A R T I C L E S
Rajca et al.
diradicals. Most importantly, we discoVered unprecedented
organic one-dimensional (1-D) antiferromagnetic chains in
polycrystalline diradicals 1 and 3, in which the antiferromag-
netic interactions between S ) 1 diradicals are mediated
through methyl nitroxide C-H- - -O contacts, including non-
classical hydrogen bonds.24
One-dimensional, highly isotropic, antiferromagnetically
coupled Heisenberg chains with integer local spins (e.g., S )
1) are of fundamental interest.25-29 In 1983, Haldane conjectured
that the ground state of the 1-D Heisenberg model strongly
depends on the value of S of local spins. In particular, a large
energy gap (Haldane gap) between the singlet spin liquid ground
state and the lowest triplet excited state was predicted for integer
spin Heisenberg antiferromagnetic chains.25 The Haldane gap
was later experimentally confirmed in the study of [Ni(ethylene
diamine)2(NO2)](ClO4), abbreviated as NENP.30 However, the
local magnetic anisotropy associated with the Ni(II) ion
introduces additional energy gaps, comparable in magnitude to
the isotropic Haldane gap,31 and therefore presents an obstacle
to the characterization of the spin excitation spectrum.32 With
the triplet ground state nitroxide diradicals as S ) 1 local spins,
we expect that the organic chains would possess much lower
local magnetic anisotropy, thus providing properties that
resemble those of isotropic Heisenberg chains. Such intrinsically
isotropic chains with S ) 1 local spins would serve as alternative
systems to better understanding of low-dimensional magnetism,
as well as the recent topic of quantum spin liquids.29
Figure 1. Annelated S ) 1 diradicals.
Annelations of the nitroxide moieties to the m-phenylene, such
that the nitroxides are conformationally constrained to copla-
narity with the m-phenylene, may provide diradicals with large
2J values that are relatively insensitive to environmental
perturbations. To our knowledge, there are only two known
annelated S ) 1 organic diradicals: the annelated polyarylmethyl
diradical18 and nitroxide diradical 1 (Figure 1).19 The annelated
polyarylmethyl diradical was reported to be stable at ambient
temperature but reacted with atmospheric oxygen to give
peroxides.20 Nitroxide diradical 1 was reported to be stable in
the solid state but decomposed slowly in solution.19 Furthermore,
contradictory magnetic properties of diradical 1, concerning the
ground state and value of 2J, were reported.21
Intrigued by these results, we set out to investigate further
magnetic properties of the annelated nitroxide diradical 1. In
addition, we designed nitroxide diradicals 2 and 3 to probe the
effect of steric hindrance on stability and magnetic properties
of annelated diradicals. In 2, spirocyclohexyl groups may
provide greater steric shielding of the nitroxides, compared to
that in 1.22 Substitution with the bulky tert-butylphenyl group
at the ortho/ortho position, where significant spin density is
expected by delocalization from both nitroxides, may provide
improved stability or persistence for diradical 3.23 However, this
steric hindrance may lead to a strained structure and decrease
the degree of planarity of the benzobisoxazine annelated system.
Therefore, both stability of the diradical and the strength of
ferromagnetic coupling may be affected.
Results and Discussion
Synthesis. The synthesis of diradicals 1, 2, and 3 is outlined
in Scheme 1.
Starting with dimethyl 4,6-dibromoisophtalate 4 (Supporting
Information), the C-N cross-coupling with excess of benzy-
lamine is followed by the reductive debenzylation to give
dimethyl 4,6-diaminoisophtalate 633 in high yield. Reaction of
6 with an excess of CH3MgBr provides a diol diamine 7. Acetic
acid-catalyzed condensations of 7 with acetone and cyclohex-
anone lead to tricyclic diamines 8 and 9, respectively.34
Functionalization of 8 with 4-tert-butylphenyl group at the
ortho position (with respect to amines) is carried out in two
steps. Bromination of 8 with NBS at low temperature gives the
monobrominated product 10 in high yield. The Suzuki cross-
We were able to optimize the synthesis to obtain 1 and 3 in
good yields and with high purity. This allowed us to carry out
a thorough magnetic and structural characterization of both
(24) The nonclassical C-H-O hydrogen bonds: (a) Taylor, R.; Kennard, O.
J. Am. Chem. Soc. 1982, 104, 5063-5070. (b) Desiraju, G. R. Acc. Chem.
Res. 1991, 24, 290-296.
