Anti-inflammatory dinuclear copper(ii) complexes with indomethacin. synthesis, magnetism and EPR spectroscopy. Crystal structure of the N,N-dimethylformamide adduct

Article abstract of DOI:10.1021/ic981100x

Veterinary anti-inflammatory Cu(II) complexes of indomethacin (1-(4-chlorobenzoyl)-5-methoxy-2-methyl-1H-indole-3-acetic acid = IndoH), of the general formula [Cu₂(Indo)₄L₂] (L = N,N-dimethylformamide (DMF), N,N-dimethylacetamide (DMA), N-methylpyrrolidone (NMP), and water), were studied by zero-field and X-band EPR spectroscopies, electronic spectroscopy, magnetic measurements, and X-ray powder diffraction. The complexes are similar to Cu(II) acetate monohydrate, with a strong antiferromagnetic exchange interaction, J, ranging from -141 to -152 cm^-1. Variable temperature magnetic susceptibility data for all of the complexes are similar, with the exception of a [Cu₂(Indo)₄(H₂O)₂] complex, which displays an unusual increase in magnetic moment with decreasing temperature from 50 to 10 K. The X-ray powder diffraction patterns of the DMF and DMA dimers show that they are isostructural. Two isostructural H₂O complexes were synthesized from different methods yet displayed different variable temperature magnetic susceptibity data. All of the [Cu₂(Indo)₄L₂] complexes crystallize in the triclinic space group P1?. Single-crystal X-ray diffraction analysis of the DMF complex, [Cu₂(Indo)₄(DMF)₂]· 1.6(DMF), shows that it is similar to the previously reported [Cu₂(Indo)₄(DMSO)₂] with a Cu-Cu bond length of 2.630(1) ?, Cu-O_RCOO of 1.960(4)-1.967(4) ?, and Cu-O_DMF of 2.143(5) ? and crystal parameters a = 10.848(3) ?,b= 13.336(6) ?, c = 16.457(4) ?, α = 104.67(3)°, β= 100.94(2)°, and γ = 107.16(3)°. The X-ray structure of the DMF dimer does not exhibit strong intermolecular interactions due to the hydrophobic nature of the exterior. This may be important in facilitating its dissolution in micelles and transport through membranes.

