Suppression of spin-spin coupling in nitroxyl biradicals by supramolecular host-guest interactions

Article abstract of DOI:10.1039/c0cc02587a

The use of supramolecular architectures to control the spatially dependent spin exchange (spin communication) between two covalently linked radical centers (biradical) has been explored. Cucurbit[8]uril, through supramolecular steric effect, completely suppresses spin exchange between two adjacent radical centers in a biradical. The Royal Society of Chemistry.

Full text of DOI:10.1039/c0cc02587a

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Suppression of spin–spin coupling in nitroxyl biradicals by
supramolecular host–guest interactionsw
Mintu Porel,a M. Francesca Ottaviani,b Steffen Jockusch,c Nithyanandhan Jayaraj,a
Nicholas J. Turro*c and V. Ramamurthy*a
Received 15th July 2010, Accepted 1st September 2010
DOI: 10.1039/c0cc02587a
The use of supramolecular architectures to control the spatially
dependent spin exchange (spin communication) between two
covalently linked radical centers (biradical) has been explored.
Cucurbit[8]uril, through supramolecular steric effect, completely
suppresses spin exchange between two adjacent radical centers in
a biradical.
Recently there has been a renaissance in the application of
Scheme 1 Binitroxyls (1–3), mononitroxyl (4) and corresponding
dynamic nuclear polarization (DNP) as a general method of
enhancement of NMR signal intensities with potential
applications for magnetic resonance imaging (MRI).1 The
‘tuning’ of the electron- and nuclear-spin systems can be
achieved by rendering their spin transition frequencies
degenerate. One such method is the use of biradicals as the
coupling source.2 In this case, the system corresponds to a
three-spin pool of two coupled electrons and a coupled nuclear
spin. The efficiency of DNP in the three-spin system can be
controlled by optimizing the interaction between the coupled
electrons in biradicals. A significant amount of work has been
reported on nitroxyl biradicals.3 Mainly these studies show
how spin–spin coupling can be controlled by the distance
between the nitroxyl moieties. In this report, we demonstrate
that control of spin–spin coupling in biradicals can also be
achieved through supramolecular effects.
diamagnetic derivative (5) examined in this study.
and completely suppresses spin exchange between the two
radical centers in 3.
To compare guests (1–3) in the absence/presence of the
various hosts, their reference EPR spectra were recorded in
DMSO–water (1 : 1) in the case of the first two (due to low
solubility in water) and in pure water in the case of 3.
Experimental and simulated spectra are displayed in Fig. 1.
The spectra were computed using the well-established procedure
of Freed and co-workers.4 The most informative input
parameters were the coupling constant between electron and
nuclear spins hANi = (Axx + Ayy + Azz)/3 which is a measure
of the environmental polarity of the nitroxides and the correlation
time (t) for the rotational diffusion motion of the spin probe
which provides information about the guest mobility.
Binitroxyls 1 and 2 exhibit three-line spectra typical of
mononitroxyls, indicative of their radical centers being too
far apart for spin–spin exchange interactions. The EPR
spectrum of compound 3 is distinctly different from that of 1
and 2. It showed a five-line spectrum, which means that each
To unravel the value of supramolecular architectures
in controlling spin communication between two nearby
radical centers, we have examined the EPR spectra of three
binitroxyls3a namely (Scheme 1), di-4-(2,2,6,6-tetramethyl-
piperidine-1-oxyl)-terephthalate (1), di-4-(2,2,6,6-tetramethyl-
piperidine-1-oxyl)-isophthalate (2) and di-4-(2,2,6,6-tetra-
methyl-piperidine-1-oxyl)-phthalate (3) in water in the presence
of hosts such as cucurbit[8]uril (CB8), cucurbit[7]uril (CB7),
b-cyclodextrin (b-CD), g-cyclodextrin (g-CD), calixarene[8]octa
sulfonic acid (CA8) and sodium dodecyl sulfate (SDS) micelle
(see ESIw, Fig. S1). We find that the host CB8 is most effective
a Department of Chemistry, University of Miami, Coral Gables,
FL 33124, USA. E-mail: murthy1@miami.edu;
Tel: +1 305 284 1534
b Department of Geological Sciences,
Chemical and Environmental Technologies (GeoTeCA),
University of Urbino, Campus ex-Sogesta, Loc. Crocicchia,
61029 Urbino, Italy
Fig. 1 EPR spectra (black lines) and their simulations (red lines) for
(a) 1 in 50% DMSO and 50% water (0.5 mM), AN = 16.5 G, t = 0.14 ns,
(b) 2 in 50% DMSO and 50% water (0.5 mM), AN = 16.4 G,
c Department of Chemistry, Columbia University, New York,
NY 10027, USA
w Electronic supplementary information (ESI) available: Structures of
hosts employed, 1H NMR and EPR spectra of 4, EPR spectra of 1, 2
and 3 under various conditions, experimental procedure including
synthesis and characterization of compounds 1–4.
