Large molecular weight nitroxide biradicals providing efficient dynamic nuclear polarization at temperatures up to 200 K

Article abstract of DOI:10.1021/ja405813t

A series of seven functionalized nitroxide biradicals (the bTbK biradical and six derivatives) are investigated as exogenous polarization sources for dynamic nuclear polarization (DNP) solid-state NMR at 9.4 T and with ca. 100 K sample temperatures. The impact of electron relaxation times on the DNP enhancement (ε) is examined, and we observe that longer inversion recovery and phase memory relaxation times provide larger ε. All radicals are tested in both bulk 1,1,2,2-tetrachloroethane solutions and in mesoporous materials, and the difference in ε between the two cases is discussed. The impact of the sample temperature and magic angle spinning frequency on ε is investigated for several radicals each characterized by a range of electron relaxation times. In particular, TEKPol, a bulky derivative of bTbK with a molecular weight of 905 g·mol^-1, is presented. Its high-saturation factor makes it a very efficient polarizing agent for DNP, yielding unprecedented proton enhancements of over 200 in both bulk and materials samples at 9.4 T and 100 K. TEKPol also yields encouraging enhancements of 33 at 180 K and 12 at 200 K, suggesting that with the continued improvement of radicals large ε may be obtained at higher temperatures.

Full text of DOI:10.1021/ja405813t

Article
pubs.acs.org/JACS
Large Molecular Weight Nitroxide Biradicals Providing Efficient
Dynamic Nuclear Polarization at Temperatures up to 200 K
Alexandre Zagdoun,† Gilles Casano,∥ Olivier Ouari,∥ Martin Schwarzwalder,‡ Aaron J. Rossini,†
̈
Fabien Aussenac,§ Maxim Yulikov,‡ Gunnar Jeschke,‡ Christophe Coperet,‡ Anne Lesage,† Paul Tordo,*,∥
́
and Lyndon Emsley*,†
†Centre de RMN a
69100 Villeurbanne, France
∥Aix-Marseille Universite,
CNRS, ICR UMR 7273, 13397 Marseille, France
̀ ̀ ́
Tres Hauts Champs, Institut de Sciences Analytiques, Universite de Lyon (CNRS/ENS Lyon/UCB Lyon 1),
́
‡Department of Chemistry, ETH Zurich, CH-8093, Zurich, Switzerland
̈
̈
§Bruker BioSpin SA, Wissembourg, 67160, France
S
* Supporting Information
ABSTRACT: A series of seven functionalized nitroxide
biradicals (the bTbK biradical and six derivatives) are
investigated as exogenous polarization sources for dynamic
nuclear polarization (DNP) solid-state NMR at 9.4 T and with
ca. 100 K sample temperatures. The impact of electron
relaxation times on the DNP enhancement (ε) is examined,
and we observe that longer inversion recovery and phase
memory relaxation times provide larger ε. All radicals are
tested in both bulk 1,1,2,2-tetrachloroethane solutions and in
mesoporous materials, and the difference in ε between the two cases is discussed. The impact of the sample temperature and
magic angle spinning frequency on ε is investigated for several radicals each characterized by a range of electron relaxation times.
In particular, TEKPol, a bulky derivative of bTbK with a molecular weight of 905 g·mol−1, is presented. Its high-saturation factor
makes it a very efficient polarizing agent for DNP, yielding unprecedented proton enhancements of over 200 in both bulk and
materials samples at 9.4 T and 100 K. TEKPol also yields encouraging enhancements of 33 at 180 K and 12 at 200 K, suggesting
that with the continued improvement of radicals large ε may be obtained at higher temperatures.
structure of the species bearing the unpaired electrons (mostly
organic free radical).24,25 Here we focus on the initial transfer of
polarization from electrons to nuclei. For this transfer at high-
magnetic fields, the cross effect (CE) and thermal mixing (TM)
are the most efficient mechanisms.3,24 They are active when
broadening of the EPR line width exceeds the nuclear Larmor
frequency, and they involve a three-spin process including two
dipolar coupled electron spins whose Larmor frequencies are
separated by the nuclear Larmor frequency (the so-called
matching condition).3,26 To make this process as efficient as
possible, over the last 10 years much effort has been devoted to
the development of increasingly efficient exogenous polarizing
agents. A major step was achieved by Griffin and co-workers
who demonstrated in 2004 that biradicals with constrained e−e
distances, and thus increased e−e dipolar couplings, yield more
efficient CE DNP.27 This has led to the introduction of stable
nitroxide-based biradicals including TOTAPOL,28 which is
today the most commonly used polarization source for DNP
solid-state NMR spectroscopy in aqueous media.
INTRODUCTION
■
Dynamic nuclear polarization (DNP)1,2 is currently one of the
most efficient methods to increase the sensitivity of NMR
experiments. Nuclear polarization can be enhanced by
microwave (MW) induced polarization transfer from unpaired
electrons, with a theoretical limit for the signal enhancement of
εmax = γe/γX, where γe and γX are the gyromagnetic ratios of the
electron and of the target nuclei, respectively (εmax = 660 for
proton, for example).3 Of particular interest is the possibility to
do high-field DNP at low temperatures in situ to increase the
sensitivity of magic-angle spinning (MAS) solid-state NMR
spectroscopy.4−6 While enhancements (ε) of 50 are today
routinely obtained at 9.4 T and 100 K, allowing the
investigation of an ever broader range of molecular systems
including biomolecules and entire cells,7−11 materials,12−22 or
bulk microcrystals,23 the signal enhancement factors are still far
from the predicted maximum values, and low sensitivity still
prevents many potential applications.
