Structure-Property Relationships in Trimethylenemethane-Type Biradicals. 2. Synthesis and EPR Spectral Characterization of Dinitroxide Biradicals

Article abstract of DOI:10.1021/jo990061j

Several trimethylenemethane-type (TMM-type) dinitroxide biradicals have been prepared that differ in tert-butylaminoxylphenyl ring torsion angles by virtue of different steric demands of their "spin-protecting groups". In addition, the synthesis of a TMM-type dinitroxide having a planar π-system is reported. The synthetic route employed for the planar TMM-type biradical is general and should therefore expand the utility of TMM-type biradicals that are susceptible to bond torsions that attenuate exchange coupling. The compounds presented are designed to be effective to map out J-coupling/conformation space in TMM-type biradicals. EPR spectroscopic parameters (hfcc and D-values) were compared to determine if their values reflected differences in conformation in the series of biradicals. Interestingly, neither N- nor H-hfcc varied within our series of biradicals/ monoradicals in a regular fashion, indicating that there is no apparent relationship between N- or H-hfcc and conformation (as judged by molecular mechanics calculations). However, D-values, estimated from relative intensities of Δm_s = 1 and Δm_s = 2 transitions, are consistent with a varied degree of delocalization in the dinitroxides: smaller D-values for biradicals having smaller aryl torsions (and greater delocalization), and larger D-values for dinitroxides having greater aryl torsions (and less delocalization). All the biradicals studied, except 5, exhibited linear Curie plots. The linear Curie plots are consistent with both triplet ground-states and singlet-triplet degeneracies. Interestingly, dinitroxide 6 exhibited a linear Curie plot, despite the lack of a π-coupling fragment. Biradical 5, however, is a ground-state singlet species with the triplet lying about 140 cal/mol above the singlet.

Full text of DOI:10.1021/jo990061j

4386
J . Org. Chem. 1999, 64, 4386-4396
Str u ctu r e-P r op er ty Rela tion sh ip s in Tr im eth ylen em eth a n e-Typ e
Bir a d ica ls. 2. Syn th esis a n d EP R Sp ectr a l Ch a r a cter iza tion of
Din itr oxid e Bir a d ica ls†
David A. Shultz,* Andrew K. Boal, Hyoyoung Lee, and Gary T. Farmer
Department of Chemistry, North Carolina State University, Raleigh, North Carolina 27695-8204
Received J anuary 12, 1999
Several trimethylenemethane-type (TMM-type) dinitroxide biradicals have been prepared that differ
in tert-butylaminoxylphenyl ring torsion angles by virtue of different steric demands of their “spin-
protecting groups”. In addition, the synthesis of a TMM-type dinitroxide having a planar π-system
is reported. The synthetic route employed for the planar TMM-type biradical is general and should
therefore expand the utility of TMM-type biradicals that are susceptible to bond torsions that
attenuate exchange coupling. The compounds presented are designed to be effective to map out
J -coupling/conformation space in TMM-type biradicals. EPR spectroscopic parameters (hfcc and
D-values) were compared to determine if their values reflected differences in conformation in the
series of biradicals. Interestingly, neither N- nor H-hfcc varied within our series of biradicals/
monoradicals in a regular fashion, indicating that there is no apparent relationship between N- or
H-hfcc and conformation (as judged by molecular mechanics calculations). However, D-values,
estimated from relative intensities of ∆ms ) 1 and ∆ms ) 2 transitions, are consistent with a varied
degree of delocalization in the dinitroxides: smaller D-values for biradicals having smaller aryl
torsions (and greater delocalization), and larger D-values for dinitroxides having greater aryl torsions
(and less delocalization). All the biradicals studied, except 5••, exhibited linear Curie plots. The
linear Curie plots are consistent with both triplet ground-states and singlet-triplet degeneracies.
Interestingly, dinitroxide 6•• exhibited a linear Curie plot, despite the lack of a π-coupling fragment.
Biradical 5••, however, is a ground-state singlet species with the triplet lying about 140 cal/mol
above the singlet.
In tr od u ction
Large amplitude bond torsions can dramatically affect
exchange couplings in organic biradicals when the bonds
connect spin-containing groups to exchange-coupling
fragments. Several examples of this structure-property
relationship have been noted for m-phenylene-type bi-
radicals.1 Trimethylenemethane-type (TMM-type) biradi-
cals having the 1,1-diarylethene moiety are an interesting
class of molecules with which to study this phenomenon
because diarylethene conformation and dynamics have
been elucidated.2
* To whom correspondence should be addressed. voice: (919) 515-
6972, e-mail: shultz@chemdept.chem.ncsu.edu, fax: (919) 515-8920,
web: http://www2.ncsu.edu/ncsu/chemistry/das.html.
1,1-Diphenylethene adopts a nonplanar geometry due
to steric interactions between the phenyl rings. The
minimum energy geometry is computed to have phenyl
torsion angles, φ, of ca. 30°.2 Increased phenyl torsion
exists in 2,2-disubstituted structures because of ad-
ditional steric interactions. The chiral, C2-symmetric
conformer of 1,1-diphenylethene rapidly equilibrates with
its enantiomer by the zero- and one-ring flip mecha-
nisms.2,3
† Part 1: Shultz, D. A.; Boal, A. K.; Farmer, G. T. J . Am. Chem.
Soc. 1997, 119, 3846; Part 3: Shultz, D. A.; Boal, A. K.; Bourbeau, M.
P.; Farmer, G. T. J . Am. Chem. Soc., to be submitted.
(1) For papers discussing the effects of torsion on exchange coupling
in m-phenylene-type biradicals, see: (a) Dvolaitzky, M.; Chiarelli, R.;
Rassat, A. Angew. Chem., Int. Ed. Engl. 1992, 31, 180. (b) Kanno, F.;
Inoue, K.; Koga, N.; Iwamura, H. J . Am. Chem. Soc. 1993, 115, 847.
(c) Silverman, S. K.; Dougherty, D. A. J . Phys. Chem. 1993, 97, 13273.
(d) Okada, K.; Matsumoto, K.; Oda, M.; Murai, H.; Akiyama, K.;
Ikegami, Y. Tetrahedron Lett. 1995, 36, 6693. (e) Adam, W.; van
Barneveld, C.; Bottle, S. E.; Engert, H.; Hanson, G. R.; Harrer, H. M.;
Heim, C.; Nau, W. M.; Wang, D. J . Am. Chem. Soc. 1996, 118, 3974.
(f) Fujita, J .; Tanaka, M.; Suemune, H.; Koga, N.; Matsuda, K.;
Iwamura, H. J . Am. Chem. Soc. 1996, 118, 9347. (g) Okada, K.;
Imakura, T.; Oda, M.; Murai, H.; Baumgarten, M. J . Am. Chem. Soc.
1996, 118, 3047. (h) See also ref 11. (i) Ab initio calculations explain
antiferromagnetic coupling in certain m-phenylene-linked structures,
see: Fang, S.; Lee, M.-S.; Hrovat, D. A.; Borden, W. T. J . Am. Chem
Soc. 1995, 117, 6727.
If the phenyl rings of 1,1-diphenylethene are substi-
1
tuted in the para-positions with S ) /2 (paramagnetic)
functional groups, as illustrated above, the biradical
produced is expected to be
a ground-state triplet
(3) Phenyl torsions in substituted 1,1-diphenylethenes are fast on
the NMR time scale even at -90 °C, see: Rabinovitz, M.; Agranat, I.;
Bergmann, E. D. Isr. J . Chem. 1969, 7, 795.
(2) Kaftory, M.; Nugiel, D. A.; Biali, S. E.; Rappoport, Z. J . Am.
Chem. Soc. 1989, 111, 8181.
10.1021/jo990061j CCC: $18.00 © 1999 American Chemical Society
Published on Web 05/25/1999

Trimethylenemethane-Type Biradicals
J . Org. Chem., Vol. 64, No. 12, 1999 4387
based on π topology,4-8 and the 1,1-diphenylethene
moiety is a TMM-type exchange coupling fragment. The
effectiveness of this exchange coupler is strongly cor-
related to the magnitude of phenyl torsion since phenyl
torsion modulates spin density in the ethene unit. That
is, the greater the delocalization of spin into the ethene
fragment, the greater the stabilization of the triplet
state.
We studied a few 1,1-diarylethene-TMM-type bis-
(phenoxy) radicals and found fundamental structure-
property relationships for spin-spin coupling in this
class of biradicals.9-11 As surmised, steric interac-
tion between structural elements of the bis(phenoxy)
biradicals has the effect of increasing the torsion
angles, φ, of the aryl groups and modulating exchange
couplings.
Since nitroxide radicals are useful for the construc-
tion of coordination polymers that order magnetically,12
we wish to study the relationship between conforma-
tion and exchange coupling in a series of isolated dini-
troxide molecules, and correlate our findings with struc-
ture-property relationships in coordination polymers
using the same dinitroxides. Ultimately, we endeavor
to determine whether crystal packing forces or coordi-
nation polymer geometric requirements can alter bi-
radical conformation, and therefore biradical exchange
coupling. If so, molecular building blocks having “unfa-
vorable” conformations (vide infra) might be more
useful for the construction of solid-state coordination
polymers with higher ordering temperatures than might
otherwise be expected. In addition to this interest,
we are excited about the possibility of generating a
series of biradicals having a variety of torsion angles
and correlating spin-spin coupling with conforma-
tion.9-11
Resu lts
Syn th esis. Dinitroxide 1•• was prepared as described
previously,13 and 3•• and 4•• were prepared in a similar
fashion as shown below. Methyl norbornane-7-carboxy-
late (8)14 and methyl adamantane-2-carboxylate (9)9 were
reacted with 4-lithiobromobenzene, and the resulting
carbinols were subsequently dehydrated to yield dibro-
mides 3-Br 2 and 4-Br 2, respectively. Fluorenyl-dibromide
6-Br 2 was prepared in a similar fashion, except that the
intermediate alcohol was reacted with poly[phosphoric
acid] to generate the fluorene moiety. All dibromides were
converted to the corresponding dinitroxides using stan-
dard protocols.13
Along these lines, we report here the synthesis and
solution EPR spectral characterization of a series of
TMM-type dinitroxides, 1••-5•• and 7••, designed to have
a range of aryl torsion angles, as well as a model
biradical, 6••.
