Copper(II) benzoate nitroxide dimers and chains: Structure and magnetic studies

Article abstract of DOI:10.1021/ic0007106

Copper(II) benzoate dimers and linear chains have been synthesized and exhibit very different magnetic behaviors. The benzoate dimers, 1a, show typical dimeric singlet-triplet transitions and strong antiferromagnetic coupling (J(ST) = -206 K (-143 cm^-1); H = -2J(ST)S(a)·S(b)). The bromobenzoate dimer can be convened into linear chains of hydrogen-bonded monomers, showing 1-D ferromagnetic coupling (6a, θ = +9 K). Copper(II) sites can also be bridged by nitroxide-substituted benzoates, 1b and 1c, that is, 2-(4'-carboxyphenyl)-4,4,5,5-tetramethylimidazoline-3-oxide-1-oxyl (NNBA, 3a), with J(ST) = -216 K (-150 cm^-1), with comparable interactions between the nitroxide and triplet Cu(II) spins, θ(T) = -157 K. A 1-D chain similar to the bromobenzoate monomers can also be produced with NNBA, also exhibiting ferromagnetic coupling (6b, θ = +0.67 K), albeit much weaker. Other nitroxides have been introduced into the Cu(II) dimer system by capping copper(II) acetate with the polydentate 2-(4'-pyridyl)-4,4,5,5-tetramethylimidazoline-3-oxide-1-oxyl (PYNN, 3b), which exhibits almost no coupling to the copper centers when both ends of the dimer are capped (1d, θ = -5.8 K). In contrast, strong coupling is observed when only one PYNN is used (2, θ = -300 K), which is the result of direct coordination of the nitroxide to the copper centers, producing a chain of the dimer units.

Full text of DOI:10.1021/ic0007106

4894
Inorg. Chem. 2000, 39, 4894-4902
Copper(II) Benzoate Nitroxide Dimers and Chains: Structure and Magnetic Studies†
Rico E. Del Sesto, Atta M. Arif, and Joel S. Miller*
Department of Chemistry, 315 South 1400 East Room 2124, University of Utah,
Salt Lake City, Utah 84112-0850
ReceiVed June 27, 2000
Copper(II) benzoate dimers and linear chains have been synthesized and exhibit very different magnetic behaviors.
The benzoate dimers, 1a, show typical dimeric singlet-triplet transitions and strong antiferromagnetic coupling
(JST ) -206 K (-143 cm-1); H ) -2JSTSa‚Sb). The bromobenzoate dimer can be converted into linear chains
of hydrogen-bonded monomers, showing 1-D ferromagnetic coupling (6a, θ ) +9 K). Copper(II) sites can also
be bridged by nitroxide-substituted benzoates, 1b and 1c, that is, 2-(4′-carboxyphenyl)-4,4,5,5-tetramethylimi-
dazoline-3-oxide-1-oxyl (NNBA, 3a), with JST ) -216 K (-150 cm-1), with comparable interactions between
the nitroxide and triplet Cu(II) spins, θT ) -157 K. A 1-D chain similar to the bromobenzoate monomers can
also be produced with NNBA, also exhibiting ferromagnetic coupling (6b, θ ) +0.67 K), albeit much weaker.
Other nitroxides have been introduced into the Cu(II) dimer system by capping copper(II) acetate with the
polydentate 2-(4′-pyridyl)-4,4,5,5-tetramethylimidazoline-3-oxide-1-oxyl (PYNN, 3b), which exhibits almost no
coupling to the copper centers when both ends of the dimer are capped (1d, θ ) -5.8 K). In contrast, strong
coupling is observed when only one PYNN is used (2, θ ) -300 K), which is the result of direct coordination
of the nitroxide to the copper centers, producing a chain of the dimer units.
Introduction
they exhibit ferro-, ferri-, or antiferromagnetic coupling or
ordering.
A wide variety of spin-bearing compounds have been studied
enroute to the development of room-temperature molecule-based
magnets.1 From the discovery of the [Fe(C5Me5)2][TCNE]2
(TCNE ) tetracyanoethylene) magnets3 to the dithiadiazolyl
radical-based weak ferromagnet,4 the approach to understanding
magnetism has covered organometallic and purely organic-based
materials. Because of the embryonic nature of organic-based
magnets, there is still much to learn about the fundamental spin-
coupling behavior from the studies of all compounds, whether
One of the oldest magnetic systems studied is the carboxylate-
1
bridged copper(II) dimers, [CuII(O2CR)2]2 (1). The S ) /2
CuII atoms exhibit strong intradimer antiferromagnetic coupling
due to superexchange via the carboxylate bridge,5 producing a
singlet ground state (Stot ) 0) and a thermally populated triplet
(Stot ) 1) excited state, and the susceptibility, ø, can be modeled
using the Bleaney-Bowers expression (eq 1) for H ) -2JSTSa‚
Sb:6
2Ng2â2
kT[3 + e-2J
† Dedicated to the memory of Leigh C. Porter (October 29, 1955-May
13, 1999).
øBB
)
(1)
ST/kT
]
(1) (a) Ovcharenko, V. I.; Sagdeev, R. Z. Russ. Chem. ReV. 1999, 68,
345. (b) Kinoshita, M. Philos. Trans. R. Soc. London, Ser. A 1999,
357, 2855. (c) Miller, J. S.; Epstein, A. J. Chem. Commun. 1998, 1319.
(d) Plass, W. Chem.-Ztg. 1998, 32, 323. (e) Day, P. J. Chem. Soc.,
Dalton Trans. 1997, 701. (f) Miller, J. S.; Epstein, A. J. Chem. Eng.
News 1995, 73, No. 40, 30. (g) Miller, J. S.; Epstein, A. J. AdV. Chem.
Ser. 1995, 245, 161. (h) Miller, J. S.; Epstein, A. J. Angew. Chem.,
Int. Ed. Engl. 1994, 33, 385. (i) Kahn, O. AdV. Inorg. Chem. 1995,
43, 179. (j) Kinoshita, M. Jpn. J. Appl. Phys. 1994, 33, 5718. (k)
Gatteschi, D. AdV. Mater. (Weinheim, Ger.) 1994, 6, 635. (l) Kahn,
O. Molecular Magnetism; VCH Publishers: New York, 1993. (m)
Caneschi, A.; Gatteschi, D.; Rey, P. Prog. Inorg. Chem. 1991, 39,
331. (n) Buchachenko, A. L. Russ. Chem. ReV. 1990, 59, 307. (o)
Caneschi, A.; Gatteschi, D.; Sessoli, R.; Rey, P. Acc. Chem. Res. 1989,
22, 392. (p) Miller, J. S.; Epstein, A. J. In New Aspects of Organic
Chemistry; Yoshida, Z., Shiba, T., Oshiro, Y., Eds.; VCH Publishers:
New York, 1989; p 237. (q) Miller, J. S.; Epstein, A. J.; Reiff, W. M.