Several factors affect enhancements since the trajectory of
polarization from the unpaired electrons to the nuclei is
complex, and the mechanism depends on parameters, such as
the main magnetic field, the temperature of the sample, the
nature of the propagating medium, or the concentration and
Received: June 10, 2013
Published: July 24, 2013
© 2013 American Chemical Society
12790
dx.doi.org/10.1021/ja405813t | J. Am. Chem. Soc. 2013, 135, 12790−12797

Journal of the American Chemical Society
Article
For nitroxide biradicals the matching condition is obtained
for a given orientation through the g tensor anisotropy and can
be fulfilled for the largest number of orientations when the two
g tensors are nearly orthogonal. In 2009, Tordo, Griffin, and
co-workers introduced the bTbK biradical29 (Figure 1, 1) for
properties. In particular, a new dinitroxide radical 7, TEKPol, is
introduced that yields an unprecedented proton enhancement
(εH) of over 200 at 9.4 T and 100 K in bulk solution as well as
in mesoporous materials. The temperature and magic angle
spinning (MAS) frequency behavior of the DNP enhancement
factors is also discussed for this new family of polarizing agents.
EXPERIMENTAL SECTION
■
NMR Spectroscopy. DNP experiments were conducted on a
commercial Bruker Avance III 400 MHz NMR spectrometer equipped
with a 263 GHz gyrotron microwave source,36 using a 3.2 mm triple
resonance MAS probe at sample temperatures around 100 K (unless
noted otherwise). The microwave power was set to maximize the DNP
enhancement (∼4 W power at the sample). The magnet sweep coil
was used to set the magnetic field to coincide with the maximum
enhancement for TEKPol (which was found to be the same as for
TOTAPOL, see Figure S1). Proton DNP enhancements were
measured on spectra acquired after a short (1 rotor cycle) spin echo
to remove probe background, and the enhancements of heteronuclei
were measured after cross-polarization (CP) from protons. More
details on the NMR parameters can be found in SI.
NMR Sample Preparation. The samples were either bulk
solutions or mesoporous silica materials impregnated with biradical
solution (prepared as previously described).12,37 Samples were topped
with a Teflon insert to minimize solvent leakage from the sapphire
rotors. A spatula tip of KBr was added at the bottom of the rotor to
enable measurement of 79Br chemical shifts for temperature
calibration.38
EPR Spectroscopy. Experiments were conducted at W Band (94
GHz) on a Bruker Elexsys E680 EPR spectrometer at a temperature of
100 K in tetrachloroethane (TCE), unless noted otherwise. The
inversion recovery times (Tir) were measured using an inversion−
recovery sequence, and the data were fitted using a stretched
exponential function. The phase memory times (Tm) were measured
using a variable-delay Hahn echo pulse sequence and were fitted using
a monoexponential function. Further details are given in SI.
To match the DNP conditions, samples were recorded at a high
radical concentration (16 mM). For comparison, bTbK, bCTbK, and
TEKPol were measured at low concentration (0.2, 0.1, and 0.1 mM,
respectively). The results are discussed below.
Glass Formation. Experimentally, we observed variations (up to
80%) of the DNP enhancement in TCE solutions for some radicals
upon freezing of the sample in the magnet. Several freezing and
thawing cycles were performed by inserting and ejecting the sample
from the probe. We report here the highest ε. Freezing the rotor in
liquid nitrogen did not yield consistently better ε. The quality of the
glass formed by freezing the sample is well-known to be an important
factor for both DNP and EPR. Poor glasses or crystalline domains
formed in the sample will lead to shortening of the relaxation times by
creating zones of high local radical concentration, therefore attenuating
the DNP process. This is why cryoprotectants, such as glycerol, are
used in DNP experiments in water,39 which crystallizes when frozen,
or why alternative sample preparation schemes are being devel-
oped.12,40 For samples yielding poor DNP enhancements, the 13C
NMR spectra display two shoulders due to residual coupling to
chlorine. They might indicate formation of crystalline domains in the
sample (see SI). In contrast reproducible enhancements were obtained
when impregnating mesoporous materials.
Figure 1. Structure, name, and molecular weight of the radicals
investigated in this study.
which the orientation of the two g tensors is fixed by a rigid
skeleton, leading to improved DNP efficiency when compared
to more flexible radicals. This radical is soluble in a wide range
of organic solvents compatible with DNP30 and as such is well
suited for applications to materials that are not compatible with
water. Water-soluble rigid biradicals have been recently
introduced, yielding enhancement factors higher than TOTA-
POL in aqueous media.31,32
Beyond the geometry, the electronic relaxation properties of
radical species are another key factor affecting the DNP
process. Initial polarization transfer depends on saturation of an
EPR transition, which is facilitated by longer electron spin
relaxation times (T1e, T2e). For instance, it was shown in 2011
by Casabianca et al. that the efficiency of paramagnetic defects
to polarize the bulk in nanoparticulate diamonds increased with
increasing T1e.33 In 2012, our group introduced 2, a bulky 2,6-
spirocyclohexyl derivative of bTbK, dubbed bCTbK,34 which
has a long electron spin−lattice relaxation time and maintains
optimal conformation for DNP. We showed that at 100 K this
biradical has a T1e twice as long as that of bTbK at 94 GHz and
that this resulted in DNP enhancement factors up to 4 times
larger. More recently it has been demonstrated that a mixture of
narrow line radicals having differential electron spin−lattice
relaxation times provides an effective alternative to biradicals
for CE DNP to polarize carbon-13 directly.35
EPR samples in small (0.5 mm) capillaries were frozen in liquid
nitrogen prior measurement. Results were reproducible, and the
samples were visually transparent and homogeneous, indicating glass
formation.
