Modulating spin delocalization in phenoxyl radicals conjugated with heterocycles

Article abstract of DOI:10.1021/ol0063407

(Equation Presented) Phenoxyl radicals were synthesized having heterocyclic substituents in the 4-position. Electron spin resonance spectroscopy shows spin density can be modulated by delocalization onto the heterocycle structure, to the extent that some

Full text of DOI:10.1021/ol0063407

ORGANIC
LETTERS
2000
Vol. 2, No. 22
3417-3420
Modulating Spin Delocalization in
Phenoxyl Radicals Conjugated with
Heterocycles
,†
Chunping Xie,† Paul M. Lahti,* and Clifford George‡
Department of Chemistry, UniVersity of Massachusetts, Amherst, Massachusetts 01003,
and Laboratory for the Structure of Matter, NaVal Research Laboratory,
Washington, D.C. 20375
lahti@chem.umass.edu
Received July 17, 2000
ABSTRACT
Phenoxyl radicals were synthesized having heterocyclic substituents in the 4-position. Electron spin resonance spectroscopy shows spin
density can be modulated by delocalization onto the heterocycle structure, to the extent that some of the largest spin density bearing sites
in 3 and 4 lies in the heterocycle, not in the phenoxyl ring.
Sterically stabilized phenoxyl radicals have been known for
almost a half-century as subjects of electron spin resonance
(ESR), UV-vis, and infrared studies. 2,6-Di-tert-butyl-
phenoxyl radicals are persistent in deoxygenated solution,
but generally dimerize and decompose in the solid state, with
the exception of the structurally related galvinoxyl radical.1
Infrared and ESR spin density studies show stabilized
phenoxyl radicals to be highly delocalized, with considerable
CdO character. Recent efforts to design organic-based
molecular magnetic materials have impelled studies of
electronic structure in delocalized open-shell organic mol-
ecules.2 In this report, we describe the generation of 2,6-di-
tert-butylphenoxyl radicals 1-4 with conjugated heterocyclic
rings attached at the 4-position and show spectroscopically
that delocalization varies considerably as a function of the
heterocycle structure and is so extensive in some cases that
major spin density bearing sites are found beyond the
phenoxyl radical moiety.
Scheme 1 shows the syntheses of the phenol precursors
5-8. In each case the key step is the palladium-catalyzed
Suzuki coupling of trimethylsilyl-protected 2,6-di-tert-
butylphenol-4-boronic acid3 9 with the appropriate hetero-
arene. Deprotection occurs during workup, directly yielding
the phenols after column chromatography. Spectral and
elemental characterizations of all new compounds were
satisfactory. The characteristically narrow OH band of a 2,6-
di-tert-butylphenol was observed in the FTIR spectrum for
KBr-pelletized solid samples of 5, 7, and 8. Curiously, a
broadened OH band was observed for 6 under the same
conditions, suggesting the presence of hydrogen bonding,
despite the steric blockading due to the tert-butyl groups.
X-ray crystallographic analysis4 confirmed the presence of
hydrogen bonding between the phenolic OH and the pyrim-
idine nitrogen of another molecule in the crystal lattice.
Figure 1 shows a schematic diagram of this hydrogen
† University of Massachusetts.
‡ Naval Research Laboratory.
(1) Altwicker, E. R. Chem. ReV. 1967, 67, 475.
(2) (a) Kahn, O. Molecular Magnetism; VCH: New York, 1993. (b)
Turnbull, M. M.; Sugimoto, T.; Thompson, L. K., Eds. Molecule-based
Magnetic Materials. Theory, Techniques, and Applications; American
Chemical Society, Washington, DC, 1996; Vol. 644. (c) Lahti, P. M., Ed.
Molecular Magnetism in Organic-Based Materials; Marcel Dekker: New
York, 1999.
(3) Satoh, Y., Shi, C. Synthesis 1994, 1146.
(4) Selected crystallographic data for 6: orthorhombic, Pna21; ortho-
rhombic (Z ) 4), a ) 12.118(1), b ) 7.514(1), c ) 18.799(2) Å. Selected
lengths, angles, torsions: O1-N′5A ) 2.797(3) Å, H-N5′A ) 2.24(4) Å,
O1-H-N5′A ) 123(3)°, C2′-C1′-C3-C4 ) 35.1°, C2-C1-O1-H )
27.1°.
10.1021/ol0063407 CCC: $19.00 © 2000 American Chemical Society
Published on Web 10/11/2000

