Crystallography and magnetism of two 1-(4-nitroxylphenyl)pyrroles

Article abstract of DOI:10.1021/jo0616299

(Figure Presented) Stable radicals 1-(4-(N-tert-butyl-N-amnoxyl)phenyl) pyrrole (BNPP) and 1-(4-(N-[para-methoxyphenyl]-N-aminoxyi)phenyl)pyrrole (MNPP) were, synthesized and characterized by crystallography and magnetism. BNPP crystals exhibit 1-D chains

Full text of DOI:10.1021/jo0616299

Crystallography and Magnetism of Two
1-(4-Nitroxylphenyl)pyrroles
Zeynep Delen and Paul M. Lahti*
Department of Chemistry, UniVersity of Massachusetts, Amherst, Massachusetts 01003
lahti@chem.umass.edu
ReceiVed August 4, 2006
Stable radicals 1-(4-(N-tert-butyl-N-aminoxyl)phenyl)pyrrole (BNPP) and 1-(4-(N-[para-methoxyphenyl]-
N-aminoxyl)phenyl)pyrrole (MNPP) were synthesized and characterized by crystallography and magnetism.
BNPP crystals exhibit 1-D chains of intermolecular nitroxide NO to nitroxide CH3 contacts, but
polycrystalline magnetic susceptibility measurements show quite small antiferromagnetic (AFM) exchange
interactions. MNPP shows stronger AFM exchange interactions that appear to be associated with a 2-D
planar mesh of crystallographic nitroxide to nitroxide (N)O‚‚‚N(O) contacts of 4.0-4.2 Å. The AFM
behavior of MNPP can be fitted to a 2-D square planar Heisenberg antiferromagnetic exchange model
with J/k ) (-)0.78 ( 0.04 K and mean field constant θ ) (-)0.77 ( 0.12 K.
Introduction
flatter molecules to allow possible closer packing. Although some
diaryl nitroxides are stable enough to be studied spectroscopical-
ly, or even to use in spin labeling experiments,3 few have been
subjected to crystallographic and bulk magnetic analyses. DANO
Nitroxides have long been targets of substantial interest, due
to their unusual stability among the family of organic radicals.1
More recently, they have been much used as building blocks for
molecular magnetic materials.2 Conjugated nitroxides are par-
ticularly interesting, since they allow significantly more spin de-
localization than nitronylnitroxides and iminoylnitroxides and
thereby more chances for intermolecular exchange interactions
in a crystal lattice between atoms bearing unpaired spin density.
Most work has concentrated on aryl tert-butylnitroxides, which
can delocalize onto one π-system. By comparison, diarylnitrox-
ides can delocalize onto two aryl rings rather than one and are
(1) (a) Forrester, A. R.; Hay, J. M.; Thomson, R. H. Organic Chemistry
of Stable Free Radicals; Academic Press: New York, 1968.
(2) (a) Rassat, A. Pure Appl. Chem. 1990, 62, 223. (b) Veciana, J.;
Cirujeda, J.; Rovira, C.; Molins, E.; Novoa, J. J. J. Phys. I. 1996, 6, 1967.
(c) Nakatsuji, S.; Anzai, H. J. Mater. Chem. 1997, 7, 2161. (d) Pilawa, B.
Ann. Phys. 1999, 8, 191-254. (e) Itoh, K.; Kinoshita, M. Molecular
Magnetism: New Magnetic Materials; Gordon and Breach: Newark, NJ,
2000. (f) Miller, J. S., Drillon, M., Eds. Magnetism: Molecules to Materials;
Volumes I-V; Wiley-VCH: Weinheim. (g) Lahti, P.M. in Carbon-based
magnetism: An oVerView of the magnetism of metal-free carbon-based
compounds and materials; Palacio F., Markova, T., Eds.; Elsevier: Am-
sterdam, 2006; pp 23-52.
has been the most studied diarylnitroxide magnetic system.4
Other examples are diphenyl nitric oxide (DNO),5 its derivatives
(3) (a) Calder, A.; Forrester, A. R. J. Chem. Soc. C. 1969, 1459. (b)
Forrester, A. R.; Hepburn, S. P.; McConnachie, G. J. Chem. Soc. Perkin
Trans. 1974, 1, 2213. (c) Forrester, A. R.; Hepburn, S. P. J. Chem. Soc.
Perkin Trans. 1, 1974, 2208. (d) Keana, J. F. W. Chem. ReV. 1978, 78, 37.