Synthesis of the Materials. The mesoporous material used as a
model here to evaluate surface NMR enhancements, Mat-Azide, was
prepared as previously described.34
Synthesis of the Radicals. The general synthetic route to prepare
compounds 1−7 involves three main steps which are illustrated in
Figure 2 for TEKPol. Experimental details for reaction procedures and
In this Article, we present a series of bTbK derivatives
suitable for high-field DNP-enhanced solid-state NMR spec-
troscopy. We establish a clear relationship between the DNP
efficiency of these new biradicals and their electronic relaxation
12791
dx.doi.org/10.1021/ja405813t | J. Am. Chem. Soc. 2013, 135, 12790−12797

Journal of the American Chemical Society
Article
characterization of all the new intermediates and biradicals are given in
SI.
the DNP process depends, as stated previously, on several
parameters. The CE is a three spin (two electrons one nucleus)
process that requires the saturation of one of the electron
resonances to create a nonequilibrium state that then transfers
polarization through a two-spin flip-flop of the other electron
spin and the nuclear spin. The efficiency of the continuous-
wave (CW) saturation of the electron spins depends on the
saturation factor s (s = T1e × T2e). The higher s, the more
efficient the saturation will be.41 We expect therefore that, if all
the other parameters (g and A tensors, structure of the radicals,
e−e dipolar couplings) are almost constant, then the DNP
efficiency of the radicals will increase with their saturation
factors. In the following we develop radicals with improved
saturation factors and thereby improved DNP performance.
Figure 1 shows the chemical structure of the biradicals
developed in this study, designed to provide longer electron
relaxation times. The aim of this series is to access a range of
radicals with the same EPR tensor parameters and relative
orientations, the same electron−electron interactions, but with
different electron spin relaxation properties. A synthetically
feasible way to obtain such radicals is to modify the derivatives
at α position (Cα) to the nitroxyl groups (N−O•) while
maintaining the rest of the skeleton. The synthesis of such
radicals is not always straightforward, and the spirocyclohexyl-
Figure 2. General synthetic route, illustrated for TEKPol, to synthesize
compounds 1−7.
RESULTS AND DISCUSSION
Longer Electronic Relaxation Times Are Expected to
Yield More Efficient DNP at 9.4 T and 100 K. Efficiency of
■
Figure 3. (A) Electron inversion recovery times (Tir), (B) electron phase memory time (Tm) as a function of molecular weight for the different
radicals, and (C) electron phase memory time (Tm) as a function of the number of methyl groups groups at the Cα carbons. (D) Saturation factor s
as a function of molecular weight for the different radicals. (All measurements were done with 16 mM solutions in 1,1,2,2-TCE at 100 K. The EPR
experiments are done at W band (94 GHz).
12792
dx.doi.org/10.1021/ja405813t | J. Am. Chem. Soc. 2013, 135, 12790−12797

Journal of the American Chemical Society
Article
based functional groups used here were chosen due to
reasonable yields. All the radicals of Figure 1 were prepared
with the goal to cover a range of molecular weights.
Since molecular weight is not the only factor in determining
relaxation times, some differences are observed in relaxation
rates for radicals with similar molecular weight in Figure 3, as is
demonstrated by 4 bCTbK, 5 (ACT)2bK, and 6 bPyTbK. Sato
et al. described electron spin relaxation for nitroxide radicals in
glassy solvents at temperatures between 100 and 300 K using
models that involved the corrected temperature T′ = TV−γ,
where T is the temperature, V the molecular volume of the
radical, and γ a parameter that characterizes the medium and is
dependent on local molecular motion.42 Their model predicts
that bulkier radicals lead to slower relaxation and also that local
motion affects relaxation. Furthermore for a biradical with
interspin distance about 9 Å these authors showed39 that
nitroxide−nitroxide interaction has little impact on electron
spin relaxation. Biradicals 4−6 have roughly the same molecular
mass, but 5 has four methyl groups at the Cα carbons and two
additional N-acetyl methyl groups. This is in line with the
observation in Figure 3 that 4 has a shorter relaxation time than
6 which is in turn shorter than that for bCTbK. Measurements
of T1e and T2e at low radical concentration were also carried out
for bTbK, bCTbK, and TEKPol. The relaxation times at low
concentration are found to be longer indicating that
intermolecular dipolar relaxation occurs at 16 mM concen-
tration, but as expected the trend remains the same, i.e. Txe bTbK
< Txe bCTbK < Txe bPhCTbK (Figure S5).
Molecular Geometry and EPR Characterization of
Biradicals 1−7. All these radicals have the same rigid trispiro
backbone as bTbK and differ only in the substitution at the
carbons in the α position (Cα) to the nitroxyl groups (N−O•)
(Figure 2). Comparison of single crystal XRD (where available)
or DFT calculations for 1, 4, 6, and 7, (Table S1), shows that as
expected, the average distance between the two unpaired
electrons (and thus the e−e dipolar coupling) as well as the
orientation of the two CαN(Oλ)Cα average planes are very
similar throughout the series. In addition EPR spectra recorded
at low temperature indicate that their g tensors are almost
identical (see Figure S6).