Acc. Chem. Res. 1988, 21, 114. (r) Miller, J. S.; Epstein, A. J.; Reiff,
W. M. Science 1988, 240, 40. (s) Miller, J. S.; Epstein, A. J.; Reiff,
W. M. Chem. ReV. 1988, 88, 201. (t) Kahn, O. Struct. Bonding 1987,
68, 89.
where N is Avogadro’s number, g is the Lande g-value, â is
the Bohr magneton, k is the Boltzmann constant, and JST is the
(3) (a) Manriquez, J. M.; Yee, G. T.; McLean, R. S.; Epstein, A. J.; Miller,
J. S. Science 1991, 252, 1415. Miller, J. S.; Yee, G. T.; Manriquez, J.
M.; Epstein, A. J. In Conjugated Polymers and Related Materials:
The Interconnection of Chemical and Electronic Structure, Proceedings
of Nobel Symposium No. NS-81; Oxford University Press: 1993; p
461. Chim. Ind. (Milan) 1992, 74, 845. Epstein, A. J.; Miller, J. S. In
Conjugated Polymers and Related Materials: The Interconnection of
Chemical and Electronic Structure, Proceedings of Nobel Symposium
No. NS-81; Oxford University Press: 1993; p 475. Chim. Ind. (Milan)-
1993, 75, 185. (b) Zhang, J.; Zhou, P.; Brinckerhoff, W. B.; Epstein,
A. J.; Vazquez, C.; McLean, R. S.; Miller, J. S. ACS Symp. Ser. 1996,
644, 311.
(4) (a) Banister, A. J.; Bricklebank, N.; Lavender, I.; Rawson, J. M.;
Gregory, C. I.; Tanner, B. K.; Clegg, W.; Elsegood, M. R. J.; Palacio,
F. Angew. Chem., Int. Ed. 1996, 35, 2533. (b) Banister, A. J.;
Bricklebank, N.; Clegg, W.; Elsegood, M. R. J.; Gregory, C. I.;
Lavender, I.; Rawson, J. M.; Tanner, B. K. J. Chem. Soc., Chem.
Commun. 1995, 679.
(2) (a) Miller, J. S.; Calabrese, J. C.; Epstein, A. J.; Bigelow, R. W.; Zhang,
J. H.; Reiff, W. M. J. Chem. Soc., Chem. Commun. 1986, 1026. (b)
Miller, J. S.; Calabrese, J. C.; Rommelmann, H.; Chittipeddi, S. R.;
Zhang, J. H.; Reiff, W. M.; Epstein, A. J. J. Am. Chem. Soc. 1987,
109, 769. Chittipeddi, S.; Cromack, K. R.; Miller, J. S.; Epstein, A. J.
Phys. ReV. Lett. 1987, 58, 2695.
(5) Carlin, R. L. Magnetochemistry; Springer-Verlag: New York, 1986;
p 82.
(6) Bleaney, B.; Bowers, K. D. Proc. R. Soc. London, Ser. A 1952, 214,
451.
10.1021/ic0007106 CCC: $19.00 © 2000 American Chemical Society
Published on Web 09/28/2000

Copper(II) Benzoate Nitroxide Dimers and Chains
Inorganic Chemistry, Vol. 39, No. 21, 2000 4895
intradimer coupling (with JST < 0 indicating antiferromagnetic
interactions, and therefore the triplet state is higher in energy
than the singlet state). These properties are also observed for
axially ligated [CuII(O2CR)2]2 dimers, that is, [CuII(O2CR)2]2L2
(L ) H2O, pyridine, 2-picoline).7 In contrast, isostructural
[MII(O2CR)2]2 (M ) Cr, Mo, Rh) dimers with their respective
single and quadruple bonds exhibit strong antiferromagnetic
coupling8 and are diamagnetic at room temperature due to
complete population of the S ) 0 ground state, whereas [RuII(O2-
and the isolated monomer 7 were unexpectedly isolated and
their characterization is reported.
+
CR)2]2 is S ) 1 (σ2π4δ2π*3δ*1)9 and [RuII/III(O2CR)2]2 has a
3
S ) /2 (σ2π4δ2δ*2π*1) ground state.10
Previous attempts to develop high-spin molecules via coupling
to the thermally populated copper-dimer triplet with an S ) 1/2
nitroxide [e.g., TEMPO (2,2,5,5-tetramethylpiperidinyl-1-oxyl,
5)11] were unsuccessful as strong antiferromagnetic interac-
tions occur between the spins, leading to a net moment near
zero.
Herein, we postulate that by (1) introducing the organic
radicals into the equatorial position of a copper dimer as a
substituent of the carboxylate bridge, that is, [CuII(O2CR)2]2L2
(R not L being spin bearing), or (2) weakening the strong Cu-
(II)-L (where L is spin bearing) antiferromagnetic coupling
by increasing the distance of the spin-containing portion on L
from the Cu(II), it may be possible that (a) isolated organic
radicals as well as the singlet-triplet transition of the copper
dimer are maintained and (b) these spins may ferromagnetically
2
2
couple if the Cu(II) magnetic orbital (dx -y ) were orthogonal
to those of R.12 Furthermore, the extent of the perturbation of
the superexchange mechanism through the carboxylate bridges
may be studied if the organic spins are delocalized onto the
carboxylate bridge.
To test these hypotheses, we targeted the preparation and
study of [CuII(3a)2L]2 [3a ) 2-(4′-carboxyphenyl)-4,4,5,5-
tetramethylimidazoline-3-oxide-1-oxyl, NNBA] and [CuII(O2-
CMe)23b]2 [3b ) 2-(4′-pyridyl)-4,4,5,5-tetramethylimidazoline-
3-oxide-1-oxyl, PYNN]. Because of the lack of direct coordination
between the nitroxide spin center (i.e., O) and Cu(II), strong
antiferromagnetic coupling between the nitroxide to the Cu(II)
is prevented. Additionally, using only 1 equiv of 3b should result
in a linear chain linked by bridging 3b as previously reported
for [RuII/III(O2CCMe3)2]2(L),13a [L ) PYNN (3b) or 2-phenyl-
,4,4,5,5-tetramethylimidazoline-4,5-dihydro-1H-imidazolyl-1-
oxy, 4b13c], and [RhII(O2CCF3)2]2(L) (L ) 4a or 2,4,4,5,5-
pentamethylimidazoline-4,5-dihydro-1H-imidazolyl-1-oxy 3-oxide,
4c).14a In addition to these new compounds, chain compound 6
Experimental Section
Synthesis. All chemicals needed in the preparations of the studied
samples were used as received. An infrared Bio-Rad FTS-40 FTIR
spectrophotometer with (1 cm-1 resolution was used. Samples were
prepared either as Nujol mulls on NaCl plates or as KBr pressed pellets
and were scanned in the wavenumber range of 400-4000 cm-1
.
Thermal properties and mass spectrometry were performed with a
TA Instruments model 2050 thermal gravimetric analyzer (TGA), with
the gas outlet connected to a TA Thermolab mass spectrometer, with
ionization potential of 70 eV. Elemental analyses (C, H, and N) were
carried out by Atlantic Microlabs (Norwalk, GA).
Magnetic susceptibility measurements were performed between 2
and 300 K at a field of 1000 Oe on a Quantum Design MPMS-5T
SQUID magnetometer with a sensitivity of 10-8 emu as previously
discussed.15 Electron paramagnetic resonance (EPR) spectra were
recorded using a Bruker EMX X-band spectrometer with 1,1-diphenyl-
2-picrylhydrazyl (DPPH, Sigma) as an external standard (g ) 2.0037).
1H NMR measurements were performed on a Varian XL-300 with
proton frequency at 300 MHz.
X-ray diffraction studies were made on a Nonius KappaCCD
diffractometer equipped with Mo KR (λ ) 0.710 73 Å) radiation.
