33
The inversion recovery experiments were performed with the
usual p–T–p/2–s–p–s-echo scheme upon varying the delay T; the
time traces were obtained by integrating the whole echo in
the fluid phase or monitoring just the top in frozen solution. In
the detection echo sequence the pulses were the same as in the
Hahn decay. p-pulses were obtained by doubling the duration of
the p/2 pulse.
The radical was then dissolved in anhydrous toluene to a
concentration of 50 mM and put into a 2 mm (i.d.) quartz tube.
The samples were carefully deaerated by 3–4 freeze–pump–thaw
cycles, and then sealed under vacuum. The same procedure was
used to prepare a sample of the CTPO (Aldrich) nitroxide spin
label (Scheme 1). The absence of other radicals was verified
by EPR.
The CW-ENDOR experiments were performed in solution
with a higher concentration of 370 mM using a Bruker ER200D-
SRC X-band EPR spectrometer equipped with a Bruker EN801
ENDOR resonator in a homemade set-up for the radio
frequency (RF) control.28 The spectrum was recorded at 270 K
by monitoring the central line of the cw-EPR spectrum; the
microwave power was set to 10 mW and the RF modulation
frequency was 12.5 kHz with an amplitude of 25 kHz.
Continuous wave (CW) and pulse X-band EPR experiments
were performed using a Bruker Elexsys E580 spectrometer
This journal is ª The Royal Society of Chemistry 2012
J. Mater. Chem., 2012, 22, 22272–22281 | 22273

View Online
W-band EPR spectra were recorded using a Bruker E600
spectrometer equipped with a 6 T split coil magnet (Oxford
Instruments) and a continuous flow He cryostat to work at low
temperatures. The microwave power was set to 50 dB (50 nW) to
exclude saturation effects, which were checked for by measuring
spectra at higher power up to 30 dB (5 mW), where saturation
starts to set in.
The starting magnetic parameters for the fitting of the exper-
imental spectra were obtained by DFT calculations. The prin-
cipal values and the principal directions of the Zeeman tensor
and of the hyperfine coupling tensors for the two nitrogen nuclei
were obtained by QM/DFT calculations with the G09 package36
using the PBE0 hybrid density functional and the 6-31+G(d,p)
basis set. Solvent effects on both the minimum energy confor-
mation and the magnetic tensors were taken into account using
the implementation of the polarizable continuum model (PCM37)
provided by G09. In the fitting procedure we considered the
hyperfine structure due to the protons as giving an inhomoge-
neous broadening of the 14N hyperfine components; the corre-
sponding inhomogeneous linewidth was allowed to be
orientation dependent, to take into account the appreciable
anisotropy of the hyperfine splitting of the two phenyl ortho
protons, which shows a maximum along the y principal axis of
the g tensor.28
Fig. 2 CW-EPR spectrum of a 50 mM solution of NitSAc in toluene at
270 K recorded for the MI ¼ 0 hyperfine component under high reso-
lution conditions.
constants: 0.47 G (2 equivalent protons), 0.28 G (2 equivalent
protons) and 0.21 G (12 equivalent protons). The three hyperfine
coupling constants (hfccs) are attributed respectively to the two
pairs of phenyl protons in ortho and meta positions with respect
to the nitronyl nitroxide residue, and to the twelve equivalent
methyl protons.38 These assignments are confirmed by the anal-
ysis of the CW-ENDOR spectrum, shown in Fig. 3: simulation of
the spectrum allowed us to extract two proton hfccs with
constants of 1.31 MHz (0.47 G) and 0.58 MHz (0.21 G). The two
lines relative to two protons with hfcc of 0.28 G are not detect-
able as they are completely covered by the pair of strong lines
relative to the twelve methyl protons. The two lines observed at
the lower frequency, due to 14N nuclei, are centered at aN/2, and
separated by twice the free 14N nuclear Larmor frequency nN.
The value of aN (7.45 G) is in agreement with the value obtained
from the nitrogen nuclei hyperfine pattern in Fig. 1 and 4, see
below.
Results
Spectral properties
Fluid solution. CW-EPR spectra (Fig. 1) of the diluted (50 mM)
solution of NitSAc in toluene were obtained at various temper-
atures in the interval 180–300 K. They show the typical
1 : 2 : 3 : 2 : 1 intensity pattern arising from the coupling of the
unpaired electron with two equivalent 14N (I ¼ 1) nuclei. In the
following we will refer to these hyperfine components by their
total nuclear spin quantum numbers MI ¼ mI1 + mI2. Each of the
five MI lines is further split into a number of components due to
hyperfine interaction with protons (Fig. 2). This hyperfine
structure consists of a complex pattern whose resolution,
dependent on linewidth, varies with the temperature and with the
14N hyperfine component. We have analyzed the central line
(MI ¼ 0) for T ¼ 270 K, showing the best resolution of the proton
splitting. A best fit simulation (see Fig. 2) gives three coupling
We obtained very intense ED-EPR spectra in fluid solution in
the range 180–300 K by using soft pulses (p/2 ¼ 48 ns) and with
an inter-pulse delay of 500 ns, see Fig. 4.
The five lines, separated by aN, do not show the proton
hyperfine structure as the experimental conditions were not
selective enough. The intensities of the lines are weighted by the
different decays of the electron spin echo, which are dependent
Fig. 3 CW-ENDOR spectrum of a 370 mM solution of NitSAc in
toluene recorded at 270 K; the two lines at lower frequency are attributed
to the 14N ENDOR transitions.
Fig. 1 CW-EPR spectra of a 50 mM solution of NitSAc in toluene
recorded in the fluid phase.
22274 | J. Mater. Chem., 2012, 22, 22272–22281
This journal is ª The Royal Society of Chemistry 2012

View Online
Fig. 6 CW-EPR spectrum of a 50 mM solution of NitSAc in toluene
recorded at W-band in the glassy matrix.