Figure 3A,B shows the inversion recovery (Tir) and phase
memory (Tm) times measured for the radicals as a function of
molecular weight. The experiment was performed near the 14
N
mI = −1 hyperfine component along gzz of the EPR spectrum,
corresponding to the optimum DNP irradiation frequency (as
the relaxation times vary along the EPR powder pattern)42 (see
Figure S3). The biradical concentration in the solution for EPR
was determined by CW EPR spin counting experiments and
was found to vary between 11.0 and 14.7 mM. At these high
concentrations, the measured inversion recovery and phase
memory times do not correspond to the electron longitudinal
and transverse relaxation times as intermolecular relaxation,
such as spin exchange or dipolar coupling plays a role in the
process.43 We note however that these data are recorded under
experimental conditions that are close to those used during the
DNP experiments, and are thus the most relevant to our results.
The most significant difference in conditions is the EPR
frequency, which here is 94 GHz. However, the electron
relaxation times of nitroxides in glassy organic solvents around
100 K are not expected to vary significantly between 94 and
263 GHz.44 Thus the values reported here will not be exactly
the same as those at the DNP frequencies (spectral diffusion
will increase at higher fields), but they are expected to follow
the same trends.44 The inversion recovery curves were fitted
using a stretched exponential function to take into account a
distribution of inversion recovery time constants. Tm was fitted
using a monoexponential function.
As expected the mean saturation factor s (s = Tir * Tm),
which reflects the ability to saturate the EPR transitions,
increases with molecular weight as shown in Figure 3D.
DNP Enhancements. Figure 4 shows the measured proton
DNP enhancement of the solvent (measured on the carbon-13
For radicals in glassy organic solvents around 100 K, the
Raman process is expected to be the dominant mechanism of
longitudinal relaxation for dilute samples. It has been shown
that the efficiency of this two phonon process depends on the
molecular weight of the radicals; the heavier the radical, the
slower the relaxation. As spin diffusion leading to the spectral
diffusion is driven by longitudinal relaxation, one still expects
positive correlation between molecular weight and inversion
recovery times at high concentrations.42,44 The electron
inversion recovery times (Figure 3A) measured for biradicals
1−7 show a roughly linear dependence on the molecular
weight of the biradicals, as discussed in detail below.
At these temperatures, and in glassy solvents, Tm relaxation is
governed by molecular motion and libration. Rotation of
methyl groups, for instance, can induce effective transverse
relaxation.42 Replacing the methyl groups at Cα carbons by
increasingly bulky and rigid functional groups is therefore
expected to lengthen Tm.45 Figure 3B,C shows, respectively, the
impact of the molecular weight and of the number of methyl
groups at Cα carbons on Tm . As expected, heavier radicals or
radicals without methyl groups exhibit longer Tm.
Figure 4. 13C enhancement after CP from protons for 16 mM bulk
solutions in TCE as a function of the mean saturation factor.
resonance of the solvent in a standard CPMAS experiment)
obtained as a function of the mean saturation factor. As
predicted, the DNP enhancement is seen to increase with the
mean saturation factor and therefore with the molecular mass
of the polarizing agent. In particular, TEKPol, with a molecular
weight of 905 g·mol−1, a 94 GHz mean saturation factor of 65.0
μs2, and an enhancement of 200, far outperforms the other
biradicals and notably the previous best bCTbK which gives an
enhancement of 80. This is the highest enhancement presented
for these conditions of temperature and field to date.
We note that it is expected that the DNP enhancement will
go through a maximum at some point for higher saturation
12793
dx.doi.org/10.1021/ja405813t | J. Am. Chem. Soc. 2013, 135, 12790−12797

Journal of the American Chemical Society
Article
factors. An optimum saturation factor is predicted since the
degree of saturation of the EPR transition is improved at a
given microwave power for larger saturation factors, whereas
for repeated polarization transfer from the same electron spin
to nuclei, the electron spin polarization must recover, so that
shorter Tir will increase turn over.24 However, we only expect to
see this effect for significantly longer electron relaxation times
than studied here. For dinitroxides at the temperatures and
fields used here we expect DNP enhancements to continue to
increase with increasing saturation factors. Other types of
radicals, such as mixed biradicals having radical moieties with
one long Tir and one short Tir have been predicted to have high
DNP enhancement.24,35
Temperature Dependence of DNP Enhancements. A
key objective in the development of solid-state DNP experi-
ments is to be able to use higher sample temperatures.46 This is
particularly relevant to proteins samples where resolution is
often (but not always) degraded at temperatures below 200 K.
ε is known to be strongly dependent on the temperature of the
sample, with a steep decrease as temperature increases. Figure
5A shows the temperature dependence of ε for bTbK, bCTbK,
and TEKPol. In these experiments, a spatula tip of crystalline
KBr (which is insoluble in TCE) was added to the rotors, and
the sample temperature was calibrated by using the 79Br
chemical shift as an internal standard.38,47 As expected, ε
decreases as the temperature increases for all three radicals.
However, whereas bTbK exhibits the typical rapid decrease
already observed for TOTAPOL in glycerol/water,36 both
bCTbK and TEKPol show a slower decrease with temperature
over the range here. As a consequence, at around 140 K, ε is
reduced by 60% from the value at 106 K for bTbK but only by
40% for both bCTbK and TEKPol. At around 176 K bTbK
yields enhancements of only 4.5 (ε/ε100K = 0.16), and bCTbK
and TEKPol have ε of 24 (ε/ε100K = 0.25) and 33 (ε/ε100K
=
0.23), respectively. Variation of the proton nuclear longitudinal
relaxation time with temperature has been shown to affect the
DNP enhancements.36,39 However, for radical solutions in
TCE, proton T1 only varies slightly (from 2.0 to 2.6 s) in the
range of temperature investigated (see SI). The temperature
dependence of the enhancement can more credibly be linked to
the temperature dependence of the electron relaxation times.