t E 0.19 ns; (c) 3 in water (1 mM) three-line component: AN
=
16.9 G, t = 0.06 ns, five-line component: 88% and (d) 3@CB8 (1 : 1),
AN = 16.7 G, t = 0.21 ns. Concentrations of host and guest: 1 mM.
c
7736 Chem. Commun., 2010, 46, 7736–7738
This journal is The Royal Society of Chemistry 2010

View Article Online
electron is coupled to both nitrogens of the binitroxyl. The
EPR spectrum of 3 is attributable to fast inter-conversion
between conformers exhibiting weak (J E 0) and strong
(J c AN) exchange interaction.
For computational convenience, we view this five-line
spectrum as the sum of a three-line signal and a two-line
signal. Both
a three-line signal and a two-line signal
(in between the three lines) were computed, and the experi-
mental line shape was fitted by adding the two components at
the proper relative percentages. The variation of the exchange
interaction between two radical centers in various systems was
estimated by calculating the intensity (in percentage) of a
five-line signal expected for very strong exchange with respect
to the overall spectral intensity obtained by double integration
of the experimental spectra.
Scheme 2 Schematic representation of the possible conformers of 3
and 3@CB8.
The three lines (1, 3 and 5) or the two lines (2 and 4) in the
five-line EPR spectra of 3 in water are separated by 16.9 G,
slightly lower than the hyperfine splitting of a 2,2,6,6-tetra-
methyl-piperidine-1-oxyl radical (monoradical) in water (17.1 G).
The correlation time (t) of 0.06 ns is characteristic of fast
rotating nitroxyl radicals.5 The estimated relative percentage
of the exchange coupling (1 : 2 : 3 : 2 : 1) is 88%, which
indicates that the two nitroxyl groups in the biradical
dynamically approach each other in water resulting in strong
spin–spin exchange.
(for details see Fig. S6 in ESIw). No nitroxyl radicals were lost
due to chemical reactions. Based on these results we conclude
that the supramolecular steric effect provided by encapsulation
of one paramagnetic moiety of 3 with CB8 is sufficient to
eliminate the spin communication between the two nitroxyl
radicals, one complexed to CB8 and one free (Scheme 2).
In Scheme 2 we have shown three of the many possible
conformations of 3. Because of the broad NMR signals
1H NMR NOESY spectra were not useful in identifying the
preferred conformation for this molecule. We conclude that
spin communication is suppressed by adoption of a conformation
where the two nitroxyl groups in 3@CB8 are separated far
enough not to exchange each other’s spin or the CB8 cage wall
is a barrier for spin–spin coupling.
Of the three compounds, 3 is clearly the best candidate to
test the influence of supramolecular hosts on spin–spin
exchange between two radical centers. It is well known that
mononitroxyls such as 2,2,6,6-tetramethyl-1-piperidinyloxyl
(TEMPO) and 4-(N,N,N-trimethylammonium)-2,2,6,6-tetra-
methylpiperidine-N-oxyl bromide (a positively charged TEMPO
derivative) can be included within the cavities of CD, CB and
CA and micelles. The question is whether such an inclusion
would affect the exchange interaction between two nearby
nitroxyl centers in molecules such as 3. In Fig. 1(d), the EPR
spectra (experimental and simulated) of 3 in the presence of
one equivalent of CB8 are displayed. Interestingly, the change
of the five-line spectrum to a three-line spectrum in the
presence of CB8 is suggestive that one equivalent of this host
is sufficient to eliminate the spin–spin exchange interaction.
