Understanding the: G-tensors of perchlorotriphenylmethyl and Finland-type trityl radicals

Article abstract of DOI:10.1039/d0cp03626a

The 285 GHz EPR spectra of perchlorotriphenylmethyl and tetrathiatriarylmethyl radicals in frozen solution have been accurately measured. The relationship between their molecular structures and their g-tensors has been investigated with the aid of DFT cal

Full text of DOI:10.1039/d0cp03626a

PCCP
View Article Online
View Journal
PAPER
Understanding the g-tensors of
perchlorotriphenylmethyl and Finland-type
Cite this:DOI: 10.1039/d0cp03626a
trityl radicals†
c
Paul Demay-Drouhard, ‡¶a H. Y. Vincent Ching, §¶b Christophe Decroos,
d
e
b
a
Regis Guillot,
Yun Li,
Leandro C. Tabares,
Clotilde Policar,
*
´
b
Helene C. Bertrand *a and Sun Un
*
The 285 GHz EPR spectra of perchlorotriphenylmethyl and tetrathiatriarylmethyl radicals in frozen
solution have been accurately measured. The relationship between their molecular structures and their
g-tensors has been investigated with the aid of DFT calculations, revealing that the degree of spin
density delocalization away from the central methylene carbon is an important determining factor of the
g-anisotropy. In particular, the small amount of spin densities on the Cl or S heteroatoms at the 2 and 6
positions with respect to the central carbon have the strongest influence. Furthermore, the amount of
spin densities on these heteroatoms and thus the anisotropy can be modulated by the protonation
(esterification) state of the carboxylate groups at the 4 position. These results provide unique insights
into the g-anisotropy of persistent trityl radicals and how it can be tuned.
Received 7th July 2020,
Accepted 2nd September 2020
DOI: 10.1039/d0cp03626a
rsc.li/pccp
Introduction
Stable triphenylmethyl (trityl) radicals have been used in a wide
variety of fields including magnetic materials,1–3 molecular
switches,4,5 donor–acceptor systems,6–8 molecular junctions,9,10
dynamic nuclear polarization (DNP),11–17 EPR distance
measurements,18–27 and in situ and in vivo applications28 such as
MRI29–31 and EPR32–37 imaging. Perchlorotriphenylmethyl (PTM)38
and tetrathiatriarylmethyl (TAM)39 radicals are two classes of trityl
radicals (Fig. 1) that have received much attention. Two of their
unique and much exploited electronic properties are their long
phase memory time and narrow EPR spectrum even at the high
Fig. 1 (left) Molecular structures of PTM and TAM radicals discussed in
this work. (right) Numbering of carbon atoms of the aromatic rings.
magnetic fields used in modern NMR and high field EPR. The
latter makes them appealing for DNP and long-range dipolar
distance measurements. The narrowness of their EPR spectra in
part arises from their lack of strong hyperfine interactions, in
particular to 1H and 14N nuclei as is the case for nitroxides radicals.
Moreover, the g-tensors of PTM and TAM radicals have small
anisotropies. For most organic radicals, the largest contribution to
their g-tensors arises from spin–orbit coupling. For a doublet-state
(S = 1/2) radical, the atom-localized approximation for this con-
tribution is given by:40
a
´
´
Laboratoire des Biomolecules, LBM, Departement de Chimie, Ecole Normale
´
´
Superieure, PSL University, Sorbonne Universite, CNRS, 75005 Paris, France.
E-mail: helene.bertrand@ens.psl.eu, clotilde.policar@ens.psl.eu
b
´
Universite Paris-Saclay, CEA, CNRS, Institute for Integrative Biology of the Cell (I2BC),
91198, Gif-sur-Yvette, France. E-mail: sun.un@cea.fr
c
´
Aix Marseille Universite, CNRS, Centrale Marseille, iSm2, Marseille, France
d
´
´
´
Institut de Chimie Moleculaire et des Materiaux d’Orsay, Universite Paris-Sud,
´
CNRS UMR 8182, Universite Paris-Saclay, 91405 Orsay, France
e
`
Laboratoire de Chimie des Processus Biologiques, CNRS UMR 8229, College de
´
France, PSL University, Sorbonne Universite, 11 Place Marcelin Berthelot,
75231 Paris Cedex 05, France
ꢀ
ꢀ
ꢀ ꢀ
D
ED
E
P
ꢀ
ꢀ
ꢀ ꢀ
zk wðkp0 Þ li wðnÞ wðknÞ li wðpÞ
† Electronic supplementary information (ESI) available. See DOI: 10.1039/d0cp03626a
‡ Current address: Department of Chemistry, York University, 4700 Keele Street,
Toronto, Ontario M3J 1P3, Canada.
ꢀ
ꢀ
ꢀ ꢀ
k0
k
k
k0
X
k;k0
ii
SO
g
¼ 2
(1)
ep ꢀ en
nap;p
§ Current address: BIMEF – Department of Chemistry, and ELCAT – Faculty of
Applied Engineering, University of Antwerp, Universiteitsplein1, B-2610 Antwerp,
Belgium.
where w(kp) and wk(n) are the atomic orbitals of atom k that has spin–
orbit coupling constant zk; its angular momentum operator along
¶ These authors contributed equally.
Phys. Chem. Chem. Phys.
This journal is © the Owner Societies 2020

View Article Online
Paper
PCCP
the i direction (where i = x, y, z) is lik. The linear combination of two different superconducting magnets is that there is also a
w(kp) atomic orbitals form the ground state singly occupied orbital degradation in calibration over time and can vary from sample
(SOMO) molecular orbital (MO) cp with energy ep and, likewise, to sample. There are also differences in the calculated magnetic
w(kn) form cn with energy en, the rest of the other molecular orbitals field from the driving current and the actual magnetic field at
and energies. The denominator of eqn (1) can be both positive or the sample that can also be significant. Although these factors
negative. The former corresponds to the excitation of an electron may be significant in determining the absolute magnetic field
from a doubly occupied orbital to the SOMO and the latter the at the sample, for the most part they have a much less
excitation of the unpaired electron to an unoccupied orbital. important role in measuring differences in magnetic fields
Positive contributions are typified by excitation of a low-lying and accuracies of better than 0.05 mT can be achieved. Without
nonbonding electron to the SOMO. This is the case for tyrosyl, specification of the method, accuracy and reproducibility
semiquinone and nitroxide radicals. Another important factor of magnetic field measurements, trityl g-values, even those
revealed by eqn (1) is the product of the ground state spin density obtained using high magnetic fields, need to be viewed with
of an atom and its spin–orbit coupling constant. Delocalization of these limitations in mind.