As seen in Figure 4C, the saturation factors indeed decrease
significantly with increasing temperature.
A key structural difference between bTbK and bCTbK or
TEKPol is the presence of methyl groups at the Cα carbons of
the former. Motion, and in particular the rotation of the methyl
groups, is the dominant process for transverse relaxation, and
the impact of temperature is therefore expected to be larger on
Tm for bTbK than for the other radicals, which could explain
the observed difference. Unfortunately, due to short Tm,
measure of relaxation parameters could not be conducted at
temperatures higher than 140 K for bTbK. This difference of
behavior in Tm was previously observed for monoradicals with
methyl groups at the Cα carbons vs radicals with cyclohexyls in
the same position.45,48 The individual temperature behavior of
Tir and Tm for the three radicals is given in Table S3. It is also
interesting to note that the widely used radical TOTAPOL,
which also has methyl groups at the Cα carbons has been
reported to present a similar temperature dependence to that of
bTbK.36,49 Most interestingly, TEKPol maintains enhance-
ments of 12 at 198 K.
Figure 5. (A) Enhancement, (B) normalized enhancement, and (C)
the mean saturation factor as a function of temperature for TEKPol
(red squares), bCTbK (blue triangles), and bTbK (black circles). The
lines are guides for the eye. Note that Tm becomes too short to detect
an EPR echo at temperatures higher than 140 K.
enhancements. Fast spinning is essential for the study of
many materials and biomolecules. The typical behavior for ε
observed and reported with spinning frequency is first to
increase and then go through a maximum (usually around 4
kHz spinning frequency) and then decrease as the spinning
frequency increases to 12 kHz.36 Rosay et al. attributed this
MAS Frequency Dependence of ε. Another important
parameter is the spinning frequency dependence of the
12794
dx.doi.org/10.1021/ja405813t | J. Am. Chem. Soc. 2013, 135, 12790−12797

Journal of the American Chemical Society
Article
behavior primarily to an increase in sample temperature due to
frictional heating with the spinning frequency.36 More recently,
Mentink-Vigier et al. and Thurber and Tycko developed a
theoretical two-electron one-nucleus model for MAS CE DNP
where the spinning frequency dependence is explained by
energy level crossings and anticrossings.50,51 We performed a
temperature-controlled experiment for a 16 mM solution of
TEKPol in TCE using the 79Br chemical shift from KBr as an
internal thermometer, as above. The 79Br chemical shift was
recorded at the highest spinning frequency (14 kHz), and for
lower spinning frequencies the temperature control setting was
increased so that the 79Br chemical shift matches the one at 14
kHz. In this way, the experiments are carried out at the same
temperature at all spinning frequencies. It turns out that the
difference with respect to the spinning frequency dependence
measured without correction for frictional heating is minor for
the radical/solvent mixtures studied here. Figure 6 shows the
spinning frequency dependence measured for the three radicals
investigated.
Table 1. Bulk Solvent and Silicon Surface Enhancements,
Quenching Factors, Overall Sensitivity Enhancements and
Silicon Transverse Dephasing Times for 1 and 3−7
a
b
e
εC CP
εSi CP
T2′(29Si)
c
d
compound
(bulk)
(mat)
θSi
ΣSi
(ms)
bTbK, 1
(CT)2bK, 3
bCTbK, 4
47
53
60
80
0.2
0.2
0.2
0.6
0.4
0.2
16
25
35
13
10
53
13
5
80
103
32
14
29
20
(ACT)2bK, 5
bPyTbK, 6
TEKPol, 7
42
115
200
10
214
7
a
b
13C enhancement measured on a bulk 16 mM TCE solution. 29Si
enhancement measured on Mat-Azide impregnated with a 16 mM
c
d
TCE solution in. 29Si quenching factor. 29Si overall sensitivity
e
enhancement. 29Si transverse dephasing time.
reported.37 θ is the fraction of nuclei that generate an
observable signal at the surface of the mesoporous silica, and
∑ represents the overall gain in sensitivity, taking into account
the quenching factor and modification in relaxation times
between regular low temperature and DNP experiments.37 For
most of the radicals, ε increases in the materials as compared to
the bulk solution, consistent with previous observations and
with the cryoprotecting role of the material.34 There are two
notable exceptions: 5 and 6, where ε drops from 42 to 32 and
from 115 to 10, respectively. This can potentially be attributed
to the differences in polarity among the radicals that causes
them to interact differently with the surface. Differences in
surface−radical interactions depending on the polarity of the
surface were observed in a previous DNP study.52 Wessig et al.
showed by EPR that matching the polarity of the solvent and
the surface minimizes radical−surface interactions.53 This
hypothesis is supported here by the measure of the 29Si
transverse dephasing time (T2′), which is a good sensor of the
proximity of the radicals to the surface silicon atoms. For the
other radicals T2′ is shorter than 14 ms, but it increases to 20
ms for 6 and to 29 ms for 5, indicating that in these cases the
radical is further away from the surface, for example, through a
partial exclusion from the pores thereby explaining the
reduction in DNP efficiency.
Figure 6. Relative enhancement dependence on the spinning
frequency of the sample for TEKPol (red square), bCTbK (blue
triangles), and bTbK (black circles).