Fig. 4 ED-EPR spectra of a 50 mM solution of NitSAc in toluene
recorded in the fluid phase with s ¼ 500 ns.
tilted with respect to the g-frame (see Table 1); the relative tilt
angle obtained between the two z principal directions of the
nitrogens is slightly larger than 9ꢁ.
on the nuclear spin configuration (mI1,mI2) of the two 14N nuclei
contributing to MI. This will be discussed further below.
The same magnetic parameters resulting from the global fit
also led to a satisfactory reproduction of the ED-EPR spectra
shown in Fig. 7. These latter spectra were recorded in order to
get information on the dependence of the phase memory time
TM on the resonance magnetic field and, therefore, on the
residual mobility of the paramagnetic probe40 also at temper-
atures giving rigid limit CW-EPR spectra. ED-EPR spectra
show effects on the spectral profile upon varying the inter-pulse
delay s down to 80 K. This observation is in agreement with
the ED-EPR studies on nitroxides in glassy solids,40 where the
orientation dependent relaxation has been attributed to low-
amplitude fast librational motions of the probe in the matrix.
We used a simple model for such a motion, considering it as a
jump of the radical between two orientations that are produced
by a rotation around an axis d.41 Under the fast motion
condition the relaxation rate due to the librational motion is
given as41
Glassy matrix. CW-EPR spectra at X-band (Fig. 5) were
recorded in the range 80–150 K. The spectrum does not show
appreciable changes up to 120 K; for higher temperatures, slight
alterations of the spectral profile indicate a residual mobility of
the radical.
To improve the reliability of the results obtained by simula-
tion, a rigid limit spectrum (100 K) at W-band was also recorded
(Fig. 6), showing a clear resolution of the g-tensor anisotropy.
Spectra at both frequencies were simulated by the EasySpin
software package.39 A global best fit procedure was performed by
a simultaneous fit of the spectra at both frequencies; the fit
required a refining of the principal values and the principal
directions of the Zeeman tensor and of the hyperfine coupling
tensors for the two nitrogen nuclei taking the starting values
from the output of a QM/DFT calculation (see Experimental); in
the refinement process the principal values of the hyperfine
tensors were kept equal. The QM calculation shows that a tilt
angle 4 between the z principal axes of the nitrogen hyperfine
tensors and the g-tensor is present. Refinement of the tensors to
fit the experimental spectra shows, furthermore, that the z prin-
cipal directions of the two nitrogen tensors are not symmetrically
1
2
¼ sL½DuðmI1; mI2; U; FÞꢂ
(1)
TMðmI1; mI2; U; FÞ
where U is the set of Euler angles defining the director of the
magnetic field in the molecular frame, F is the librational angle
specifying the displacement of the director of the magnetic field,
Du(mI1,mI2,U,F) is the difference between the resonance
frequencies corresponding to the two orientations for a given
nuclear spin configuration (mI1,mI2), and sL is the correlation
time for the exchange process.
Table 1 Principal values of g, nitrogen hyperfine (AN) magnetic tensors
and 14N isotropic hfccs. The tilt angle 4 between the z principal directions
of the AN and that of the g tensors is reported
Magnetic tensor/unit
Principal values
4
g
2.01111, 2.00671, 2.00211
2.00663
—
hgi ¼ 1/3Tr{g}
ꢁ
ꢁ
A
N1/G
0.01, 1.61, 18.61
6.73
7.455
7.75
hAN1i ¼ 1/3Tr{AN1}/G
a0,N1 (fluid solution)/G
A
N2/G
0.01, 1.61, 18.61
6.73
7.455
1.65
Fig. 5 CW-EPR spectra of a 50 mM solution of NitSAc in toluene
recorded in the glassy matrix. For the 80 K spectrum the power was
reduced to 2 mW to avoid microwave power saturation.
hAN2i ¼ 1/3Tr{AN2}/G
a0,N2 (fluid solution)/G
This journal is ª The Royal Society of Chemistry 2012
J. Mater. Chem., 2012, 22, 22272–22281 | 22275

View Online
Fig. 7 ED-EPR spectra of a 50 mM solution of NitSAc in toluene
recorded at 80 K. Soft pulses (p/2 ¼ 48 ns) were used with s ¼ 200 ns
(uppermost trace), s ¼ 1000 ns and s ¼ 2500 ns (lowermost trace). The
arrow denotes the magnetic field position where relaxation measurements
were performed.
The best fit between experimental and calculated spectra
(Fig. 7) was obtained leaving as free parameters the orientation
of the axis d and the amplitude a of the rotation angle, on the
basis of typical correlation times sL for low-amplitude librational
motions.42 The calculations were performed using a home-
written software, which evaluates the resonance field shift due to
the libration making use of second-order perturbation
theory.41,43 The best fitting between calculated and experimental
profiles was found for an orientation of the axis d tilted by 24ꢁ
from the z towards the y axis of the g tensor eigenframe (see
Fig. 8) and a ¼ 2.5ꢁ assuming sL ¼ 30 ps.42
Fig. 8 Magnetic tensor eigenframes (blue: x axis; green: y axis; red: z
axis); based upon the local symmetry, the x axes of the nearly axial
nitrogen hyperfine coupling tensors have been oriented parallel to the N–
O bonds.
Relaxation properties
In the following we present the results of the relaxation proper-
ties of NitSAc and of the nitroxide radical CTPO for
comparison.