10.1021/jo0616299 CCC: $33.50 © 2006 American Chemical Society
Published on Web 11/04/2006
J. Org. Chem. 2006, 71, 9341-9347
9341

Delen and Lahti
BTAO,6 DMAO, DEAO, DBAO and DPAO,7 and bis[3-tert-
butyl-5-(N-tert-butylaminoxyl)phenyl]nitroxide (bNPhNO).8
The attachment of conjugating heteroarene substituents
provides an important strategy for modulating spin density
distributions, crystal packing, and magnetic exchange in solid
nitroxide radicals.9 This article describes the synthesis and
magnetostructural characterization of two para-nitroxylphen-
ylpyrroles, 1-(4-(N-tert-butyl-N-aminoxyl)phenyl)pyrrole (BP-
NPP) and 1-(4-(N-[para-methoxyphenyl]-N-aminoxyl)phenyl)-
pyrrole (MNPP).
SCHEME 1. Syntheses of BNPP and MNPP
Results
Synthesis. BNPP and MNPP were synthesized as shown in
Scheme 1. Initially, BNPP was pursued by palladium diacetate
catalyzed C-N bond formation between pyrrole and BrPhNIT-
BDMS to give BNPPTBDMS, followed by deprotection to
BNPPH and oxidation to the radical (route 1). This route did
afford milligram quantities of BNPP, but the initial coupling
step gave lower yields than was desirable, given the effort of
making10 BrPhNITBDMS. Route 2 turned out to be better for
us, via a high yield condensation of 4-bromobenzaldehyde with
2,5-dimethoxytetrahydrofuran to give pyrrole BrPP, followed
by metalation, reaction with 2-methyl-2-nitrosopropane11 to give
BNPPH, and oxidation with PbO2 to yield BNPP. Similarly,
MNPP was made by metalation of BrPP and reaction with
4-nitrosoanisole;12 the presumed MNPPH was not isolated but
oxidized directly in air to give the radical. Both product radicals
are deep red-purple, crystalline solids that show no evidence
of decomposition after months of exposure to air at room
temperature.
ESR Spectroscopy. Table 1 summarizes hyperfine coupling
constants (hfc) derived from the X-band solution ESR spec-
troscopy at 200-300 K and includes comparisons to compu-
tational hfc estimates. Spin density populations computed13 using
the UB3LYP/6-31G* hybrid density functional14 (at crystal-
lographic molecular geometries described in the following
section) were converted to predicted hfc using the McConnell
relationship,15 eq 1. Here, aX is the predicted hfc from atom X,
F is the spin density on nitrogen for aN or on the attached
π-carbon for an aryl hydrogen atom: values of QN ) 30 G and
QH ) (-)22 G were used.
Crystallography. The radicals were studied by single-crystal
X-ray diffraction at room temperature. Figure 1 shows ORTEP
style diagrams16 of the molecular structures. Further crystal-
lographic details are summarized in the experimental section
and in supporting material, including selected structural and
intermolecular contact parameters.
Magnetic Measurements. Figures 2 and 3 show temperature
versus paramagnetic susceptibility (ø) data for the radicals. The
solid and dashed line functions in the figures come from analyses
that are described in the discussion section below. Magnetization
versus field measurements also were carried out for the samples
at 1.8 K; details are given in the Experimental Section and in
the Supporting Information.
Discussion
Molecular Spin Density Distributions. An important reason
for choosing arylnitroxide building blocks for molecular mag-
netism is their ability to delocalize unpaired spin to some extent.
The ESR spectra and computational spin density analyses show
where the significant spin density populations are located in
BNPP and MNPP. BNPP shows significant delocalization onto
its phenyl ring. Hfc from the tert-butyl group on BNPP was
not resolved: comparison to hyperfine deduced for di-tert-
butylnitroxide and structurally related nitroxides suggests17 that
this would be ∼0.1 G or less. MNPP shows delocalization onto
both aryl rings, somewhat more on the phenylene than the anisyl
ring. Neither radical shows significant spin delocalization onto
the pyrrole rings.
aX ) QXF
(1)
(4) (a) Duffy, W. J.; Strandburg, D. L.; Deck, J. F. Phys. ReV. 1969,
183, 567. (b) Takizawa, O. Bull. Chem. Soc. Jpn. 1976, 49, 583.
(5) Takizawa, O.; Yamauchi, J.; Ohya-Nishiguchi, H.; Deguchi, Y. Bull.
Chem. Soc. Jpn. 1971, 44, 3188.
(6) Imachi, R.; Ishida, T.; Suzuki, M.; Yasui, M.; Iwasaki, F.; Nogami,
T. Chem. Lett. 1997, 8, 743.
(7) Nakagawa, M.; Ishida, T.; Suzuki, M.; Hashizume, D.; Yasui, M.;
Iwasaki, F.; Nogami, T. Chem. Phys. Lett. 1999, 302, 125.
