Molecular recognition in a heteromolecular radical pair system with complementary multipoint hydrogen-bonding

Article abstract of DOI:10.1039/b800779a

Stable radicals 2-(6-uradinyl)-4,4,5,5-tetramethyl-4,5-dihydro-1H- imidazole-1-oxyl (Ur6IN) and 4-(p-tert-butylaminoxylphenyl)-2,6-di(propylamido) pyridine (DAPPN) form heterospin radical pair complexes due to complementary multi-point hydrogen-bonds. The Royal Society of Chemistry.

Full text of DOI:10.1039/b800779a

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Molecular recognition in a heteromolecular radical pair system with
complementary multipoint hydrogen-bondingw
Hidenori Murata,ab Paul M. Lahti*a and Safo Aboakua
Received (in Cambridge, UK) 22nd January 2008, Accepted 22nd April 2008
First published as an Advance Article on the web 2nd June 2008
DOI: 10.1039/b800779a
Stable radicals 2-(6-uradinyl)-4,4,5,5-tetramethyl-4,5-dihydro-
1H-imidazole-1-oxyl (Ur6IN) and 4-(p-tert-butylaminoxyl-
phenyl)-2,6-di(propylamido)pyridine (DAPPN) form heterospin
radical pair complexes due to complementary multi-point hydro-
gen-bonds.
methane by slow evaporation. The DAPPN X-band EPR
spectrum in toluene at 298 K shows hyperfine coupling (hfc)
from the nitroxide N and from phenyl protons ortho and meta
to the nitroxide: a(N) = 1.15 mT, a(ortho) = 0.22 mT, a(meta)
= 0.09 mT. Unresolved, additional hfc must be o0.051 mT,
based on the spectral linewidth. Multiple attempts at analyzing
the X-ray diffraction (XRD) of single crystals of DAPPN did
not allow us to determine all structural coordinates unequi-
vocally, even at 100 K, so only the unit cell data from these
attempts are given.y However, single crystals of DAPPN give
very strong EPR signals from the nitroxide spin, and FAB-MS
shows the mass spectrum expected for DAPPN. These results
confirmed the identity of DAPPN for the Ur6INꢀDAPPN
complementary hydrogen-bonding experiments descri-
bed below.
Strongly directional, complementary hydrogen-bonding inter-
actions in biology have provided much inspiration for mole-
cular recognition and assembly strategies in chemistry.1 In
2005, Taylor et al. from our group showed that stable radical
2-(6-uradinyl)-4,4,5,5-tetramethyl-4,5-dihydro-1H-imidazole-1-
oxyl, Ur6IN, binds with the complementary three-point hydro-
gen-bonding receptor 2,6-di(propylamido)pyridine, DAP,3 giv-
ing prospects for supramolecular assemblies of different stable
radicals (heterospin assembly). Both biological and materials-
based uses of radicals can be at least partially controlled and
predicted with the assistance from supramolecular assembly
strategies. This article reports the synthesis and characteriza-
tion of a radical-functionalized DAP complement to Ur6IN:
(4-(p-tert-butylaminoxyl-phenyl)-2,6-di(propylamido)pyridine,
DAPPN. Studies of heterospin complementary complexation
of Ur6IN with DAPPN in solution and solid state are also
described.
Because both Ur6IN and DAPPN are paramagnetic, it was
not convenient to monitor complexation between them by the
NMR method used2 to investigate Ur6INꢀDAP complexation.