Structures of 1a, 6a, and 7 (Table 1) were solved by direct methods
using the program SIR-97 and were refined by the full-matrix least-
squares method on F2 with SHELXL 97.16 Anisotropic thermal
parameters were assigned to Cu, C, N, and O atoms, and hydrogen
atoms and their positions were assigned using the Riding model. ORTEP
(7) (a) Harada, A.; Tsuchimoto, M.; Ohba, S.; Iwasawa, K.; Tokii, T.
Acta Crystallogr., Sect. B 1997, 53, 654. (b) Mei, W.; Yu, X.; Na,
Y.; Man, J. Indian J. Chem. 1997, 36A, 887. (c) Meln´ık, M. J. Inorg.
Nucl. Chem. 1979, 41, 779.
(8) Boudreaux, E. A.; Mulay, L. N. Theory and Applications of Molecular
Paramagnetism; John Wiley and Sons: New York, 1976; p 406.
(9) (a) Lindsay, A. J.; Wilkinson, G.; Motevalli, M.; Hursthouse, M. B.
J. Chem. Soc., Dalton Trans. 1985, 2321. (b) Lindsay, A. J.; Wilkinson,
G.; Motevalli, M.; Hursthouse, M. B. J. Chem. Soc., Dalton Trans.
1987, 2723. (c) Cogne, A.; Belorizky, E.; Laugier, J.; Rey, P. Inorg.
Chem. 1994, 33, 3364.
(10) Bino, A.; Cotton, F. A.; Felthouse, T. R. Inorg. Chem. 1979, 18, 2599.
(11) (a) Porter, L.; Dickman, M.; Doedens, R. Inorg. Chem. 1983, 22, 1962.
(b) Porter, L.; Doedens, R. Inorg. Chem. 1985, 24, 1006. (c) Porter,
L.; Dickman, M.; Doedens, R. Inorg. Chem. 1986, 25, 678. (d) Porter,
L. Ph.D. Thesis, University of California, Irvine, CA, 1984.
(12) Luneau, D.; Rey, P.; Laugier, J.; Fries, P.; Caneschi, A.; Gatteschi,
D.; Sessoli, R. J. Am. Chem. Soc. 1991, 113, 1245.
(13) (a) Sayama, Y.; Handa, M.; Mikuriya, M.; Hiromitsu, I.; Kasuga, K.
Chem. Lett. 1998, 777. (b) Sayama, Y.; Handa, M.; Mikuriya, M.;
Hiromitsu, I.; Kasuga, K. Chem. Lett. 1999, 453. (c) Handa, M.;
Sayama, Y.; Mikuriya, M.; Hukada, R.; Hiromitsu, I.; Kasuga, K. Bull.
Chem. Soc. Jpn. 1998, 71, 119.
(14) (a) Cogne, A.; Grand, A.; Rey, P.; Subra, R. J. Am. Chem. Soc. 1989,
111, 3230. (b) Cogne, A.; Grand, A.; Rey, P.; Subra, R. J. Am. Chem.
Soc. 1987, 109, 7927.
(15) Brandon, E. J.; Rittenberg, D. K.; Arif, A. M.; Miller, J. S. Inorg.
Chem. 1998, 37, 3376.
(16) (a) Sheldrick, G. M. SHELX97-2 Programs for Crystal Structure
Analysis, (release 97-2); Institu¨t fu¨r Anorganische Chemie der Uni-
versita¨t, Go¨ttingen, Germany, 1998. (b) Farrugia, L. J. J. Appl.
Crystallogr. 1997, 30, 565.

4896 Inorganic Chemistry, Vol. 39, No. 21, 2000
Del Sesto et al.
Table 1. Crystallographic Data Collection and Structure Refinement Parameters for [Cu(O2CPh)2(DMF)]2 (1a), [Cu(O2CC6H4Br)2Py2]‚H2O
(6a), and [Cu(O2CPh)2Py2] (7)
1a
6a
7
chemical formula
formula mass, Da
space group
a, Å
b, Å
c, Å
C34H34N2O10Cu2
757.71
P21/n
10.5055(2)
10.0650(4)
16.2460(5)
90
93.764(2)
90
C24H20N2O5Br2Cu
639.78
C2
15.8610(9)
5.8690(a)
14.9720(8)
90
121.3811(15)
90
C24H18N2O5Br2Cu
621.76
P1h
5.9662(2)
8.1860(3)
12.8681(5)
94.1200(19)
99.512(2)
993.944(2)
607.25(4)
1.700
R, deg
â, deg
γ, deg
V, Å3
1714.11(9)
1.468
1189.85(10)
1.786
F
calcd, mg/m3
T, K
λ, Å
200(0.1)
0.71073
0.0258
0.0720
2
200(0.1)
0.71073
0.0221
0.0582
2
200(0.1)
0.71073
0.0770
0.2413
1
R1(Fo )a )
2
wR2(Fo )b )
2
Z
µ, mm-1
1.299
4.315
4.222
a R1 ) ∑(||Fo| - |Fc||)/∑|Fo|. b wR2 ) [∑(w(Fo - Fc2)2)/∑(Fo )2]1/2
.
2
2
diagrams were generated using ORTEP 3.16 Disorder was not observed
in any of the structures.
and dried in vacuo. The product was dissolved in warm dimethylfor-
mamide (DMF) and allowed to cool slowly to 0 °C, upon which time
green-blue crystals formed. IR ν (Nujol): 1757m, 1719s, 1617m,
2-(4′-Carboxyphenyl)-4,4,5,5-tetramethyl-1,3-dihydroxyimidazo-
line (NNBAH).17 To a solution of 2,3-bis(hydroxyamine)-2,3-dimeth-
ylbutane (ACROS) (2.42 g, 0.010 mol) and sodium bicarbonate (0.020
mol) in water (30 mL) was added 4-carboxybenzaldehyde (Aldrich)
(1.48 g, 0.010 mol) in small portions at 0 °C. The solution was allowed
to stir for 48 h at room temperature. The resulting white precipitate
was filtered, washed with cold water, and then dried in vacuo over
P2O5 (yield: 2.04 g, 68%). IR, ν (Nujol): 3402m, 3181m, 1623m,
1592m, 1439s, 1260s, 1173s cm-1. 1H NMR (300 MHz, DMSO): 7.81
(s, 3H), 7.55-7.98 (dd, 4H), 4.60 (s, 1H), 1.09 (2s, 12H).
2-(4′-Carboxyphenyl)-4,4,5,5-tetramethylimidazoline-3-oxide-1-
oxyl (NNBA), 3a.17 To a suspension of NNBAH (510 mg, 1.68 mmol)
in water (35 mL), a solution of sodium periodate (Aldrich) (600 mg,
2.80 mmol) in water (10 mL) was added, and the reaction mixture was
allowed to stir at room temperature for 4 h. The deep blue precipitate
was filtered out, resuspended in water, and extracted with ethyl acetate
(4 × 75 mL). The ethyl acetate solution was then washed with water
(4 × 50 mL), dried with magnesium sulfate, and then removed under
reduced pressure. The resulting purple powder was then dried in vacuo
over P2O5 (yield: 225 mg, 49%). IR ν (Nujol): 2928m, 1702w, 1355s
(NO), 1313m, 1260m, 1171m, 1018w cm-1. EPR (ethyl acetate, DPPH
ref): g(NO) ) 2.0013, A(14N) ) 0.79 mT.