Scheme 1
bonding arrangement in 6. Additional data is given in the
698 nm, respectively. Extinction coefficients and radical
concentrations were not determined due to the instability of
1-4 over extended time periods. Radical 3 showed no longer
Supporting Information. Further support for the presence of
this somewhat surprising hydrogen bonding motif is given
by the fact that a 1:1 coordination complex between 6 and
copper(II) chloride shows the return of a typical, narrow OH
band at 3620 cm-1 in the FTIR, presumably because the
nitrogen acceptor sites are now at least partly coordinated
by copper.
The phenols were oxidized in degassed benzene by stirring
with lead dioxide recently prepared by hydrolysis of lead
tetraacetate. The solutions rapidly became colored, as is
characteristic for formation of phenoxyl radicals. The longest
wavelength UV-vis bands of 1-4 were 496, 614, 420, and
Figure 1. Hydrogen bonding in phenol 6, taken from a single-
crystal X-ray analysis. N5′A is symmetry generated. See Supporting
Information for further structural details.
Figure 2. Room-temperature ESR spectra of 1 (curve a, 9.791
GHz), 2 (curve b, 9.791 GHz), 3 (curve c, 9.800 GHz), 4 (curve d,
9.798 GHz).
3418
Org. Lett., Vol. 2, No. 22, 2000