(4) Borden, W. T.; Davidson, E. R. J . Am. Chem. Soc. 1977, 99, 4587.
(5) Berson, J . A. Diradicals; Wiley: New York, 1982.
(6) Dougherty, D. A. Acc. Chem. Res. 1991, 24, 88.
(7) Iwamura, H. Pure Appl. Chem. 1987, 59, 1595.
(8) Rajca, A. Chem. Rev. 1994, 94, 871.
(9) Shultz, D. A.; Boal, A. K.; Farmer, G. T. J . Am. Chem. Soc. 1997,
119, 3846.
(10) Shultz, D. A.; Boal, A. K.; Bourbeau, M. P.; Farmer, G. T. J .
Am. Chem. Soc., to be submitted.
(11) Shultz, D. A. In Magnetic Properties of Organic Materials; P.
Lahti, Ed.; Marcel Dekker: New York, in press.
(12) For papers on nitroxide complexes, see the following: (a)
Caneschi, A.; Gatteschi, D.; Sessoli, R.; Rey, P. Acc. Chem. Res. 1989,
22, 392. (b) Caneschi, A.; Gatteschi, D.; Rey, P.; Sessoli, R. Chem.
Mater. 1992, 4, 204. (c) Gatteschi, D.; Sessoli, R. J . Magn. Magn. Mater.
1992, 104, 2092. (d) Stumpf, H. O.; Ouahab, L.; Pei, Y.; Grandjean,
D.; Kahn, O. Science 1993, 261, 447. (e) Gatteschi, D. Adv. Mater. 1994,
6, 635. (f) Luneau, D.; Laugier, J .; Rey, P.; Ulrich, G.; Ziessel, R.; Legoll,
P.; Drillon, M. J . Chem. Soc., Chem. Commun. 1994, 741. (g) Kitano,
M.; Koga, N.; Iwamura, H. J . Chem. Soc., Chem. Commun. 1994, 447.
(h) Kitano, M.; Ishimaru, Y.; Inoue, K.; Koga, N.; Iwamura, H. Inorg.
Chem. 1994, 33, 6012. (i) de Panthou, F. L.; Belorizky, E.; Calemczuk,
R.; Luneau, D.; Marcenat, C.; Ressouche, E.; Turek, P.; Rey, P. J . Am.
Chem. Soc. 1995, 117, 11247. (j) Inoue, K.; Hayamizu, T.; Iwamura,
H.; Hashizume, D.; Ohashi, Y. J . Am. Chem. Soc. 1996, 118, 1803. (k)
Pei, Y.; Kahn, O.; Ouahab, L. Inorg. Chem. 1996, 35, 193. (l) Iwamura,
H.; Inoue, K.; Hayamizu, T. Pure Appl. Chem. 1996, 68, 243. (m)
Ottaviani, M. F.; Turro, C.; Turro, N. J .; Bossman, S. J .; Tomalia, D.
J . Phys. Chem. 1996, 100, 13667. (n) Sokolowsky, A.; Bothe, E.; Bill,
E.; Weyhermueller, T.; Wieghardt, K. Chem. Commun. 1996, 6, 1671.
(o) Nakatsuji, S.; Anzai, H. J . Mater. Chem. 1997, 7, 2161. (p) Fegy,
K.; Luneau, D.; Ohm, T.; Paulsen, C.; Rey, P. Angew. Chem., Int. Ed.
Engl. 1998, 37, 1270.
Biradicals 2•• and 5•• required a different approach due
to the greater steric demands of the di-tert-butylmeth-
ylene and bicyclo[4.4.1]undecyl groups, and we chose the
(13) Matsumoto, T.; Ishida, T.; Koga, N.; Iwamura, H. J . Am. Chem.
Soc. 1992, 114, 9952.
(14) Kwart, H.; Kaplan, L. J . Am. Chem. Soc. 1954, 76, 4072.

4388 J . Org. Chem., Vol. 64, No. 12, 1999
Shultz et al.
Barton-Kellogg reaction15,16 for synthesis of dibromides
2-Br 2 and 5-Br 2. This route required the preparation of
diazo compounds 14 and di-tert-butyldiazomethane,17 as
well as 4,4′-dibromothiobenzophenone,18 16, as outlined
below. Diazoalkane 14 was prepared from bicyclic ketone
13, which in turn was prepared by intramolecular alky-
lation of the enolate formed from 2-(4′-iodobutyl)cyclo-
heptanone. This iodide was prepared from the corre-
sponding chloride, 12, which was the alkylation product
of cycloheptane enolate reacted with 4-chloro-1-iodobu-
tane.
(1•-7•) in all cases. Mass spectral analysis indicated that
these molecules lacked bromine atoms.
Cross conjugated dinitroxides such as 1••-5•• are red
solids, giving red-colored solutions, and the electronic
absorption spectrum of 2•• in Figure 1 is typical. Interest-
ingly, 7•• is a beautiful purple color and gives purple
solutions. Consistent with the electronic and conforma-
tional changes resulting from the planarized chro-
mophore, the longest wavelength visible absorption band
of 7•• is red-shifted ca. 30 nm (ca. 1825 cm-1) and
hyperchromically shifted compared to that of 2•• as shown
in Figure 1.
Molecu lar Mech an ics Calcu lation s.11,22 MM223 mini-
mized structures for i-propyldiphenylmethane, 1,1-
diphenyl-2,2-di-tert-butylethene, diphenylmethylenenor-
bornane, diphenylmethyleneadamantane, diphenyl-
methylenebicyclo[4.4.1]undecane, and 9,9-diphenylfluo-
rene as well as pertinent bond angles and torsion angle
values are shown in Figure 2. For the diphenylethylenes
listed, phenyl torsion increased in the series: diphenyl-
methylenenorbornane < i-propyldiphenylmethane <
diphenylmethyleneadamantane < diphenylmethylene-
bicyclo[4.4.1]undecane < 1,1-diphenyl-2,2-di-tert-butyl-
ethene. The steric demand of the bicyclic or tricyclic
“protecting group” increases as the valence angle θ
increases causing φ to increase in the same order. Of
course these arguments and results do not apply to
3,12-dihydro-3,3,12,12-tetramethylindeno[1,2-a]indene,
a polycyclic molecule in which phenyl torsion is pre-
cluded by fusion of the gem-alkyl groups to the phenyl
rings.
F lu id -Solu tion EP R Sp ectr oscop y. Figure 3 shows
the EPR spectra of 1••-7•• as solutions in toluene recorded
at 298 K along with simulations.24 The experimental
spectra are best reproduced when a small amount of
monoradical is included in the simulation. We presume
that the monoradical impurity appears during sample
preparation and storage since satisfactory magnetic
moments were found for the biradicals. The EPR spectra
were also concentration independent, ruling out the
formation of dimers as the impurities. Nominal hyperfine
coupling constants (hfcc) and biradical purities, estimated
by simulation, are given in Table 1. Temperature-
dependent spectral changes due to conformational modu-
The planar TMM-type biradical, 7••, was prepared as
shown below. Bis(4-bromophenyl)methanol, 17, was re-
acted with acetylacetone in the presence of anhydrous
cobaltous chloride,19 followed by bromination with NBS,
and workup-induced elimination to give conjugated ene-
dione 18. Dione 18 was converted into an intermediate
diol by reaction with 2 equiv of tetramethylzirconium.20,21
Ring-closure to indeno[1,2-a]indene derivative 7-Br 2 was
affected by stirring the diol with poly[phosphoric acid]
at 147 °C. The biradical was prepared using conditions
described for the other dinitroxides.
The procedures used to generate the hydroxylamine/
nitroxide functionality also gave mononitroxide radicals
(15) Buter, J .; Wassenaar, S.; Kellog, R. M. J . Org. Chem. 1972, 37,
4045.
(16) Barton, D. H. R.; Guziec, F. S.; Shahak, I. J . J . Chem. Soc.
Perkin Trans. 1 1974, 1794.
(17) Hartzler, H. D. J . Am. Chem. Soc. 1971, 93, 4527.
(18) Scheeren, J . W.; Ooms, P. H. J .; Nivard, R. J . F. Synthesis 1973,
149.
(19) Marquez, J .; Moreno-Man˜as, M. Chem. Lett. 1981, 173. Direct
formation of ene-dione 18 by condensation of ketone 15 and acetylac-
etone according to the method of Lehnert failed (Lehnert, W. Tetra-
hedron 1973, 29, 635). However, this method was for Knovenagel
reaction of diethyl malonate and benzophenone derivatives.
(20) Majetich, G.; Lowery, D.; Khetani, V.; Song, J .-S.; Hull, K.;
Ringold, C. J . Org. Chem. 1991, 56, 3988. Reaction of ene-dione 18
with the methylcerium reagent formed from methyllithium and ceric
chloride (Imamoto, T.; Takiyama, N.; Nakamura, K. Tetrahedron Lett.
1985, 26, 4763) gave only very small amounts of the desired diol.
(21) J ung, A.; Reetz, M. T. J . Am. Chem. Soc. 1983, 105, 4833.
(22) Crystal structures of 1-Br 2-6-Br 2 have been determined and
will be reported elsewhere. The crystal structures of the dibromides
corroborate the MM2 results presented here. For
account, see ref 11.
a preliminary
(23) Allinger, N. L.; Yuh, Y. H. QCPE 1981, 13, 395.
(24) Fluid solution EPR spectra were simulated using the EPR
Calculations for MS-Windows NT/95, Version 0.96, Public EPR Soft-
ware Tools, National Institute of Environmental Health Sciences,
National Institutes of Health, Research Triangle Park, NC.; Dave
Duling, 1996. The correlation coefficient for the simulations are R g
0.995.

Trimethylenemethane-Type Biradicals
J . Org. Chem., Vol. 64, No. 12, 1999 4389
F igu r e 1. Electronic absorption spectra of biradicals 2••
(- - -)and 7•• (__) as solutions in CH2Cl2.
lation of exchange25 could account for the spectral ap-
pearance, but were not observed.26 All biradical spectra
are typical for S ) 1 species in which the exchange
interactions are much greater than hyperfine interactions
(|J | . |a|).27,28 Hyperfine coupling constants for the
monoradicals, 1•-7•, are given in Table 2.