Table 2 Linear extrapolation to null concentration (T02) of the T2 values
obtained as a function of the concentration for the five MI components of
NitSAc in toluene at T ¼ 292 K
Fluid solution. The transverse spin relaxation time T2 at 292 K
was measured by Hahn echo decay (p/2 ¼ 48 ns) in a series of
fluid solutions of NitSAc at decreasing concentrations in order to
separate the contribution from the inter-radical Heisenberg
exchange. The relaxation rates show a linear dependence on the
concentration of the radical with non-zero intercept; the recip-
rocals of the intercepts of the fitting lines give the T02 values (T2
values at null concentration). The five T02 values are different,
ranging from 0.5 to 1.1 ms at 292 K (see Table 2). We see that the
components with MI ¼ +1 and MI ¼ 0 have similar relaxation
times, whereas the decay of the MI ¼ ꢀ1 component is faster
and similar to that of MI ¼ +2 component (lower field line); the
MI ¼ ꢀ2 component (higher field line) has the shortest T02.
From the slope of the central line we obtain a value for
the exchange rate constant kHE ¼ (3.00 ꢃ 0.03) ꢄ 109 Mꢀ1 sꢀ1 z
200 mGpp mMꢀ1, which is quite similar to the one reported in the
literature for perdeuterated Tempone in toluene-d8 (ref. 44)
MI
T02 [ms]
+2
+1
0
0.795
1.045
1.125
0.775
0.505
ꢀ1
ꢀ2
TM) for NitSAc, the former by echo-detected inversion recovery
and the latter by Hahn decay, in fluid and in glassy toluene
phases, in the range 180–300 K and 80–120 K respectively. In the
temperature range 120–180 K it was impossible to detect the
primary echo because of the phase memory time being too short.
The relaxation measurements were also performed for CTPO
in the same solvent and concentration at 300, 180, 100 and 80 K
to compare the values of the relaxation times for the two radicals.
In fluid solutions T1 and T2 were measured for both NitSAc
and CTPO with the magnetic field at the central lines of their
EPR spectra. The inversion recovery traces were all fitted with a
monoexponential decay function. The echo decay was best fit as
(kHE ¼ (3.20 ꢃ 0.03) ꢄ 109 Mꢀ1 sꢀ1 z 210 mGpp mMꢀ1
,
extrapolated from graph). On the basis of these results we can
conclude that at the concentration of 50 mM the exchange
contribution is negligible. We measured therefore T1 and T2 (or
22276 | J. Mater. Chem., 2012, 22, 22272–22281
This journal is ª The Royal Society of Chemistry 2012

View Online
a monoexponential in the range 240–300 K and as a biexpo-
nential in the range 180–210 K (see Table 3).
As we measured T2 on the central line with hard pulses (cor-
responding to a spectral range with a half amplitude value of
about ꢃ30 MHz (ref. 45)) we expect contributions from the
neighbouring lines with MI ¼ +1 and ꢀ1. In the fluid phase at
low temperatures we find indeed a biexponential decay of the
echo, merging to a monoexponential one at T ¼ 240 K as
expected when the spin rotational contribution dominates the
relaxation for all the hyperfine components (see Discussion). An
estimate of the decay times for the neighbouring components can
be obtained by the analysis of the relative intensities of the cor-
responding lines in the ED-EPR spectra (Fig. 4). The fast
relaxing components of the biexponential fittings (see Table 3)
are consistent with this estimate, hence we attribute the compo-
nent of the biexponential decay with the longest time and the
largest weight to T2 relative to MI ¼ 0.
Fig. 9 T1 and T2 (TM) spin relaxation times; the dashed line separates
the measurements in the fluid solution (right) from that in the glassy
phase (left). In fluid solution the measurements were performed on the
central (MI ¼ 0) component of the EPR spectrum both for NitSAc and
for CTPO; in the glassy phase the magnetic field was set to the position
shown in Fig. 7 for NitSAc and to the position with the largest echo
intensity for CTPO.
In Fig. 9 we report the temperature dependence of T1 and T2 in
fluid solution for NitSAc (only the longest relaxation time in the
range 180–210 K), and the corresponding values at 180 and
300 K for CTPO (at this temperature T1 ¼ T2).
hyperfine tensors of the two 14N nuclei have different orientations
(see above). Therefore at each field one sees a superposition of
many echoes characterized, in general, by different relaxation
times (see eqn (1)). In order to disentangle the relaxation behav-
iours we should be able to excite a single type of transition. The
deconvolution of the cw-EPR powder spectrum (see Fig. 10)
shows that single transitions can be excited by selective pulses only
at its wings; however, the echo corresponding to these positions is
expected to be very weak. In order to gain intensity we set the field
position as shown by the arrow in Fig. 7. For this field, in our
experimental conditions (p ¼ 96 ns) we excite the (+1,0) and the
(0,+1) transitions for all the radicals in any orientation. The
distributions of resonance fields for these two transitions are the
narrowest ones (approximately 6 G wide), giving a quite large
signal intensity. The Hahn decay shows a biexponential behav-
iour with a slow component which is largely dominant. In Fig. 9
we report this latter component as TM in the glassy phase.
For NitSAc, T1 increases monotonically upon decreasing the
temperature, whereas T2 passes through a maximum at around
240 K and drops to very low values as the temperature gets close
to the melting point of toluene (178 K (ref. 46)).
Glassy matrix. In the glassy phase T1 and TM were measured
by the same procedures for NitSAc in the range 80–120 K and for
CTPO at 80 K and 100 K. In this phase we checked that the
contribution from instantaneous diffusion to the echo dephas-
ing47 was negligible. We used generally longer pulses (p/2 ¼
48 ns) than in fluid solutions to suppress the echo modulation due
to the interaction with nuclei (ESEEM). Both the inversion and
the Hahn decay time traces were best fit with biexponential
functions.