2-(4′-Pyridyl)-4,4,5,5-tetramethyl-1,3-dihydoxyimidazoline.17 To
a solution of 2,3-bis(hydroxyamine)-2,3-dimethylbutane (2.49 g, 11
mmol) and sodium bicarbonate (1.80 g, 22 mmol) in water (30 mL)
was added 4-pyridinecarboxaldehyde (Aldrich) (1.1 mL, 1.122 g/mL,
0.012 mol) in small portions at 0 °C. The solution was allowed to stir
for 40 h at room temperature. The resulting white precipitate was
filtered, washed with cold water, and then dried in vacuo over P2O5
(yield: 1.05 g, 42%). 1H NMR (300 MHz, DMSO): 7.95 (s, 2H), 7.45-
8.52 (dd, 4H), 4.50 (s, 1H), 1.05 (2s, 12H).
1590m, 1252m, 1112w cm-1
.
[Cu2(NNBA)3(O2CMe)(EtOH)2], 1b. [Cu(O2CMe)2]2‚2H2O (183
mg, 0.92 mmol) was dissolved in 100% ethanol (40 mL). A solu-
tion of p-NNBAH (550 mg, 1.98 mmol) in ethanol (75 mL) was added
to the copper solution, and the mixture was stirred at room tempera-
ture for 20 h. The resulting deep blue precipitate was filtered out and
dried in vacuo (yield: 530 mg, 89%). IR ν (Nujol): 3434w, 1676m,
1618s, 1560m, 1360s, 1308m, 1229m, 1171s cm-1. Anal. Calcd for
C48H64N6O16Cu2: C, 52.03; H, 5.71; N, 7.58%. Found: C, 51.74; H,
5.78; N, 7.81%.
[Cu2(NNBA)3(O2CMe)py2], 1c. To the ethanol-capped dimer, 1b,
suspended in methanol was added exactly 2 equiv of pyridine, and the
mixture was allowed to stir for 10 min. The solvent was removed under
reduced pressure until a precipitate appeared, resulting in deep blue
powder. IR ν (Nujol): 1681w, 1623m, 1560m, 1366s, 1302w, 1224m,
1171m cm-1. Anal. Calcd for C54H64N8O14Cu2: C, 55.10; H, 5.44; N,
9.55%. Found: C, 54.71; H, 5.78; N, 9.91%.
[Cu(O2CMe)2(PYNN)]2, 1d. To [Cu(O2CMe)2]2‚2H2O (85 mg, 0.22
mmol) in ethanol (10 mL) was added 2 equiv of p-PYNN (100 mg,
0.44 mmol) per dimer, and the mixture was stirred for 1 h at room
temperature. The resulting deep blue powder was filtered and dried in
vacuo (yield: 75 mg, 59%). IR ν (Nujol): 1623s, 1600s, 1550m, 1345s,
1306m, 1213m, 1166m cm-1. Anal. Calcd for C32H44N6O12Cu2: C,
46.21; H, 5.37; N, 10.10%. Found: C, 46.17; H, 5.51; N, 9.89%.
[Cu(O2CMe)2(PYNN)]2, 2.13 To [Cu(O2CMe)2]2‚2H2O (85 mg, 0.22
mmol) dissolved in ethanol (10 mL) was added 1 equiv of p-PYNN
(50 mg, 0.22 mmol) per dimer, and the mixture was stirred for 1 h at
room temperature. The resulting deep blue powder was filtered and
dried in vacuo. IR ν (Nujol): 1631s, 1592m, 1535m, 1325m, 1201w,
1166w cm-1
.
2-(4′-Pyridyl)-4,4,5,5-tetramethylimidazoline-3-oxide-1-oxyl(PYNN),
3b.17 The oxidation of the pyridine analogue was carried out in the
exact same manner as for NNBAH (yield: 242 mg, 54%). IR ν
(Nujol): 2946m, 1625s, 1427m, 1361s, 1290m, 1010m cm-1. EPR
(ethyl acetate, DPPH ref): g(NO) ) 2.0014.
[Cu(O2CPh)2(DMF)] 2, 1a. To a solution of [Cu(O2CMe)2]2‚2H2O
(Aldrich) (4.0 g, 20 mmol) in warm mixed xylenes (100 mL, 45 °C)
was added benzoic acid (6.1 g, 50 mmol). The solution was allowed
to reflux (118 °C, bp for acetic acid) for 30 min, followed by heating
at 130 °C for 4 h. After this time, the solution was allowed to cool to
room temperature, and the resulting light blue precipitate was filtered
[Cu(O2CC6H4Br)2py2].H2O, 6a, and [Cu(O2CC6H4Br)2py2], 7.
These compounds were made according to the same procedure for the
copper benzoate, using 4-bromobenzoic acid (Aldrich) instead of
benzoic acid. Also, the products were crystallized from a mixture of
DMF/pyridine (5:1), because the initial crude product would not
dissolve in DMF alone. Both compounds cocrystallize from this solvent.
IR ν (Nujol) 6a: 3076s, 3018m, 1676m, 1602s, 1492s, 1118m, 1050s,
1024m cm-1. The limited amount of 7 crystallized from solution was
insufficient to carry out any characterization other than X-ray crystal
structure. The crystals were separated by hand under a microscope in
order to determine its structure.
[Cu(NNBA)2py2]‚H2O, 6b. To the ethanol-capped dimer, 1b,
suspended in methanol was added exactly 2 equiv of pyridine, and the
(17) Ullman, E.; Osiecki, J.; Boocock, D.; Darcy, R. J. Am. Chem. Soc.
1972, 94, 7049.

Copper(II) Benzoate Nitroxide Dimers and Chains
Inorganic Chemistry, Vol. 39, No. 21, 2000 4897
Table 2. Summary of TGA-MS Fragment Loss
decomposition
mixture was allowed to stir for 12 h. The solvent was removed under
reduced pressure until a precipitate appeared, resulting in deep blue
powder. IR ν (Nujol): 3355w, 1602m, 1355s, 1302m, 1218w, 1172m
cm-1. Anal. Calcd for C38H46N6O9Cu: C, 57.46; H, 5.84; N, 10.58%.
Found: C, 56.52; H, 5.58; N, 10.41%.
fragments lost,
mass (amu)
compound
T (°C)
Cu2(O2CMe)4(H2O)2
1a
1b
300
295
200
290
270
215
300
44, 60
44, 58, 76
46, 60
44, 76, 103, 104
44, 60, 76, 79, 103, 104
44, 58, 60
X-ray Crystal Structure Analysis. Crystallographic data (exclud-
ing structure factors) for the structure reported herein have been
deposited with the Cambridge Crystallographic Data Centre as supple-
mentary publication no. CCDC-143665 (1a), CCDC-143666 (6a),
and CCDC-143667 (7). Copies of the data can be obtained free of
charge on application to The Director, CCDC, 12 Union Road,
Cambridge CB2 1EZ, U.K. (fax: int code + (1223) 336 033; e-mail:
deposit@ccdc.cam.ac.uk).
1c
3a
60, 76, 103, 104
(60 amu). A second decomposition at 290 °C, when the NNBA
ligand began to decompose, was also observed (Table 2). For
comparison, [Cu(O2CMe)2(OH2)]2 begins to decompose at ∼300
°C, with the fragments lost containing mass values of 44 (carbon
dioxide) and 60 amu (acetic acid). NNBAH shows decomposi-
tion at ∼215 °C, with formation of fragments with mass values
of 44 (CO2) and 60 amu (HOAc), and a second loss of mass at
300 °C with mass values of 44, 76, 103, and 104 amu. These
peaks correspond to loss of carbon dioxide, C6H4, C6H4CdNH,
and C6H4CdNH2. The reaction of 1b in methanol with exactly
2 equiv of pyridine resulted in the pyridine-capped dimer 1c.