CONCLUSION
For clarity, to compare the behavior, only the normalized
enhancements are displayed. Interestingly, the curves show a
decrease in efficiency from about 6 to 12 kHz, as expected, but
they all then show an increase to a second maximum at around
14 kHz. We do not have an explanation for this interesting
behavior. The three radicals show largely the same MAS
dependence despite their different T1e. The differences in T1e
may not be large enough to produce an observably different
functional form.50 Note that enhancements of over 160 are
obtained at high spinning frequencies (15 kHz) for TEKPol.
Performance of the Radicals in Materials. One of the
most recent applications of high-field DNP is the study of
surfaces and materials.12,13,15,17−22,34 In this approach, a
material is impregnated with a minimal amount of the
radical-containing solution. Interestingly, the material itself
usually serves as a cryoprotectant and helps to form a good
glass. DNP performance of the new radicals was investigated
here for a mesoporous silica-containing propyl azide moieties
bound to the silica surface via a C−Si linkage (Mat-Azide).34
The results are summarized in Table 1. Along with ε, the
quenching factor θ and the overall silicon enhancements ∑ are
■
The study of the effect of electron spin relaxation times on
DNP enhancements for a series of bTbK analogs leads us to
establish clear design principles for efficient dinitroxide
biradicals. The structure and thus the g and A tensors as well
as the e−e dipole couplings of these derivatives are very similar,
and therefore the main differences in observed DNP efficiency
are attributed to electron relaxation properties of the radicals.
The saturation factor of the electron spin transitions was found
to be highly correlated with the DNP enhancements.
Dinitroxide biradicals which have high molecular weights
(large rigid volumes) and no methyl groups close to the
nitroxyl groups (N−O•) are predicted to have high DNP
efficiency.
A new heavy analog of bTbK, TEKPol is shown to yield
proton polarization enhancements of over 200 at 9.4 T and 100
K, both for bulk and materials applications, and is the most
efficient polarizing agent found so far for these conditions.
Furthermore, TEKPol retains high enhancements at higher
temperatures, here providing ∑ = 33 at 180 K and 12 at 200 K.
It also provides substantial enhancements at high MAS
12795
dx.doi.org/10.1021/ja405813t | J. Am. Chem. Soc. 2013, 135, 12790−12797

Journal of the American Chemical Society
Article
(13) Lelli, M.; Gajan, D.; Lesage, A.; Caporini, M. A.; Vitzthum, V.;
frequencies (over 150 at 15 kHz). We expect that water-soluble
derivatives of this biradical, or more generally water-soluble
dinitroxide biradicals with bulky substituents and no methyl
groups at the Cα carbons, will retain high enhancements at high
temperatures and high spinning frequencies, thus opening up
the perspective of achieving higher resolution DNP spectros-
copy in biosolids, currently one of the main drawbacks of CE
DNP in biological applications.
Miev
Boualleg, M.; Veyre, L.; Bodenhausen, G.; Coper
Am. Chem. Soc. 2011, 133, 2104−2107.
(14) Lafon, O.; Rosay, M.; Aussenac, F.; Lu, X.; Treb
O.; Kinowski, C.; Touati, N.; Vezin, H.; Amoureux, J.-P. Angew. Chem.,
Int. Ed. 2011, 50, 8367−8370.
(15) Rossini, A. J.; Zagdoun, A.; Lelli, M.; Canivet, J.; Aguado, S.;
́
ille, P.; Her
́
oguel, F.; Rascon
́
, F.; Roussey, A.; Thieuleux, C.;
́
et, C.; Emsley, L. J.
́
osc, J.; Cristini,
Ouari, O.; Tordo, P.; Rosay, M.; Maas, W. E.; Coper
D.; Emsley, L.; Lesage, A. Angew. Chem., Int. Ed. 2012, 51, 123−127.
(16) Vitzthum, V.; Mieville, P.; Carnevale, D.; Caporini, M. A.;
Gajan, D.; Coperet, C.; Lelli, M.; Zagdoun, A.; Rossini, A. J.; Lesage,
́
et, C.; Farrusseng,
ASSOCIATED CONTENT
* Supporting Information
́
■
́
S
A.; Emsley, L.; Bodenhausen, G. Chem. Commun. 2012, 48, 1988−
Additional experimental details of preparation of materials and
radicals, measurement of quenching and enhancement factors,
impact of the number of methyl groups and of temperature on
the relaxation times, comparison of relaxation times at high and
low concentration, and detailed MAS curves are available. This
material is available free of charge via the Internet at http://
pubs.acs.org.
1990.
(17) Lafon, O.; Thankamony, A. S. L.; Kobayashi, T.; Carnevale, D.;
Vitzthum, V.; Slowing, I. I.; Kandel, K.; Vezin, H.; Amoureux, J.-P.;
Bodenhausen, G.; Pruski, M. J. Phys. Chem. C 2013, 117, 1375−1382.
(18) Lee, D.; Takahashi, H.; Thankamony, A. S. L.; Dacquin, J.-P.;
Bardet, M.; Lafon, O.; Paepe, G. D. J. Am. Chem. Soc. 2012, 134,
̈
18491−18494.
(19) Blanc, F.; Sperrin, L.; Jefferson, D. A.; Pawsey, S.; Rosay, M.;
Grey, C. P. J. Am. Chem. Soc. 2013, 135, 2975−2978.
(20) Samantaray, M. K.; Alauzun, J.; Gajan, D.; Kavitake, S.; Mehdi,
AUTHOR INFORMATION
Corresponding Author
paul.tordo@univ-amu.fr; lyndon.emsley@ens-lyon.fr
■
A.; Veyre, L.; Lelli, M.; Lesage, A.; Emsley, L.; Coper
C. J. Am. Chem. Soc. 2013, 135, 3193−3199.
(21) Gruning, W. R.; Rossini, A. J.; Zagdoun, A.; Gajan, D.; Lesage,
́
et, C.; Thieuleux,
̈
Notes
A.; Emsley, L.; Coper
13274.