(8) Ishida, T.; Iwamura, H. J. Am. Chem. Soc. 1991, 113, 4238.
(9) Cf. a review in Lahti, P. M. “Magneto-structural Correlations in
π-Conjugated Nitroxide-based Radicals: Hydrogen-bonds and Related
Interactions in Molecular Organic Solids”, in Carbon-based magnetism:
An oVerView of the magnetism of metal-free carbon-based compounds and
materials; Palacio, F., Markova, T., Eds.; Elsevier: Amsterdam, The
Netherlands, 2006; pp 23-52.
From Table 1, the largest delocalized spin populations lie on
the protons ortho to the nitroxides (7-9%); the second largest
lie on the meta protons (4-7%). The anisole proton hfc are
smaller than those on the phenyl ring of MNPP, but still
(10) Field, L. M.; Lahti, P. M. Chem. Mater. 2003, 15, 2861.
(11) Stowell, J. C. J. Org. Chem. 1971, 36, 3055.
(12) Defoin, A. Synthesis, 2004, 5, 706.
(13) Computations were carried out using UNIX Spartan 2000 from
Wavefunction, Irvine, CA.
(14) Becke, A. D. J. Chem. Phys. 1993, 98, 5648.
(15) McConnell, H. M. J. Chem. Phys. 1956, 24, 764.
(16) Farrugia, L. J. J. Appl. Cryst. 1997, 30, 565.
(17) (a) Faber, R. J.; Markeley, F. W.; Weil, J. A. J. Chem. Phys. 1967,
46, 1652. (b) Bales, B. L.; Peric, M.; Lamy-Freund, M. T. J. Magn. Res.
1998, 132, 279.
9342 J. Org. Chem., Vol. 71, No. 25, 2006

Crystallography and Magnetism of BNPP and MNPP
TABLE 1. Experimental and Computed hfc for BNPP and MNPPa
BNPP
MNPP
experimental
11.74
estimated
experimental
estimated
11.8 (nitroxide N)
9.73
1.72
10.0 (nitroxide N)
1.5 (anisole CH ortho to NO)
0.8 (anisole CH meta to NO)
2.0 (phenyl CH ortho to NO)
1.1 (phenyl CH meta to NO)
0.04 (pyrrole 2,5 CH)
0.15 (pyrrole 3,4 CH)
<0.01 (pyrrole N)
<0.001 (methoxy CH)
2.22
0.90
2.0 (CH ortho to NO)
1.1 (CH meta to NO)
0.04 (pyrrole 2,5 CH)
0.13 (pyrrole 3,4 CH)
-0.01 (pyrrole N)
1.98
0.89
<0.0025 (tert-Bu C-H)
a All hfc in gauss. Experimental hfc from ESR spectroscopy; non-italic estimated hfc from UB3LYP/6-31G* spin density computations with conversion
using eq 1 (QN ) 30 G, QH ) -22 G); italic estimated hfc for sp3 CH taken directly from computations.
The computed spin density patterns show that the radicals
conform to typical expectations of spin polarization behavior,
with alternating signs of spin density on the phenyl rings that
are in direct resonance conjugation with the nitroxide groups.
The nitroxides themselves are computed to retain over 80% of
the total spin density, and the rest is mostly delocalized onto
the phenyl rings, with quite small populations on other moieties.
No more than 0.7% spin population is predicted on any CH of
the pyrroles (Table 1), and e0.25% spin population is predicted
on any BNPP tert-butyl group proton (depending on confor-
mational placement in the latter). These results are in good
accord with the experimental observations and provide a
reasonable basis for attempting to understand intermolecular
exchange in the crystal structures of the two radicals.
Crystallographic Close Contacts. BNPP crystallizes in the
orthorhombic Pna21 space group. The closest intermolecular
contacts between NO units are all >5.4 Å, sufficiently long to
rule out significant direct interaction between these major spin
density sites. The closest intermolecular contacts with the NO
units are to the CH bonds on nearby tert-butyl, as shown in
Figure 4, with r[(N)O‚‚‚H(C,tert-butyl)] ∼ 2.6 Å. There are no
FIGURE 1. ORTEP diagrams for BNPP and MNPP. Thermal
ellipsoids shown at 50% probability.
close contacts between the nitroxides and any aryl CH bonds.