Fr ozen -Solu tion EP R Spectr oscopy. Figure 4 shows
the ∆ms ) 1 EPR spectra of 1••-7•• recorded at 77 K in a
toluene matrix. Spectra are consistent with randomly
oriented triplet species,29 and are typical for dinitroxides
having small D-values and hyperfine structure in the ∆ms
) 1 region.30,31 These spectral features, unfortunately,
preclude our simulating the spectra. As shown in Figure
5, all biradicals exhibited ∆ms ) 2 transitions, signatures
of high-spin states, in which the parallel component of
the nominal N-hfcc was approximately one-half that of
a nitroxide monoradical (monoradical aN,| ≈ 56 G; 7•• does
not exhibit hfcc in the ∆ms ) 2 transition), again
consistent with |J | . |a|. At this time, we cannot explain
the lack of hyperfine structure in the ∆ms ) 2 transition
of 7••. The zero-field-splitting (ZFS) parameters, |D/hc|,
calculated from the relative intensities of ∆ms ) 1 and
∆ms ) 2 transitions32 are given in Table 1.
Curie plots for the doubly integrated ∆ms ) 2 signals
of 1••-4••, 6••, and 7•• in toluene are shown in Figure 6.
The signal intensities increase linearly with increasing
inverse absolute temperature, consistent with both high-
spin coupling of the nitroxide units, and singlet-triplet
degeneracies, i.e., J g 0.33
F igu r e 2. MM2-minimized structures for i-propyldiphenyl-
methane, 1,1-diphenyl-2,2-di-tert-butylethene, diphenylmeth-
ylenenorbornane, diphenylmethyleneadamantane, diphenyl-
methylenebicyclo[4.4.1]undecane, 9,9-diphenylfluorene, and
3,3,12,12-tetramethyl-3,12-dihydroindeno[1,2-a]indene. θ de-
3
2
3
The Curie plot for biradical 5•• was quite different than
those of the other biradicals studied, and is shown in
scribes the Csp -Csp -Csp “protecting group” bond angle, and
φ is the phenyl torsion angle.
Figure 7. The data were fit according to the equation:33
(25) J -modulation can cause the “alternating linewidth effect,” see
refs 46-48.
(26) We thank Professor Malcolm Forbes for a copy of his program
for simulating J -modulated EPR spectra.
-2J
RT
3 exp
(
)
(27) Atherton, N. M. Principles of Electron Spin Resonance; Ellis
Horwood PTR Prentice Hall: New York, 1993.
(28) Wertz, J . E.; Bolton, J . R. Electron Spin Resonance; Chapman
and Hall: New York, 1986.
(29) Wasserman, E.; Snyder, L. C.; Yager, W. A. J . Chem. Phys.
1964, 41, 1763.
(30) Mitsumori, T.; Inoue, K.; Koga, N.; Iwamura, H. J . Am. Chem.
Soc. 1995, 117, 2467.
C
T
IEPR
)
-2J
RT
[
]
1 + 3 exp
(
)
where C is a constant and J is the exchange parameter.
Best fit results give J ) -68 ( 6 cal/mol.
(31) Akita, T.; Mazaki, Y.; Kobayashi, K. Tetrahedron Lett. 1995,
36, 5543.
(32) Weissman, S. I.; Kothe, G. J . Am. Chem. Soc. 1975, 97, 2537.
(Ims)1)/(Ims)2) ) [(Hms)2)/D]2, where Ims)1 (Ims)2) is the intensity of the
∆ms ) 1 (∆ms ) 2) transition, Hms)2 is the field at which the ∆ms ) 2
transition occurs, and D is the zero-field-splitting parameter.
(33) A linear Curie plot is also consistent with degenerate singlet
and triplet states. For an excellent discussion of the salient features
of the Curie plot, see: The Chemistry of Quinonoid Compounds; Berson,
J . A., Ed.; J ohn Wiley & Sons: New York, 1988; Vol. II, p 482. In
addition, linear Curie plots can also result for small negative J -values.
Discu ssion
Molecu la r Design Elem en ts. TMM-type spin-con-
taining units have been used to explore exchange cou-
pling through sp3-hybridized coupling fragments and
aromatic rings;6,34-37 however, high-spin molecules hav-
ing a TMM-type exchange coupling unit are rare com-

4390 J . Org. Chem., Vol. 64, No. 12, 1999
Shultz et al.
F igu r e 4. X-band EPR spectra (∆ms ) 1 transitions) of
biradicals 1••-7•• as frozen solutions in toluene at 77 K.
F igu r e 3. X-band EPR spectra (thick trace) and simulations
(thin trace) of biradicals 1••-7•• as solutions in toluene at 298
K.
Ta ble 1. EP R Sp ectr a l P a r a m eter s for Bir a d ica ls 1••-7••
e
biradical %biradicala aN/Gb aH(ortho)/Gb,c aH(meta)/Gb,d |D/hc|/cm-1
1••f
2••
3••
4••
5••
6••
7••
100
96.35
90.65
92.01
95.73
89.10
5.95
5.82
5.95
5.99
5.98
5.88
6.02
-
-
0.0042
0.0064
0.0035
0.0041
0.0062
0.0046
0.0035
1.37
0.63
1.01
1.10
1.13
1.12
0.40
0.44
0.90
0.58
0.50
100
1.16, 0.97
a
Simulation includes corresponding monoradical if % biradical
b
* 100. Hyperfine values are actually a/2; estimated by spectral
simulation.24 c Protons ortho to nitroxide group. Protons meta
d
to nitroxide group. e D-value parameters estimated from relative
intensities of ∆ms ) 1 and ∆ms ) 2 transitions.32 f H-hyperfine
not observed in biradical; quality of spectral simulation not
improved by additional hyperfine.
Ta ble 2. EP R Sp ectr a l P a r a m eter s for Mon or a d ica ls
1•-7•a
F igu r e 5. X-band EPR spectra (∆ms ) 2 transitions) of
biradicals 1••-7•• as frozen solutions in toluene at 77 K.
monoradical
aN/G
aH(ortho)/Gb
aH(meta)/Gc
1•
2•
3•
4•
5•
6•
7•
12.00
11.93
11.98
12.03
12.08
12.00
12.13
2.10
2.09
2.09
2.09
2.08
2.05
2.03, 2.16
0.90
0.92
0.91
0.90
0.88
0.89
0.91
a
b
Hyperfine values determined by spectral simulation.24 Pro-
tons ortho to nitroxide group. c Protons meta to nitroxide group.
pared to those based on m-phenylene linkers.38,39 The
major advantage to using the TMM coupler is that the
parent biradical has a larger singlet-triplet gap (14.5
kcal/mol)40 than m-xylylene (9.2 ( 0.2 kcal/mol)41. Thus,
the TMM coupler is an inherently stronger exchange
coupler.
F igu r e 6. Curie plots for biradicals 1••-7••.The data points
were collected between ca. 8 and ca. 50 K and represent
normalized signal intensities obtained by double integration
of the ∆ms ) 2 signal.
(34) J acobs, S. J .; Dougherty, D. A. Angew. Chem., Int. Ed. Engl.
1994, 33, 1104.
(35) J acobs, S. J .; Shultz, D. A.; J ain, R.; Novak, J .; Dougherty, D.
A. J . Am. Chem. Soc. 1993, 115, 1744.
(36) West, A. P., J r.; Silverman, S. K.; Dougherty, D. A. J . Am.
Chem. Soc. 1996, 118, 1452.
(37) Silverman, S. K.; Dougherty, D. A. J . Phys. Chem. 1993, 97,
13273.
Design elements of a generalized diphenyl-TMM-type
biradical are shown below. First, spin-containing units
are attached geminal to an ethene moiety to complete
the TMM π-topology. Second, if stable molecules are
desired, a “spin-protecting group” that sterically protects
unpaired electrons is necessary for biradicals having
spin-delocalization into the ethene group. However, this
(38) Rajca, A. Chem. Rev. 1994, 94, 871.
(39) Nau, W. M. Angew. Chem., Int. Ed. Engl. 1997, 36, 2445.
(40) Wenthold, P. G.; Squires, R. R.; Lineberger, W. C. J . Am. Chem.
Soc. 1996, 118, 475.
(41) Wenthold, P. G.; Kim, J . B.; Lineberger, W. C. J . Am. Chem.
Soc. 1997, 119, 1354.

Trimethylenemethane-Type Biradicals
J . Org. Chem., Vol. 64, No. 12, 1999 4391
groups might be superior to tertiary alkyl groups for
kinetic protection of the unpaired electrons in TMM-type
biradicals and should be a particularly useful structural
element for studying the effect of bond torsion on
exchange coupling in TMM-type biradicals.
F igu r e 7. Curie plot for biradical 5••. The data points were
collected between 7 and 100 K and represent normalized signal
intensities obtained by double integration of the ∆ms ) 2
signal.
To continue exploring this structure-property rela-
tionship, the TMM-type dinitroxides in this study are
designed to have a range of aryl torsion angles: large
torsions (>60°, 2••,5••), moderate torsions (≈60°, 1••,4••),
small torsions (<60°, 3••), and no aryl torsions (7••). Of
course the tert-butylnitroxide groups in these dinitroxides
are free to rotate and assume any torsion angles, but this
torsional motion should not be coupled to the structure
of the spin-protecting group. Introduction of structural
elements to prohibit rotation of the tert-butylnitroxide
groups, analogous to those that prohibit phenyl torsions
in biradical 7••, is more challenging to implement syn-
thetically. Moreover, unless there are substituents ortho
to the tert-butylnitroxide group, it usually adopts a
conformation having little torsion with respect to the
phenyl group, as shown in crystal structures of several
nitroxides and their complexes.43 That is, tert-butylni-
troxide torsion is typically less than aryl group torsions
in 1,1-diarylethenes, and more importantly there is no
reason to expect that tert-butylnitroxide torsion will vary
in our series of dinitroxides. Nonetheless, tert-butylni-
troxide torsion is as important as aryl torsion, but we
cannot deconvolute the two torsions in the present study.