EPR spectra of 1 : 1 mole ratio solutions of the two in toluene
did not show significant deviation from a sum of their individual
spectra, even when cooled to about 180 K. This indicates either
that the exchange J between the radical sites is much smaller
than the hfc (J { a(N)), or that the amount of the Ur6INꢀ
DAPPN complex in the solution is too small to detect. EPR
hfc changes have been reported for hydrogen-bonded complexes
when there is significant spin density on the atoms at the site
being complexed.5 That is not the case here, however. Separate
UB3LYP/6-31G* computations6 for a somewhat simplified
Ur6IN was made as previously4 described. DAPPN was
synthesized as shown in Scheme 1 to give a brick-red powderz
that forms clear red prisms when crystallized from dichloro-
a Department of Chemistry, University of Massachusetts, Amherst,
MA 01003, USA. E-mail: lahti@chem.umass.edu;
Fax: +1-413-545-4490
b Department of Pure & Applied Chemistry, Faculty of Science &
Technology, Tokyo University of Science, Noda, Chiba, 278-8510,
Japan
w Electronic supplementary information (ESI) available: Synthesis,
EPR, MS, computed spin densities of DAPPN, ESI-MS of Ur6INꢀ
DAPPN and a spin density map of the Ur6INꢀDAPPN model. See
DOI: 10.1039/b800779a
Scheme 1 Synthesis of DAPPN; see ESI for details.w
Chem. Commun., 2008, 3441–3443 | 3441
ꢁc
This journal is The Royal Society of Chemistry 2008

View Article Online
either the zero-field splitting of any solution complexes is
small, or the amount of complex is small, or the complex
favors a singlet state by so large an energy gap that a triplet
state is not thermally populated (this last seems an unlikely
scenario). But, ESI-MS analysis of a 10 micromolar solution
of a 1 : 1 Ur6IN : DAPPN mixture in methanol shows the m/z
= 662 peak of the heterospin dyad complex Ur6INꢀDAPPN,
and also a weak peak at m/z = 1072 attributable to a Ur6INꢀ
(DAPPN)2 triad complex. Analogous 1 : 1 and 1 : 2 complexes
were found2 in the ESI-MS analyses of Ur6IN–DAP solution
mixtures. The structures of these complexes are not directly
obvious from the ESI-MS data, but are hypothesized as shown
above, based on the likely multi-point donor–acceptor hydro-
gen-bond interactions.
Fig. 1 UB3LYP/6-31G* Mulliken spin density populations for
Ur6IN and a simplified model structure for DAPPN. Ur6IN data
from ref. 2. Details for DAPPN and a spin density map for the pair are
given in the ESI.w
model of DAPPN and Ur6IN give the Mulliken spin density
populations shown in Fig. 1. Although the computations con-
firm that the nitroxide spin in DAPPN delocalizes more than the
iminoylnitroxide in Ur6IN, there is still little overall spin density
at the sites of 3-point hydrogen-bonding. If one applies a
McConnell type relationship between hfc (a[N]) and computed
spin density (r[N]), a[N] = Qꢀr(N) with Q = 3 mT, the
hyperfine on the pyridyl nitrogen is predicted to be only 0.06 mT,
which would be at or below the limits of resolution of the
observed spectra.7 The hfc in the uradinyl portion of the spin
localized Ur6IN is even smaller, so there is little close-contact
spin density interaction at the actual sites of hydrogen-bonding
in a complementary hydrogen-bonded Ur6INꢀDAPPN complex.
The 3-point hydrogen-bonding interactions in a dyad
between DAP and a thymine/uradinyl moiety are energetically
worth about 12–25 kJ molꢂ1. The binding constant found
earlier2 for the Ur6INꢀDAP complex in methanol yielded
DG[binding] D 12 kJ molꢂ1 at 33 1C. Assuming similar
thermodynamics for Ur6INꢀDAPPN, only a small amount of
a 1 : 1 complex is expected under solution conditions for equal
amounts of the two radicals at typical concentrations. This
would explain why it is difficult to detect the complexes by
solution EPR.
Although observing high degrees of complexation in 1 : 1
solutions of donor : receptor dyads like Ur6INꢀDAPPN can be
difficult, solid state assembly should be more favourable, so
long as the separate crystallization of the components does not
occur. Shiomi et al. demonstrated heterospin assembly of two
nitronylnitroxide radicals functionalized with a pyridine and a
monocarboxylic acid,8 using a one-site hydrogen-bond, and a
nucleic-acid inspired two-site hydrogen-bonded donor–accep-
tor solid incorporating two different, non-delocalized nitro-
nylnitroxides.9
In order to form the Ur6INꢀDAPPN complex in the solid
state, an equimolar solution of the component radicals in
toluene was slowly evaporated in a 4 mm o.d. EPR tube under
nitrogen to give a powdery orange solid. The EPR spectrum of
this solid does not show dipolar features from triplet state zero
field splitting down to 5 K; neither do neat solid samples of the
individual components. However, the precipitated mixed solid
exhibits an EPR half-field transition at 167 mT, characteristic
for a triplet state (Fig. 2). Notably, neither of the individual
component solids shows a similar peak. The half-field peak
shows linear Curie law temperature dependence of doubly-
integrated intensity as a function of reciprocal inverse tem-
perature over 7–50 K, so the triplet state from which it arises is
nearly degenerate with the corresponding singlet state, or is
robustly favored over the singlet state.