Since there are no close intermolecular contacts between the
groups bearing significant spin densities, any intermolecular
exchange would result from through-space dipole-dipole
interactions or spin-polarization interactions involving the very
small spin densities on the tert-butyl or pyrrole groups. Aryl
CH/π-cloud herringbone T-type contacts between phenyl and
pyrrole rings help to hold BNPP molecules together in the lattice,
but they do not provide additional close contacts between sites
bearing significant spin density.
MNPP crystallizes in the monoclinic Cc space group. The
vectors of all NO bonds point essentially along the crystal-
lographic b-axis. Unlike BNPP, MNPP has both nitroxide to
aryl CH (N)O‚‚‚HC contacts at distances of 2.7-2.9 Å, and
(N)O‚‚‚N(O) contacts at moderate distances of about 4.1 Å. Both
types of contacts are shown in Figure 5. (CH3)O‚‚‚H3C(O)
dipolar interactions as well as phenyl/phenyl and pyrrole/ pyrrole
CH/π-cloud herringbone T-type contacts bring the MNPP
molecules into close contact throughout the lattice, forming
unusual 2-D sheets of nitroxide-nitroxide contacts shown in
Figure 6. Pyrrole-anisole CH‚‚‚(O)CH3 dipolar contacts as-
sociate neighboring sheets at a plane to plane distance of 15 Å.
The combination of these various relatively weak dipolar and
induced dipolar interactions in MNPP causes it to crystallize to
form sheet-like regions of high spin density sites (the nitroxides)
that are widely separated by regions with very little spin density.
FIGURE 2. øT versus T plot for polycrystalline BNPP at 1000 Oe.
Inset shows full temperature range of the experiment, 1.8-300 K.
Broken line shows spin-pairing fit, and solid line shows 1-D AFM
chain fit to data.
significant at the position ortho to the nitroxide. The phenyl
rings therefore bear enough spin density to provide potential
intermolecular exchange interactions at close enough contact
distances. Since spin density in MNPP is spread onto two aryl
rings, it has more possibilities for intermolecular solid-state
exchange contacts.
J. Org. Chem, Vol. 71, No. 25, 2006 9343

Delen and Lahti
FIGURE 3. ø versus T plots for polycrystalline MNPP at 1000 Oe, 1.8-300 K (left) and 1.8-10 K (right). Inset shows full temperature range plot
of the øT(T) data, 1.8-300 K. Broken gray line shows spin-pairing fit, solid gray line shows 1-D AFM chain fit, and solid black line shows 2-D
square planar AFM model fit to data.
FIGURE 4. Close contacts in BNPP involving nitroxide groups. Part
a shows nitroxide-HC contacts, and part b shows nitroxide-nitroxide
contacts, with relation to nitroxide-HC contacts from part a. See also
Supporting Information.
FIGURE 5. Close contacts in MNPP involving nitroxide groups. Part
Magnetostructural Analysis of BNPP: Curie-Weiss analy-
a shows nitroxide-HC contacts, and part b shows nitroxide-nitroxide
contacts. See also Supporting Information.
sis (1/ø versus T, see Supporting Information) and the high-
temperature limiting value of the øT(T) plot (Figure 2) for
Heisenberg 1-D chain19 expressions, with variation of the sample
Lande´ constant, exchange constant, and a generalized mean field
constant; g, J/k, and θ, respectively. The fitted curves are shown
in Figure 2, and the model equations given in Supporting
Information.
For the spin-pairing fit, g ) 1.861 ( 0.008, J/k ) (-)3.7 (
0.2 K, and θ ) (-)2.35 ( 0.03 K; for the 1-D chain fit, g )
1.881 ( 0.002, J/k ) (-)1.3 ( 0.2 K, and θ ) (-)0.5 ( 0.2
multiple samples of BNPP yields a Curie constant of C ) 0.33
emu‚K/Oe‚mol. Although this value is a bit small (possibly due
to hygroscopic water uptake during sample preparation), it is
quite consistent with S ) 1/2 spin carriers. The Weiss constant
is virtually zero, showing near isolation of paramagnetic spins.
The øT(T) data decrease to about 0.18 emu‚K/Oe‚mol as the
temperature decreases below ∼30 K, showing weak antiferro-
magnetic (AFM) exchange interactions. Nonlinear least-squares
fits were carried out using spin-pairing18 and antiferromagnetic
(19) (a) Bonner, J. C.; Fisher, M. E. Phys. ReV. A 1964, 135, 650. (b)
Bonner, J. C.; Ph.D. Dissertation, University of London, U.K., 1968.
(18) Bleaney, B.; Bowers, K. D. Proc. R. Soc. London A 1952, 214.