range band in the >500 nm range, but has a long absorbance
tail that makes it appear yellow in dilute solution but red at
higher concentrations.
of these systems appears to be somewhat lower than that of
the 4-(1H-benzimidazol-2-yl)-2,6-di-tert-butylphenoxyl sys-
tems that we have described previously.7 For example,
system 2 shows the appearance of bands in the CdO region
of the FTIR in KBr pellets, implying considerable dimer-
ization. Systems 3 and 4 appear not to dimerize so readily
and also show longer-lived coloration in solution upon
exposure to air.
ESR hyperfine analysis shows the extent of delocalization
of the unpaired spin from the phenoxyl radical moiety. In
systems 3 and 4, the largest hydrogen hyperfine coupling
constant is found on the heterocyclic ring system, not on
the phenoxyl unit. This demonstrates the extent to which
the spin density in a conjugated radical can be modulated
by appropriate choice of structure. We can use the ap-
proximate McConnell model8 relating hyperfine coupling to
spin density and assume that the McConnell proportionality
constant is (-)22 G, a typical value for delocalized radicals.
Using the experimental hfc a(4) ) 1.8-2.1 G in 1 and 2
and a(2) ) 4.2 and 3.3 G in 3 and 4, the amount of π-spin
at these positions is estimated to be 8-10%, 19%, and 15%,
respectively.
Figure 2, curves a-d, show the X-band ESR spectra of
the radical solutions. The hyperfine coupling constants (hfc)
were obtained by line shape analysis of the spectra using
the WINSIM program of Duling.5 To assist with the hfc
assignments, density functional computations were carried
out using Gaussian 986 at the BLYP/cc-pVDZ level, using
B3LYP/6-31G* optimized geometries for model systems
where the tert-butyl groups were replaced by hydrogen
atoms. All computed geometries gave essentially coplanar
rings. All computed hfc were taken directly from the
Gaussian computations, which are based upon the computed
Fermi contact couplings. Table 1 summarizes the experi-
Table 1. Experimental and Computed hfc for 1-4
The experimental estimates of spin density distributions
also compare well to the Mulliken spin populations predicted
by Gaussian 98 at the 4′-positions for 1 and 2 and the 2′-
positions for 3 and 4. Figure 3 summarizes the experimental
and computed percent spin density populations for the hfc
data in Table 1. The furan ring substituent in the 4-position
of the phenoxyl radical leads to the largest amount of
delocalization, with the thiophene also supporting a large
delocalization.
2,4,6-Tri-tert-butylphenoxyl has a 32% spin density popu-
lation at the 4-position.9 By comparison, 4-phenyl-2,6-di-
tert-butylphenoxyl has no spin density population that is
larger than 7% on any ortho and para position of the
4-phenyl substituent.10 (The percent spin populations ignore
small contributions of negative spin density on some sites.)
So, only a modest amount of delocalization occurs from the
phenoxyl 4-position into the conjugated phenyl ring in
4-phenyl-2,6-di-tert-butylphenoxyl. By comparison to these
phenoxyl radicals that lack heteroatom substituents π-con-
jugated to the 4-position, 1-4 all have spin sites with 10-
19% population on the 4-substituent ring. The difference in
delocalization is partly attributable to the strong desire of
the radicals to place large spin density populations on sites
a Hfc are absolute values based on WINSIM simulations. b Signs are
based on expected negative hfc for a(m), for model systems with tert-butyl
groups replaced by hydrogen, computed at the BLYP/cc-pVDZ//B3LYP/
6-31G* level.
(5) Duling, D. R. J. Magn. Reson. 1994, B104, 105.
(6) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Gill, P. M. W.; Johnson,
B. G.; Robb, M. A.; Cheeseman, J. R.; Keith, T.; Petersson, G. A.;
Montgomery, J. A.; Raghavachari, K.; Al-Laham, M. A.; Zakrzewski, V.
G.; Ortiz, J. V.; Foresman, J. B.; Cioslowski, J.; Stefanov, B. B.;
Nanayakkara, A.; Challacombe, M.; Peng, C. Y.; Ayala, P. Y.; Chen, W.;
Wong, M. W.; Andres, J. L.; Replogle, E. S.; Gomperts, R.; Martin, R. L.;
Fox, D. J.; Binkley, J. S.; Defrees, D. J.; Baker, J.; Stewart, J. P.; Head-
Gordon, M.; Gonzalez, C.; Pople, J. A.; Gaussian Inc.: Pittsburgh, PA,
1998.
mental hfc for 1-4 and the computed hfc for the model
systems. The brackets around a(m) emphasize that these
computed hfc are likely to be influenced by the replacement
of the tert-butyl groups, relative to the actual experimental
values.
(7) Xie, C.; Lahti, P. M. Tetrahedron Lett. 1999, 40, 4305.
(8) a ) QF, where A is hfc, F is π-spin density, and Q is the McConnell
constant. See McConnell, H. M. J. Chem. Phys. 1956, 28, 1188 and Pople,
J. A.; Beveridge, D. L. Approximate Molecular Orbital Theory; McGraw-
Hill: New York, 1970, p 146 ff.
(9) Prabhanada, B. S. J. Chem. Phys. 1983, 79, 5752.
(10) Mukai, K.; Inagaki, N. Bull. Chem. Soc. Jpn. 1980, 53, 2695.
The solution ESR spectra and colors of 1-4 persisted for
up to 3 days under nitrogen, but faded swiftly when exposed
to air. When solvent was removed, the resultant solids
lightened in color upon exposure to air: the color change is
irreversible and the radicals nonrecoverable. The persistence
Org. Lett., Vol. 2, No. 22, 2000
3419