The magnetic field was set at the position giving the largest echo
intensity for CTPO, whereas for NitSAc its value was chosen as
shown by the arrow in Fig. 7, on the basis of the following
considerations. The powder ED-EPR spectrum of NitSAc is a
superposition of the powder spectra obtained for all the nine
nuclear spin configurations (mI1,mI2), the degeneracy in fluid
solution being lost in the glassy matrix as the principal axes of the
We measured T1 at the same resonance field as TM. We also
observed in this case a biexponential behaviour; the two
components, reported in Fig. 9, have similar weights. We attri-
bute the fast component to spectral diffusion and the slow one to
spin lattice relaxation.48 In the same graph we also show the
Table 3 Relaxation times as obtained from inversion recovery (T1) and two pulse echo decay (T2) experiments for NitSAc and CTPO in toluene
solution. In parentheses the weight of the relative component in biexponential fits is reported
NitSAc
CTPO
T1,A [ms]
0.561
T [K]
T1,A [ms]
T1,B [ms]
T2,A [ms]
T2,B [ms]
T1,B [ms]
T2,A [ms]
T2,B [ms]
300
270
240
210
195
180
120
100
80
1.051
1.341
2.031
4.141
6.081
7.954
921
0.951
1.0355
1.0555
1.002 (50%)
0.672 (60%)
0.271 (70%)
0.431
0.5505
0.401
0.261
0.112
5.34
0.091
271
813
2034
2405
58010
1.72 (35%)
5.415
0.61
1316 (60%)
2755 (55%)
101
442
0.672
2.794 (85%)
a
0.123
a
Obtained by fitting with a stretched exponential54 with x ¼ 3/2.
This journal is ª The Royal Society of Chemistry 2012
J. Mater. Chem., 2012, 22, 22272–22281 | 22277

View Online
sensitive enough to monitor a small distortion from C2
symmetry. On the other hand at high temperatures the confor-
mation shows the twofold symmetry, as proved by the identical
isotropic hfccs. We conclude that in fluid solution there is a fast
intramolecular motion between two equivalent non-planar
conformations, giving rise to a symmetric average conformation
of the fragment. On lowering the temperature the interconver-
sion rate decreases, eventually leading to two equivalent but
asymmetric conformations. A similar type of interconversion
between two equivalent twisted-crossover conformations,
becoming ‘‘fast’’ for T > 230 K, has been previously reported for
the nitroxide radical Tempone in a solid host matrix.54
Finally, we note that the proton experimental hfccs in fluid
solution, obtained by the simulation of the central EPR line at
270 K and further checked by ENDOR, are consistent with
proton hfccs reported in the literature for similar molecules,38,55,56
apart from the ortho phenyl protons. We attribute this latter
difference to the effect induced by the CH2–S–R residue in the
para position with respect to the nitronyl nitroxide ring.
Fig. 10 Sum (black thick line) and deconvoluted (coloured lines) powder
CW-EPR spectrum of NitSAc in frozen toluene (T ¼ 80 K); the
components are indexed according to their (mI1,mI2) values (see text).
corresponding values for CTPO at T ¼ 80 K and 100 K, both of
which are shorter than the corresponding values for NitSAc (at
80 K respectively, 577 and 207 ms for the latter and 270 and 44 ms
for CTPO).
Relaxation properties
Fluid solution. In fluid solution the transverse relaxation times
at zero concentration, T02, of NitSAc are different for the five
lines, as reported above (see Table 2). This is nicely and directly
shown by the spectral profile of the ED-EPR spectrum of the
diluted fluid solution of the radical reported in Fig. 4. This is in
agreement with the expected effect of the time modulation of the
Discussion
Let us discuss separately the results of the magnetic tensors of
NitSAc and of the relaxation properties of NitSAc and CTPO.
magnetic interactions on the relaxation rates of the different 14
N
Magnetic tensors
hyperfine components, in analogy with the usual results on
organic radicals.57 However, the usual simple Redfield analysis of
the 1/T2 dependence on the nuclear quantum number MI,
customary for nitroxide radicals, is not suitable in this case due to
the presence of degenerate spin states; in order to get relaxation
parameters which are comparable with the experimental ones,
one should obtain the whole relaxation matrix and diagonalize
it.58 Furthermore, the effect of anisotropic rotational diffusion
should be taken into account.27,59
Information as accurate as possible on the principal values of the
magnetic tensors and on the mutual orientation between their
principal axes is of fundamental importance when using the
radical as a spin probe in advanced EPR applications26,49 and for
implementation in QIP.50 For these reasons, we used a hybrid
theoretical/experimental approach for their determination,27–29
which consists, as explained in the Results section, of the calcu-
lation of the principal axes and principal values of the g and
hyperfine tensors by QM/DFT methods and their refinement via
best fitting of the experimental spectra in the glassy phase. The
reliability of the results lies on the fact that (a) the principal
values have been obtained from the best fit of the X-band and
the W-band spectra together (see Fig. 5 and 6, Table 1) (b) the
obtained principal values compare well with those reported in the
literature for similar radicals27,51–53 (c) the principal directions
and values of the 14N hyperfine tensors obtained by this global fit
have been also used to simulate the ED-EPR spectra, which are
extremely sensitive to small changes in the z principal directions
of the hyperfine coupling tensors. This is a relevant result, since –
to the best of our knowledge – this is the first time that the
nitrogen nuclei hyperfine coupling tensors and the mutual
orientation of the magnetic tensors of a NitR radical have been
both experimentally determined.
A comparison with the results obtained at 300 K for T1 and T2
on the central (MI ¼ 0) line of CTPO shows that NitSAc has
relaxation times longer by a factor two with respect to the nitroxide
(see Fig. 9 and Table 3). At 180 K the ratio between the relaxation
times of NitSAc and CTPO is 3 and 1.5 for T2 and T1, respectively.
As the relevant contribution to spin relaxation for both radi-
cals comes from the reorientational motion of the molecule,60
a
meaningful comparison between the relaxation times of the two
radicals should take into account the different dimensions of the
molecules. Let us discuss in the following this issue.