The balance of evidence is most consistent with the half-field
transition arising from formation of Ur6INꢀDAPPN com-
plexes in the solid. Although we have not been able to make
diffraction grade single crystals of Ur6INꢀDAPPN, the
ESI-MS evidence clearly shows formation of the complex in
solution, and only the mixed solid-state results show forma-
tion of a radical–radical interaction strong enough to yield a
triplet state. The results do not rule out p-stacking or inter-
radical close contact from less specific crystal packing inter-
actions, but it is unclear how such interactions would be
Ur6IN–DAPPN binary solutions frozen to 77 K showed no
dipolar features attributable to a triplet state radical pair;
ꢁc
This journal is The Royal Society of Chemistry 2008
3442 | Chem. Commun., 2008, 3441–3443

View Article Online
Notes and references
z For DAPPN: mp: 84–86 1C. IR (KBr pellet, cmꢂ1): 3424 (NH str),
1676 (CQO str). MS (EI): calc for C23H31N4O3 m/z = 411.5, found
m/z = 412. Analysis: calc for C23H31N4O3: C, 67.13; H, 7.59; N, 13.61.
Found: C 67.93, H 7.86, N 12.07. EPR (toluene, 9.64723 GHz): g =
2.00549, a(N) = 1.15 mT, a(ortho) = 0.22 mT (2H), a(meta) = 0.09
mT (2H). HPLC (C18 column, 1.2 mL minꢂ1, 7 : 3 MeOH : H2O):
Rt = 8.4 min.
y Crystal unit cell data. Formula C23H31N4O3, formula weight =
411.5, temperature = 100 K, Monoclinic, P21/c, a = 10.7221(2) A,
b = 25.4626(4) A, c = 16.9760(3) A, b = 96.8291(6)1, V = 4602(3)
A3, Z = 8, D(calc) = 1.188 g cmꢂ3, F(000) = 1768.
1. (a) J. Bernstein, M. C. Etter and L. Leiserowitz, Struct. Correl.,
1994, 2, 431; (b) P. A. Giguere, Chem. Phys. Lett., 1981, 80, 207;
(c) P. Schuster, NATO ASI Ser., Ser. C, 1978, C39, 229;
(d) K. Morokuma, Acc. Chem. Res., 1977, 10, 294;
(e) M. C. Etter, Acc. Chem. Res., 1990, 23, 120.
Fig. 2 Curie plot of doubly integrated spectral intensity of the half
field EPR band (inset) observed in the powder solid Ur6IN–DAPPN
mixture. Inset spectrum obtained at 4.8 K, 9.37359 GHz, background
subtracted after multiscan signal averaging.
2. P. Taylor, P. M. Lahti, J. B. Carroll and V. M. Rotello, Chem.
Commun., 2005, 895.
3. For use of DAP in molecular recognition, see for example
(a) J.-M. Lehn, Angew. Chem., Int. Ed. Engl., 1990, 29, 1304;
(b) J.-M. Lehn, Supramolecular chemistry: concepts and perspec-
tives, Weinheim, VCH, New York, 1995; (c) A. Niemz and
V. M. Rotello, Acc. Chem. Res., 1999, 32, 44.
present in the mixed solid if not in the individual component
solids. The complementarity of hydrogen-bonding would seem
to encourage strongly the formation of the Ur6INꢀDAPPN
complex as postulated above, or a structure very similar to
this. The three-point hydrogen-bonding set of interactions
should be a strong driving force for molecular assembly in
the solid.
4. P. Taylor, P. Serwinski and P. M. Lahti, Chem. Commun., 2003,
1400.
5. A. Niemz and V. M. Rotello, J. Am. Chem. Soc., 1997, 119, 6833.
6. Computations were carried out using the Spartan 2002 software
package for SGI computer, Wavefunction Inc., Irvine, CA.