Scheme 2
An important consideration in the design of molecular
magnetic materials is interaction of spin density in the solid
state or in very high spin polymeric materials. The ability
to manipulate sites of major spin density by structural
variation gives potential electronic control in such materials.
For example, our group and Yamamoto’s have synthesized
2,5-thienediyl polymers based upon radical 4.13-14 This
design was based upon the expectation that delocalization
from the pendant phenoxyl groups through the polythiophene
backbone would be sufficient to allow effective exchange
between spin-bearing units. The present results for 4 show
that spin delocalization is qualitatively sufficient to meet the
original design intent.
This study demonstrates some of the possibilities for
modulation of spin density in organic radicals by a “mix-
and-match” strategy of attaching heterocyclic systems with
π-conjugation to delocalizable radical sites. In particular, the
furan and thiophene substituents allow a substantial degree
of spin density delocalization away from the phenoxyl ring.
DFT computations turn out to be very good predictors of
the spin density distributions in these systems and should
be very useful for prediction of delocalization extents in
related phenoxyl-based open-shell molecules. Reasonable
strategies to incorporate such organic radicals into molecular
magnetic materials will combine tuning spin density distribu-
tion by this strategy and coordination of sites of higher spin
density with paramagnetic ions to strengthen exchange
interactions between organic and inorganic units in hybrid
material. By these methods we hope to improve the stability
and magnetic ordering characteristics of new organic and
hybrid organic-inorganic materials.
Figure 3. Experimental and [BLYP computed] percent spin
populations. Experimental spin densities for 4-phenyl-2,6-di-tert-
butylphenoxyl from ref 9. BLYP computed values were done on
model systems with replacement of tert-butyl groups by hydrogen
atom substituents. Negative (-) spin populations are deduced by
spin polarization parity guidelines.
stabilized by adjacent nitrogen, oxygen, or sulfur atoms.
There is also an effect of having five-member rings in 3 and
4, by comparison to a six-membered ring in 4-phenyl-2,6-
di-tert-butylphenoxylswith one less π-atom, there is more
likelihood of a single site with a larger spin population.
Some comparisons to related delocalizable radicals with
heteroring substitution are available. 2-Pyridyl-tert-butyl
nitroxide11-12 and 5-pyrimidyl-tert-butyl nitroxide are con-
nectivity analogues of 1 and 2, respectively. The pyridyl
compound has previously been generated in trapping experi-
ments, and both have been synthesized by us12 through
metalation of the corresponding bromo compounds, treatment
with 2,2-dimethyl-2-nitrosopropane, and oxidation. The
nitroxide group is inherently less delocalized than the
phenoxyl group, but the qualitative spin density patterns in
the nitroxides (Scheme 2, experimental hfc given in gauss)
are analogous to those in the phenoxyl systems.
Acknowledgment. This work was supported by the
National Science Foundation (CHE 9809548). The opinions
expressed in this paper are solely those of the authors and
not necessarily those of the Foundation. We also acknowl-
edge assistance from the Naval Research Laboratory for
assistance with crystallographic analysis.
Supporting Information Available: Experimental pro-
cedures for syntheses of 1-8, Gaussian archive files for spin
density and hfc computations of 1-4, tables of crystal-
lographic parameters for 6, FTIR spectra of 5-8, 2, 4, and
6-CuCl2. This material is available via the Internet free of
charge from http://pubs.acs.org.
(11) Bentley, T. W.; John, J., A.; Johnstone, R. A. W. J. Chem. Soc.,
Perkin Trans. 2 1973, 1039-1044.
(12) (a) Lahti, P. M.; Esat, B.; Ferrer, J. R.; Liu, Y.; Marby, K. A.; Xie,
C.; George, C.; Antorena, G.; Palacio, F. Mol. Cryst. Liq. Cryst. Sci.
Technol., Sect. A 1999, 334, 285-294. (b) Field, L. M.; Marby, K. A.;
Lahti, P. M. Unpublished results.
(13) Xie, C.; Lahti, P. M. J. Polym. Sci. Part A: Polym. Chem. 1999,
37, 779.
(14) Yamamoto, T.; Hiyashi, H. J. Polym. Sci. Part A, Polym. Chem.
1997, 35, 463.
OL0063407
3420
Org. Lett., Vol. 2, No. 22, 2000