Studies of exchange coupling in the solid state in con-
junction with molecular structures determined by X-ray
crystallography are needed to deconvolute the effects of
the two torsions. Efforts along these lines will be the
subject of future reports. Dinitroxide 7•• should combine
the best of both worlds: a planar TMM-type coupler (no
phenyl torsions) having steric protection near both the
aryl ring and the ethene coupler fragment.
necessity creates a potential problemssteric interactions
between atoms in the spin-protecting functionality and
those in the spin-containing unit might disrupt conjuga-
tion between the spin-containing units and the ethene
group. As noted in the Introduction, this inhibition of
resonance delocalization attenuates the effectiveness of
the coupling fragment to high-spin couple unpaired
electrons.1,9-11
Structures considered for the spin-protecting group
include tertiary alkyl groups and polycyclic (bicyclic or
tricyclic) ring systems. Simple primary or secondary alkyl
groups are less attractive since they contain allylic
hydrogens that are excellent candidates for abstraction
and subsequent loss of an unpaired electron (below,
left).42 Tertiary alkyl spin-protecting groups introduce
obvious steric congestion problems (below, center). How-
ever, polycyclic ring systems combine the more desirable
sterics of a secondary alkyl group with the low chemical
reactivity provided by a tertiary alkyl group (below,
right). As shown below, the bridgehead hydrogens of a
polycyclic spin-protecting group are orthogonal to the
TMM π-system and are therefore much less susceptible
to abstraction. However, these same hydrogens experi-
ence van der Waals repulsions with aryl rings at small
torsion angles, as shown below. Thus, biradicals having
polycyclic spin-protecting groups are not free from de-
bilitating steric interactions.
Regardless of the structure of the spin-protecting
group, this steric interaction could result in reduced
exchange coupling. However, polycyclic spin-protecting
groups are unique in that the steric interaction can be
regulated and should decrease as the bond angle θ
decreases, as implied above. Moreover, a gradual change
in θ should cause a gradual change in φ thereby providing
a series of biradicals having incrementally different
singlet-triplet gaps. Thus, polycyclic spin-protecting
Biradical 6•• has phenyl-tert-butylnitroxide groups ar-
ranged in a geminal fashion, as in the TMM-type biradi-
cals, but joined to an sp3-hybridized carbon and is
therefore geometrically similar to 1••-5••. This is not to
say that there is no through-bond coupling in 6••, rather
that the through-space interaction of the radical halves
of 6•• are similar to those in 1••-5••, and that the coupling
is not regulated by a π-bond. Thus, 6•• is a model
compound, and comparison of its properties to those of
the TMM-type dinitroxides serves as a check of the utility
of the ethene coupler of the latter. Our studies of TMM-
type bis(semiquinone)s included a model biradical analo-
gous to 6•• (having a fluorenyl coupler) that promoted
(43) Aryl-tert-butylnitroxide torsion angles in the following citations
vary from ca. 2° to ca. 38°: Kitano, M.; Ishimaru, Y.; Inoue, K.; Koga,
N.; Iwamura, H. Inorg. Chem. 1994, 33, 6012; Kanno, F.; Inoue, K.;
Koga, N.; Iwamura, H. J . Phys. Chem. 1993, 97, 13267; Inoue, K.;
Iwamura, H. Angew. Chem., Int. Ed. Engl. 1995, 34, 927; Nakazono,
S.; Karasawa, S.; Iwamura, H. Angew. Chem., Int. Ed. Engl. 1998, 37,
1550.
(42) Anderson, K. K.; Shultz, D. A.; Dougherty, D. A. J . Org. Chem.
1997, 62, 7575.

4392 J . Org. Chem., Vol. 64, No. 12, 1999
Shultz et al.
antiferromagnetic coupling in opposition to the TMM
couplers studied.44
For organic biradicals lacking significant spin-orbit
coupling, |D/hc| is inversely proportional to the cube of
the average interelectronic separation.51,52 As can be seen
in Table 2, for the TMM-type biradicals, D(2••) > D(5••)
> D(4••) ≈ D(1••) > D(3••) ≈ D(7••)sthe same order as
decreasing aryl ring torsion angles. Since smaller torsion
angles correspond to greater spin delocalization, and
therefore greater interelectronic separation, the trend in
Molecu la r Mech a n ics Ca lcu la tion s. The results of
the MM2 calculations on the model compounds in Figure
2 clearly indicate an increase in phenyl torsion as the
valence angle increases from diphenylmethylenenorbor-
nane to diphenylmethyleneadamantane to diphenyl-
methylenebicyclo[4.4.1]undecane, supporting our hypoth-
esis concerning the relationship between θ and φ.11 The
results of the calculations were also useful in comparing
the geometries of bicyclic/tricyclic spin-protecting groups
with the noncyclic spin-protecting groups. Thus, i-pro-
pyldiphenylmethane is sterically similar to diphenyl-
methyleneadamantane, while 1,1-diphenyl-2,2-di-tert-
butylethene is similar to diphenylmethylenebicyclo[4.4.1]-
undecane. This illustrates the impact of the bridgehead
hydrogen on phenyl torsion in the compounds having
bicyclic or tricyclic spin-protecting groupssthat is, the
bicyclic or tricyclic groups can be sterically demanding
substituents due to the orientation of the bridgehead
hydrogens.
F lu id Solu tion EP R Sp ectr oscop y. The N- and
H-hfcc do not vary in a regular fashion in either series
of biradicals or monoradicals and are therefore not an
indicator of any differences in conformation as hfcc of
substituted biphenylnitroxides were shown to be.45
Since the spin-protecting group modulates phenyl
torsion angles, one might expect a dependence of tor-
sional dynamics on the structure of the spin-protecting
group. The torsional dynamics is expected to manifest
itself in EPR spectral shape for the biradicals since
torsion affects spin density in the ethene fragment, and
therefore torsion modulates the exchange parameter, J .
That is, the biradicals might be expected to exhibit
J -modulated EPR spectra.46-48 However, even though |J |
should be rather small due to both (a) bond torsions that
limit spin delocalization, and (b) the fact that the phenyl
rings for tert-butylphenylnitroxides carry spin densities
of only ca. 0.1 s 0.15,45,49 the average J might be much
greater than in biradicals known to exhibit J -modulation.
In addition, the torsional correlation time, τ is expected
to be very small.2,3 Therefore, we expect the spectral
manifestations of J -modulation to be subtle, and indeed
we found no evidence for J -modulation.50
D supports greater delocalization from 2•• f 5•• f 4••
f
1•• f 3••/7•, in good agreement with the MM2-derived
geometries of the corresponding hydrocarbons shown in
Figure 2. Examination of the bond and torsion angles
given in Figure 2 shows that the spatial orientation of
the phenyl rings in 9,9-diphenylfluorene most closely
matches that of diphenylmethylenenorbornane. The cor-
responding biradicals, 6•• and 3••, respectively, have
similar D-values. We believe the difference in D-values
between 6•• and 3•• is consistent with 3•• having additional
delocalization (into the ethene coupler) that is not pos-
sible in 6•• which has an sp3-hybridized coupler. Finally,
the D-values for 3•• and 7•• are nearly equal, but based
solely on differences in delocalization, one would expect
the D-value for 7•• to be less than that of 3••. We feel that
the similarity of the two values could be a result of the
limitation of the method used to estimate D-values.
As seen in Figure 6, the temperature variation of the
∆ms ) 2 transition intensities for 1••-4••, 6••, and 7•• follow
the Curie law from ca. 8 to ca. 50 K, consistent with J g
0, i.e., consistent with both a triplet ground-state and a
singlet-triplet degeneracy.33 Thus, there is no evidence
from the Curie plots of dramatically different exchange
couplings as a function of the spin-protecting groups in
this series of biradicals. Interestingly, biradical 6••, in
which the spin-coupling fragment is an sp3-hybridized
carbon, exhibits a linear Curie plot, unlike its bis-
(semiquinone) analogue.44 The linearity of the Curie plot
is most likely evidence of a very small J -value, although
ferromagnetic coupling for biradicals having sp3-hybrid-
ized couplers is precedented.6 Dougherty’s group pre-
sented computational evidence that, as a result of a
balance of through-bond and through-space effects, cou-
pling in the 1,3-biradical trimethylene (a biradical with
an sp3-hybridized coupler) ranges from ferromagnetic to
antiferromagnetic depending both on the central bond
angle and on terminal methylene torsions.35,53 The J -
values in trimethylene-type biradicals are ca. 1 order of
magnitude smaller than J for biradicals coupled through
π-systems.6,8 This intrinsically weaker coupling and the
fact that the spin density in the phenyl ring of a phenyl-
tert-butylnitroxide is ca. 0.1-0.15.45,49 leads us to suggest
that J for 6•• is weaker than in the TMM-type biradicals.
In addition, the fluid-solution EPR spectral appearance
of 7•• is temperature-independent. Thus, in a molecule
where aryl torsion is precluded, the free to rotate tert-
butylnitroxide groups have no apparent effect on spectral
appearance.
F r ozen -Solu tion EP R Sp ectr oscop y. The calculated
D-values illustrate the geometric differences in the series
of biradicals in a manner consistent with the results of
MM2 calculations on the model compounds in Figure 2.
Also noteworthy is that biradical 2•• exhibits a linear
Curie plot despite the fact that di-tert-butyl might be
expected to exhibit greater steric demand than bicyclo-
[4.4.1]undecane. We point out that the aryl groups of 2••
are capable of “gearing” with the tert-butyl groups,
whereas in 5•• such a conformation is not possible. Geared
conformations in 2•• should give a range of more positive
J -values, hence the linear Curie plot.
(44) Shultz, D. A.; Boal, A. K.; Lee, H.; Farmer, G. T. J . Org. Chem.
1998, 63, 9462.
(45) Shultz, D. A.; Gwaltney, K. P.; Lee, H. J . Org. Chem. 1998, 63,
769.
(46) Luckhurst, G. R. Mol. Phys. 1966, 10, 543.
(47) Hudson, A.; Luckhurst, G. R. Chem. Rev. 1969, 69, 191.
(48) Marshall, J . H. J . Chem. Phys. 1967, 47, 1374-1378.
(49) Aurich, H. G.; Hahn, K.; Stork, K.; Weiss, W. Tetrahedron 1977,
33, 969.
(50) Simulation of the spectra provided no evidence for differential
J -modulation within the series of biradicals. We used a correlation
time, τ, of ca. 5 ps, in accord with the reported torsional barrier of
1,1-diphenylethene.2,3 Also, the planar dinitroxide, 7••, does not behave
differently from the other dinitroxides. Thus, the absence of J -
modulation is most likely a combination of a large average J , and fast
correlation times of the nitroxide groups themselves.
Overall, steric bulkiness does manifest itself in ex-
change coupling of TMM-type dinitroxides. Unlike the
(51) Gleason, W. B.; Barnett, R. E. J . Am. Chem. Soc. 1976, 98, 2701.