The TGA-MS, showing peaks for pyridine, acetate, and NNBA
fragments, and the magnetic properties (vide infra) support this
structure. However, if the solution is allowed to stir for 12 h, a
product similar to that of the linear chain of the bromobenzoate
monomers is believed to result, in which two NNBA ligands,
two pyridines, and one water molecule create the square-
pyramidal environment around the copper center, 6b. Elemental
analysis agrees with this monomeric form, and IR shows that
Results and Discussion
Synthesis. (a) [Cu(O2CC6H4X)2L]2. The model compound
[Cu(O2CPh)2]2 was synthesized by refluxing [Cu(O2CMe)2-
(OH2)]2 with benzoic acid at 118 °C in xylenes, boiling off acetic
acid as the reaction progressed. This synthesis was developed
in order to drive the reaction in favor of the products, as material
of reasonable purity in practical yields could not be obtained
using known syntheses.18 Typical syntheses of the dimers
involve the reaction of Cu(II) salts with the corresponding
carboxylic acid and a coordinating ligand (e.g., py); however,
high-purity materials were difficult to obtain from these
reactions.
The reaction in refluxing xylenes forms the linear chain of
copper monomers similar to those previously reported,19 which
upon dissolution in DMF forms the DMF-capped dimers, 1a.
In contrast, using 4-bromobenzoic acid the initial linear chain
product did not redissolve in DMF but did dissolve in 5:1 DMF/
pyridine solution. Upon standing at room temperature, two
different products cocrystallized: (1) deep-blue crystals, 6a, of
composition [Cu(O2CC6H4Br)2py2(OH2)] and (2) a purple
crystalline coproduct, 7, of composition [Cu(O2CC6H4Br)2py2],
both of which were structurally determined.
the nitroxide is still present with the νNO stretch at 1355 cm-1
.
The magnetic data also show ferromagnetic coupling similar
to that of 6a (vide infra).
(c) [Cu(O2CMe)2]2(PYNN)x. In the attempts to place the
nitroxide radical distant from the Cu(II) centers in the apical
position, [Cu(O2CMe)2]2‚2H2O was dissolved in ethanol and 2
equiv of PYNN per copper dimer was added, expecting the
pyridyl N of the PYNN to be the strongest donor atom and
thus to displace ethanol and water. This resulted in discrete bis-
PYNN-capped copper acetate dimers 1d, as suggested by
elemental analysis, the presence of the nitroxide stretch in the
IR, and magnetic properties. In contrast, reaction of only 1 equiv
of PYNN with the [Cu(O2CMe)2]2 results in the formation of
[Cu(O2CMe)2]2 (PYNN), which is assigned to be a chain of
copper dimers linked by the bridging PYNN ligands, as
(b) [Cu2(O2CMe)(NNBA)3(EtOH)2]. To prevent decomposi-
tion of the radical-based ligands at elevated temperatures, [Cu-
(O2CMe)2(OH2)]2 was stirred with the NNBAH in absolute
ethanol for 1 day at room temperature. The deep blue color of
the precipitate was consistent with a dimeric structure that was
confirmed from its magnetic behavior (vide infra). On the basis
of infrared spectra, the nitroxide was still present, as evidenced
17
by the 1360 cm-1 peak assigned to the νNO
.
The expected
percent composition predicted for the anticipated [Cu(NNBA)2-
(EtOH)]2 (C, 54.41; H, 5.29; N, 8.46%) was at variance with
the observed values (C, 51.73; H, 5.82; N, 7.81%), and
quantitative agreement of the magnetic data could not be
obtained. Hence, this expected composition was deemed incor-
rect, and using a computer analysis20 the composition was
determined to be [Cu2(O2CMe)(NNBA)3(EtOH)2] (1b) (C,
52.07; H, 5.69; N, 7.59%). This unexpected composition also
enabled the quantitative fitting of the magnetic data (vide infra).
TGA-MS also supports the presence of both acetate and NNBA
ligands in 1b. In the thermal analysis of 1b, decomposition is
seen at ∼200 °C, with loss of both ethanol (46 amu) and acetate
evidenced by the shift of the νNO bands from 1345 to 1325 cm-1
.
Hence, the copper dimers could be bridged by a PYNN with a
pyridyl N and nitroxide O bound to each dimer forming a
uniform chain, 2a. This motif has been reported for [RuII/III
-
(O2CCMe3)2]2(PYNN)13 and [RhII(O2CMe3)2]2(L) (L ) 4c).14
Alternatively, a chain can form with one copper dimeric unit
axially coordinated to two pyridyl moieties while the adjacent
dimeric unit is axially coordinated by two nitroxide groups,
producing a nonuniform chain of dimers, 2b.21 This motif has
been very recently reported for [CuII(O2CCMe3)2]2(m-PYNN)21a
as well as [RuII/III(O2CCMe3)2]2(L) (L ) 4a,13b 522). Either 2a
(18) (a) Bateman, W.; Conrad, D. J. Am. Chem. Soc. 1915, 37, 2553. (b)
Lewis, J.; Lin, Y.; Royston, L.; Thompson, R. J. Chem. Soc. 1965,
6464. (c) van Niekerk, J.; Schoening, F. Acta Crystallogr. 1953, 6,
227.
(19) (a) Deakin, L.; Arif, A. M.; Miller, J. S. Inorg. Chem. 1998, 38, 5072.
(b) Bakalbassis, E.; Bergerat, P.; Kahn, O.; Jeannin, S.; Jeannin, Y.;
Dromzee, Y.; Guillot, M. Inorg. Chem. 1992, 31, 625.
(20) Miller, J. S.; Goedde, A. O. J. Chem. Educ. 1975, 50, 43; Quant.
Chem. Prog. Exch. 1976, 10, 299.
(21) (a) Chung, Y.-H.; Wei, H.-H. Inorg. Chem. Commun. 1999, 2, 269.
(b) The authors report attempting to fit their susceptibility data using
a nonuniform chain, resulting in a poor fit. However, we believe that
this model would be inappropriate for our compound 2 as the
nonuniform chain model is for interacting (S ) 1/2)-(S ) 1/2) systems,
which is not consistent with the (S ) 1/2 (nitroxide))-(S ) 1) copper
dimer systems we are studying.
(22) Handa, M.; Sayama, Y.; Mikuriya, M.; Hiromitsu, I.; Kasuga, K. Bull.
Chem. Soc. Jpn. 1995, 68, 1647.

4898 Inorganic Chemistry, Vol. 39, No. 21, 2000
Del Sesto et al.
Figure 1. ORTEP (50% electron density) and atom labeling diagram
for [Cu(O2CPh)2(DMF)]2, 1a, with a Cu‚‚‚Cu distance of 2.63 Å.
or 2b results in direct Cu(II)-nitroxide interactions and thus a
large antiferromagnetic coupling (vide infra).