The reorientational motion of a spherical molecule undergoing
Brownian rotational diffusion is described by means of the
rotational correlation time, which is given by61
h
sc ¼ V
(2)
kBT
For a spin distribution on a planar ONCNO fragment with a
C2 symmetry axis one would expect that the z axes of the 14N
hyperfine tensors should be symmetric with respect to the
symmetry axis. However, as reported in Fig. 8 and in Table 1, the
two 14N hyperfine coupling tensors do not follow this expected
symmetry, indicating that at low temperatures the technique is
where h is the dynamic viscosity and V is the molecular volume.
Equivalent expressions can be defined also for molecules with
non-spherical shapes;62 the main point is that the dependence on
the dimensions of the molecule, the temperature and the viscosity
is retained.
22278 | J. Mater. Chem., 2012, 22, 22272–22281
This journal is ª The Royal Society of Chemistry 2012

View Online
Among the various contributions to electron spin relaxation,
the spin-rotational interaction – arising from the coupling
between the angular momentum of the tumbling molecule and
the spin63 – gives equal terms to both the transverse and the
longitudinal relaxation rates:64
nitroxides in glasses,69,70 showing a contribution to relaxation
due to fast librations. In particular, for T > 80 K the Hahn decay
shows an exponential behaviour typical of a dominant relaxation
due to the presence of residual motions. Only at T ¼ 80 K the
decay is a stretched exponential, typical of relaxation due to the
interaction with surrounding paramagnetic centers, either nuclei
or electrons, at low temperatures.54
1
1
1 ðDg : DgÞ
¼
¼
(3)
T1SR T2SR
9
sc
The more suitable method to evidence the presence of residual
intramolecular motions is the ED-EPR powder-like spectrum. In
the presence of motionsthe ED-EPRspectrumshows an increasing
distortion on increasing the inter-pulse delay s with respect to the
integrated CW-EPR profile. This is due to the dependence of the
phase memory time TM on the orientation of the principal axes of
the anisotropic interactions with respect to the magnetic field. As
can be obtained from eqn (1), the relaxation rate becomes faster on
increasing the angle between the field and the principal axes.
We obtained satisfactory simulations of the ED-EPR spectra
of NitSAc in the glassy matrix by taking into account a libra-
tional motion around an axis whose components in the g tensor
eigenframe are approximately given by (0.10, 0.40, 0.91);
assuming a correlation time equal to 30 ps,42 the amplitude of the
motion is restricted to a few degrees.
where (Dg:Dg) ¼ (gxx ꢀ ge)2 + (gyy ꢀ ge)2 + (gzz ꢀ ge)2 and ge is
the free electron g value. At high temperatures the contribution
to T1 and T2 from spin-rotation has been shown to become
dominant, hence making T1 ¼ T2,60 as in the present case. This
consideration allows us to estimate the ratio between the
hydrodynamic volumes of the two radicals in toluene. The values
of (Dg:Dg) are 9.54 ꢄ 10ꢀ5 for NitSAc (see Table 1) and 5.85 ꢄ
10ꢀ5 for CTPO.65 From eqn (2) and (3) we get V(NitSAc)/
V(CTPO) z 3. This ratio compares well with the one derived
from hydrodynamic modeling of the rotational diffusion
tensor;66 in this case we find V(NitSAc)/V(CTPO) ¼ 2.4.
At lower temperatures in fluid solution the spin-rotational
contribution is negligible and the transverse spin relaxation rate
is dominated by the secular and pseudosecular contributions;60
these are proportional to sc,57 a smaller sc producing a slower
relaxation. However we find that NitSAc has a longer T2 by a
factor of around 3 with respect to the smaller CTPO. We
reasonably attribute this behaviour to the different magnitudes
of the dipolar hyperfine coupling tensors, which affects the
secular and pseudosecular spin relaxation rates.60
Above 120 K the echo decay becomes too fast with respect to
the instrumental dead time; at the same temperature the CW-
EPR spectrum shows slight modifications with respect to the
80 K spectrum, indicating the activation of a new motion. A
candidate for this motion could be the intramolecular confor-
mational interconversion of the five atom ring, discussed above.
Finally, when the phase memory time of NitSAc is compared
with that of the CTPO at 80 K, where no specific intramolecular
motions are active, the former is still longer by a factor of two, as
TM is equal to 2.8 ms for CTPO and to 5.4 ms for NitSAc.
These observations lead us to conclude that NitRs are prom-
ising candidates in view of potential applications, both in QIP
and as spin probes. Indeed, one of the prerequisites for the use of
these systems in quantum information applications is that the
coherence times exceed by a large factor the time needed for
coherent manipulations of electron spins, the ratio between the
two being defined as the figure of merit for single-qubit opera-
tions.14,18 With present technology the latter time scale is of the
order of 10 ns for electron spins, thus leading, for NitSAc, to a
figure of merit of about one hundred at room temperature and of
several hundred at liquid nitrogen temperature. This is compa-
rable to decoherence times reported for metal-based molecular
In summary, at high temperatures, where the spin rotational
term dominates the spin relaxation, nitroxides with similar
dimensions can in general be competitive with NitSAc, due to
their smaller g anisotropy; at lower temperatures, where secular
and pseudosecular terms dominate, nitroxides are relaxing much
faster than NitSAc, reasonably due to their larger A anisotropy.
Here we note that for NitSAc the opposite dependence on sc of
the dominant contribution to transverse spin relaxation is
responsible for the maximum in T2 observed at 240 K.