unpaired electron density will tend to reduce g-anisotropy. When it
Using 285 GHz 10 T measurements on TAM and PTM type
is constrained to carbon atoms through a p-bonding network, the trityl radicals, we have examined the relationship between their
two angular momentum quantum mechanical terms will only g-tensors and their electronic structures. Both the relative and
have significant values if the excited-state is neither p nor p* in absolute accuracy of g-tensors were ensured by using Mn(II)
character which typically lead to larger absolute values of ep ꢀ en. and Gd(III) as calibration standards. The theoretical g-tensors
By contrast, delocalization onto heavy atoms will increase derived from DFT calculations were in good agreement with
g-anisotropy, since spin–orbit constants increase with Z (nuclear the measured values and provide the means to understanding
charge). As just discussed, this is more so if the atoms in question what electronic structural factors influenced the g-tensors.
also have electrons that occupy low-lying non-bonding MO. Both This knowledge will be helpful in tuning the structure of trityl
TAM and PTM trityl radicals (Fig. 1) have extensive electron radicals for a wide range of important applications.
delocalization, but also peripheral heavy atoms. However, despite
having similar structures and heavy atoms that have spin–orbit
coupling constants that are less than a factor two different,41 their
Results and discussion
EPR spectra exhibit significant differences.13–16,42–44
Although it is straightforward to ascertain the relative differ- Fig. 2 shows the 285 GHz cw-HFEPR spectra of FT, H3FT,
ences in the broadness of the different trityl EPR spectra, PTMTC, H3PTMTC and PTMTE. Above them, examples of two
high magnetic field EPR (HFEPR) has been generally required spectra used to calibrate the magnetic field are also shown: one
to quantitatively measure the small g-anisotropies of trityl of two physically separated frozen solutions of the FT radical
radicals. For example, 3 T/95 GHz EPR measurements have and Gd(III) coaxially mounted in the magnet and another of two
shown that the principal g-values of the tricarboxylate derivative frozen aqueous solutions of Mn(II) and Gd(III) in the same
of PTM (PTMTC) were gx = 2.00271 and gy = gz = 2.0000515 and physical arrangement. The six Mn(II) hyperfine resonances of
those of the TAM Finland Trityl (FT) derivative were [2.0030, the latter spectrum were separated on average by 9.523 mT with
2.0027, 2.0021]44 (here and in the following we will use this a standard deviation of 0.033 mT. This demonstrated that
compact [gx, gy, gz] notation and use the convention gx
Z
the field sweep was highly linear over this range. We have
gy Z gz. The subscripts in this notation do not necessarily carry previously demonstrated that this linearity extends over a much
any physical meaning nor are they necessarily related to the larger field range, including where Gd(III) resonates, by incre-
directions indicated in eqn (1). Similarly, 8.5 T/239 GHz mea- menting the microwave frequency to produce an overlapping
surements determined that those of the TAM OX063 derivative comb of Mn(II) magnetic field markers.46,47 The g-value of Mn(II)
were [2.00319, 2.00319, 2.00258].16
in water has been previously measured to be g = 2.00107 ꢂ
However, the use of high magnetic field presents challenges 0.00003.48 Based on this value, the g-value of Gd(III) was deter-
in measuring the magnetic field at the sample in a reproducible mined to be 1.99191 with comparable precision. In light of the
and accurate manner. For example, specification of a given broadness of the spectral features of the trityl radicals that were
g-value to 4-decimal places implies the magnetic field has been much larger than this level of precision and the assumptions
measured to better than 1.7 ꢁ 10ꢀ4 T (near ge) at a nominal made in simulating and fitting the g-tensors of the radicals, in the
field of 3.38983 T at 95 GHz and 5.1 ꢁ 10ꢀ4 T at a nominal field following, the geff (= f/(gB) where g = 13.99623 GHz Tꢀ1 and
of 10.16951 T at 285 GHz at the sample. Typically, the magnetic f = 285.0915 GHz) are reported to only 5-decimal places.
fields are measured as a function of the current driving
The spectra were indicative of axial g-tensors. However, the
the superconducting solenoid (for an exception see ref. 45). direction of axiality was different with the TAM radicals having
In principle, current measurements can afford accuracies to gx B gy and the PTM radicals gy B gz. The overall g-anisotropy,
less than 0.1 mT. In practice, there are factors that limit the defined as Dg = gx ꢀ gz, was significantly smaller for the former,
accuracy. For example, there are inductive effects that limit 10 ꢁ 10ꢀ4, compared to 25 ꢁ 10ꢀ4 for the latter. These values
current measurements that depend on not only sweep rates but were significantly narrower than 70 ꢁ 10ꢀ4 for stable nitroxide49,50
also magnet design (number of solenoids). Our experience with and 55 ꢁ 10ꢀ4 for tyrosyl radicals,51 but still considerably broader
Phys. Chem. Chem. Phys.
This journal is © the Owner Societies 2020

View Article Online
PCCP
Paper
parameters and a scaling factor used to simulate the measured
spectra. This model was fitted to the experimental data using a
standard nonlinear least squares algorithm that minimized the
root-mean-squared difference between the model and data.
As can be seen in Fig. 2 with the exception of the spectrum of
the FT radical, simulations based on this model fit the experi-
mental data extremely well. The Gaussian convolution widths
closely matched the amplitude of magnetic field modulation
used to record the data. The anisotropic contributions to the
inhomogeneous lineshape were, if any, small. The best fits to
this simple model were rhombic. The size of these distributions
and the fitted principal g-values are tabulated in Table 1.
The model could not be fitted to the entirety of the FT
radical spectrum. However, as Fig. 3 shows, the model did fit
the rising and falling edges when the minimization was con-
strained to these regions. The difference between the spectrum
and this constrained fit was very similar to the spectrum of
1 mM radical in water. This suggested that the FT radical
spectrum was composed of two components, a resolved com-
ponent with a g-anisotropy of 0.0011 and another with sub-
stantial smaller g-anisotropy. One possible explanation for the
multiple spectral contribution was the presence of aggregates
in the sample due to the lipophilicity of the molecule. Under
conditions that were similar to those previously reported for
DNP measurements17,53,54 (15 mM H2O/glycerol 1 : 1 solution),
the spectrum of FT also appeared to contain an unresolved
aggregation component. Compared to the 9 : 1 H2O/glycerol
sample no changes in the g-values were expected, but rather
the amount of aggregation component. The spectral features
observed in the 1 : 1 H2O/glycerol sample were similar to those
reported for a FT derivative, OX63, obtained at 239.2 GHz under
Fig. 2 (top) 285 GHz cw-HFEPR spectra of coaxially mounted frozen
aqueous solutions of: (black) FT radical and Gd(III) and (red) Mn(II) and
Gd(III). (bottom) Spectra of FT, H3FT, PTMTC, H3PTMTC, and PTMTE (black),
fitted (dashed-red) and DFT simulations (blue). The dashed line indicates the same sample conditions.16 Taken together, our observations
the free electron g-value (ge) and the dotted black lines the trend in gx.