(52) Keana, J . F.; Dinerstein, R. J . J . Am. Chem. Soc. 1971, 93, 2808.
(53) Goldberg, A. H.; Dougherty, D. A. J . Am. Chem. Soc. 1983, 105,
284.

Trimethylenemethane-Type Biradicals
J . Org. Chem., Vol. 64, No. 12, 1999 4393
Curie plots for 1••-4••, 6••, and 7•• the Curie plot for 5••,
shown in Figure 7, is characteristic of a low-spin ground-
state. A multiparameter fit of the data gives J ) -68 (
6 cal/molsa rather large negative value considering that
J ) +13 cal/mol for 1••.13 Thus, polycyclic spin-protecting
groups can permit greater spin delocalization into the
TMM spin coupler (as in 3••) or they can very effectively
inhibit spin delocalization and give rise to a singlet
ground state as in 5••.
ments. In addition, solid-state studies are needed to
separate effects due to phenyl-tert-butylnitroxide torsion
from aryl torsion. Studies of solid-state structure and
properties of 1••-7•• are underway.
Exp er im en ta l Section
Compounds 1••,13 8,44 9,9 10,54 and di-tert-butyl ketone
hydrazone17 were prepared according to literature procedures.
Carbinol 17 was prepared by NaBH4 reduction of the com-
mercially available 4,4′-dibromobenzophenone (15) and used
without purification. Di-tert-butyldiazomethane was prepared
in excellent yield by oxidation of the corresponding hydrazone
as described below, in contradiction to an earlier report.17
Solvent distillations, synthetic procedures, and EPR sample
preparation were carried out under an argon or nitrogen
atmosphere, often in combination with the use of a glovebox.
THF and toluene were distilled from sodium benzophenone-
ketyl prior to use. NMR spectra were recorded at 300 MHz
(1H NMR) and 75 MHz (13C NMR). Elemental analysis was
performed by Atlantic Microlab, Inc, Norcross, GA. X-Band
EPR spectroscopy was performed as described previously.55
Room-temperature magnetic moments were measured on a
Cahn Faraday Balance, calibrated with Hg[Co(SCN)4].
Nitr oxid e Ch r om a togr a p h y. Following removal of solids
from the oxidation reaction, the crude reaction mixture is first
chromatographed on a silica gel column using gradient elution
(0 to 1% THF/CH2Cl2). This column separates mononitroxide
from other products. The remaining red material is subjected
to a second column again using gradient elution (0 to 25%
ether/petroleum ether) to obtain pure dinitroxide. Nitroxides
from 1-Br 2-6-Br 2 were purified by this procedure. All biradi-
cals had room-temperature magnetic moments of 2.6 ( 0.2
Bohr magnetonssconsistent with two unpaired electrons.
2-(4′-Ch lor o-n -bu tyl)cycloh eptan on e, 12. Cycloheptanone
(2.85 g, 25.5 mmol) was dissolved in 75 mL of THF in a 250
mL Schlenk flask and cooled to -78 °C. LDA (prepared from
the addition of 10.2 mL of a 2.5 M BuLi solution to 2.59 g,
25.5 mmol, i-Pr2NH in 10 mL THF) was then added dropwise,
and the reaction was warmed to room temperature followed
by recooling to -78 °C. 4-Chloro-1-iodobutane (5.0 g, 22.9
mmol) was then added as a solution in 5 mL of THF, and the
reaction was stirred 12 h while warming to room temperature.
The reaction was quenched with 1 mL of a 1 M aqueous HCl
solution, transferred to a separatory funnel, washed once 1 M
aqueous HCl and once with saturated aqueous NaCl, and dried
over Na2SO4. Following filtration and evaporation, the residue
was subjected to chromatography (SiO2, gradient elution: 0
to 2% ether/petroleum ether) to give the product as a colorless
oil (2.32 g, 50%). 1H NMR (CDCl3) δ (ppm): 3.52 (t, J ) 6.7
Hz, 2H), 2.5 (m, 3H), 2.0-1.0 (m, 14H). 13C NMR (CDCl3) δ
(ppm): 215.9, 51.0, 44.3, 43.8, 32.6, 31.5, 31.3, 29.5, 32.6, 25.6,
24.5. IR (film from CH2Cl2) ν (cm-1): 1700. HRMS (EI) calcd
for C11H19ClO (M+): 202.1124. Found: 202.1130.
Con clu sion s
Several dinitroxide biradicals have been prepared that
differ in tert-butylaminoxylphenyl ring torsion angles by
virtue of different steric demand of their spin-protecting
groups. In addition, the synthesis of a TMM-type dini-
troxide having a planar π-system was prepared. The
synthetic route employed for the planar TMM-type
biradical should be general and should therefore expand
the utility of TMM-type biradicals that are susceptible
to bond torsions that attenuate exchange coupling. These
dinitroxides are designed to be an effective series of
molecules to map out J -coupling/conformation space in
TMM-type biradicals.
Molecular mechanics was used to estimate the differ-
ences in phenyl torsion in a series of 1,1-diphenylethenes
that serve as conformational models for the correspond-
ing dinitroxides. Bicyclic/tricyclic spin-protecting groups
can be sterically demanding at large bond angles, θ.
Conversely, bicyclic/tricyclic spin-protecting groups hav-
ing a small bond angle θ allow for greater delocalization
in the diphenylethene unit by permitting smaller phenyl
torsions, φ.
EPR spectroscopic parameters (hfcc and D-values) were
examined to determine if their values reflected differ-
ences in conformation in the series of biradicals. Interest-
ingly, neither N- nor H-hfcc varied within our series of
biradicals/monoradicals in a regular fashion indicating
that there is no apparent relationship between N- or
H-hfcc and conformation as in biphenyl-tert-butylnitrox-
ides.45 Also, the biradicals studied do not exhibit J -
modulated EPR spectra. However, D-values, estimated
from relative intensities of ∆ms ) 1 and ∆ms ) 2
transitions,32 are consistent with a varied degree of
delocalization in the dinitroxides: smaller D-values for
biradicals having smaller aryl torsions, and larger D-
values for dinitroxides having greater aryl torsions.
All the biradicals studied, except 5••, exhibited linear
Curie plots. These results are consistent with both triplet
ground-states and singlet-triplet degeneracies. Interest-
ingly, dinitroxide 6•• exhibited a linear Curie plot, despite
the lack of a π-coupling fragment. We feel these results
indicate that the singlet triplet gaps are probably small,
and that of 6•• is smaller than the TMM-type dinitroxides.
Indeed, the triplet ground state of 1•• lies only 13 cal/mol
below the singlet state,13 and so our series of biradicals
probably have similar J -values. The notable exception
is biradical 5••, a ground-state singlet species with the
triplet lying about 140 cal/mol above the singlet.
In summary, the affect of bond torsions on exchange
coupling the series of TMM-type dinitroxides as deter-
mined by variable-temperature EPR was manifested only
in biradical 5••, presumably due to large phenyl torsions.
Differences in exchange couplings in TMM-type dinitrox-
ides other than 5•• are probably subtle and should be
revealed in solid-state magnetic susceptiblity experi-
Bicyclo[4.4.1]u n d eca n -11-on e, 13. 2-(4′-Chloro-n-butyl)-
cycloheptanone, 12 (1.44 g, 7.11 mmol), and NaI (5.3 g, 36
mmol) were dissolved in 100 mL of reagent grade acetone in
a 250 mL round-bottom flask and refluxed 12 h. Acetone was
evaporated and the yellow/brown solid mass partitioned
between H2O and Et2O. The organic portion was transferred
to a separatory funnel, washed once with H2O and once with
saturated aqueous NaCl, and dried over Na2SO4. Solvent
removal yielded the iodide as a viscous, air-sensitive yellow
oil which was used without purification. The crude iodide was
dissolved in 40 mL of THF in a 250 mL Schlenk flask and
cooled to -78 °C. LiHDMS (prepared from the addition of 3.0
mL of 2.5 M BuLi solution to 1.43 g, 1.25 mmol, (SiMe3)2NH
in 10 mL of THF) was added dropwise, and the reaction
mixture was allowed to warm to room temperature im-
mediately after addition. We found it to be important to add
(54) Davies, D. I.; Hey, D. H.; Summers, B. J . Chem. Soc. C 1970,
2653.
(55) Shultz, D. A.; Farmer, G. T. J . Org. Chem. 1998, 63, 6254.

4394 J . Org. Chem., Vol. 64, No. 12, 1999
Shultz et al.
LiHDMS slowly and warm the reaction immediately after
addition in order to maximize yield of the desired ketone. The
reaction was stirred 12 h, quenched with 1 mL of a 1 M
aqueous HCl solution, transferred to a separatory funnel,
washed once 1 M aqueous HCl and once with saturated
aqueous NaCl, dried over Na2SO4, filtered, and evaporated to
dryness. Chromatography (SiO2, gradient elution: 0 to 10%
ether/petroleum ether) gave the product as a colorless oil (770
mg, 65%). 1H NMR (CDCl3) δ (ppm): 2.77 (m, 2H), 1.81 (m,
8H), 1.65 (m, 4H), 1.48 (m, 4H). 13C NMR (CDCl3) δ (ppm):
219.3, 34.9, 27.1, 26.4. IR (film from CH2Cl2) ν (cm-1): 1690.
Anal. Calcd for C11H18O: C, 79.46; H, 10.91. Found; C, 79.33;
H, 10.85.
Bicyclo[4.4.1]u n d eca n -11-on e Hyd r a zon e. Bicyclo[4.4.1]-
undecan-11-one, 13 (1.2 g, 7.25 mmol), hydrazine monohydrate
(10 g, 200 mmol, 10 mL), and hydrazine sulfate (1.0 g, 9 mmol)
were combined in a 25 mL round-bottom flask. The resulting
biphasic mixture was refluxed for 4 days, and upon cooling a
solid precipitated. This mixture was transferred to a separa-
tory funnel containing H2O and Et2O, washed twice with H2O
and once with saturated brine, and dried over Na2SO4. Solvent
removal gave a white solid which was used without further
purification. 1H NMR (CDCl3) δ (ppm): 4.9 (broad s, 2H), 3.13
(m, J ) 7.6 Hz, 1H), 2.76 (m, J ) 5.7 Hz, 1H), 1.8-1.1 (m,
16H). 13C NMR (CDCl3) δ (ppm): 162.6, 48.0, 36.3, 30.7, 28.7,
26.6, 26.2. IR (film from CH2Cl2) ν (cm-1): 3331, 3216, 1638.