The absence of any effect on the spectrum on addition of more
than one equivalent of CB8 suggests that only one of the two
nitroxyls is included within CB8. From simulation of the EPR
spectrum, it is observed that t increases from 0.06 ns in pure
water to 0.21 ns in the presence of one equivalent of CB8. The
conclusion that only one nitroxyl group of 3 is complexed to
CB8 is supported by two observations: (a) almost no change
was observed for t between one and two equivalents of CB8
(0.21 ns vs. 0.24 ns); (b) estimated polarities of the nitroxyl
environment based on the hyperfine splitting constant (AN)
decreases slightly in the presence of CB8 from that in pure
water (16.9 G for 3 in water and 16.7 G for 3@CB8 in water).6
Further addition of CB8 did not change AN. Similarities in
polarity experienced by the nitroxyl radical in 3 and 3@CB8 in
water suggest that one of the two nitroxyl groups in 3@CB8
remains uncomplexed and is exposed to water. It was
confirmed by double integration of the EPR signal of 3 free
in solution and 3@CB8 that both nitroxyl radicals in 3 in the
presence of CB8 are contributing to the 3-line signal intensity
1H NMR and EPR experiments with monoradical 4, a
molecule analogue to 3 (Scheme 1) but with one diamagnetic
N–CH3 group and one paramagnetic NO group supported our
conclusion about the formation of a 1 : 1 (guest to host)
complex with CB8. NMR spectra of 4 in CDCl3, D2O and
in the presence of one equivalent of CB8 are shown in Fig. S2
(ESIw). In water, the paramagnetic nitroxyl probably broadens
the 1H NMR signals of the tetramethyl groups of the adjacent
N-methylpiperidinyl suggesting the existence of electron–1H
spin communication between the two groups. Addition of one
1
equivalent of CB8 to 4 in water resulted in sharper H NMR
signals as well as an upfield shift of the signals of N-methyl-
piperidinyl group. Addition of more than one equivalent of
CB8 resulted in a turbid solution that prevented recording of
the NMR spectrum. Thus, inclusion of one of the two groups
within CB8 was sufficient to arrest the influence of para-
magnetic nitroxyl on the diamagnetic N-methylpiperidine
group. To probe which of the two groups is included within
CB8 we recorded the EPR spectrum (experimental and
simulated spectra are provided in ESIw; Fig. S3). Computation
of the EPR spectra of 4 in the absence/presence of CB8
showed that the hyperfine splitting is almost the same (free:
A
N = 16.9 G; and complexed to CB8: AN = 17 G) suggesting
that the paramagnetic nitroxyl moiety is located in water
outside the CB8 cage. However, the rotational correlation
time of 4 in the presence of CB8 was slightly larger compared
to 4 free in solution (complexed: 0.14 ns; free: 0.04 ns) which
is probably caused by steric hindrance of the rotational
c
This journal is The Royal Society of Chemistry 2010
Chem. Commun., 2010, 46, 7736–7738 7737

View Article Online
mobility of the paramagnetic moiety by the complexation
of the diamagnetic moiety with CB8. The above results thus
suggest that the less hydrophobic piperidine moiety is included
within CB8.
reserved for SDS micelle. In spite of 3 being solubilized within
the micelle as indicated by the decrease of the hyperfine
splitting, AN (see Fig. S9a in ESIw), and by the increase of
the rotational correlation time (Fig. S9b, ESIw), the percentage
of exchange interaction is close to that in water. This is most
likely due to the soft flexible nature of the micellar interior that
allows the molecule to rotate freely within its core.8
To further confirm that only one of the two groups in 3 was
included within CB8 we synthesized diamagnetic equivalent
compound 5 and recorded its 1H NMR spectra in water in the
presence and absence of CB8 (Fig. S4 and S5 in ESIw). Upon
addition of one equivalent of CB8 only the N–CH3 signal was
upfield shifted indicating that only this group is included
within CB8. Consistent with this, signals due to the four
methyl groups of the N-methylpiperidyl ring alone were
upfield shifted. Addition of more than one equivalent of
CB8 resulted in no change in the spectra suggesting that 5
similar to 3 and 4 forms 1 : 1, not 1 : 2 (guest : host), complex
with CB8. Since 5 and 5@CB8 complex had a very limited
solubility in water we could not carry out NOESY studies to
ascertain the conformation of complexed and uncomplexed 5.