Glassy matrix. For paramagnetic probes in solid matrices the
spectral diffusion is competitive with spin lattice relaxation in the
recovery of inverted magnetization due to the inescapable
selectivity of the microwave pulses burning a hole in the spec-
trum. The disentanglement of the two characteristic times has
been studied in detail and various methods have been proposed
for the scope.48,67,68 In the present case our main scope was a
comparison between the relaxation behaviours of the radicals
NitSAc and CTPO, therefore in the glassy matrix we limited
ourselves to measure the inversion recovery time traces for the
two radicals in the same conditions. All the recoveries are given
by biexponential functions, and we attribute for all of them the
long time to the electronic T1 of the inverted spin packet, and the
short time constant to spectral diffusion to other spin packets. T1
for CTPO is about one half that of NitSAc (270 ms vs. 580 ms),
while the spectral diffusion time is, in relative terms, much
shorter (45 ms vs. 200 ms).
magnets such as Cr7Ni below liquid He temperature (TM
¼
15.3 ms at 1.5 K)23 and Fe3 (TM ¼ 2.6 ms up to 7 K)71 and much
longer than those observed in an Fe4 single molecule magnet
(TM ¼ 630 ns at 4.3 K)72 and in V15, the first molecular magnet
for which Rabi oscillations have been detected.73,74
As for their potential use as spin probes, the slower relax-
ation – especially of the transverse magnetization – makes
possible the use of relatively long pulse experiments, as those
used to measure distances between pairs of radicals (DEER75–77).
Conclusions
In the temperature range explored in this work for relaxation
studies in frozen solution (80–120 K) we have evidence that the
relaxation processes are also due to residual motions of the
molecule, as discussed below. This is in line with the results on
This study allowed the determination of the relaxation properties
of a NitR molecule, NitSAc, which was explicitly designed to
favor well-organized self-assembled monolayers.
This journal is ª The Royal Society of Chemistry 2012
J. Mater. Chem., 2012, 22, 22272–22281 | 22279

View Online
For the first time the hyperfine interaction tensors for the 14
N
12 H. Atsumi, K. Maekawa, S. Nakazawa, D. Shiomi, K. Sato,
M. Kitagawa, T. Takui and K. Nakatani, Chem.–Eur. J., 2012, 18,
178–183.
nuclei of a NitR radical have been determined. The results show
that the two nuclei have the same hyperfine isotropic coupling at
high temperatures, but at low temperatures their pseudo-axial
HF tensors are tilted by 9ꢁ. We conclude, also on the basis of the
relaxation behaviour, that a fast interconversion motion of the
five atoms ring in fluid solution averages the two 14N hyperfine
interactions, becomes slow in the soft glassy phase – preventing
the detection of the echo in a restrict range of temperatures
between 120 and 180 K – and brings to non-equivalent hyperfine
tensors at lower temperatures.
13 K. Sato, S. Nakazawa, R. Rahimi, T. Ise, S. Nishida, T. Yoshino,
N. Mori, K. Toyota, D. Shiomi, Y. Yakiyama, Y. Morita,
M. Kitagawa, K. Nakasuji, M. Nakahara, H. Hara, P. Carl,
P. Hofer and T. Takui, J. Mater. Chem., 2009, 19, 3739–3754.
14 M. Affronte, J. Mater. Chem., 2009, 19, 1731–1737.
15 A. Ardavan and S. J. Blundell, J. Mater. Chem., 2009, 19, 1754–1760.
16 J. Lehmann, A. Gaita-Arino, E. Coronado and D. Loss, J. Mater.
Chem., 2009, 19, 1672–1677.
17 P. C. E. Stamp and A. Gaita-Arino, J. Mater. Chem., 2009, 19, 1718–
1730.
18 F. Troiani and M. Affronte, Chem. Soc. Rev., 2011, 40, 3119–3129.
ꢁ
19 G. Aromı, D. Aguila, P. Gamez, F. Luis and O. Roubeau, Chem. Soc.
Rev., 2012, 41, 537–546.
The analysis of the relaxation data shows that both the T1 and
the T2 (or TM) relaxation times for NitSAc are longer by a factor
of two than for a typical nitroxide (CTPO), both in fluid and in
frozen toluene solution.
20 M. Mannini, L. Sorace, L. Gorini, F. M. Piras, A. Caneschi,
A. Magnani, S. Menichetti and D. Gatteschi, Langmuir, 2007, 23,
2389–2397.
21 P. Messina, M. Mannini, A. Caneschi, D. Gatteschi, L. Sorace,
P. Sigalotti, C. Sandrin, S. Prato, P. Pittana and Y. Manassen,
J. Appl. Phys., 2007, 101, 053916.
The detailed characterization of both static and dynamic
paramagnetic properties of the former system, together with the
availability of recent software for the interpretation of CW-EPR
spectra,78,79 paves the way toward the use of this and similar
radicals as potential spin labels in advanced EPR applications.
Further, the long T2 (or TM) and T1 observed are particularly
appealing in view of the use of this system as a potential qubit,
with figure of merit larger than one hundred at both room and
liquid nitrogen temperatures. In perspective it is then of much
relevance to investigate whether these properties are maintained
once the radical is organized in ordered addressable arrays of
qubits obtained e.g. by exploiting the thioacetyl functions to
promote self-assembling of monolayers.
~
ꢁ
22 C. Simao, M. Mas-Torrent, N. Crivillers, V. Lloveras, J. M. Artes,
P. Gorostiza, J. Veciana and C. Rovira, Nat. Chem., 2011, 3, 359–
364.
23 C. J. Wedge, G. A. Timco, E. T. Spielberg, R. E. George, F. Tuna,
S. Rigby, E. J. L. McInnes, R. E. P. Winpenny, S. J. Blundell and
A. Ardavan, Phys. Rev. Lett., 2012, 108, 107204.
24 F. Meier, J. Levy and D. Loss, Phys. Rev. B: Condens. Matter Mater.
Phys., 2003, 68, 134417.
25 A. Ardavan, O. Rival, J. J. L. Morton, S. J. Blundell,
A. M. Tyryshkin, G. A. Timco and R. E. P. Winpenny, Phys. Rev.
Lett., 2007, 98, 057201.