Figure 2. (a) ORTEP (50% electron density) and atom labeling
diagram for [Cu(O2CPh)2py2]‚H2O, 6a; (b) looking approximately down
the b axis of 6a.
bromobenzoate analogue of 1a, two other products cocrystal-
lized from the solution. The first was determined to be a linear
chain of psuedo-square-pyramidal copper centers, with trans-
benzoate and trans-pyridine ligands and one water molecule,
6a (Figures 2and 3). The CuO, CuO3H O, and CuN distances
2
are 1.94, 2.20, and 1.98 Å, respectively, and the O3-Cu-N
and O3-Cu-O1 angles are 98.9 and 87.5°, respectively. The
water hydrogen bonds (1.98 Å) between the water of one
monomer and the carboxylate of the adjacent monomer form
linear chains (Figure 3). The second bromobenzoate product
was an anhydrous pseudo-octahedral [CuII(O2CC6H4Br)2py2]
complex (7). There is a center of inversion about the Cu(II)
(Figure 4), with two bromobenzoate ligands creating a plane
about the copper center. The oxygens of the carboxylates are
asymmetrically coordinated to the copper center, with distances
from the copper of 1.97 (Cu-O1) and 2.55 Å (Cu-O2).
Pyridine occupies both apical positions, with the Cu-N distance
of 2.01 Å.
Structures. [Cu(O2CC6H4X)2L]2. The structure of the [Cu-
(O2CPh)2(DMF)]2 (1a) model compound was determined. The
intradimer Cu(II)‚‚‚Cu(II) distance is 2.63 Å, which is very
similar to 2.64 Å in copper(II) acetate.18 The angle of the plane
of the benzene ring to that of the copper-carboxylate plane is
twisted by 35° (Figure 1). In the attempt to produce the
Magnetic Properties. (a) [Cu(O2CC6H4Br)2py2]‚H2O. The
susceptibility, ø, of the ligands 3a and 3b as well as those of
the monomeric copper(II) complexes can be modeled by the
1
Curie-Weiss expression, eq 2. The S ) /2 nitroxide ligands
are magnetically well behaved with Weiss constants, θ, of -3.5

Copper(II) Benzoate Nitroxide Dimers and Chains
Inorganic Chemistry, Vol. 39, No. 21, 2000 4899
Figure 3. Hydrogen bonding structure of [Cu(O2CPh)2py2]‚H2O, 6a.
Figure 5. øT(T) and 1/ø(T) for (a) [Cu(O2CPh)2py2]‚H2O, 6a, with
gCu ) 2.16 and θ ) +9 K and (b) [Cu(NNBA)2py2]‚H2O, 6b, with gCu
) 2.16, gorg ) 2.001, and θ ) +0.7 K.
Figure 4. ORTEP (50% electron density) and atom labeling diagram
for [Cu(O2CPh)2py2] (7).
( 0.3 K and gorg ) 2.00, characteristic of weak antiferromag-
netic coupling between radicals in the solid state.
Ng2â2S(S + 1)
3k(T - θ)
Figure 6. øT(T) (n) and ø(T) (B) for [Cu(O2CPh)2(DMF)]2, 1a. Fit
to eq 3 with JST ) -206 K (-143 cm-1), gCu ) 2.16, and F ) 0.05.
øCW
)
(2)
(-148 cm-1), gCu ) 2.09, and F ) 0.0085)].23,25
[Cu(O2CC6H4Br)2py2]‚H2O. The Cu(II) complexes, 6a and
6b, obey the Curie-Weiss expression above 30 K with gCu
2.16 and θ ) +9 and 0.7 K, respectively (Figure 5). The
relatively high θ value for 6a suggests significant ferromagnetic
coupling arising from the 1-D chains, and this is the subject of
further study.
2Ng2â2
kT[3 + e-2J
Ng2â2
2kT
)
ø )
(1 - F) +
F
(3)
ST/kT
]
[Cu2(O2CMe)(NNBA)3L2], 1b, 1c. By modification of the
substituents of the copper benzoate dimers with organic radicals,
for example, NNBA (3a), the Bleaney-Bowers equation, eq
1, is not sufficient to account for the introduction of the organic
radicals onto the dimer. The data can be fit to a summation of
Curie-Weiss equation (2) for each quantum spin (due to the
organic radicals, eq 2), and the Bleaney-Bowers equation (1)
for the copper singlet-triplet dimer, resulting in eq 4, where n
is the number of organic radicals per dimer and JST is the
intradimer spin coupling between the singlet-triplet states of
the copper dimer. The e-2J/kT term accounts for the population
of the triplet state. The middle term of eq 4 reflects the
intermolecular coupling, θS, between the S ) 1/2 organic radicals
via either the singlet state of the copper dimer, which is
[Cu(O2CC6H4X)2L]2, 1a. The magnetic susceptibility typical
for copper dimers is exhibited by 1a, where the triplet state is
increasingly thermally populated with increasing temperature
with ø(T) reaching a maximum around 230 K (Figure 6). The
large increase in ø at very low temperatures is attributed to a
small fraction (F) of the monomeric S ) 1/2 Cu(II) spin impurity
being present,23 and the total ø can be modeled by adding in a
Curie component (eq 2) for that fraction of the impurity to eq
1, resulting in eq 3. Using H ) -2JSTSa‚Sb, a fit to the observed
data for 1a has JST ) -206 K (-143 cm-1),24 gCu ) 2.16, and
F ) 0.05, which is comparable to [Cu(OAc)2]2 [JST ) -213 K
(23) Kahn, O. Molecular Magnetism; VCH Publishers: New York, 1993;
p 107.
(25) Felthouse, T. R.; Dong, T.-Y.; Hendrickson, D. N.; Shieh, H.-S.;
(24) J values are reported in cm-1 or as J/k in units of K.
Thompson, M. R. J. Am. Chem. Soc. 1986, 108, 8201.

4900 Inorganic Chemistry, Vol. 39, No. 21, 2000
Table 3. Summary of Magnetic Data
Del Sesto et al.
θT, K (cm-1
) eq
complex
gCu
gorg
JST, K (cm-1
)
θS, K (cm-1
)
[CuII(O2CMe)2]2‚2H2O 23‚25
2.16
2.16
2.18
2.18
2.18
2.18
-213 (-148)
-206 (-143)
-216 (-150)
-216 (-150)
-223 (-155)
-238 (-165)
3
3
4
4
4
4
2
2
2
2
1a
1b
2.001
2.001
2.001
2.001
2.0013
2.0014
-1.5 (-1.0)
-1.8 (-1.3)
-5.8 (-4.0)
-5.9 (-4.1)
-3.8 (-2.7)
-3.2 (-2.2)
+9 (+6.4)
-157 (-109)
-169 (-117)
-5.8 (-4.0)
-300 (-205)
1c
1d
2
3a
3b
6a
2.16
2.16
6b
2.001
+0.7 (+0.5)
[CuII(O2CCCl3)2]2(5)11b c
[CuII(O2CCCl3)2]2(m-3b)11b c
very largea
2.00
2.00
2.00
2.23
2.117
2.00
2.20
2.00
2.00
2.00
2.00
-253 (-176)
-2.0 (-1.4)
3
[RuII(O2CCF3)2(5)]2
-379 (-263)
-337 (-234)
14 (20)
ref 9c
9c
9c
[RuII(O2CC6F5)2(5)]2
ref 9c
[RuII/III(O2CCMe3)2]2(3b)13a
[RuII/III(O2CCMe3)2(5)]21
[RuII/III(O2CCMe3)2]2(4b)13c c
[RuII/III(O2CCMe3)2(4a)]13b
-0.3 (0.45)
b
-187 (-130)
-144 (-100)
14 (20)
ref 26
2.00
ref 13c
2.00
-35 (-50)
-0.2 (-0.14)
-344 (-239)
-387 (-269)
-265 (-184)
-120 (-83.6)
-142 (-98.8)
-11.8 (-5.2)
3.3 (2.3)
b
2
1
1
1
1
[MoII(O2CCF3)2(5)]2
2.00
very largea
25
25
[RhII(O2CCF3)2(5)]2
2.00
[RhII(O2CC3F7)2(5)]2
2.00
25
25
[RhII(O2CC6F5)2(5)]2
2.00
[RhII(O2CCF3)2(4b)]2
1.998
2.114
1.976
2.122
14a
[RhII(O2CCF3)2]2(4a)14a c
[RhII(O2CCF3)2(4c)]2
1
14a
[RhII(O2CCF3)2]2(4c)14a c
ref 14a
a Diamagnetic. b Unknown. c Extended structure similar to that of 2.