9344 J. Org. Chem., Vol. 71, No. 25, 2006

Crystallography and Magnetism of BNPP and MNPP
SCHEME 3. Close Contacts and Presumed Spin
Polarization in MNPP
FIGURE 6. Planar sheets of spin density in MNPP, showing edge-on
view of packing between two sheets. Red atoms are nitroxide oxygens;
cyan atoms are nitrogens.
SCHEME 2. Close Contacts and Presumed Spin
Polarization in BNPP
Magnetostructural analysis of MNPP. For øΤ(T > 100 K),
MNPP has a Curie constant of 0.37 emu‚K/Oe‚mol, in excellent
1
agreement with the value for S ) /2 spin carriers. The øT(T)
data below 100 K show significant deviation from isolated
paramagnetic behavior, dropping to ∼0.12 emu‚K/Oe‚mol at
1.8 K. ø(T) also shows a maximum at about 2 K, near the
minimum temperature accessible in these experiments; this
indicates significantly stronger exchange interactions than
observed in BNPP. The ø(T) plots are virtually identical at 300,
1000, and 10 000 Oe, showing no field-dependent saturation
or metamagnetic effects.
The ø(T) data were fit using the same models as were used
for BNPP; the results are shown in Figure 3. For the spin-pairing
fit, g ) 2.065 ( 0.086, J/k ) (-)7.5 ( 3.6 K, and θ ) (-)5.9
( 0.7 K; for a 1-D chain fit, g ) 1.994 ( 0.006, J/k ) (-)2.09
( 0.05 K, and θ ) (-)0.41 ( 0.04 K. The fits were much
poorer if mean-field terms θ were not included. The spin-pairing
model deviates significantly from the data for T < 4.5 K, and
the statistical uncertainties in its fitted parameters are large. This
is consistent with a lack of crystallographic dimer contacts
between nitroxide groups that could yield spin-pairing
K; uncertainties are 95% confidence limits. Both fits deviate
somewhat from the data in the 5-30 K region, and the large
value of the Weiss constant for the spin-pairing fit suggests that
this model is imprecise. Also, BNPP has no nitroxide dimeric
contacts that are reasonably associated with spin-pairing ex-
change.
BNPP does have two chain contact motifs involving high
spin density sites (Figure 4): (1) nitroxide/nitroxide contacts
and (2) nitroxide‚‚‚HC(tert-butyl) contacts. Scheme 2 shows a
simplified representation of AFM exchange through chains of
NO‚‚‚NO contacts. These should be very weak due to the distan-
ces involveds5.7-5.8 Å for O(1)‚‚‚O(1′b) and O(1′)‚‚‚N(1′b)s
in qualitative accord with the observed magnetic behavior. The
closest contacts of the large nitroxide spin densities are with
tert-butyl groups at (N)O‚‚‚H(C) distances of 2.6-2.8 Å (based
on O(1)‚‚‚C′(2) and O(1)‚‚‚C′(3) distances; see Supporting
Information). The computed spin populations of the tert-butyl
CH groups are very small (Table 1), with absolute sign
depending on conformation relative to the NO group. Simplistic
connectivity-only spin polarization considerations suggest FM
exchange, as shown in the Scheme 2. However, the very small
spin density populations on the nitroxide tert-butyl group make
such chains at best a weak pathway for exchange. Since the
experimental AFM exchange is so weak, a confident assignment
to one of these mechanismssor other, less well-defined through-
space interaction such as dipole-dipole exchangesis not
possible. Overall, BNPP is best described as having nearly
paramagnetic behavior with only weak AFM interactions
between molecules.
The 1-D chain fit is better, although it shows some deviation
from the data below about 3 K. As described in the crystal-
lography section (Figure 5), there are two chain contact motifs
in MNPP: nitroxide to phenyl NO‚‚‚HC contacts, and nitrox-
ide-nitroxide NO‚‚‚NO contacts. Applying the alternant spin
density signs found in the computational modeling, an NO‚‚‚HC
chain is predicted to produce FM exchange, while through space
NO‚‚‚NO contacts favor AFM exchange; this is summarized
in Scheme 3. The latter mechanism is in accord with the
experimental result. The through space interaction between large
spin density nitroxides at the (N)O‚‚‚N(O) distances of 4.0-
4.2 Å apparently outweighs interaction between the nitroxide
and the (much smaller) aryl H-C spin density, even though
the latter is a much closer contact at 2.7-2.9 Å.20 Since the
nitroxide spin density sites in MNPP are significantly closer
together than in BNPP, the greater AFM exchange constant in
(20) Cf. a similar case of close contact NO‚‚‚NO interactions overcoming
closer contact NO‚‚‚HC interactions in Ferrer, J. R.; Lahti, P. M.; George,
C.; Oliete, P.; Julier, M.; Palacio, F. Chem. Mater. 2001, 13, 2447.