are consistent with the spontaneous formation of spherical
nanoparticles and fibers by FT radical that have been recently
reported.55
than 5 ꢁ 10ꢀ4 for a methyl radical in CO below 4.2 K.52 Equally
The principal g-values of the FT and PTMTC radicals have
significant was that the TAM and PTM g-tensors were sensitive been previously reported.15,44 For the former, the values
to the protonation state with protonation increasing the g-aniso- reported in this study and the previous ones were in good
tropy. The cw-HFEPR spectrum of the deprotonated FT radical agreement, the main difference being 3 ꢁ 10ꢀ4 in the isotropic
dissolved in water appeared to be narrower and less resolved than g-values. For the latter, isotropic g-values differed by 26 ꢁ 10ꢀ4
.
in 9 : 1 H2O/glycerol. At the higher radical and glycerol concentra- However, there was closer agreement in Dg values. This sug-
tions (15 mM in 1 : 1 H2O/glycerol), the spectral features were gested the differences between the previously reported values
broader leading to lower resolution, but with essentially the same and those reported here were due to accuracy in the absolute
g-anisotropy.
magnetic field calibration. Neither of these previous reports
A simple model was used to simulate and fit the spectra. addressed the accuracy of the g-values.
This model consisted of the three principal g-values and three
Although the variation among the five trityl radicals is small,
parameters that described the Gaussian distribution of each of it was evident that their g-tensors were sensitive to the differ-
these three principal g-values. This meant that the distribution ences in electronic structure and environment. In addition to
in g-tensor and any other anisotropic inhomogeneous contri- the theory developed by Stone,40,56 g-tensors of organic radicals
butions to the lineshape were assumed to be independent of were analyzed using Density Functional Theory (DFT) calculations.
each other and coaxial with the three principal directions of the The B3LYP/6-31+G(D,P) combination of hybrid density functional
g-tensor. The powder spectra generated from these values were and basis-set was used to geometry optimize the model structures
then convolved with a derivative Gaussian function to obtain of the five trityl radicals, as well as to calculate the g-tensors of the
derivative powder patterns and account for the magnetic field radicals.57 These calculations took into account the solvent using
modulation. The only significant factor in this convolution was the Polarizable Continuum Model.58 They did not account for any
the width of the Gaussian function. Together there were seven specific solvent–radical interactions, such as hydrogen bonding
Phys. Chem. Chem. Phys.
This journal is © the Owner Societies 2020

View Article Online
Paper
PCCP
Table 1 Measured and calculated B3LYP/6-31+G(d,p) DFT g-tensors of the TAM and PTM and radicals. The measured values and their corresponding
distributions (in parenthesis) were obtained from simulations
Radical
gx
gy
gz
giso
2.00291
2.0026
2.00286
ꢀ0.00005
Dg
0.00106
0.0009
0.00092
ꢀ0.00014
FT
Measureda
Measuredc
DFTa
2.00339 (0.00002)
2.0030
2.00320
ꢀ0.00019
2.00301 (0.00004)
2.0027
2.00309
2.00233 (0.00008)
2.0021
2.00228
ꢀ0.00005
Db
0.0008
H3FT
Measureda
2.00358 (0.00002)
2.00361
0.00003
2.00340 (0.00004)
2.00342
0.00002
2.00262 (0.00004)
2.00262
0.00000
2.00320
2.00322
0.00002
0.00096
0.00099
0.00003
DFTa
D
Protonation shift
0.00019
0.00039
0.00029
PTMTC
Measureda
Measuredd
DFTa
2.00410 (0.00001)
2.00271
2.00443
2.00236 (0.00026)
2.00005
2.00149
ꢀ0.00087
2.00176 (0.00021)
2.00005
2.00148
ꢀ0.00028
2.00274
2.00009
2.00247
0.00027
0.00234
0.00266
0.00295
0.00061
D
0.00033
H3PTMTC
Measureda
DFTa
2.00429 (0.00000)
2.00498
0.00069
2.00168 (0.00007)
2.00120
2.00132 (0.00007)
2.00116
2.00243
2.00245
0.00002
0.00297
0.00382
0.00085
D
ꢀ0.00048
ꢀ0.00016
Protonation shift
0.00019
ꢀ0.00068
ꢀ0.00044
PTMTE
Measureda
2.00458 (0.00000)
2.00493
0.00035
2.00156 (0.00011)
2.00120
2.00125 (0.00010)
2.00117
2.00246
2.00243
ꢀ0.00003
0.00333
0.00377
0.00044
DFTae
D
Esterification shift
ꢀ0.00036
ꢀ0.00008
0.00048
ꢀ0.00080
ꢀ0.00051
Methyl
Measuredf (CO matrix)
DFTa
2.0027
2.0029
2.0027
2.0029
2.0022
2.0022
2.0025
2.0027
0.0005
0.0007
a
b
c
d
e
f
This work. Difference between DFT and measured. Ref. 44. Ref. 15. Modeled as the methyl ester derivative. Ref. 52.
between water molecules and carboxylate groups in the case of FT
and PTMTC. The optimized structures were found to be in good
agreement with previously reported X-ray crystallographic
structures.59–61 Most of the unpaired electron spin density was
localized on the central carbon, but the amount was only about
42 to 51%. As can be seen from Table 2, the rest of spin density was
localized onto the 2, 6 and 4 carbons and the heteroatoms bonded
to the 2 and 6 carbons. Protonation and esterification of the
peripheral carboxylate groups lead to an increase in spin density
at the center carbon and position 2 and 6 heteroatoms and to a
decrease at C2,6 and C4.
The DFT derived principal g-values are reported in Table 1
and simulations based on them are shown in Fig. 2. The DFT
calculations captured the differences in the g-tensors of PTM
and TAM radicals. The calculated isotropic g-values of four of
Table
2
The total B3LYP/6-31+G(D,P) Hirshfeld spin densities for
selected atom types (X = S or Cl)
Center
C2,6
C4
X–(C2,6)
FT
H3FT
PTMTC
H3PTMTC
PTMTE
0.42
0.43
0.49
0.51
0.50
0.36
0.35
0.33
0.33
0.33
0.17
0.14
0.15
0.14
0.14
0.09
0.10
0.07
0.08
0.08
Fig. 3 285 GHz cw-HFEPR spectra of FT: (top) in 9 : 1 H2O/glycerol
(black) its fit (red) and the difference between the two (green); (middle)
in aqueous solution without glycerol; (bottom) under typical DNP
conditions.
Phys. Chem. Chem. Phys.
This journal is © the Owner Societies 2020

View Article Online
PCCP
Paper
the five radicals were within 0.00005 of their measured values.
The largest difference of 0.0003 for the PTMTC was in part
associated with the noticeably large distribution in g-values.
Except for FT (see above), the calculated g-anisotropies were
over-estimated by as much as 0.0009. However, they nearly
quantitatively predicted the influence of the peripheral carboxyl
groups.