(22) Rossini, A. J.; Zagdoun, A.; Lelli, M.; Lesage, A.; Coper
Emsley, L. Acc. Chem. Res. 2013, DOI: 10.1021/ar300322x.
(23) Rossini, A. J.; Zagdoun, A.; Hegner, F.; Schwarzwalder, M.;
́
et, C. Phys. Chem. Chem. Phys. 2013, 15, 13270−
The authors declare no competing financial interest.
́
et, C.;
ACKNOWLEDGMENTS
■
A.J.R. acknowledges support from a EU Marie-Curie IIF
Fellowship (PIIF-GA-2010-274574). M.S. and C.C. thank SNF
for financial support (200021_134775/1). Financial support is
acknowledged from EQUIPEX contract ANR-10-EQPX-47-01,
̈
́
Gajan, D.; Coperet, C.; Lesage, A.; Emsley, L. J. Am. Chem. Soc. 2012,
134, 16899−16908.
(24) Hu, K.-N. Solid State Nucl. Magn. Reson. 2011, 40, 31−41.
(25) Ni, Q. Z.; Daviso, E.; Can, T. V.; Markhasin, E.; Jawla, S. K.;
Swager, T. M.; Temkin, R. J.; Herzfeld, J.; Griffin, R. G. Acc. Chem. Res.
2013, DOI: 10.1021/ar300348n.
(26) Hovav, Y.; Feintuch, A.; Vega, S. J. Magn. Reson. 2012, 214, 29−
41.
ERC Advanced grant no. 320860, and the ETH Zurich. We
̈
acknowledge Dr. Moreno Lelli and Dr. David Gajan for their
insightful conversations and advice and Mr. W. Gruning and
̈
Dr. M. Conley for preparation of materials.
(27) Hu, K.-N.; Yu, H.-H.; Swager, T. M.; Griffin, R. G. J. Am. Chem.
Soc. 2004, 126, 10844−10845.
REFERENCES
(1) Overhauser, A. Phys. Rev. 1953, 92, 411−415.
(2) Carver, T. Phys. Rev. 1956, 102, 975−981.
(3) Maly, T.; Debelouchina, G. T.; Bajaj, V. S.; Hu, K.-N.; Joo, C.-G.;
Mak Jurkauskas, M. L.; Sirigiri, J. R.; van der Wel, P. C. A.; Herzfeld, J.;
Temkin, R. J.; Griffin, R. G. J. Chem. Phys. 2008, 128, 052211.
(4) Becerra, L.; Gerfen, G.; Temkin, R.; Singel, D.; Griffin, R. Phys.
Rev. Lett. 1993, 71, 3561−3564.
(5) Gerfen, G.; Becerra, L.; Hall, D.; Griffin, R. J. Chem. Phys. 1995,
102, 9494−9496.
(6) Rosay, M.; Lansing, J. C.; Haddad, K. C.; Bachovchin, W. W.;
Herzfeld, J.; Temkin, R. J.; Griffin, R. G. J. Am. Chem. Soc. 2003, 125,
13626−13627.
(7) Andreas, L. B.; Barnes, A. B.; Corzilius, B.; Chou, J. J.; Miller, E.
A.; Caporini, M.; Rosay, M.; Griffin, R. G. Biochemistry 2013, 52,
2774−2782.
(8) Sergeyev, I. V.; Day, L. A.; Goldbourt, A.; McDermott, A. E. J.
Am. Chem. Soc. 2011, 133, 20208−20217.
■
(28) Song, C.; Hu, K.-N.; Joo, C.-G.; Swager, T. M.; Griffin, R. G. J.
Am. Chem. Soc. 2006, 128, 11385−11390.
(29) Matsuki, Y.; Maly, T.; Ouari, O.; Karoui, H.; Le Moigne, F.;
Rizzato, E.; Lyubenova, S.; Herzfeld, J.; Prisner, T.; Tordo, P.; Griffin,
R. G. Angew. Chem., Int. Ed. 2009, 48, 4996−5000.
(30) Zagdoun, A.; Rossini, A. J.; Gajan, D.; Bourdolle, A.; Ouari, O.;
Rosay, M.; Maas, W. E.; Tordo, P.; Lelli, M.; Emsley, L.; Lesage, A.;
́
Coperet, C. Chem. Commun. 2012, 48, 654−656.
(31) Dane, E. L.; Corzilius, B.; Rizzato, E.; Stocker, P.; Maly, T.;
Smith, A. A.; Griffin, R. G.; Ouari, O.; Tordo, P.; Swager, T. M. J. Org.
Chem. 2012, 77, 1789−1797.
(32) Kiesewetter, M. K.; Corzilius, B.; Smith, A. A.; Griffin, R. G.;
Swager, T. M. J. Am. Chem. Soc. 2012, 134, 4537−4540.
(33) Casabianca, L. B.; Shames, A. I.; Panich, A. M.; Shenderova, O.;
Frydman, L. J. Phys. Chem. C 2011, 115, 19041−19048.
(34) Zagdoun, A.; Casano, G.; Ouari, O.; Lapadula, G.; Rossini, A. J.;
Lelli, M.; Baffert, M.; Gajan, D.; Veyre, L.; Maas, W. E.; Rosay, M.;
́
Weber, R. T.; Thieuleux, C.; Coperet, C.; Lesage, A.; Tordo, P.;
(9) Renault, M.; Pawsey, S.; Bos, M. P.; Koers, E. J.; Nand, D.;
Tommassen-van Boxtel, R.; Rosay, M.; Tommassen, J.; Maas, W. E.;
Baldus, M. Angew. Chem., Int. Ed. 2012, 51, 2998−3001.