J. Org. Chem, Vol. 71, No. 25, 2006 9345

Delen and Lahti
SCHEME 4. Simplified Representation of the Proposed 2-D
Exchange Lattice That Can Arise from the Crystallography
of MNPPa
(Figure 3). Caution is appropriate, since the maximum in the
ø(T) plot is nearly at the low-temperature limit of our
magnetometersthis may contribute to the fact that the exchange
constant and the mean field constant are found to be nearly the
same. Still, the network of crystallographic contacts between
nitroxides (Scheme 4) and the good fit of a quasi 2-D square
planar AFM exchange model to the susceptibility data both
support describing MNPP exchange behavior in 2-D terms.
Additional experiments will be required to find whether MNPP
actual undergoes magnetic ordering as temperature decreases
further.
Conclusions
BNPP and MNPP are both stable radicals, MNPP unusually
so for a diarylnitroxide. BNPP is less spin delocalized than
MNPP by ESR, based on both experimental hyperfine coupling
data and computational studies; neither has significant spin
density on the pyrrole rings. Because MNPP is more delocalized,
it potentially can form more exchange-inducing contacts in a
crystal lattice than BNPP. Although the pyrrole units do not
bear spin density, they do contribute to the solid-state crystal
lattice packing in both, primarily by participating in CH-π type
contacts. Both radicals form noncentrosymmetric crystal lattices.
The closest intermolecular contacts involving the nitroxide
groups are NO‚‚‚tert-Bu contacts in BNPP, NO‚‚‚HC(aryl)
contacts in MNPP. The antiferromagnetic behaviors of both
systems are not consistent with a simple spin-polarization
approach that emphasizes these closest contacts. Instead,
exchange appears to be dominated by through space interactions
between the nitroxide groups, quite weakly in BNPP at an inter-
nitroxide distance of about 5.7 Å, more strongly in MNPP at
about 4.0 Å. The variable temperature magnetism for BNPP
fits a 1-D Heisenberg antiferromagnetic chain behavior with
weak exchange. The magnetism and crystallographic packing
of MNPP strongly suggests that its AFM exchange behavior is
due to quasi 2-D square planar antiferromagnetic interactions.
The unusual packing and magnetism of MNPP shows that an
organic molecule that lacks strongly directional crystal assembly
interactions (such as hydrogen bonds) can still form magnetically
interesting crystal lattices.
a Distances from Supporting Information.
MNPP makes sense if nitroxide-nitroxide contacts are the main
mechanisms of exchange interaction.
Although the 1-D chain fit to the MNPP magnetic behavior
is mathematically justifiable, crystallography suggests that the
exchange interactions could propagate in a more complex
manner. As shown in Figure 6, the MNPP NO‚‚‚NO contacts
form sheet-like planar meshes, with the planes strongly spin-
isolated at distances of 15 Å. Scheme 4 shows one plane from
Figure 6, with a top view of the unusual mesh of nearly equal
N-to-O and O-to-N inter-nitroxide distances within the plane.
The 2-π-center/3-electron electronic structure of the nitroxide
units is not spatially isotropic in terms of spin density distribu-
tion, complicating the application of simplified magnetic lattice
models. But, the crystallographic symmetry of the MNPP lattice
places each nitroxide unit close to four other nitroxides in the
planar mesh, with all interactions essentially occurring within
the plane. This strongly suggests quasi 2-D square planar
exchange between the spins. Effectively square planar magnetic
behavior has been observed in other organic molecules having
complex crystallographic close contact interactions. For ex-
ample, 2-(N-tert-butyl-N-aminoxyl)benzimidazole (BABI) ex-
Experimental
4-Nitrosoanisole. This compound was synthesized by a literature
procedure to yield a green liquid at room temperature (lit. mp 22
°C12). H NMR (CDCl3): δ 3.95 (s, 3 H), 7.05 (d, 2 H, J ) 9.1
1
hibits21 quasi 2-D square planar type AFM magnetic exchange
behavior, and thermodynamic measurements show it to behave21c
as a 2-D Heisenberg square planar bilayer antiferromagnet with
an ordering temperature of 1.7 K.
Hz), 7.92 (broad d, 2 H).