HRMS (EI) calcd for C11H20N2 (M+): 180.1626. Found: 180.1630.
11-Diazobicyclo[4.4.1]u n decan e, 14. Bicyclo[4.4.1]undecan-
11-one hydrazone (300 mg, 1.67 mmol) was dissolved in 30
mL of Et2O and added to yellow HgO (2.0 g, 9.23 mmol). Na2-
SO4 (2.0 g, 14.08 mmol) was added, followed by a few drops of
a saturated ethanolic solution of NaOH. The resulting yellow
suspension was stirred for 14 h, filtered, and evaporated to
yield a crystalline orange solid. The IR spectrum of this
material was consistent with nearly complete conversion to
the product (no NH stretches at 3331 and 3216 cm-1), and the
material was used without further purification. IR (film from
CH2Cl2) ν (cm-1): 2024.
Di-ter t-bu tyld ia zom eth a n e. Di-tert-butyl ketone hydra-
zone (490 mg, 3.1 mmol) was dissolved in 30 mL of Et2O and
added to yellow HgO (1.5 g, 6.9 mmol). Na2SO4 (2.0 g, 14 mmol)
was added, followed by a few drops of a saturated ethanolic
solution of NaOH, and the resulting yellow suspension was
stirred for 4 h. After this time, the mixture was filtered and
evaporated to yield a thick orange liquid. This material was
distilled (25 °C, 100 mTorr) with a Kugelrohr apparatus to
give the pure diazoalkane, having the signature IR stretch
(2027 cm-1) cited in the literature.16
Ad a m a n tylbis(4′-br om op h en yl)ca r bin ol. In a 100 mL
Schlenk flask, 1,4-dibromobenzene (2.10 g, 8.88 mmol) was
dissolved in 20 mL of THF and cooled to -78 °C. tert-
Butyllithium (11.3 mL of a 1.5 M solution) was added dropwise
and the reaction stirred for 1 h at -78 °C, yielding a turbid,
colorless solution. Ester 9 (750 mg, 3.86 mmol) was then added
as a solution in 10 mL of THF, and the reaction allowed to
stir 12 h at ambient temperature. The reaction was quenched
with saturated aqueous NH4Cl (2.5 mL), transferred to a
separatory funnel, and washed once with saturated aqueous
NaCl. The organic portion was collected and dried over Na2-
SO4, filtered, and evaporated to dryness. The residue was
subjected to radial chromatography (SiO2, gradient elution: 1%
to 10% Et2O/petroleum ether) to give the product as a colorless
solid (1.57 g, 87%). 1H NMR (CDCl3) δ (ppm): 7.41 (d, J ) 7.5
Hz, 4H), 7.29 (d, J ) 7.6 Hz, 4H), 2.61 (s, 1H), 2.24-2.20 (broad
m, 3H), 1.73-1.92 (broad m, 10H), 1.54-1.59 (broad m, 2H).
13C NMR (CDCl3) δ (ppm): 146.2, 131.4, 127.5, 120.6, 82.5,
49.7, 41.2, 38.4, 33.5, 29.2, 28.6, 27.1. IR (film from CH2Cl2) ν
(cm-1): 3614. Anal. Calcd for C23H24Br2O: C, 59.04; H, 4.96.
Found; C, 58.81; H, 5.09.
The reaction was cooled, transferred to a separatory funnel,
and washed once with saturated aqueous NaCl. The organic
portion was collected, dried over Na2SO4, filtered, and evapo-
rated to dryness. The resulting solid was subjected to radial
chromatography (SiO2, petroleum ether) to give the product
as a colorless solid (1.12 g, 96%). 1H NMR (CDCl3) δ (ppm):
7.39 (d, J ) 8.1 Hz, 4H), 6.97 (d, J ) 7.9 Hz, 4H), 2.73 (broad
s, 2H), 7.99 (broad m, 2H), 1.85 (broad m, 10H). 13C NMR
(CDCl3) δ (ppm): 148.4, 141.3, 131.3, 128.5, 120.3, 39.6, 37.0,
34.6, 28.1. Anal. Calcd for C23H22Br2: C, 60.28; H, 4.83. Found;
C, 60.35; H, 4.85.
Bis(4′-ter t-b u t yla m in oxylp h en yl)m et h ylen ea d a m a n -
ta n e, 4•• (a n d 4•). 4-Br 2 (500 mg, 1.09 mmol) was dissolved
in 20 mL of THF in a 100 mL Schlenk flask and cooled to -78
°C. tert-Butyllithium (2.91 mL of a 1.5 M solution) was then
added dropwise, and the resulting pink to light red solution
was stirred for 3 h at -78 °C. A solution of 2-methyl-2-
nitrosopropane dimer (210 mg, 1.20 mmol) was then added as
a solution in 15 mL of THF and the reaction stirred 12 h. The
resulting emerald green solution was quenched with NH4Cl
(3 mL), transferred to a separatory funnel, and washed once
with saturated aqueous NaCl. The organic portion was col-
lected and dried over Na2SO4. When the solvent was removed,
either a viscous oil or solid remained, and the product was
often slightly colored due to aerial oxidation of the hydroxy-
lamine. The resulting material was oxidized without purifica-
tion by stirring with AgO (1.5 g, 12.1 mmol) in 15 mL of THF
for 3 h under argon and protected from light. The resulting
deep red solution was filtered through Celite and the solvent
removed to give a dark red solid. This solid was chromato-
graphed to give monoradical (107 mg, 26%) and 4•• (272 mg,
53%). Monoradical: HRMS (FAB) calcd for C27H33NO (M+
+
H): 387.2562. Found: 387.2569. 4••: HRMS (FAB) calcd for
C
31H42N2O2 (M+ + 2H): 474.3246. Found: 474.3260.
Nor bor n ylbis(4′-br om op h en yl)ca r bin ol. In a 100 mL
Schlenk flask, 1,4-dibromobenzene (1.76 g, 7.47 mmol) was
dissolved in 20 mL of THF and cooled to -78 °C. tert-
Butyllithium (9.53 mL of a 1.5 M solution) was then added
dropwise and the reaction stirred for 1 h at -78 °C, yielding
a turbid, colorless solution. Ester 8 (500 mg, 3.25 mmol) was
then added as a solution in 10 mL of THF, and the reaction
allowed to stir 12 h at ambient temperature. The reaction was
quenched with saturated aqueous NH4Cl (2.5 mL), transferred
to a separatory funnel, and washed once with saturated
aqueous NaCl. The organic portion was collected and dried
over Na2SO4, filtered, and evaporated to dryness. The residue
was purified by radial chromatography (SiO2, gradient elu-
tion: 1% to 10% Et2O/petroleum ether) to give the product as
a colorless solid (1.21 g, 85%). 1H NMR (CDCl3) δ (ppm): 7.42
(d, J ) 8.8 Hz, 4H), 7.28 (d, J ) 8.7 Hz, 4H), 2.41 (broad s,
1H), 1.97 (broad m, 3H), 1.66 (broad m, 4H), 1.18-1.26 (broad
m, 6H). 13C NMR (CDCl3) δ (ppm): 147.0, 131.2, 127.9, 120.6,
78.5, 57.5, 38.1, 31.2, 30.0. IR (film from CH2Cl2) ν (cm-1):
3586. Anal. Calcd for C20H20Br2O: C, 55.07; H, 4.62. Found:
C, 54.99; H, 4.71.
Bis(4′-br om op h en yl)m eth ylen en or bor n a n e, 3-Br 2. The
carbinol from the previous reaction (1.0 g, 2.29 mmol) and
p-toluenesulfonic acid hydrate (25 mg) were added to a 250
mL round-bottom flask with 125 mL of benzene. A Dean-
Stark trap/reflux condenser assembly was attached to the
flask, and the reaction mixture was refluxed for 3 h. The
solution was cooled, transferred to a separatory funnel, and
washed once with saturated aqueous NaCl. The organic
portion was dried over Na2SO4, filtered, and evaporated to
dryness. The resulting solid was purified by radial chroma-
tography (SiO2, petroleum ether) to give the product as a
1
colorless solid (910 mg, 95%). H NMR (CDCl3) δ (ppm): 7.39
(d, J ) 8.6 Hz, 4H), 6.98 (d, J ) 8.5 Hz, 4H), 2.63 (broad m,
2H), 1.73 (broad m, 4H), 1.44 (broad m, 4H). 13C NMR (CDCl3)
δ (ppm): 152.0, 141.3, 131.5, 131.1, 124.5, 120.6, 38.2, 28.9.
Anal. Calcd for C20H18Br2: C, 57.45; H, 4.34. Found; C, 57.55;
H, 4.41.
Bis(4′-br om op h en yl)m eth ylen ea d a m a n ta n e, 4-Br 2. The
carbinol from the previous reaction (1.21 g, 2.56 mmol) and
p-toluenesulfonic acid hydrate (25 mg) were added to a 250
mL round-bottom flask containing 125 mL of benzene. The
flask was connected to a Dean-Stark trap/reflux condenser
assembly and refluxed for 3 h with azeotropic removal of water.
B i s (4′-t er t -b u t y la m i n o x y lp h e n y l)m e t h y le n e n o r -
bor n a n e, 3•• (a n d 3•). 3-Br 2 (500 mg, 1.12 mmol) was
dissolved in 20 mL of THF in a 100 mL Schlenk flask and

Trimethylenemethane-Type Biradicals
J . Org. Chem., Vol. 64, No. 12, 1999 4395
cooled to -78 °C. tert-Butyllithium (3.12 mL of a 1.5 M
solution) was then added dropwise and the resulting solution
stirred for 3 h at -78 °C. A solution of 2-methyl-2-nitrosopro-
pane dimer (230 mg, 1.3 mmol) was then added as a solution
in 15 mL of THF and the reaction stirred 12 h. The resulting
solution was quenched with NH4Cl (3 mL), transferred to a
separatory funnel, and washed once with saturated aqueous
NaCl. The organic portion was collected and dried over Na2-
SO4. When the solvent was removed, either a viscous oil or
solid remained, and the product was often slightly colored due
to aerial oxidation of the hydroxylamine. The resulting mate-
rial was oxidized without purification by stirring with AgO
(1.5 g, 12.1 mmol) in 15 mL of THF for 3 h under argon and
protected from light. The resulting deep red solution was
filtered through Celite and the solvent removed to give a dark
red solid which was chromatographed to give monoradical (78
mg, 19%) and 3•• (326 mg, 63%). Monoradical: HRMS (FAB)
calcd for C24H29NO (M+ + H): 347.2249. Found: 347.2240.