The main contribution to the g-tensors is the orbital
Zeeman/spin–orbit contribution (the contributions to each
of the DFT derived g-tensors are shown in the ESI†). The
relativistic and diamagnetic components were not only small,
but also essentially cancelled each other leaving a difference of
5 ꢁ 10ꢀ5. Hence, the departure from the ge, as depicted in
Fig. 2, was a measure of the larger remaining orbital Zeeman/
spin–orbit contribution. In the framework of Stone’s theory,
shifts to values lower than ge arise from mixing of the ground
state with excited states that results from the excitation of the
unpaired electron in the singly occupied molecular orbital
(SOMO) to a higher unoccupied state. By contrast, g-values
above ge originate from mixing of the ground state with an
excited state arising from the excitation of an electron from a
lower doubly occupied state to the SOMO.
The former is the case for the PTM gy and gz values and the
latter, for the FT g-tensor as a whole. These differences between
the TAM and PTM radicals can be qualitatively understood
from their ground-state SOMO and lowest unoccupied mole-
cular orbitals (LUMO). The significant p*-like interactions of
the C2,6 sulfur pz orbitals with the center carbon pz orbital make
the FT SOMO more spherically distributed about the central
carbon giving the SOMO no significant directionality. This is
apparent in the region delimited by the black circles in Fig. 4.
This spherical distribution results in lower g-anisotropy. This is
not so for the PTMTC SOMO which has three distinct directions,
one along the pz and two in the perpendicular plane. Although
pz orbitals of the C2,6 chlorine atoms do participate in p*-like
interactions with the center carbon pz orbital, their contributions
are much smaller than those of the FT radicals. The SOMO
heteroatoms circled in yellow in Fig. 4 demonstrate this. In the
case of LUMOs, the FT C2,6 sulfur atoms pz orbitals are involved in
anti-bonding interactions. By contrast for the PTM radicals, the
p-orbitals of the C2,6 chlorine atoms are non-bonding in character.
In this case, the spin–orbit mixing of SOMO and LUMO would
lead to shift the g-values to lower values since ep ꢀ en is negative
and would explain why the PTM gy and gz values are below ge. The
Fig. 4 The LUMO (top row) and two orthogonal views of the SOMO of the
FT (left) and PTMTC (right) radicals obtained from the DFT calculations
(isosurface cutoff: 0.02, figure created with Avogadro,83 see ESI† for more
detailed graphics material).
Fig. 5 Orientation of the g-tensors of the FT (left), H3FT (center) and
PTMTC (right). The view is perpendicular to the plane defined by the center
C and the three C1 atoms. Hydrogens have been omitted for clarity (figure
created with PyMOL82).
Not only were the g-anisotropies of FT radicals significantly
comparably important FT orbitals were further in energy with smaller but the DFT calculations predicted that the gx,y
-
many of the low-lying excited states involving sulfur p and p* directions are inverted upon deprotonation (Fig. 5). By contrast,
bonding interactions. The nearest excited state(s) that contribute neither protonation nor esterification affected the g-tensor
to the g-tensor evidently were those involving the excitation of an orientation in the PTM radicals (Fig. 5). The trityl radicals are
electron from a doubly occupied state to the SOMO having more complex than p-planar radicals such as semiquinones
significant contribution from the px or py-orbitals of the 2 and 6 and tyrosyl radicals. For such planar radicals, the magnitude
position sulfur atoms. These would result in a positive spin–orbit and direction of effects such as hydrogen bonding can be
contribution. Although the ground state spin densities of the C2,6 understood and rationalized in terms of spin polarization,
heteroatoms are significantly larger for FT radicals, their effect on n - p* transition energy and atomic orbit coupling terms
the g-values is largely diminished by the energy difference term such as hpx|ly|pzi, the directionality of which is well defined.
while the opposite is true for the PTM radicals with the energy As can be seen from the spin densities in Table 2, protonation
difference having a larger effect.
increases the ground state spin density of the C2,6 heteroatoms.
Phys. Chem. Chem. Phys.
This journal is © the Owner Societies 2020

View Article Online
Paper
PCCP
This increases their atomic orbital spin–orbit contributions to or replacing the CO2R groups with more electron-withdrawing
the g-tensors. This is consistent with the measurements which groups are likely to decrease spin density at the 2 and 6 position
show that for the FT radical, all three principal g-values are heteroatoms.
further shifted away from ge upon protonation. The same is
Trityl derivatives with smaller g-anisotropies, and thus
true for PTM radicals except the shifts are in both directions narrower spectra, would be of interest for high field EPR
with respect to ge. Hence, the larger g-anisotropies of proto- dipolar distance measurements. The inherent sensitivity gains
nated (and esterified) forms appear to be in part due to increase by going to higher field are offset by complications due to the
in spin densities at the heteroatoms. The energy differences broadening of the trityl EPR spectrum. The resulting partial
between the coupled excited and ground states (ep ꢀ en) are also excitation of the trityl EPR spectra can lead to orientation
likely to contribute, but not in a direct way since the carboxylate selection effects, which are not always desirable and make
groups carry virtually no spin density (o0.01 in total). The more sensitive single frequency experiments such as DQC or
effect of protonation on the direction of g-tensors appears to SIFTER challenging.21 Trityl radicals with lower g-anisotropies
reflect the three-dimensional nature of the unpaired spin would help circumvent these limitations and increase their
density of the trityl radicals. In the case of PTM radicals that utility and appeal. The ability to rationally control the trityl
have well-defined directions, the g-tensor orientation remained g-anisotropies will help engineer and analyze novel trityl con-
unchanged. The more spherically distributed FT radicals also taining multi-radical species that display complex EPR spectra
have smaller g-anisotropies causing the directionality of the with dipolar and exchange interactions, such as trityl-nitroxide
g-tensor to be sensitive to small changes.
biradicals that have recently emerged as very promising para-
In general, the heteroatoms in the 2 and 6 position appear to magnetic polarizing agents for MAS DNP at high field.54
play an important role in the FT and PTM radical g-tensors.
The more than a factor of two larger g-anisotropies of the PTM
radicals compared to those of the FT’s mirrors the 1.5 ratio of
Conclusions
atomic spin–orbit coupling constants of sulfur and chlorine.41
By contrast, the carbon atoms are not likely to be as important. Using cw-HFEPR at 285 GHz we have accurately determined the
Although they carry most of the unpaired electron spin density, g-tensors of FT, H3FT, PTMTC, H3PTMTC and PTMTE trityl
the central carbons of the TAM and PTM trityl radicals do not radicals in frozen solution. The differences between their
play a significant role in determining the g-tensors of either of g-tensors were rationalized using DFT calculations. Analysis
these radicals. The reasons for this are the same as those that of the SOMO and LUMO orbitals of the trityl radicals provided
result in a small g-anisotropy for the methyl radical (Table 1), a qualitative explanation of the differences between their
the unpaired spin density of which resides almost entirely on g-tensors. Our results suggest that the preparation of trityl
the pz orbital of the carbon atom. In spite of the highly localized radicals where the spin density is more delocalized over the
spin density, the spin–orbit contribution is very small because triphenylmethyl core and away from heteroatoms would poten-
the contribution from the terms hpx|ly|pzi and hpy|lx|pzi in tially yield derivatives with even lower g-anisotropies, making
eqn (1) involve orbitals that are involved in covalent bonds them more advantageous for many applications.
and are small, and (ep ꢀ en) of these terms will be large.