Emsley, L. J. Am. Chem. Soc. 2012, 134, 2284−2291.
(35) Michaelis, V. K.; Smith, A. A.; Corzilius, B.; Haze, O.; Swager, T.
M.; Griffin, R. G. J. Am. Chem. Soc. 2013, 135, 2935−2938.
(36) Rosay, M.; Tometich, L.; Pawsey, S.; Bader, R.; Schauwecker, R.;
Blank, M.; Borchard, P. M.; Cauffman, S. R.; Felch, K. L.; Weber, R.
T.; Temkin, R. J.; Griffin, R. G.; Maas, W. E. Phys. Chem. Chem. Phys.
2010, 12, 5850.
(10) Takahashi, H.; Ayala, I.; Bardet, M.; De Paepe, G.; Simorre, J.-
̈
P.; Hediger, S. J. Am. Chem. Soc. 2013, 135, 5105−5110.
(11) Jacso, T.; Franks, W. T.; Rose, H.; Fink, U.; Broecker, J.; Keller,
S.; Oschkinat, H.; Reif, B. Angew. Chem., Int. Ed. 2012, 51, 432−435.
(12) Lesage, A.; Lelli, M.; Gajan, D.; Caporini, M. A.; Vitzthum, V.;
(37) Rossini, A. J.; Zagdoun, A.; Lelli, M.; Gajan, D.; Rascon
Rosay, M.; Maas, W. E.; Coperet, C.; Lesage, A.; Emsley, L. Chem. Sci.
2012, 3, 108−115.
́
, F.;
Miev
Bodenhausen, G.; Coper
15459−15461.
́
ille, P.; Alauzun, J.; Roussey, A.; Thieuleux, C.; Mehdi, A.;
et, C.; Emsley, L. J. Am. Chem. Soc. 2010, 132,
́
́
12796
dx.doi.org/10.1021/ja405813t | J. Am. Chem. Soc. 2013, 135, 12790−12797

Journal of the American Chemical Society
Article
(38) Thurber, K. R.; Tycko, R. J. Magn. Reson. 2009, 196, 84−87.
(39) Rosay, M. Ph.D. Thesis, Massachuchetts Institute of Technology,
Cambridge, MA, 2001.
(40) Ravera, E.; Corzilius, B.; Michaelis, V. K.; Rosa, C.; Griffin, R.
G.; Luchinat, C.; Bertini, I. J. Am. Chem. Soc. 2013, 135, 1641−1644.
(41) Weil, J. A.; Bolton, J. R. Electron Paramagnetic Resonance; Wiley-
Interscience: Hoboken, NJ, 2007.
(42) Sato, H.; Kathirvelu, V.; Fielding, A.; Blinco, J. P.; Micallef, A. S.;
Bottle, S. E.; Eaton, S. S.; Eaton, G. R. Mol. Phys. 2007, 105, 2137−
2151.
(43) Sato, H.; Kathirvelu, V.; Spagnol, G.; Rajca, S.; Rajca, A.; Eaton,
S. S.; Eaton, G. R. J. Phys. Chem. B 2008, 112, 2818−2828.
(44) Eaton, S.; Harbridge, J.; Rinard, G.; Eaton, G.; Weber, R. Appl.
Magn. Reson. 2001, 20, 151−157.
(45) Kathirvelu, V.; Smith, C.; Parks, C.; Mannan, M. A.; Miura, Y.;
Takeshita, K.; Eaton, S. S.; Eaton, G. R. Chem. Commun. (Camb.)
2009, 454−456.
̈
(46) Akbey, U.; Linden, A. H.; Oschkinat, H. Appl. Magn. Reson.
2012, 43, 81−90.
́
(47) Mieville, P.; Vitzthum, V.; Caporini, M. A.; Jannin, S.; Gerber-
Lemaire, S.; Bodenhausen, G. Magn. Reson. Chem. 2011, 49, 689−692.
(48) Rajca, A.; Kathirvelu, V.; Roy, S. K.; Pink, M.; Rajca, S.; Sarkar,
S.; Eaton, S. S.; Eaton, G. R. Chemistry 2010, 16, 5778−5782.
̈
(49) Akbey, U.; Franks, W. T.; Linden, A.; Lange, S.; Griffin, R. G.;
van Rossum, B.-J.; Oschkinat, H. Angew. Chem., Int. Ed. 2010, 49,
7803−7806.
̈
(50) Mentink-Vigier, F.; Akbey, U.; Hovav, Y.; Vega, S.; Oschkinat,
H.; Feintuch, A. J. Magn. Reson. 2012, 224, 13−21.
(51) Thurber, K. R.; Tycko, R. J. Chem. Phys. 2012, 137, 084508.
(52) Zagdoun, A.; Rossini, A. J.; Conley, M. P.; Gruning, W. R.;
̈
́
Schwarzwalder, M.; Lelli, M.; Franks, W. T.; Oschkinat, H.; Coperet,
̈
C.; Emsley, L.; Lesage, A. Angew. Chem., Int. Ed. 2013, 52, 1222−1225.
(53) Wessig, M.; Drescher, M.; Polarz, S. J. Phys. Chem. C 2013, 117,
2805−2816.
12797
dx.doi.org/10.1021/ja405813t | J. Am. Chem. Soc. 2013, 135, 12790−12797