1-(para-Bromophenyl)-1H-pyrrole (BrPP). This compound was
made by a literature procedure. Mp: 86-90 °C, lit.24 mp 86-90
1
°C. H NMR (CDCl3) δ 6.35 (t, 2 H, J ) 2.1 Hz), 7.04 (t, 2 H, J
) 2.1 Hz), 7.28 (para AA′BB′, 2 H, J ) 9.7 Hz), 7.52 (para
AA′BB′, 2 H, J ) 9.7 Hz).
The magnetic data for MNPP were fit to a literature
expression22 for a Heisenberg 2-D AFM square planar system
1-(4-(N-tert-butyl-N-hydroxylamino)phenyl)pyrrole (BNPPH).
A solution of n-BuLi in hexanes (9.6 mL, 16.9 mmol) was slowly
added to a solution of BrPP in 80 mL of anhydrous ether at
-78 °C under nitrogen. The reaction was stirred for 1 h, during
which it was allowed to warm to about -10 °C. A solution of
2-methyl-2-nitrosopropane11 (1.5 g, 16.9 mmol) in 20 mL of
anhydrous ether was then added dropwise to the reaction mixture.
The mixture was allowed to stir under nitrogen overnight while
1
of S ) /2 spin units with a mean-field term; using the same
variables as were described above, g ) 2.027 ( 0.004, J/k )
(-)0.78 ( 0.04 K, θ ) (-)0.77 ( 0.12 K. The fit to the data
is noticeably improved over the spin-pairing and 1-D chain fits
(21) (a) Ferrer, J. R.; Lahti, P. M.; George, C.; Antorrena, G.; Palacio,
F. Chem. Mater. 1999, 11, 2205. (b) Lahti, P. M.; Ferrer, J. R.; George, C.;
Oliete, P.; Julier, M.; Palacio, F. Polyhedron 2001, 20, 1465. (c) Miyazaki,
Y.; Sakakibara, T.; Ferrer, J. R.; Lahti, P. M.; Antorrena, G.; Palacio, F.;
Sorai, M. J. Phys. Chem. B 2002, 106, 8615.
(23) Duling, D. R. J. Magn. Res. 1994, B104, 105.
(22) Baker, G., Jr.; Gilbert, H. E.; Eve, J.; Rushbrooke, G. S. Phys. Lett.,
1967, 25A(3), 207.
(24) Nakazaki, J.; Chung, I.; Matsushita, M. M.; Sugawara, T.; Watanabe,
R.; Izuoka, A.; Kawada, Y. J. Mat. Chem. 2003, 13, 1011.
9346 J. Org. Chem., Vol. 71, No. 25, 2006

Crystallography and Magnetism of BNPP and MNPP
warming to room temperature, then quenched by careful, dropwise
addition of excess saturated aqueous ammonium chloride The layers
were separated, and the organic layer was washed with water. The
combined aqueous layers were extracted with ether, and the
combined organic layers were dried over anhydrous magnesium
sulfate. The organic layers were evaporated, and the residue was
subjected to column chromatography (silica, dichloromethane) to
give the product as off yellow solid (0.59 g, 18%). This material
was used as soon as possible in the next synthetic step. Mp: 145-
147 °C. 1H NMR (CDCl3): δ 1.17 (s, 9 H), 5.11 (s, 1 H), 6.34 (t,
2 H, J ) 2.6 Hz), 7.06 (t, 2 H, J ) 2.6 Hz), 7.29 (s, 4 H). FT-IR
(KBr, cm-1): 3337-3016 (broad, O-H str), 2982-2900 (str, C-H
str).
1-(4-(N-tert-butyl-N-aminoxyl)phenyl)pyrrole (BNPP). PbO2
(1.13 g, 47 mmol) was added to a solution of BNPPH (0.27 g, 12
mmol) in dichloromethane. The reaction was monitored by TLC
and found to be complete after half an hour. The reaction was
filtered and evaporated, and the residue purified by column
chromatography (silica, dichloromethane) to yield the radical as a
red solid (0.20 g, 73%). Mp: 94-96 °C. Anal. calcd for
C14H17N2O: C, 73.33; H, 7.47; N, 12.22. Found: C, 73.07; H, 7.32;
of anhydrous ether at -78 °C under nitrogen. The reaction was
stirred for 1 h, during which it was allowed to warm to about
-10 °C. A solution of 4-nitrosoanisole12 (0.8 g, 5.63 mmol) in 7
mL of anhydrous ether was then added to the reaction. The mixture
was allowed to stir under nitrogen overnight while warming to room
temperature, then quenched by careful, dropwise addition of excess
saturated aqueous ammonium chloride. The layers were separated
and the organic layer was washed with water, then the combined
aqueous layers were extracted with ether. The combined organic
layers were dried over anhydrous magnesium sulfate. Column
chromatography (silica, dichloromethane) yielded the product as a
red solid (0.26 g, 21% yield), which was recrystallized from
dichloromethane. Mp: 184-185 °C Anal. calcd for C17H15N2O2:
C, 73.10; H, 5.41; N, 10.03. Found: C, 72.40; H, 5.41; N, 9.94.