3••: HRMS (FAB) calcd for C28H38N2O2 (M+ + 2H): 434.2933.
Found: 434.2928.
transferred to a separatory funnel containing Et2O and H2O,
and the resulting blue organic layer was washed three times
with saturated aqueous NaHCO3, washed once with saturated
aqueous NaCl, and then dried over Na2SO4. The solvents were
removed to give a blue solid which was recrystallized (argon
was blown over the solution during boiling and cooling) from
MeOH to give dark blue microcrystals of the product. This
material was contaminated (ca. 25% by 1H NMR) with starting
material; however, the subsequent reaction was unaffected by
this, so no further purification was attempted.
B i s (4′,4′′-b r o m o p h e n y l)m e t h y le n e b i c y c lo [4.4.1]-
u n d eca n e Ep isu lfid e. Thioketone 16 (ca. 1.14 g, ca. 3.25
mmol) was dissolved in 5 mL of THF in a 50 mL round-bottom
flask to give a deep blue solution. Diazoalkane 14 (580 mg,
3.25 mmol) was then added as an orange solution in 15 mL of
THF, causing gas evolution, and the resulting blue solution
protected from light and refluxed 12 h. THF was removed, and
the material was subjected to radial chromatography (SiO2,
petroleum ether) to give the episulfide as a colorless solid (1.06
1
g, 66%). H NMR (CDCl3) δ (ppm): 7.50 (d, J ) 8.7 Hz, 4H),
9,9-Bis(4′-b r om op h en yl)flu or en e, 6-Br 2.
A 100 mL
7.39 (d, J ) 8.6 Hz, 4H), 2.07 (pent, J ) 6.6 Hz, 2H), 1.26-
1.92 (broad m, 18H). 13C NMR (CDCl3) δ (ppm): 140.7, 132.6,
130.7, 120.9, 69.1, 66.6, 43.0, 33.9, 31.1, 26.6, 25.4. HRMS
(FAB) calcd for C24H26Br2S (M+): 504.0122. Found: 504.0125.
Schlenk flask was charged with 1,4-dibromobenzene (2.41 g,
9.43 mmol). This material was dissolved in 15 mL of THF and
cooled to -78 °C, tert-butyllithium (12.6 mL of a 1.5 M
solution) was then added dropwise, and the resulting turbid
white solution was stirred for 1 h at -78 °C. Methyl 2-phe-
nylbenzoate, 10 (1.0 g, 4.71 mmol), in 5 mL of THF was then
added, and the reaction was allowed to stir at room temper-
ature 12 h. The reaction was quenched with 1 mL of a
saturated NaCl solution, transferred to a separatory funnel,
washed with a saturated NaCl solution, dried over MgSO4,
filtered, and evaporated to dryness. The crude product was
added to a 25 mL beaker containing poly[phosphoric acid] (10
g) and heated to 80 °C for 10 min during which time the brown
solution produced a white precipitate. This mixture was then
cooled and quenched with crushed ice, poured into a separatory
funnel, and extracted with Et2O. The organic solution was
washed once with a saturated NaCl solution, filtered, dried
over MgSO4, and evaporated to dryness. The resulting material
was purified by radial chromatography (SiO2, petroleum ether)
to give the product as a colorless solid (220 mg, 92%). 1H NMR
(CD2Cl2) δ (ppm): 7.80 (d, J ) 7.5 Hz, 2H), 7.36 (m, 10H), 7.06
(d, J ) 8.6 Hz, 4H). 13C NMR (CDCl3) δ (ppm): 150.13, 144.50,
140.04, 131.42, 129.75, 127.96, 127.90, 120.92, 120.39, 64.58.
Anal. Calcd for C25H16Br2: C, 63.06; H, 3.39. Found; C, 63.18;
H, 3.44.
B i s ( 4 ′-b r o m o p h e n y l ) m e t h y l e n e b i c y c l o [ 4 .4 .1 ] -
u n d eca n e, 5-Br 2. The episulfide from the previous reaction
(1.06 g, 2.13 mmol) and triphenylphosphine (1.12 g, 4.26 mmol)
were dissolved in 15 mL of toluene, and the resulting solution
was refluxed for two nights. After cooling, the reaction mixture
was filtered through SiO2, evaporated, and the residue was
subjected to radial chromatography (SiO2, petroleum ether)
to give the product as a colorless solid (815 mg, 87%). 1H NMR
(CDCl3) δ (ppm): 7.40 (d, J ) 8.4 Hz, 4H), 7.07 (d, J ) 8.3 Hz,
4H), 2.95 (pent, J ) 5.8 Hz, 2H), 1.5 (broad m, 16H). 13C NMR
(CDCl3) δ (ppm): 148.4, 142.9, 136.7, 131.7, 130.5, 120.1, 41.2,
32.2, 27.0. Anal. Calcd for C24H26Br2: C, 60.78; H, 5.52. Found;
C, 60.54; H, 5.52.
Bis(4′-ter t-b u t yla m in oxylp h en yl)m et h ylen eb icyclo-
[4.4.1]u n d eca n e, 5•• (a n d 5•). 5-Br 2 (200 mg, 0.43 mmol) was
dissolved in 10 mL of THF in a 50 mL Schlenk flask and cooled
to -78 °C. tert-Butyllithium (1.15 mL of a 1.5 M solution, 1.72
mmol) was then added dropwise and the resulting solution
stirred for 3 h at -78 °C. 2-Methyl-2-nitrosopropane dimer
(83 mg, 0.47 mmol) was then added as a solution in 5 mL of
THF, the cooling bath was removed, and the reaction mixture
was allowed to stir for 12 h at room temperature. Saturated
aqueous NH4Cl solution (1 mL) was added to quench the
reaction, and the mixture was transferred to a separatory
funnel, washed once with both saturated aqueous NH4Cl and
saturated aqueous NaCl, separated, filtered, and dried over
Na2SO4. Following evaporation, the resulting crude bis(hy-
droxylamine) was oxidized without purification by stirring over
AgO (1.5 g, 12.1 mmol) in 15 mL of THF for 3 h under argon
and protected from light. The resulting red suspension was
filtered through Celite and the solvent removed to give the
crude biradical as a dark red solid. This solid was chromato-
graphed in the usual manner, yielding monoradical (19 mg,
10%) and 5•• (148 mg, 72%). Monoradical: HRMS (FAB) calcd
for C28H37NO (M+ + H): 403.2875. Found: 403.2888. 5••:
HRMS (FAB) calcd for C32H46N2O2 (M+ + 2H): 490.3559.
Found: 490.3548.
9,9-Bis(4′-ter t-bu tyla m in oxylp h en yl)flu or en e, 6•• (a n d
6•). 6-Br 2 (500 mg, 1.05 mmol) was dissolved in 20 mL of THF
in a 100 mL Schlenk flask and cooled to -78 °C. tert-
Butyllithium (2.8 mL of a 1.5 M solution) was then added
dropwise and the resulting pink to light red solution stirred
for 3 h at -78 °C. A solution of 2-methyl-2-nitrosopropane
dimer (200 mg, 1.3 mmol) was then added as a solution in 15
mL of THF and the reaction stirred 12 h. The resulting
emerald green solution was quenched with NH4Cl (3 mL),
transferred to a separatory funnel, and washed once with
saturated aqueous NaCl. The organic portion was collected and
dried over Na2SO4. When the solvent was removed, either a
viscous oil or solid remained, and the product was often slightly
colored due to aerial oxidation of the hydroxylamine. The
resulting material was oxidized without purification by stirring
with AgO (1.5 g, 12.1 mmol) in 15 mL of THF for 3 h under
argon and protected from light. The resulting deep red solution
was filtered through Celite and the solvent removed to give a
dark red solid which was chromatographed to give mononi-
troxide (320 mg, 40%) and dinitroxide, 6•• (270 mg, 33%).
Mononitroxide: HRMS (FAB) calcd for C29H27NO (M+ + H):
405.2093. Found: 405.2087. 6••: HRMS (FAB) calcd for
Bis(4′-b r om op h en yl)m et h ylen ed i-ter t-b u t ylm et h a n e
Ep isu lfid e. Thioketone 16 (1.02 g, 2.92 mmol) was dissolved
in 5 mL of THF in a 50 mL round-bottom flask to give a deep
blue solution. Di-tert-butyldiazomethane (450 mg, 2.92 mmol)
was then added as an orange solution in 15 mL of THF, at
which point the color dissipated. The resulting solution,
protected from light, was refluxed for 12 h during which time
some of the blue color returned. THF was removed, and the
material was subjected to radial chromatography (SiO2, 5%
Et2O/petroleum ether) to give the episulfide as a light yellow
C
33H36N2O2 (M+ + 2H): 492.2777. Found: 492.2751.
4,4′-Dibr om oth ioben zop h en on e, 16. 4,4′-Dibromoben-
zophenone (2.0 g, 6.03 mmol) and P4S10 (4.02 g, 9.05 mmol)
were suspended in 20 mL of acetonitrile. NaHCO3 (3.1 g, 36
mmol) was then added and the mixture heated to 50 °C. The
resulting mixture was then allowed to stir 12 h, during which
time it became dark blue. After cooling, the mixture was
1
solid (770 mg, 56%). H NMR (d6-acetone) δ (ppm): 8.29 (dd,
J ) 8.4, 2.3 Hz, 2H), 7.8 (m, 6H), 1.6 (s, 18H). 13C NMR (CDCl3)
δ (ppm): 145.2, 132.1, 131.8, 131.3, 130.2, 120.3, 76.7, 70.5,

4396 J . Org. Chem., Vol. 64, No. 12, 1999
Shultz et al.
42.0, 34.3. HRMS (FAB) calcd for C22H26Br2S (M+): 480.0122.
Found: 480.0117.
glovebox. Freshly distilled THF (9 mL) was added, and after
removal from the glovebox, the mixture was cooled to -20 °C.