Having established CB8’s ability to fully suppress the
spin–spin exchange between adjacent nitroxyl groups in 3 we
were interested in comparing this feature with hosts such as
CB7, b-CD, g-CD, CA8 and SDS micelles.7 The EPR (experi-
mental and simulated) spectra obtained upon addition of two
equivalents (guest to host: 1 : 2) of the first four hosts and in
the presence of excess of SDS surfactant are provided in
Fig. S8 (ESIw). A large excess of SDS (1 : 1200) was necessary
to avoid the presence of two or more molecules of 3 in one
SDS micelle, which causes intermolecular spin–spin inter-
action due to Heisenberg exchange. Even in the presence of
a large excess of SDS a ‘broad’ single line EPR Heisenberg
component (Fig. S8g, inset, ESIw) contributes to the overall
EPR signal for about 45%. This component was present even
in the case of g-CD complex to the extent of 10%, which is
probably caused by the poor solubility of g-CD complex. The
% of five-line spectrum (strong spin–spin exchange component)
thus obtained for various hosts are displayed in the form of
histograms in Fig. 2. The exchange interaction that totally
disappeared in the presence of CB8 persisted to various
degrees (28–79%) in other hosts. While the most exciting
result was obtained with CB8, the most surprising result was
The quenching of spin communication due to internalization of
one nitroxyl radical into a cage may arise from two effects:
(a) the cage wall creates a barrier to spin exchange; (b)
conformational modifications due to complexation that keeps
the two nitroxyls apart. Due to the paramagnetic nature of 3
we could not perform NOESY experiments to gain an insight
into the conformational change that might have occurred
upon complexation. We are currently pursuing our belief that
the ‘supramolecular steric effect’ that has been exploited in this
study to control the spin–spin exchange in biradicals can be
extended to other polyradicals. Given the role of nitroxyl
polyradicals for use in dynamic nuclear polarization (DNP)
a method to enhance NMR signal intensities by two to three
orders of magnitude by controlling their spin–spin interaction
is likely to gain momentum in the near future. In this context
supramolecular assemblies such as the ones described here
could play a significant role.
The authors thank the National Science Foundation for
its generous support of this research through grants
NSF-CHE-07-17518 (NJT) and NFS-CHE-08-48017 (VR).
Notes and references
1 R. G. Griffin and T. F. Prisner, Phys. Chem. Chem. Phys., 2010, 12,
5737.
2 (a) C. Song, K. N. Hu, C. G. Joo, T. M. Swager and R. G. Griffin,
J. Am. Chem. Soc., 2006, 128, 11385; (b) Y. Matsuki, T. Maly,
O. Ouari, H. Karoui, F. L. Moigne, E. Rizzato, S. Lyubenova,
J. Herzfeld, T. Prisner, P. Tordo and R. G. Griffin, Angew. Chem.,
Int. Ed., 2009, 48, 4996.
3 (a) N. J. Turro, I. V. Khudyakov and S. H. Bossmann, J. Phys.
Chem., 1993, 97, 1138; (b) P. Ferruti, D. Gill, M. P. Klein,
H. H. Wang, G. Entine and M. Calvin, J. Am. Chem. Soc., 1970,
92, 3704; (c) V. A. Tran, A. I. Kokorin, G. Grampp and
K. Rasmussen, Appl. Magn. Reson., 2009, 35, 389;
(d) J. Szyd"owska, K. Pietrasik, L. G"az and A. Kaim, Chem. Phys.
Lett., 2008, 460, 245; (e) C. Song, K.-N. Hu, C.-G. Joo,
T. M. Swager and R. G. Griffin, J. Am. Chem. Soc., 2006, 128,
11385; (f) C. D. Smith, R. C. Bott, S. E. Bottle, A. S. Micallef and
G. Smith, J. Chem. Soc., Perkin Trans. 2, 2002, 533; (g) G. L. Closs,
M. D. E. Forbes and P. Piotrowiak, J. Am. Chem. Soc., 1992, 114,
3285; (h) M. D. E. Forbes, V. P. McCaffrey and E. J. Harbron,
Spectrum, 2005, 18, 12.
4 D. E. Budil, S. Lee, S. Saxena and J. H. Freed, J. Magn. Reson., Ser.
A, 1996, 120, 155.
5 M. F. Ottaviani, J. Phys. Chem., 1987, 91, 779.
6 M. F. Ottaviani, G. Martini and L. Nuti, Magn. Reson. Chem.,
1987, 25, 897.
7 (a) P. Franchi, M. Lucarini and G. F. Pedulli, Curr. Org. Chem.,
2004, 8, 1831; (b) G. Ionita, V. Meltzer, E. Pincu and V. Chechik,
Org. Biomol. Chem., 2007, 5, 1910.
Fig. 2 Comparison of percentage of spin–spin exchange interaction
in 3; note the exchange interaction in CB8 was 0%. For g-CD and SDS
micelles the broad Heisenberg component was subtracted before
calculation of % exchange interaction. 3 : host = 1 : 2 and in SDS
(1 : 1200).
8 N. J. Turro, V. Ramamurthy and J. C. Scaiano, Modern Molecular
Photochemistry of Organic Molecules, University Science Books,
Sausalito, CA, 2010, ch. 13.
c
7738 Chem. Commun., 2010, 46, 7736–7738
This journal is The Royal Society of Chemistry 2010