essentially completely populated at low temperature, or through-
space either inter- or intramolecularly. The right-hand term of
eq 4 reflects the intramolecular coupling, θT, between the
organic radicals and the triplet state of the copper dimer, which
is appreciably populated at higher temperatures (e.g., ∼20% at
room temperature). Both θT and θS along with JST are
determined from a fit of the data to eq 4.
2NgCu2â2
Ngorg2â2S(S + 1)
3kT(T - θS)
Ngorg2â2S(S + 1)
ø )
+ n 1 - e(-2J/kT)
+
ST/kT)
kT[3 + e(-2J
]
[
(26) A separate model which the authors unsuccessfully attempted to fit
1
1
e(-2J/kT)
(4)
the data was for an (S ) /2)-(S ) 1)-(S ) /2) trimer, where the S
1
) 1 is the copper dimer at room temperature and the two S )
spins are for the axially bound nitroxides. For the [Cu2(O2CMe)4-
/
2
]
3kT(T - θT)
(PYNN)]2 system, 1d, the following model was attempted:
øtotal ) øBB + [(1 - e-2(J /kT))øj ] + [(e-2(J /kT))øJ j ] +
ST
ST
When three NNBA radicals are positioned equatorially around
the copper dimer in 1b, the susceptibility can be fit between 2
and 300 K with eq 4 with n ) 3, gCu ) 2.18, gorg ) 2.001, JST
) -216 K (-150 cm-1), θS ) -1.5 K, and θT ) -157 K
(Table 3, Figure 7a). Only unacceptably poor fits are obtained
for n ) 2 or 4, consistent with the formulation of the
composition of 1b as [Cu2(O2CMe)(NNBA)3(EtOH)2] (vide
supra). The intradimer Cu-Cu coupling (JST) of -150 cm-1
shows little perturbation from the presence of the nitroxide, as
1
2 2
[(e-2(J /kT))øj ]
ST
3
where JST is the copper-copper intradimer interactions as defined by
the Bleaney-Bowers equation (1). The exp(-2JST/kT) represents the
population of the singlet and triplet states. Also,
(2e-2(J /kT) + 2e-2(j /kT) + 10e2(J /kT))Nâ2g2
2
2
2
øJ j
)
2
2
(e-4(J /kT) + 3e-2(J /kT) + 3e-2(j /kT) + 5e2(J /kT))kT
2
2
2
2
where J2 is the intradimer copper-nitroxide interactions, which can
only occur at the higher temperatures as the triplet state of the copper
dimer is populated; thus the exp(-2JST/kT) weighting factor is needed.
The j2 represents the intradimer nitroxide-nitroxide interactions, which
should be ferromagnetic if each nitroxide is antiferromagnetically
coupled to the triplet copper dimer. At the lower temperatures, where
the singlet state of the copper dimer is most populated, J2 f 0 as the
copper-nitroxide interactions no longer exist. Thus,
J
ST is essentially the same as that of 1a (-143 cm-1). The small
value of the nitroxide-nitroxide spin coupling (θS ≈ -1.5 K)
is attributed to either weak intermolecular nitroxide-nitroxide
interactions or weak intramolecular nitroxide-nitroxide coupling
via the singlet state of the copper dimer and is typical of
nitroxides, for example, 3a and 3b. However, poor intradimer
nitroxide coupling is expected because the pathway necessary
for interaction would be through the copper dimer in the singlet
state. Also, similar values (ca. -5 K) have been reported for
nitroxide-nitroxide intermolecular interactions in other metal-
nitroxide compounds, suggesting further that the low-temper-
ature interactions are solely intermolecular.14 In contrast, the
large negative θT value of -157 K reflects strong antiferro-
magnetic coupling between the S ) 1/2 nitroxide ligand and the
triplet state of the copper dimer. When the ethanol capping
ligands of 1b are replaced by pyridine ligands, again two
products result depending on the reaction time. The pyridine-
capped dimer structure, 1c, shows similar magnetic behavior
to that of 1b (Figure 7b), with the same coupling between the
copper centers (JST ) -150 cm-1) and small changes in the
Nâ2g2
øj )
1
(e-2(j /kT) + 3)kT
1
where j1 is the intermolecular nitroxide-nitroxide interactions, which
should be the only interactions at the lower temperatures as the
nitroxides are not expected to interact intramolecularly through the
singlet state copper dimer. This same fraction of the equation was
also applied to the high-temperature fit, as there should be intermo-
lecular nitroxide-nitroxide interactions at higher temperatures as well.
Therefore,
Nâ2g2
øj )
(e-2(j /kT) + 3)kT
3
3
where j3 is the intermolecular nitroxide-nitroxide interactions at
high temperatures, but j3 should be different than j1 as the
nitroxides should now be coupled intramolecularly at high
temperatures as well. For a more detailed analysis of the øJ2j2 term,
see: Gruber, S. J.; Harris, C. M.; Sinn. E. J. Chem. Phys. 1968,
49, 2183.

Copper(II) Benzoate Nitroxide Dimers and Chains
Inorganic Chemistry, Vol. 39, No. 21, 2000 4901
Figure 8. øT(T) for (a) [Cu(O2CMe)2]2(PYNN)2, 1d, and its fit to eq
4 with n ) 2, JST ) -223 K (-155 cm-1), gCu ) 2.18, gorg ) 2.001,
θS ) -5.8 K, and θT ) -5.8 K and (b) [Cu(O2CMe)2]2(PYNN), 2,
and its fit to eq 4 with n ) 1, JST ) -238 K (-165 cm-1), gCu ) 2.18,
gorg ) 2.001, θS ) -5.9 K, and θT ) -300 K. The Bleaney-Bowers
term of eq 4 is represented by dashed lines, the θS term by smaller
dashes, the θT term by the dotted line, and the full fit by the solid
line.