26 A. Schweiger and G. Jeschke, Principles of Pulse Electron
Paramagnetic Resonance, Oxford University Press, Oxford, 2001.
27 V. Barone, M. Brustolon, P. Cimino, A. Polimeno, M. Zerbetto and
A. Zoleo, J. Am. Chem. Soc., 2006, 128, 15865–15873.
28 A. Collauto, M. Zerbetto, M. Brustolon, A. Polimeno, A. Caneschi
and D. Gatteschi, Phys. Chem. Chem. Phys., 2012, 14, 3200–3207.
29 V. Barone and A. Polimeno, Phys. Chem. Chem. Phys., 2006, 8, 4609–
4629.
Acknowledgements
ꢀ
This work is supported by Ministero dell’Istruzione, Universita e
Ricerca (MIUR), by grant 2008FZK5AC PRIN 2008. M. M., L.
S. and D. G. acknowledge the financial support of EC through
the FP7-People Marie Curie Action IAPP project 286196 and of
MIUR through the PRIN 20097X44S7 ‘‘Record’’ Project.
30 C.-J. Zhong and M. D. Porter, J. Am. Chem. Soc., 1994, 116, 11616–
11617.
31 H. Takiguchi, K. Sato, T. Ishida, K. Abe, K. Yase and K. Tamada,
Langmuir, 1999, 16, 1703–1710.
32 G. Rajaraman, A. Caneschi, D. Gatteschi and F. Totti, J. Mater.
Chem., 2010, 20, 10747–10754.
33 L. Poggini, Master thesis, University of Florence, 2011.
34 L. Gorini, PhD thesis, University of Florence, 2006.
35 L. Gorini, A. Caneschi and S. Menichetti, Synlett, 2006, 948–950.
36 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone,
B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li,
H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng,
J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda,
J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao,
H. Nakai, T. Vreven, J. A. Montgomery, Jr, J. E. Peralta,
F. Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin,
V. N. Staroverov, T. Keith, R. Kobayashi, J. Normand,
K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar,
J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox,
J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts,
R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli,
J. W. Ochterski, R. L. Martin, K. Morokuma, V. G. Zakrzewski,
G. A. Voth, P. Salvador, J. J. Dannenberg, S. Dapprich,
A. D. Daniels, O. Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski
and D. J. Fox, Gaussian 09, Revision B.01, Gaussian, Inc.,
Wallingford, CT, 2010.
References
1 J. H. Osiecki and E. F. Ullman, J. Am. Chem. Soc., 1968, 90, 1078–
1079.
2 E. F. Ullman, J. H. Osiecki, D. G. B. Boocock and R. Darcy, J. Am.
Chem. Soc., 1972, 94, 7049–7059.
3 A. Caneschi, D. Gatteschi, R. Sessoli and P. Rey, Acc. Chem. Res.,
1989, 22, 392–398.
4 A. Caneschi, D. Gatteschi and P. Rey, in Progress in Inorganic
Chemistry, John Wiley & Sons, Inc., New York, 1991, vol. 39, pp.
331–429.
5 M. Tamura, Y. Nakazawa, D. Shiomi, K. Nozawa, Y. Hosokoshi,
M. Ishikawa, M. Takahashi and M. Kinoshita, Chem. Phys. Lett.,
1991, 186, 401–404.
6 A. Caneschi, D. Gatteschi, N. Lalioti, C. Sangregorio, R. Sessoli,
G. Venturi, A. Vindigni, A. Rettori, M. G. Pini and M. A. Novak,
Angew. Chem., Int. Ed., 2001, 40, 1760–1763.
7 L. Bogani, C. Sangregorio, R. Sessoli and D. Gatteschi, Angew.
Chem., Int. Ed., 2005, 44, 5817–5821.
8 K. Bernot, L. Bogani, A. Caneschi, D. Gatteschi and R. Sessoli,
J. Am. Chem. Soc., 2006, 128, 7947–7956.
37 J. Tomasi, B. Mennucci and R. Cammi, Chem. Rev., 2005, 105, 2999–
3094.
9 N. Ishii, Y. Okamura, S. Chiba, T. Nogami and T. Ishida, J. Am.
Chem. Soc., 2008, 130, 24–25.
€
38 J. Cirujeda, J. Vidal-Gancedo, O. Jurgens, F. Mota, J. J. Novoa,
10 G. Poneti, K. Bernot, L. Bogani, A. Caneschi, R. Sessoli,
W. Wernsdorfer and D. Gatteschi, Chem. Commun., 2007, 1807–
1809.
11 X.-L. Wang, L.-C. Li and D.-Z. Liao, Inorg. Chem., 2010, 49, 4735–
4737.
C. Rovira and J. Veciana, J. Am. Chem. Soc., 2000, 122, 11393–11405.
39 S. Stoll and A. Schweiger, J. Magn. Reson., 2006, 178, 42–55.
40 E. Kirilina, S. Dzuba, A. Maryasov and Y. Tsvetkov, Appl. Magn.
Reson., 2001, 21, 203–221.
22280 | J. Mater. Chem., 2012, 22, 22272–22281
This journal is ª The Royal Society of Chemistry 2012

View Online
41 A. Barbon, M. Bortolus, A. L. Maniero and M. Brustolon, Phys.
Chem. Chem. Phys., 2005, 7, 2894–2899.
59 S. A. Goldman, G. V. Bruno, C. F. Polnaszek and J. H. Freed,
J. Chem. Phys., 1972, 56, 716–735.
42 A. Barbon, M. Bortolus, M. Brustolon, A. Comotti, A. L. Maniero,
U. Segre and P. Sozzani, J. Phys. Chem. B, 2003, 107, 3325–3331.
43 J. A. Weil, J. Magn. Reson., 1975, 18, 113–116.
44 R. N. Schwartz, L. L. Jones and M. K. Bowman, J. Phys. Chem.,
1979, 83, 3429–3434.