Consequently, the spin–orbit contribution will be small. With
less localized spin density than the methyl radical, the atomic
spin–orbit contribution from the center C atom in the TAM and
PTM radicals would be expected to be less and therefore
Material and methods
General
unlikely to be the determining factor in the g-anisotropy. This 1H and 13C NMR spectra were recorded on a Bruker Avance 300
suggests that in terms of designing trityl radicals it is likely to spectrometer using solvent residuals as internal references.
be more useful to focus on the nature of heteroatoms and The following abbreviations are used: singlet (s), triplet (t),
R(C4)-groups. In this context the TAM and PTM molecules can quadruplet (q). Mass spectrometry services were provided by
´
be considered as different radicals rather than derivatives of the ICMMO (Universite Paris Sud, Orsay, France). The following
the simple Gomberg triphenylmethyl radical. abbreviations are used: HRMS (high resolution mass spectro-
With these results in hand it is possible to formulate metry), atmospheric pressure chemical ionization (APCI), electro-
approaches for obtaining persistent trityl radicals with even spray (ESI). TLC analysis was carried out on silica gel (Merck
smaller g-anisotropies. The spin–orbit constants of heteroa- 60F-254) with visualization at 254 and 366 nm. Preparative flash
toms at the 2 and 6 positions should be as small as possible chromatography was carried out with Merck silica gel (Si 60,
and their spin densities minimal. Extended spin delocalization 40–63 mm). All reactions were performed under an Ar inert
over the carbon backbone trityl core is desirable. The mesityl- atmosphere unless otherwise stated. Dry tetrahydrofuran (THF)
substituted tri(9-anthryl)methyl radical62 is an example of this. was purchased from Sigma and used without further purification.
Another possibility would be to tune the spin delocalization MilliQt water was used for all experiments requiring water.
using the 4 positions of the trityls. Extended delocalization All other solvents and reagents were purchased from commercial
would also reduce any contributions from proton hyperfine sources and used without further purification. The tripotassium
interactions. Incorporating aromatic moieties at these positions salt of Finland trityl (FT) was synthesized according to a literature
Phys. Chem. Chem. Phys.
This journal is © the Owner Societies 2020

View Article Online
PCCP
Paper
procedure.63 The synthesis of PTMTE and H3PTMTC is summar- washed with Et2O (3ꢁ) and acidified with aq. 1 M HCl to obtain
ized in Scheme S1 (ESI†).
a red precipitate. This mixture was extracted with Et2O (3ꢁ) and
the combined organic layers were dried over Na2SO4, filtered
and concentrated to afford H3PTMTC (200 mg, 0.253 mmol,
Synthesis
Tris(2,3,5,6-tetrachlorophenyl)methane (2). 1,2,4,5-Tetra- 85%) as a red solid.
chlorobenzene 1 (9.6 g, 44 mmol, 9.0 eq.), AlCl3 (730 mg,
EPR sample preparation
5.2 mmol, 1.06 eq.) and chloroform (0.4 mL, 4.9 mmol, 1.0 eq.)
were mixed in a sealed high-pressure tube. The resulting mixture Unless otherwise stated the concentration of all trityl EPR
was heated at 160 1C for 45 min and cooled to rt. CH2Cl2 was samples were 1 mM. H3FT, H3PTMTC and PTMTE were dis-
added and the black suspension was sonicated for 30 min and solved in 2-Me-THF and FT and PTMTC in 9 : 1 H2O/glycerol. FT
washed with aq. 0.1 M HCl (1ꢁ), H2O (1ꢁ), and sat. aq. NaCl (1ꢁ). samples (1 or 15 mM) were prepared by dissolving K3FT in H2O
The organic layer was dried over Na2SO4, filtered and concen- with 10% v/v glycerol, H2O, or H2O with 50% v/v glycerol. The
trated. Column chromatography (SiO2, 1300 g, f 9 cm, 100% H3FT sample was prepared by addition of one drop (E1 mL) of
cyclohexane) followed by pentane washings afforded tris(2,3,5,6- concentrated HCl (10 M) to an aqueous solution of K3FT (1 mM,
tetrachlorophenyl)methane 2 (1.181 g, 1.796 mmol, 37%) as a 500 mL). The precipitate was isolated by centrifugation and
white solid. 1H-NMR (CDCl3, 300 MHz): d 7.65 (s, 3H), 6.98 (s, 1H); dissolved in 2-Me-THF (500 mL). The PTMTC sample was
13C-NMR (CDCl3, 75 MHz): d 138.7 (Cq), 134.5 (Cq), 133.7 (Cq), prepared in situ by deprotonation of H3PTMTC using excess
133.4 (Cq), 132.6 (Cq), 130.5 (CH), 56.2 (CH); HRMS (APCI): NaOH (20 eq.) in H2O with 10% v/v glycerol. The H3PTMTC and
+
m/z = 616.6886 [M ꢀ Cl]+ (found), 616.6881 calcd for C19H4Cl11
.
PTMTE samples were prepared by dissolving in 2-Me-THF. The
experimental setup consists of a 0.6 mL Eppendorf tube coaxially
Characterization in accordance with literature.64,65
Triethyl 4,40,400-methanetriyltris(2,3,5,6-tetrachlorobenzoate) mounted within a 2 mL Nalgenes cryogenic vial that are typically
(3). Compound 2 (730 mg, 1.11 mmol, 1.0 eq.) was dissolved used as sample tubes. A 85 mL solution of 50 mM GdCl3 in H2O
in dry THF (70 mL). TMEDA (530 mL, 3.55 mmol, 3.2 eq.) was with 10% v/v glycerol was added to the Eppendorf tube, while the
added and the resulting solution was cooled to ꢀ78 1C. n-BuLi cryotube contained 300 mL solution of either 50 mM Mn(ClO4)2 in
(1.6 M in THF, 2.36 mL, 3.77 mmol, 3.4 eq.) was added in one H2O with 20% v/v glycerol or the radical samples. Spectra were
portion and the resulting mixture was stirred at ꢀ78 1C for 2 h. obtained at 15 K with modulation of 5 G under non-saturating
Ethyl chloroformate (1.06 mL, 11.1 mmol, 10.0 eq.) was added conditions except for PTMTC, for which the spectrum was obtained
dropwise and the resulting solution was allowed to warm to rt for at 100 K and 10 G modulation to achieve non-saturation.