An aliquot of 1.4 M methyllithium (10.4 mL, 14.60 mmol) was
added via syringe, and the reaction mixture was stirred for 1
h at -20 °C and 10 min at 25 °C. Dione 18 (770 mg, 1.82 mmol)
as a solution in THF (9 mL) was added by cannulation, keeping
the temperature at -20 °C. The reaction mixture was stirred
for 10 min, after which time the reaction was judged complete
by TLC, and quenched by pouring into saturated aqueous
brine. The resulting mixture was extracted with ether (3×),
and the extracts were combined and dried over Na2SO4. After
decanting and removing the solvent, the crude product was
rinsed with CH2Cl2 to give the intermediate diol as a white
solid (597 mg, 72%). The diol decomposes (to produce an
alkene) in solution at a rate that precludes recording a 13C
NMR spectrum. 1H NMR (CDCl3) δ (ppm): 7.37 (d, J ) 8.2
Hz, 4H) 6.98 (d, J ) 8.2 Hz, 4H), 3.70 (s, 2H), 1.33 (s, 12H).
IR (film from CH2Cl2) ν (cm-1): 3270. Anal. Calcd for C20H22O2-
Br2: C, 52.88; H, 4.88. Found: C, 52.88; H, 4.91.
A 100 mL round-bottom flask containing the diol from the
previous reaction (500 mg, 1.1 mmol) and excess poly[phos-
phoric acid] (10 g) was heated to 147 °C for ca. 1 h. When
judged complete by TLC, the reaction was cooled and carefully
poured onto water/crushed ice with stirring. The resulting
mixture was extracted with ether (3×), and the combined
organic extracts were dried with Na2SO4, filtered, and evapo-
rated to dryness. The residue was crystallized from ether to
give 7-Br 2 as long, white needles (431 mg, 63%). 1H NMR
(CDCl3) δ (ppm): 7.48 (s, 2H) 7.44 (m, 4H), 1.50 (s, 12H). 13C
NMR (CDCl3) δ (ppm): 170.5, 160.5, 139.7, 135.4, 129.6, 125.2,
120.9, 119.3, 48.1, 24.7. Anal. Calcd for C20H18Br2: C, 57.44;
H, 4.33. Found: C, 57.45; H, 4.35.
5,10-Di(ter t-bu tyla m in oxyl)-3,12-d ih yd r o-3,3,12,12-tet-
r a m eth ylin d en o[1,2-a ]in d en e, 7•• (a n d 7•). To a solution of
7-Br 2 (275 mg, 0.66 mmol) in THF (14 mL) at -78 °C in a 100
mL Schlenk flask was added tert-butyllithium (1.5 M in
pentane, 1.77 mL, 2.66 mmol, 4.05 equiv) slowly via syringe.
After 4 h at -78 °C, a solution of 2-methyl-2-nitrosopropane
dimer (126 mg, 0.72 mmol, 1.1 equiv) in THF (10 mL) was
added by cannulation. After stirring 12 h, the reaction was
quenched with saturated aqueous NH4Cl. The resulting mix-
ture was diluted with ether and washed with saturated
aqueous NaCl. The organic layer was dried with Na2SO4,
filtered, and concentrated under reduced pressure to give the
bis(hydroxylamine) as a light yellow solid (285 mg, 99%). IR
(film from CH2Cl2) ν (cm-1): 3348, 3177.
A mixture of the bis(hydroxylamine) (91.5 mg, 0.211 mmol)
and excess AgO (300 mg, 2.42 mmol) in THF (5 mL) was
stirred for 4 h. The reaction mixture was passed through a
plug of Celite and rinsed with ether. After evaporation, the
residue was purified by radial chromatography (SiO2, gradient
elution: 10% to 30% ether/petroleum ether) to give biradical
7•• as a purple solid (67 mg, 73%), and monoradical 7• (3.5 mg,
5%). UV-Vis (CH2Cl2) λmax (log ꢀ): 230 (4.53), 290 (4.39), 349
(4.75), 520 (3.11), 620 (2.94). Monoradical: HRMS (FAB) calcd
for C24H29NO (M+ + H): 347.2249. Found: 347.2263. 7••:
HRMS (FAB) calcd for C28H38N2O2 (M+ + 2H): 434.2933.
Found: 434.2933.
1,1-Bis(4′-br om op h en yl)-2,2-d i-ter t-bu tyleth en e, 2-Br 2.
The episulfide from the previous reaction (770 mg, 1.62 mmol)
and triphenylphosphine (850 mg, 3.24 mmol) were dissolved
in 15 mL of toluene, and the resulting solution was refluxed
for 48 h. After cooling, the reaction mixture was filtered
through SiO2 and evaporated, and the residue was subjected
to radial chromatography (SiO2, petroleum ether) to give the
1
alkene as a colorless solid (590 mg, 82%). H NMR (CDCl3) δ
(ppm): 7.31 (d, J ) 8.5 Hz, 4H), 6.84 (d, J ) 8.4 Hz, 4H), 1.56
(s, 18H). 13C NMR (CDCl3) δ (ppm): 156.0, 148.0, 141.7, 132.0,
130.8, 120.2, 40.5, 35.4. Anal. Calcd for C22H26Br2: C, 58.69;
H, 5.82. Found; C, 58.57; H, 5.86.
1,1-Bis(4′-t er t -b u t yla m in oxylp h e n yl)-2,2-d i-t er t -b u -
tyleth en e, 2•• (a n d 2•). 2-Br 2 (200 mg, 0.45 mmol) was
dissolved in 10 mL of THF in a 50 mL Schlenk flask and cooled
to -78 °C. tert-Butyllithium (1.2 mL of a 1.5 M solution, 1.8
mmol) was then added dropwise and the resulting solution
stirred for 3 h at -78 °C. 2-Methyl-2-nitrosopropane dimer
(88 mg, 0.50 mmol) was then added as a solution in 5 mL of
THF. The cooling bath was removed, and the reaction mixture
was stirred for 12 h. Saturated aqueous NH4Cl solution (1 mL)
was added to quench the reaction, and the mixture was
transferred to a separatory funnel. The organic layer was
washed once with both saturated aqueous NH4Cl and satu-
rated aqueous NaCl, separated, dried over Na2SO4, filtered,
and evaporated. The resulting material was oxidized without
purification by stirring with AgO (1.5 g, 12.1 mmol) in 15 mL
of THF for 3 h under argon and protected from light. The
resulting red suspension was filtered through Celite and the
solvent removed to give the crude biradical as a dark red solid.
This solid was chromatographed in the usual manner, yielding
mononitroxide (17 mg, 8%) and 2•• (100 mg, 50%). Monoradi-
cal: HRMS (FAB) calcd for C26H37NO (M+ + H): 379.2875.
Found: 379.2898. 2••: HRMS (FAB) calcd for C30H46N2O2 (M+
+ 2H): 466.3559. Found: 466.3589.
3-Acet yl-4,4-b is(4′-b r om op h en yl)b u t a n -2-on e. In the
glovebox, carbinol 17 (1.98 g, 5.80 mmol), 2,4-pentadione (1.2
mL, 11.59 mmol), and anhydrous CoCl2 (300 mg, 2.32 mmol)
were added to a 25 mL Schlenk flask. After a condenser was
attached, the flask was removed from the glovebox, connected
to a Schlenk line, and heated to reflux for 1 h. The reaction
mixture was then cooled to room temperature, CH2Cl2 was
added, and the mixture was filtered to remove inorganic salts.
Solvent was removed by rotary evaporation and the residue
was rinsed with petroleum ether. The resulting solid was
recrystallized from ethanol to give the product as white needle
1
crystals (2.31 g, 94%). H NMR (CDCl3) δ (ppm): 7.40 (d, J )
8.2 Hz, 4H) 7.10 (d, J ) 8.2 Hz, 4H), 4.76 (d, J ) 12.2 Hz,
1H), 4.63 (d, J ) 12.2 Hz, 1H), 2.01 (s, 6H). 13C NMR (CDCl3)
δ (ppm): 202.0, 139.8, 132.2, 129.4, 121.3, 74.3, 49.8, 29.6. IR
(film from CH2Cl2) ν (cm-1): 1736, 1698. Anal. Calcd for
C
18H16O2Br2: C, 50.97; H, 3.80. Found: C, 51.08; H, 3.80.
3-Acetyl-4,4-bis(4′-br om op h en yl)-3-bu ten -2-on e, 18. A
100 mL Schlenk flask containing 3-acetyl-4,4-bis(4′-bromophe-
nyl)butan-2-one (2.17 g, 5.13 mmol), benzoyl peroxide (372 mg,
1.54 mmol), K2CO3 (1.06 g, 7.69 mmol), NBS (912 mg, 5.13
mmol), and 48 mL of CCl4 was heated to 78 °C for 10-15 min,
during which time reaction progress was monitored by TLC.
When the reaction was judged complete, and after cooling the
reaction mixture, saturated brine was added and the organic
layer was separated and evaporated to dryness. The crude
product was purified by column chromatography (SiO2, 2%
ether/petroleum ether) and then crystallized from ether to give
Ack n ow led gm en t. We thank the National Science
Foundation (CHE-9634878) for support of this work.
D.A.S. also thanks the Camille and Henry Dreyfus
Foundation for a Camille Dreyfus Teacher-Scholar
Award. Partial funding for the Mass Spectrometry
Laboratory for Biotechnology Facility was obtained from
the North Carolina Biotechnology Center and the
National Science Foundation (CHE-9111391).
1
18 as white crystals (1.36 g, 63%). H NMR (CDCl3) δ (ppm):
7.51 (d, J ) 8.4 Hz, 4H) 7.04 (d, J ) 8.4 Hz, 4H), 1.99 (s, 6H).
13C NMR (CDCl3) δ (ppm): 202.6, 146.0, 143.9, 137.7, 132.1,
131.5, 31.3. IR (film from CH2Cl2) ν (cm-1): 1697, 1667. Anal.
Calcd for C18H14O2Br2: C, 51.21; H, 3.34. Found: C, 51.21; H,
3.38.
Su p p or tin g In for m a tion Ava ila ble: Spectral data. This
material is available free of charge via the Internet at
http://pubs.acs.org.
5,10-Dibr om o-3,12-dih ydr o-3,3,12,12-tetr am eth ylin den o-
[1,2-a ]in d en e, 7-Br 2. A 50 mL Schlenk flask containing ZrCl4
(893 mg, 3.83 mmol) and a stir bar was prepared in the
J O990061J