(15) Fang, S.; Lee, M.-S.; Hrovat, D. A.; Borden, W. A. J. Am. Chem. Soc.
1995, 117, 6727-6731.
(16) Rajca, A.; Lu, K.; Rajca, S.; Ross, C. R., II. J. Chem. Soc., Chem. Commun.
1999, 1249-1250.
(17) There is no detailed quantitative experimental data on the dependence of
the exchange coupling vs torsional angles in m-phenylene moiety. However,
a “Karplus-Conroy-type” relationship for exchange coupling in trimeth-
ylene-based bis(semiquinone) diradicals was proposed: Shultz, D. A.; Fico,
R. M., Jr.; Bodnar, S. H.; Kumar, K.; Vostrikova, K. E.; Kampf, J. W.;
Boyle, P. D. J. Am. Chem. Soc. 2003, 125, 11761-11771.
(18) Rajca, A.; Utamapanya, S. J. Org. Chem. 1992, 57, 1760-1767.
(19) Rassat, A.; Sieveking, U. Angew. Chem., Int. Ed. Engl. 1972, 11, 303-
304.
(20) Rajca, A.; Rajca, S.; Desai, S. R.; Day, V. W. J. Org. Chem. 1997, 62,
6524-6528.
(21) Chiarelli, R.; Gambarelli, S.; Rassat, A. Mol. Cryst. Liq. Cryst. 1997, 305,
455-478.
(22) (a) Rozantsev, E. G. Free Nitroxyl Radicals; Plenum Press: New York,
1970; pp 135-139. (b) For the related nitroxide monoradicals, stability
might be improved by replacement of the dimethylmethylene group at the
nitroxide with a spirocyclohexyl group, though the evidence is not definitive.
(23) (a) Forrester, A. R.; Thompson, R. H. Nature 1964, 203, 74-75. (b) Calder,
A.; Forrester, A. R. J. Chem. Soc. C 1969, 1459-1464.
(25) (a) Haldane, F. D. M. Phys. Lett. A 1983, 93, 464-468. (b) Haldane, F. D.
M. Phys. ReV. Lett. 1983, 50, 1153-1156.
(26) Antiferromagnetic spin-1 chains of [Ni(1,3-diamino-2,2-dimethylpropane)2-
(µ-N3)](PF6): Tsujii, H.; Honda, Z.; Andraka, B.; Katsumata, K. Phys. ReV.
B 2005, 71, 014426-1-6.
(27) Synthesis of [Ni(1,3-diamino-2,2-dimethylpropane)2(µ-N3)](PF6): Monfort,
M.; Ribas, J.; Solans, X.; Font-Bard´ıa, M. Inorg. Chem. 1996, 35, 7633-
7638.
(28) Zheludev, A.; Honda, Z.; Chen, Y.; Broholm, C. L.; Katsumata, K.; Shapiro,
S. M. Phys. ReV. Lett. 2002, 88, 077206-1-4.
(29) Stone, M. B.; Zaliznyak, I. A.; Hong, T.; Broholm, C. L.; Reich, D. H.
Nature 2006, 440, 187-190.
(30) Renard, J. P.; Verdaguer, M.; Regnault, L. P.; Erkelens, W. A. C.; Rossat-
Mignod, J.; Stirling, W. G. Europhys. Lett. 1987, 3, 945-952.
(31) Anisotropy of Haldane gap in [Ni(1,3-diamino-2,2-dimethylpropane)2(µ-
N3)](PF6): Zheludev, A.; Chen, Y.; Broholm, C. L.; Honda, Z.; Katsumata,
K. Phys. ReV. B 2001, 63, 104410-1-5.
(32) Spin-excitation spectrum in relatively more isotropic CsNiCl3: Zaliznyak,
I. A.; Lee, S. H.; Petrov, S. V. Phys. ReV. Lett. 2001, 87, 017202-1-4.
(33) 4,6-Diaminoisophthalate: Bogert, M. T.; Kropff, A. H. J. Am. Chem. Soc.
1909, 31, 841-848.
(34) Syntheses of benzobisoxazines by silica catalysis: Spagnol, G.; Rajca, A.;
Rajca, S. J. Org. Chem. 2007, 72, 1867-1869.
9
10160 J. AM. CHEM. SOC. VOL. 129, NO. 33, 2007