15 h and concentrated. H2O was added and the mixture was
EPR measurements
extracted with DCM (3ꢁ). The combined organic layers were
dried over Na2SO4, filtered and concentrated. Column chroma- The 285 GHz cw-HFEPR spectra were recorded on a locally
tography (99 : 1 to 90: 10 cyclohexane/EtOAc) afforded triethyl constructed HFEPR spectrometer which has been previous
4,40,400-methanetriyltris(2,3,5,6-tetrachlorobenzoate) 3 (640 mg, described.67 The magnetic power supply had a digital accuracy
0.732 mmol, 66%) as a white solid. H-NMR (CDCl3, 300 MHz): of 0.5 G.67 To insure non-saturating conditions, the signal
1
d 7.00 (s, 1H), 4.49 (q, J = 7.1 Hz, 6H), 1.43 (t, J = 7.1 Hz, 9H); amplitude was monitored as a function power of the 95 GHz
HRMS (APCI): m/z = 868.7303 [M + H]+ (found), 868.7282 synthesizer which drives the frequency tripler. The output
+
calcd for C28H17Cl12O6 . Characterization in accordance with power of the tripler was monotonic with respect to the 95 GHz
literature.65,66
input power that we used in these experiments. The relative change
PTMTE. Compound 3 (820 mg, 0.938 mmol, 1.0 eq.) was in 285 GHz output power was also monitored at the bolometer
dissolved in a mixture of dry DMSO (30 mL) and dry Et2O detector. A second criterion for non-saturating conditions was that
(135 mL), and then powdered NaOH (750 mg, 18.8 mmol, the integration of spectra, after baseline correction, yielded a proper
20.0 eq.) was added. The resulting mixture was stirred at rt for absorption line shape, in particular one where the baseline was
24 h protected from light and filtered to a solution containing zero on either side of the resonance. This criterion was found to be
iodine (1.05 g) dissolved in dry Et2O (55 mL). The resulting solution more stringent and could detect passage and saturation effects
was left undisturbed protected from light for 24 h and washed with associated with modulation that have been documented in early
sat. aq. NaHCO3 (1ꢁ), sat. aq. NaCl (1ꢁ), and H2O (2ꢁ). The works of Portis,68,69 Hyde70,71 and Weger.72 For this reason, suffi-
organic layer was dried over Na2SO4, filtered and concentrated. ciently large enough modulation amplitudes were used to fulfill
Column chromatography (95 : 5 to 80 : 20 cyclohexane/EtOAc) this integrability condition. The largest modulation amplitude used
afforded PTMTE (672 mg, 0.770 mmol, 82%) as a red solid. was 10 G. The apparent inhomogeneous broadening was substan-
Rf (SiO2): 0.65 (80 : 20 cyclohexane/EtOAc); HRMS (ESI): m/z = tially larger and, thus, had little effect on the quality of spectra with
+
+
ꢃ
889.7008 [M + Na]ꢃ (found), 889.7023 calcd for C28H15Cl12NaO6
.
respect to resolution.
Characterization in accordance with literature.65,66
Density functional calculations
H3PTMTC. PTMTE (260 mg, 0.298 mmol, 1.0 eq.) was
dissolved in MeOH (5 mL). KOH (836 mg, 14.9 mmol, Density functional calculations were carried out with Gaussian 09
50.0 eq.) was added and the resulting mixture was refluxed (revision B.01).73 The B3LYP hybrid density functional74–77 and
for 48 h and cooled to rt. H2O was added, the mixture was 6-31+G(d,p) basis set78–81 were used for geometry optimization
Phys. Chem. Chem. Phys.
This journal is © the Owner Societies 2020

View Article Online
Paper
PCCP
12 J. Leggett, R. Hunter, J. Granwehr, R. Panek, A. J. Perez-
Linde, A. J. Horsewill, J. McMaster, G. Smith and
and subsequent calculation of g-values. For FT and PTMTC, the
calculation included water solvation and for H3FT, H3PTMTC
and PTMTE, solvation in THF using the default polarizable
continuum model.58
¨
W. Kockenberger, Phys. Chem. Chem. Phys., 2010, 12,
5883–5892.
13 C. Gabellieri, V. Mugnaini, J. C. Paniagua, N. Roques,
M. Oliveros, M. Feliz, J. Veciana and M. Pons, Angew. Chem.,
Int. Ed., 2010, 49, 3360–3362.
Conflicts of interest
14 J. C. Paniagua, V. Mugnaini, C. Gabellieri, M. Feliz,
N. Roques, J. Veciana and M. Pons, Phys. Chem. Chem. Phys.,
2010, 12, 5824–5829.
There are no conflicts to declare.
15 D. Banerjee, J. C. Paniagua, V. Mugnaini, J. Veciana,
A. Feintuch, M. Pons and D. Goldfarb, Phys. Chem. Chem.
Phys., 2011, 13, 18626–18637.
16 L. Lumata, Z. Kovacs, A. D. Sherry, C. Malloy, S. Hill, J. Van
Tol, L. Yu, L. Song and M. E. Merritt, Phys. Chem. Chem.
Phys., 2013, 15, 9800–9807.
17 G. Mathies, M. A. Caporini, V. K. Michaelis, Y. Liu, K.-N. Hu,
D. Mance, J. L. Zweier, M. Rosay, M. Baldus and
R. G. Griffin, Angew. Chem., Int. Ed., 2015, 54, 11770–11774.
18 Z. Yang, Y. Liu, P. Borbat, J. L. Zweier, J. H. Freed and
W. L. Hubbell, J. Am. Chem. Soc., 2012, 134, 9950–9952.
19 G. W. Reginsson, N. C. Kunjir, T. Sigurdsson and
O. Schiemann, Chem. – Eur. J., 2012, 18, 13580–13584.
20 G. Y. Shevelev, O. A. Krumkacheva, A. A. Lomzov, A. A.
Kuzhelev, O. Y. Rogozhnikova, D. V. Trukhin, T. I.
Troitskaya, V. M. Tormyshev, M. V. Fedin and D. V.
Pyshnyi, et al., J. Am. Chem. Soc., 2014, 136, 9874–9877.
Acknowledgements
This work was financially supported by ANR MnHFPELDOR,
ANR-Deutsche Forschungsgemeinschaft (DFG) Chemistry (2011-
INTB-1010-01, P. D.-D. PhD studentship and H. Y. V. C. post-
doctoral fellowship) and has benefited from the biophysics platform
of the Institute for Integrative Biology of the Cell supported by
French Infrastructure for Integrated Structural Biology (FRISBI) ANR-
10-INBS-05-05. We are very grateful to Drs Dmitry Akhmetzyanov
(University of Waterloo) and Melissa Van Landeghem (University of
Antwerp) for fruitful and constructive discussions.
Notes and references
1 D. Maspoch, D. Ruiz-Molina, K. Wurst, N. Domingo,
M. Cavallini, F. Biscarini, J. Tejada, C. Rovira and
J. Veciana, Nat. Mater., 2003, 2, 190–195.
¨
21 D. Akhmetzyanov, P. Schops, A. Marko, N. C. Kunjir,
2 D. Maspoch, N. Domingo, D. Ruiz-Molina, K. Wurst,
G. Vaughan, J. Tejada, C. Rovira and J. Veciana, Angew.
Chem., Int. Ed., 2004, 43, 1828–1832.