ESR (9.6335 GHz, toluene): g ) 2.00666, aN ) 9.73 G, aH1
1.98 G (2 H), aH2 ) 1.72 G (2 H), aH3 ) 0.89 G (2 H).
)
Crystallography: red needle from dichloromethane, using λ(Mo
KR) ) 0.7107 Å, µ ) 0.090 mm-1, formula ) C17H15N2O2, MW
) 279.3, monoclinic, space group Cc, T ) 293 K, Z ) 4, a )
30.002(2) Å, b ) 7.3588(4) Å, c ) 6.2049(4) Å, â ) 90.467(2)°,
V ) 1369.88(14) Å,3 Dcalc ) 1.354 g/cm3, F(000) ) 558. A total
of 1314 reflections were recorded at a threshold intensity of 2σ(I).
A total of 1217 independent reflections were analyzed with 192
parameters using SHELXL-97.25 The structure was checked for
missing symmetry using the ADDSYM function of PLATON.26
For 1016 reflections with I > 2σ(I), R1(I > 2σ) ) 0.0608, wR2(I >
2σ) ) 0.1029; R1(all) ) 0.0766, wR2(all) ) 0.1089, goodness of
fit on F2 ) 1.279. CCDC deposition no. 606968.
N, 12.10. ESR (9.592 GHz, toluene, 230 K): g ) 2.00678, aN
11.74 G, aH1 ) 2.22 G (2 H), aH2 ) 0.90 G (2 H).
)
Crystallography: red needle from dichloromethane, using λ(Mo
KR) ) 0.7107 Å, µ ) 0.077 mm-1, formula ) C14H17N2O, MW
) 229.3, orthorhombic, space group Pna21, T ) 298 K, Z ) 4, a
) 27.3232(12), b ) 5.8054(2), c ) 7.9936(4) Å, V ) 1267.96(8)
Å,3 Dcalc ) 1.201 g/cm3, F(000) ) 492. A total of 2213 reflections
were recorded at a threshold intensity of 2σ(I). A total of 2060
independent reflections were analyzed with 156 parameters using
SHELXL-97.25 The structure was checked for missing symmetry
using the ADDSYM function of PLATON.26 For 1642 reflections
with I > 2σ(I), R1(I > 2σ) ) 0.0422, wR2(I > 2σ) ) 0.1022; R1-
(all) ) 0.0613, wR2(all) ) 0.1140, goodness of fit on F2 ) 1.063.
An absolute structure parameter of 0(2) was used. CCDC deposition
no. 606967.
Acknowledgment. This material is based upon work sup-
ported by the National Science Foundation under Grant CHE-
0415716. We thank Dr. Peter Khalifah and Dr. Gregory
Dabkowski for assistance with X-ray crystallography and
microanalysis, respectively.
Supporting Information Available: Experimental procedures
not included in the main text; HPLC traces, ESR spectra, and
computational spin density summaries for BNPP and MNPP;
magnetization versus field and Curie-Weiss data for BNPP and
MNPP; magnetic models and hamiltonians used to fit variable
temperature magnetic susceptibility data; a figure of nitroxide-
nitroxide contacts in MNPP crystals; CIF crystallographic summary
files for BNPP and MNPP. This material is available free of charge
via the Internet at http://pubs.acs.org.
1-(4-(N-[para-Methoxyphenyl]-N-hydroxylaminoxyl)phenyl)-
pyrrole (MNPP). A solution of n-BuLi in hexanes (2.6 mL, 5.6
mmol) was added to a solution of BrPP (1.0 g, 4.5mmol) in 10 mL
(25) Sheldrick, G. M. SHELXTL97 Program for the Refinement of
Crystal Structures, University of Go¨ttingen, Germany, 1997.
(26) (a) Le Page, Y. J. Appl. Crystallogr. 1987, 20, 264. (b) Le Page,
Y. J. J. Appl. Crystallogr. 1988, 21, 983. (c) Spek, A. L. J. Appl. Crystallogr.
1988, 21, 578. (d) Spek, A. L. Acta Crystallogr. 1990, A46, C34.
JO0616299
J. Org. Chem, Vol. 71, No. 25, 2006 9347