60 A. Collauto, A. Barbon and M. Brustolon, J. Magn. Reson., 2012,
223, 180–186.
61 P. J. W. Debye, Polar Molecules, Dover, New York, 1929.
62 J. H. Freed, J. Chem. Phys., 1964, 41, 2077–2083.
63 P. S. Hubbard, Phys. Rev., 1963, 131, 1155–1165.
64 P. W. Atkins and D. Kivelson, J. Chem. Phys., 1966, 44, 169–174.
65 A. Capiomont, B. Chion, J. Lajzerowicz-Bonneteau and H. Lemaire,
J. Chem. Phys., 1974, 60, 2530–2535.
€
45 P. Hofer, Bruker Application Note no. 117.
46 E. A. Weiss, M. J. Tauber, M. A. Ratner and M. R. Wasielewski,
J. Am. Chem. Soc., 2005, 127, 6052–6061.
47 M. Marrale, M. Brai, A. Barbon and M. Brustolon, Radiat. Res.,
2009, 171, 349–359.
66 V. Barone, M. Zerbetto and A. Polimeno, J. Comput. Chem., 2009, 30,
2–13.
48 J. R. Harbridge, S. S. Eaton and G. R. Eaton, J. Magn. Reson., 2002,
159, 195–206.
49 D. E. Budil, K. A. Earle and J. H. Freed, J. Phys. Chem., 1993, 97,
1294–1303.
50 (a) T. Yoshino, S. Nishida, K. Sato, S. Nakazawa, R. D. Rahimi,
K. Toyota, D. Shiomi, Y. Morita, M. Kitagawa and T. Takui,
J. Phys. Chem. Lett., 2011, 2, 449–453; (b) S. Nakazawa,
S. Nishida, T. Ise, T. Yoshino, N. Mori, R. D. Rahimi, K. Sato,
Y. Morita, K. Toyota, D. Shiomi, M. Kitagawa, H. Hara, P. Carl,
67 M. Brustolon and U. Segre, Appl. Magn. Reson., 1994, 7, 405–413.
68 J. R. Harbridge, S. S. Eaton and G. R. Eaton, J. Phys. Chem. A, 2003,
107, 598–610.
69 H. Sato, S. E. Bottle, J. P. Blinco, A. S. Micallef, G. R. Eaton and
S. S. Eaton, J. Magn. Reson., 2008, 191, 66–77.
70 N. P. Isaev, L. V. Kulik, I. A. Kirilyuk, V. A. Reznikov,
I. A. Grigor’ev and S. A. Dzuba, J. Non-Cryst. Solids, 2010, 356,
1037–1042.
71 G. Mitrikas, Y. Sanakis, C. P. Raptopoulou, G. Kordas and
G. Papavassiliou, Phys. Chem. Chem. Phys., 2008, 10, 743–748.
72 C. Schlegel, J. van Slageren, M. Manoli, E. K. Brechin and
M. Dressel, Phys. Rev. Lett., 2008, 101, 147203.
73 S. Bertaina, S. Gambarelli, T. Mitra, B. Tsukerblat, A. Muller and
B. Barbara, Nature, 2008, 453, 203–206.
74 J. Yang, Y. Wang, Z. Wang, X. Rong, C.-K. Duan, J.-H. Su and
J. Du, Phys. Rev. Lett., 2012, 108, 230501.
75 M. Pannier, S. Veit, A. Godt, G. Jeschke and H. W. Spiess, J. Magn.
Reson., 2000, 142, 331–340.
76 G. Jeschke and Y. Polyhach, Phys. Chem. Chem. Phys., 2007, 9, 1895–
1910.
€
P. Hofer and T. Takui, Angew. Chem., Int. Ed., 2012, DOI:
10.1002/anie.201204489.
51 J. A. D’Anna and J. H. Wharton, J. Chem. Phys., 1970, 53, 4047–
4052.
52 V. I. Gulin, S. A. Dikanov, Y. D. Tsvetkov, I. A. Grigor’ev and
I. A. Kirilyuk, J. Struct. Chem., 1988, 29, 472–475.
53 S. A. Dikanov, V. I. Gulin, Y. D. Tsvetkov and I. A. Grigor’ev,
J. Chem. Soc., Faraday Trans., 1990, 86, 3201–3205.
54 A. Barbon, M. Brustolon, A. L. Maniero, M. Romanelli and
L.-C. Brunel, Phys. Chem. Chem. Phys., 1999, 1, 4015–4023.
55 J. Goldman, T. E. Petersen, K. Torssell and J. Becher, Tetrahedron,
1973, 29, 3833–3843.
56 T. Takui, Y. Miura, K. Inui, Y. Teki, M. Inoue and K. Itoh, Mol.
Cryst. Liq. Cryst. Sci. Technol., Sect. A, 1995, 271, 55–66.
57 A. Carrington, A. Hudson and G. R. Luckhurst, Proc. R. Soc.
London, Ser. A, 1965, 284, 582–593.
77 R. Dastvan, B. E. Bode, M. P. R. Karuppiah, A. Marko,
S. Lyubenova, H. Schwalbe and T. F. Prisner, J. Phys. Chem. B,
2010, 114, 13507–13516.
78 S. K. Misra, J. Magn. Reson., 2007, 189, 59–77.
79 M. Zerbetto, A. Polimeno and V. Barone, Comput. Phys. Commun.,
2009, 180, 2680–2697.
58 J. H. Freed and G. K. Fraenkel, J. Chem. Phys., 1963, 39, 326–348.
This journal is ª The Royal Society of Chemistry 2012
J. Mater. Chem., 2012, 22, 22272–22281 | 22281