S. T. Sigurdsson and T. F. Prisner, Phys. Chem. Chem. Phys.,
2015, 17, 24446–24451.
22 B. Joseph, V. M. Tormyshev, O. Y. Rogozhnikova,
D. Akhmetzyanov, E. G. Bagryanskaya and T. F. Prisner,
Angew. Chem., Int. Ed., 2016, 55, 11538–11542.
3 A. Datcu, N. Roques, V. Jubera, I. Imaz, D. Maspoch, J.-P. Sutter,
C. Rovira and J. Veciana, Chem. – Eur. J., 2011, 17, 3644–3656.
4 C. Sporer, I. Ratera, D. Ruiz-Molina, Y. Zhao, J. Vidal-
Gancedo, K. Wurst, P. Jaitner, K. Clays, A. Persoons and
C. Rovira, et al., Angew. Chem., Int. Ed., 2004, 43, 5266–5268.
¨
23 J. J. Jassoy, A. Berndhauser, F. Duthie, P. K. Sebastian,
G. Hagelueken and O. Schiemann, Angew. Chem., Int. Ed.,
2017, 56, 177–181.
˜
5 C. Simao, M. Mas-Torrent, N. Crivillers, V. Lloveras,
´
24 A. A. Kuzhelev, O. A. Krumkacheva, G. Y. Shevelev,
M. Yulikov, M. V. Fedin and E. G. Bagryanskaya, Phys. Chem.
Chem. Phys., 2018, 20, 10224–10230.
25 A. Giannoulis, Y. Yang, Y. J. Gong, X. Tan, A. Feintuch,
R. Carmieli, T. Bahrenberg, Y. Liu, X. C. Su and D. Goldfarb,
Phys. Chem. Chem. Phys., 2019, 21, 10217–10227.
J. M. Artes, P. Gorostiza, J. Veciana and C. Rovira, Nat.
Chem., 2011, 3, 359–364.
6 J. Guasch, L. Grisanti, V. Lloveras, J. Vidal-Gancedo, M. Souto,
D. C. Morales, M. Vilaseca, C. Sissa, A. Painelli and I. Ratera,
et al., Angew. Chem., Int. Ed., 2012, 51, 11024–11028.
7 J. Guasch, L. Grisanti, M. Souto, V. Lloveras, J. Vidal-
Gancedo, I. Ratera, A. Painelli, C. Rovira and J. Veciana, 26 Y. Yang, B. Bin Pan, X. Tan, F. Yang, Y. Liu, X. C. Su and
J. Am. Chem. Soc., 2013, 135, 6958–6967.
D. Goldfarb, J. Phys. Chem. Lett., 2020, 11, 1141–1147.
8 M. Souto, M. V. Solano, M. Jensen, D. Bendixen, F. Delchiaro, 27 N. Fleck, C. A. Heubach, T. Hett, F. R. Haege, P. P. Bawol,
A. Girlando, A. Painelli, J. O. Jeppesen, C. Rovira and I. Ratera,
et al., Chem. – Eur. J., 2015, 21, 8816–8825.
H. Baltruschat and O. Schiemann, Angew. Chem., Int. Ed.,
2020, 59, 9767–9772.
˜
9 N. Crivillers, C. Munuera, M. Mas-Torrent, C. Simao, 28 V. V. Khramtsov, A. A. Bobko, M. Tseytlin and B. Driesschaert,
S. T. Bromley, C. Ocal, C. Rovira and J. Veciana, Adv. Mater.,
2009, 21, 1177–1181.
10 L. Yuan, C. Franco, N. Crivillers, M. Mas-Torrent, L. Cao,
C. S. Suchang Sangeeth, C. Rovira, J. Veciana and C. A.
Nijhuis, Nat. Commun., 2016, 7, 12066.
11 J. H. Ardenkjær-Larsen, B. Fridlund, A. Gram, G. Hansson,
L. Hansson, M. H. Lerche, R. Servin, M. Thaning and K. Golman,
Proc. Natl. Acad. Sci. U. S. A., 2003, 100, 10158–10163.
Anal. Chem., 2017, 89, 4758–4771.
29 J. H. Ardenkjær-Larsen, I. Laursen, I. Leunbach, G. Ehnholm,
L.-G. Wistrand, J. S. Petersson and K. Golman, J. Magn. Reson.,
1998, 133, 1–12.
30 M. C. Krishna, S. English, K. Yamada, J. Yoo, R. Murugesan,
N. Devasahayam, J. A. Cook, K. Golman, J. H. Ardenkjær-
Larsen and S. Subramanian, et al., Proc. Natl. Acad. Sci. U. S. A.,
2002, 99, 2216–2221.
Phys. Chem. Chem. Phys.
This journal is © the Owner Societies 2020

View Article Online
PCCP
Paper
31 S. J. Nelson, J. Kurhanewicz, D. B. Vigneron, P. E. Z. Larson, 54 F. Mentink-Vigier, G. Mathies, Y. Liu, A. L. Barra,
A. L. Harzstark, M. Ferrone, M. Van Criekinge, J. W. Chang,
M. A. Caporini, D. Lee, S. Hediger, R. Griffin and G. De
¨
Paepe, Chem. Sci., 2017, 8, 8150–8163.
R. Bok and I. Park, et al., Sci. Transl. Med., 2013, 5, 198ra108.
32 P. Kuppusamy, P. Wang, M. Chzhan and J. L. Zweier, Magn. 55 I. Martin-Montesinos, J. C. Paniagua, A. Peman, M. Vilaseca,
Reson. Med., 1997, 37, 479–483.
33 G. He, Y. Deng, H. Li, P. Kuppusamy and J. L. Zweier, Magn.
Reson. Med., 2002, 47, 571–578.
F. Luis, S. Van Doorslaer and M. Pons, Phys. Chem. Chem.
Phys., 2016, 18, 3151–3158.
56 A. J. Stone, Mol. Phys., 1963, 6, 509–515.
34 M. Elas, B. B. Williams, A. Parasca, C. Mailer, C. A. Pelizzari, 57 F. Neese, J. Chem. Phys., 2001, 115, 11080–11096.
M. A. Lewis, J. N. River, G. S. Karczmar, E. D. Barth and 58 J. Tomasi, B. Mennucci and R. Cammi, Chem. Rev., 2005,
H. J. Halpern, Magn. Reson. Med., 2003, 49, 682–691.
35 K. Matsumoto, S. Subramanian, N. Devasahayam, 59 Y. Liu, F. A. Villamena, J. Sun, Y. Xu, I. Dhimitruka and
T. Aravalluvan, R. Murugesan, J. A. Cook, J. B. Mitchell J. L. Zweier, J. Org. Chem., 2008, 73, 1490–1497.
and M. C. Krishna, Magn. Reson. Med., 2006, 55, 1157–1163. 60 D. Maspoch, D. Ruiz-Molina, K. Wurst, G. Vaughan,
105, 2999–3093.