Figure 7. øT(T) for (a) [Cu2(O2CMe)(NNBA)3](EtOH)2, 1b, and its
fit to eq 4 with n ) 3, JST ) -216 K (-150 cm-1), gCu ) 2.18, gorg
)
2.001, θS ) -1.5 K, and θT ) -157 K and (b) [Cu2(O2CMe)(NNBA)3]-
py2, 1c, with its fit to eq 4 with n ) 3, JST ) -216 K (-150 cm-1),
gCu ) 2.18, gorg ) 2.001, θS ) -1.8 K, and θT ) -169 K. The
Bleaney-Bowers term of eq 4 is represented by dashed lines, the θS
term by smaller dashes, the θT term by the dotted line, and the full fit
by the solid line.
coupling between the S ) 1/2 ligands and the copper dimer triplet
state (θS ) -1.8 K; θT ) -169 K) (Table 3).
comparable to the -5.8 K observed for 1d. However, direct
copper(II)-nitroxide CuII‚‚‚O interaction causes a significant
increase in the antiferromagnetic interaction between the ni-
troxide and copper dimer triplet state compared to the discrete
PYNN-pyridyl-capped dimers. The θT value of -300 K reflects
strong antiferromagnetic coupling predominantly arising from
As discussed above, after the reaction is allowed to proceed
for 12 h, it is proposed that a linear chain of copper monomers
is produced, 6b, similar to that of the bromobenzoate monomers
6a. As noted above, the susceptibility in this case can be
modeled to Curie-Weiss behavior, eq 2, with gCu ) 2.16 and
a weak ferromagnetic coupling of θ ) +0.7 K (Figure 5b).
1
direct interactions between the S ) /2 ligand and the triplet
state of the copper dimer. Note that for 2b, J is an average of
the J values for two differently coordinated dimers.21b
[Cu(O2CMe)2(PYNN)]2, 1d. The addition of two PYNN
radicals to the apical position of [Cu(O2CMe)2]2, that is, [Cu-
(O2CMe)2(PYNN)]2, 1d, as observed for 1b and 1c, which differ
by substituent, does not significantly alter the behavior of the
dimer (Figure 8a), and the data are qualitatively similar to that
observed for 1b and 1c; therefore the behavior was also modeled
with eq 4.25 The dimeric unit exhibits the singlet-triplet
transition with JST ) -223 K (-155 cm-1), only slightly higher
in energy with respect to 1b and 1c. Hence, in contrast to spin
residing on a bridging carboxylate ligand [θT(1b, 1c) ≈ -163
K], substantially weaker antiferromagnetic coupling (θT ) θS
) -5.8 K) is observed with a similar ligand axially bound to
the copper dimer. Because θT ) θS, it appears that only
intermolecular interactions are present and the nitroxide does
not couple to the triplet state of the dimer. The distant spin-
bearing nitroxide moiety is probably twisted with respect to the
pyridyl group and thus breaks the conjugation, reducing the
coupling. Similarly poor, albeit ferromagnetic, coupling was
reported for [Ru(O2CMe)2]2(PYNN).13a
In contrast to the Cu(II) dimer, which lacks metal-metal
bonding, the spin coupling via related diamagnetic dimers with
metal-metal bonding is substantially greater. Spins on aryl
nitronyl and imidazolyl nitroxides axially bound to the S ) 0
[RhII(O2CCF3)2]2 can interact intramolecularly through the Rh-
Rh σ-bond,14a,25 resulting in an observed nitroxide-only singlet-
triplet transition, with J ) -120 K (-84 cm-1) for the phenyl
nitronyl nitroxide (4b) substituted Rh(II) dimer. Likewise, strong
coupling between nitroxides axially bound to S ) 0 metal-
metal quadruple-bonded [MoII(O2CCF3)2]2 has been reported
for [MoII(O2CCF3)2(5)]2.25 Thus, metal-metal bonding facili-
tates spin coupling via bypassing the poorer spin-coupling
pathway of the carboxylate bridges as occurs for the Cu(II)
dimers.5
In addition to the diamagnetic metal ion mediated nitroxide-
nitroxide coupling, strong S ) 1 [RuII(O2CCF3)2]2-nitroxide
coupling has been reported for [RuII(O2CCF3)2(5)]2,9c (J ) -379
3
K; -263 cm-1), and slightly weaker S )
/
[RuII/III(O2-
2
CCF3)2]2-nitroxide coupling has been reported for [RuII/III
-
[Cu(O2CMe)2]2(PYNN), 2. Addition of only 1 equiv of
PYNN to [Cu(O2CMe)2]2 results in the formation of chain-
structured [Cu(O2CMe)2]2(PYNN), 2a or 2b, with ø(T) being
fit to eq 4 with J ) -238 K (-165 cm-1), gCu ) 2.18, θS )
-5.9 K, and θT ) -300 K (Figure 8b). The θS of -5.9 K is
(O2CCMe3)2(5)]2 (J ) -187 K; -130 cm-1), although the
strong withdrawing effect of the CF3 with respect to the CMe3
groups makes the former a stronger Lewis acid, resulting in
stronger Ru-O bonding and spin coupling.9c,22

4902 Inorganic Chemistry, Vol. 39, No. 21, 2000
Del Sesto et al.
as PYNN, or when a coordinating solvent is used, as dimers
may form (d).
The magnetic properties of these copper carboxylates vary
with the type of structures formed. The singlet-triplet transition
is still observed in all of the dimeric structures with JST ranging
from -200 to -240 K (-140 to -170 cm-1) characteristic of
antiferromagnetic coupling. The singlet-triplet coupling is
minimally altered by introduction of nitroxide-based radicals
at either the bridging equatorial or terminal apical positions.
However, direct coupling between the triplet copper dimer and
the nitroxide is strongly antiferromagnetic, with stronger
coupling observed for direct terminal coordination with respect
to occupying bridging sites or for terminal coordination but with
broken conjugation. The strong triplet Cu(II) dimer-nitroxide
spin coupling has also been observed for related Ru(II) and Ru-
(II/III) dimer-nitroxide couplings.9b,c,22 Furthermore, the Cu-
(II) dimers can be converted to hydrogen-bonded linear chains
upon reaction with water, and these chains exhibit weak 1-D
ferromagnetic coupling. Formation of chains composed of
copper dimers bridged with the PYNN ligand leads to enhanced
antiferromagnetic coupling.
Figure 9. General mechanism for the formation of copper dimers and
chains. (a) Reflux in xylenes or room temperature in ethanol. (b)
Addition of 2 equiv of a capping ligand, L. (c) Addition of capping
ligand and water, let stand. (d) Addition of multidentate carboxylate
ligand, e.g., NNBA ligand.
Conclusion
In attempts to synthesize the copper carboxylate-based dimers,
a general procedure for the pure dimers as well as linear chains
was identified (Figure 9). Reaction of dimeric [Cu(O2CMe)2]2‚
2H2O with the acid of the desired carboxylate bridge leads to
direct substitution (a). However, the product usually results in
a chain structure, similar to those reported by Deakin et al.19a
and Bakalbassis et al.19b Addition of a coordinating ligand to
this chain compound, as a suspension or in solution, breaks apart
the chains, and the ligand-capped dimers result (b). If water is
present and the reaction is allowed to stand for long, the
carboxylate bridges are broken and monomeric species (c) and
in some cases ferromagnetically coupled one-dimensional chain
structures form. An exception to this general mechanism may
occur if a multidentate carboxylate-based ligand is utilized, such
Acknowledgment. The authors thank Drs. Laura Deakin,
Jamie L. Manson, Kostantin I. Pokhodnya, Leigh C. Porter, and
Professor Arthur J. Epstein (The Ohio State University) for their
helpful discussions on the syntheses and magnetic properties
of these compounds. Funding provided by the Department of
Energy DMS (Grant No. DE FG 03-93ER45504) is gratefully
acknowledged.
Supporting Information Available: The X-ray crystallographic
files for Cu(O2CPh)2(DMF)]2, 1a, [Cu(O2CC6H4Br)2py2]‚H2O, 6a, and
[Cu(O2CPh)2py2], 7 in CIF format. This material is available free of
charge via the Internet at http://pubs.acs.org.
IC0007106