7. Use of Q = 3 mT in the McConnell relationship gives reasonable
agreement with experimental nitrogen hfc; see ref. 4 and the
discussion in J. A. Pople and D. A Beveridge, Approximate
Molecular Orbital Theory, McGraw-Hill, New York, NY, 1970,
pp. 128–149.
8. D. Shiomi, M. Nozaki, T. Ise, K. Sato and T. Takui, J. Phys.
Chem. B, 2004, 108, 16606.
9. T. Ise, D. Shiomi, K. Sato and T. Takui, Chem. Commun., 2006,
4832.
In recent years, nitronylnitroxide and nitroxide type radicals
with attached nucleosides and related moieties have been
studied for a number of reasons.2–4,8–12 However, few of these
have been crystallized to explore their solid state assembly
behaviour as molecular one-component solids,3,4,8–10 so this
remains a relatively new area of endeavour among studies of
organic radical materials. By extension, the design of heterospin
interactions by inducing different radicals (or other open-shell
molecules) to co-crystallize offers many additional possibilities
for new solid-state behaviours, especially behaviours that are
unlikely or unable to occur in one-component solids. The
present work and related studies using complementary multi-
point hydrogen-bonding interactions have much potential
scope for solid state assembly of heteromolecular complexes
of spin-bearing organic molecules. Further work is ongoing to
study other heterospin solids related to Ur6INꢀDAPPN.
This work was supported in part by the US National Science
Foundation by grants CHE 0415716, CHE-9974648 (UMass-
Amherst X-ray Structural Characterization facility), CHE-
0443180 (UMass-Amherst Electron Paramagnetic Resonance
Facility). We thank Dr Gregory Dabkowski of the University
of Massachusetts Amherst Microanalytical Laboratory for
elemental microanalyses, and Prof. H. Nishide of Waseda
University for assistance with ESI-MS measurements.
10. K. Das, M. Pink, S. Rajca and A. Rajca, J. Am. Chem. Soc., 2006,
128, 5334.
11. (a) T. Kaneko, T. Katsurabara, Y. Matsui, M. Aso, N. Koga and
H. Suemune, Nucleic Acids Symp. Ser., 2004, 48, 5; (b) M. Aso,
T. Kaneko, M. Nakamura, N. Koga and H. Suemune, Chem.
Commun., 2003, 9, 1094; (c) M. Aso, T. Ikeno, K. Norihisa,
M. Tanaka, N. Koga and H. Suemune, J. Org. Chem., 2001, 66,
3513; (d) M. Aso, K. Norihisa, M. Tanaka, N. Koga and
H. Suemune, J. Chem. Soc., Perkin Trans. 2, 2000, 8, 1637;
(e) M. Aso, T. Ikeno, T. Kaneko, N. Koga and H. Suemune,
Nucleic Acids Res. Suppl., 2001, 1, 115.
12. (a) N. Barhate, P. Cekan, A. P. Massey and S. Th. Sigurdsson,
Angew. Chem., Int. Ed., 2007, 46, 2655–2658; (b) A. Okamoto,
T. Inasaki and I. Saito, Tetrahedron Lett., 2005, 46, 791; (c) J. M.
J. Tronchet, G. Zosimo-Landolfo, M. Balkadjian, A. Ricca,
M. Zsely, F. Barbalat-Rey, D. Cabrini, P. Lichtle and
M. Geoffroy, Tetrahedron Lett., 1991, 32, 4129; (d) M. J.
M. Tronchet, K. Mekhael, J. Graf-Poncet, R. Benhamza and
M. Geoffroy, Helv. Chim. Acta, 1985, 68, 1893; (e) J. L. Duh and
A. M. Bobst, Helv. Chim. Acta, 1991, 74, 739; (f) J. Gajewska,
A. Bierzynski, K. Bolewska, K. L. Wierzchowski, A. I. Petrov and
B. I. Sukhorukov, Biophys. Chem., 1982, 15, 191; (g) A. Hakam,
I. E. Thomas and A. M. Bobst, Int. J. Biol. Macromol., 1980, 2, 49.
ꢁc
This journal is The Royal Society of Chemistry 2008
Chem. Commun., 2008, 3441–3443 | 3443