36 A. Samouilov, G. L. Caia, E. Kesselring, S. Petryakov,
T. Wasowicz and J. L. Zweier, Magn. Reson. Med., 2007, 58,
156–166.
37 H. B. Elajaili, J. R. Biller, M. Tseitlin, I. Dhimitruka,
V. V. Khramtsov, S. S. Eaton and G. R. Eaton, Magn. Reson.
Chem., 2014, 53, 280–284.
N. Domingo, J. Tejada, C. Rovira and J. Veciana, CrystEng-
Comm, 2004, 6, 573–578.
61 D. Maspoch, N. Domingo, N. Roques, K. Wurst, J. Tejada,
C. Rovira, D. Ruiz-Molina and J. Veciana, Chem. – Eur. J.,
2007, 13, 8153–8163.
62 T. Nishiuchi, S. Aibara and T. Kubo, Angew. Chem., Int. Ed.,
2018, 57, 16516–16519.
˜
´
´
38 M. Ballester, J. Riera, J. Castaner, C. Badıa and J. M. Monso,
J. Am. Chem. Soc., 1967, 93, 2215–2225.
39 M. Jørgensen, F. Rise, S. Andersson, T. Almen, A. Aabye, L.-G.
63 T. J. Reddy, T. Iwama, H. J. Halpern and V. H. Rawal, J. Org.
Chem., 2002, 67, 4635–4639.
´
¨
˜
Wistrand, H. Wikstrom, K. Golman, R. Servin and P. Michelsen, 64 M. Ballester, J. Riera, J. Castaner, C. Rovira and O. Armet,
Hafslund Nycomed Innovation AB, WO91/12024, 1991. Synthesis, 1986, 64–66.
40 A. J. Stone, Proc. R. Soc. London, Ser. A, 1963, 271, 424–434. 65 J. Frank, M. Elewa, M. M. Said, H. A. El Shihawy, M. El-
41 A. Carrington and A. D. McLachlan, Introduction to Magnetic
Resonance, with Applications to Chemistry and Chemical
Physics, Harper & Row, New York and London, 1967.
42 F. M. Vigier, D. Shimon, V. Mugnaini, J. Veciana,
A. Feintuch, M. Pons, S. Vega and D. Goldfarb, Phys. Chem.
Chem. Phys., 2014, 16, 19218–19228.
Sadek, D. Mu¨ller, A. Meister, G. Hause, S. Drescher and
H. Metz, et al., J. Org. Chem., 2015, 80, 6754–6766.
66 V. Dang, J. Wang, S. Feng, C. Buron, F. A. Villamena,
P. G. Wang and P. Kuppusamy, Bioorg. Med. Chem. Lett.,
2007, 17, 4062–4065.
67 S. Un, P. Dorlet and A. W. Rutherford, Appl. Magn. Reson.,
2001, 21, 341–361.
68 A. M. Portis, Phys. Rev., 1953, 91, 1071–1078.
43 D. Banerjee, D. Shimon, A. Feintuch, S. Vega and
D. Goldfarb, J. Magn. Reson., 2013, 230, 212–219.
44 A. J. Fielding, P. J. Carl, G. R. Eaton and S. S. Eaton, Appl. 69 A. M. Portis, Tech. Note No. 1, Sarah Mellon Scaife Radia-
Magn. Reson., 2005, 28, 231–238. tion Laboratory, University of Pittsburgh, 1955.
45 S. Un, J. Bryant and R. G. Griffin, J. Magn. Reson., Ser. A, 70 J. S. Hyde, Phys. Rev., 1960, 119, 1483–1492.
1993, 101, 92–94. 71 J. S. Hyde, Phys. Rev., 1960, 119, 1492–1495.
46 S. Un, L. C. Tabares, N. Cortez, B. Y. Hiraoka and 72 M. Weger, Bell Syst. Tech. J., 1960, 39, 1013–1112.
F. Yamakura, J. Am. Chem. Soc., 2004, 126, 2720–2726. 73 M. J. Frisch, Gaussian 09, Revis. B0.01, 2010.
47 C. Hureau, G. Blondin, M. Cesario and S. Un, J. Am. Chem. 74 A. D. Becke, Phys. Rev. A: At., Mol., Opt. Phys., 1988, 38, 3098–3100.
Soc., 2003, 125, 11637–11645.
48 S. Un, Inorg. Chem., 2013, 52, 3803–3813.
75 C. Lee, W. Yang and R. G. Parr, Phys. Rev. B: Condens. Matter
Mater. Phys., 1988, 37, 785–789.
49 F. Kurdzesau, B. Van Den Brandt, A. Comment, P. Hautle, 76 A. D. Becke, J. Chem. Phys., 1993, 98, 5648–5652.
S. Jannin, J. J. Van Der Klink and J. A. Konter, J. Phys. D: 77 A. D. Becke, J. Chem. Phys., 1993, 98, 1372–1377.
Appl. Phys., 2008, 41, 155506.
50 J. Soetbeer, P. Gast, J. J. Walish, Y. Zhao, C. George, C. Yang,
78 R. Ditchfield, W. J. Hehre and J. A. Pople, J. Chem. Phys.,
1971, 54, 724–728.
T. M. Swager, R. G. Griffin and G. Mathies, Phys. Chem. 79 W. J. Hehre, R. Ditchfield and J. A. Pople, J. Chem. Phys.,
Chem. Phys., 2018, 20, 25506–25517. 1972, 56, 2257–2261.
51 S. Un, A. W. Rutherford, M. Atta, M. Atta and M. Fontecave, 80 V. A. Rassolov, J. A. Pople, M. A. Ratner and T. L. Windus,
J. Am. Chem. Soc., 1995, 117, 10713–10719. J. Chem. Phys., 1998, 109, 1223–1229.
52 Y. A. Dmitriev and R. A. Zhitnikov, J. Low Temp. Phys., 2001, 81 P. C. Hariharan and J. A. Pople, Theor. Chim. Acta, 1973, 28,
122, 163–170. 213–222.
53 J. Wolber, F. Ellner, B. Fridlund, A. Gram, H. Johannesson, 82 The PyMOL Molecular Graphics System, Version 2.0
´
¨
Schrodinger, LLC.
G. Hansson, L. H. Hansson, M. H. Lerche, S. Månsson and
R. Servin, et al., Nucl. Instrum. Methods Phys. Res., Sect. A, 83 M. D. Hanwell, D. E. Curtis, D. C. Lonie, T. Vandermeersch,
2004, 526, 173–181.
E. Zurek and G. R. Hutchison, J. Cheminf., 2012, 4, 1–17.
Phys. Chem. Chem. Phys.
This journal is © the Owner Societies 2020