Full text of DOI:10.1021/ic981100x

1736
Inorg. Chem. 1999, 38, 1736-1744
Anti-Inflammatory Dinuclear Copper(II) Complexes with Indomethacin. Synthesis,
Magnetism and EPR Spectroscopy. Crystal Structure of the N,N-Dimethylformamide
Adduct
Jane E. Weder,† Trevor W. Hambley,*,† Brendan J. Kennedy,*,† Peter A. Lay,*,†
Dugald MacLachlan,†,‡ Richard Bramley,§ Christopher D. Delfs,§ Keith S. Murray,|
Boujemaa Moubaraki,| Barry Warwick, John R. Biffin, and Hubertus L. Regtop
School of Chemistry, University of Sydney, Sydney, NSW 2006, Australia, Research School of
Chemistry, Australian National University, Canberra, ACT 0200, Australia, Department of Chemistry,
Monash University, Melbourne, VIC 3168, Australia, and Biochemical Veterinary Research Pty. Ltd.,
Braemar, NSW 2575, Australia
ReceiVed September 11, 1998
Veterinary anti-inflammatory Cu(II) complexes of indomethacin (1-(4-chlorobenzoyl)-5-methoxy-2-methyl-1H-
indole-3-acetic acid ) IndoH), of the general formula [Cu2(Indo)4L2] (L ) N,N-dimethylformamide (DMF), N,N-
dimethylacetamide (DMA), N-methylpyrrolidone (NMP), and water), were studied by zero-field and X-band EPR
spectroscopies, electronic spectroscopy, magnetic measurements, and X-ray powder diffraction. The complexes
are similar to Cu(II) acetate monohydrate, with a strong antiferromagnetic exchange interaction, J, ranging from
-141 to -152 cm-1. Variable temperature magnetic susceptibility data for all of the complexes are similar, with
the exception of a [Cu2(Indo)4(H2O)2] complex, which displays an unusual increase in magnetic moment with
decreasing temperature from 50 to 10 K. The X-ray powder diffraction patterns of the DMF and DMA dimers
show that they are isostructural. Two isostructural H2O complexes were synthesized from different methods yet
displayed different variable temperature magnetic susceptibity data. All of the [Cu2(Indo)4L2] complexes crystallize
in the triclinic space group P1h. Single-crystal X-ray diffraction analysis of the DMF complex, [Cu2(Indo)4(DMF)2]‚
1.6(DMF), shows that it is similar to the previously reported [Cu2(Indo)4(DMSO)2] with a Cu-Cu bond length
of 2.630(1) Å, Cu-ORCOO of 1.960(4)-1.967(4) Å, and Cu-ODMF of 2.143(5) Å and crystal parameters a )
10.848(3) Å, b ) 13.336(6) Å, c ) 16.457(4) Å, R ) 104.67(3)°, â ) 100.94(2)°, and γ ) 107.16(3)°. The
X-ray structure of the DMF dimer does not exhibit strong intermolecular interactions due to the hydrophobic
nature of the exterior. This may be important in facilitating its dissolution in micelles and transport through
membranes.
Introduction
uncomplexed Cu salt.3,8 Furthermore, there are suggestions that
Cu may play a role in preventing the gastrointestinal damage
associated with use of the NSAIDs.3,9
Over the past 30 years, numerous reports have appeared on
the benefits of Cu(I) and Cu(II) compounds in the treatment of
inflammation.1-5 A range of Cu complexes have been studied
as anti-inflammatory agents, ranging from simple Cu salts6,7 to
monomeric and dimeric Cu(II)-carboxylato complexes.8 In
addition, Cu(II) complexes of clinically used nonsteroidal anti-
inflammatory drugs (NSAIDs) are often more potent anti-
inflammatory agents than either the parent NSAID drug or the
The mechanism of the analgesic and anti-inflammatory action
of the NSAIDs is still a matter of some debate;10,11 however,
these compounds have been shown to be potent inhibitors of
prostaglandin synthesis.10,12-14 Indomethacin (I) is one of the
most potent of the clinically used NSAIDs15 and interferes with
prostaglandin synthesis by direct inhibition of the two cyclooxy-
genase enzyme systems, Cox-1 and Cox-2.16,17 Inhibition of the
Cox-2 system results in anti-inflammatory action, while inhibi-
tion of the Cox-1 enzyme system results in anti-inflammatory
† University of Sydney.
‡ Current address: National Registration Authority for Agricultural and
Veterinary Chemicals, PO Box E240, Kingston, ACT 2604, Australia.
§ Australian National University.
(9) Boyle, E.; Freeman, P. C.; Goudie, A. C.; Mangan, F. R.; Thomson,
M. J. Pharm. Pharmacal. 1976, 28, 865-868.
(10) Abramson, S. B.; Weissmann, G. Arthritis Rheum. 1989, 32, 1-9.
(11) Thomas, J. Australian Presciption Products Guide, 26th ed.; Australian
Pharmaceutical Publishing Co. Ltd.: Hawthorn, Victoria, 1997; Vol.
1, pp 1327-1336.
(12) Vane, J. R. NATUA 1971, 231, 232-235.
(13) Vane, J. R.; Ferreira, S. H. Anti-inflammatory Drugs; Springer-
Verlag: New York, 1979.
(14) Moncada, S.; Vane, J. R. AdV. Intern. Med. 1979, 24, 1-22.
(15) Dukes, M. N. G. Meyler’s Side Effects of Drugs, An Encyclopedia of
AdVerse Reactions and Interactions, 13th ed.; Elsevier: Amsterdam,
1996.
| Monash University.
Biochemical Veterinary Research Pty. Ltd.
(1) Bonta, I. L. Acta Physiol. Pharmacol. Neerl. 1969, 15, 188-222.
(2) Adams, S. S.; Cobb, R. Prog. Med. Chem. 1976, 5, 59-138.
(3) Sorenson, J. R. J. Inflammatory Diseases and Copper, 1st ed.; Humana
Press: Clifton, NJ, 1982.
(4) Sorenson, J. R. J.; Kishore, V.; Pezeshk, A.; Oberley, L. W.;
Leuthauser, S. W. C. Inorg. Chim. Acta 1984, 91, 285-294.
(5) Sorenson, J. R. J. A Physiological Basis for Pharmacological ActiVities
of Copper Complexes: An Hypothesis; Humana Press: Clifton, NJ,
1987; pp 3-16.
(6) Sorenson, J. R. J. J. Med. Chem. 1976, 19, 135-148.
(7) Lewis, A. J. Agents Actions 1978, 8, 244-250.
(8) Sorenson, J. R. J. Prog. Med. Chem. 1989, 26, 437-568.
(16) Ouellet, M.; Percival, M. D. J. Biochem. 1995, 306, 247-251.
(17) Vane, J. R.; Botting, R. M. Semin. Arthritis Rheum. 1997, 26, 2-10.
10.1021/ic981100x CCC: $18.00 © 1999 American Chemical Society
Published on Web 03/31/1999

Anti-Inflammatory Cu(II)-Indomethacin Complexes
Inorganic Chemistry, Vol. 38, No. 8, 1999 1737
action as well as gastric irritation.15,18 Gastrointestinal irritation
and ulceration are recognized as significant side effects of
NSAIDs.11,19,20
(14.3 g, 0.04 mol) was added a solution of Cu(II) acetate monohydrate
(4 g, 0.02 mol) in DMF (25 mL), and the mixture was heated to 80
°C. Ethanol (250 mL) was added to the mixture with vigorous shaking,
and the resulting deep green solution was set aside for about 1 day,
during which time the Cu complex separated as a microcrystalline light
green powder. The mixture was filtered under vacuum, and the green
product was washed exhaustively with ethanol (100 mL) and dried
overnight at room temperature. Yield: 13.9 g (76%). Anal. (Cu2C85H85-
Cl4N7O21) C, H, N, Cu: calcd, 56.42, 4.73, 5.42, 7.02; found, 56.41,
4.55, 5.41, 7.05.
Bis(N,N ′-dimethylformamide)tetrakis-µ-(O,O ′-Indo)dicopper-
(II) 1.6-N,N ′-Dimethylformamide, [Cu2(Indo)4(DMF)2]‚1.6(DMF).
The complex was crystallized from a 36% w/v solution of [Cu2(Indo)4-
(DMF)2] dissolved in a mixture of 20% DMF and 80% ethanol. Anal.
(Cu2C86.8H85.2Cl4N7.6O19.6) C, H, N: calcd, 57.4, 4.7, 5.9; found, 57.8,
4.8, 5.9.
I
Interest in developing potent yet less irritating NSAIDs has
led to several studies on the preparation, characterization, and
veterinary and medical use of divalent metal salts of IndoH and
a number of other NSAIDs.3,21-25 Carboxylate-type NSAIDs
Bis(N,N ′-dimethylacetamide)tetrakis-µ-(O,O ′-Indo)dicopper-
(II) Trihydrate, [Cu2(Indo)4(DMA)2]‚3H2O. The complex was pre-
pared in a manner similar to that for [Cu2(Indo)4(DMF)2], except that
N,N-dimethylacetamide was used as the solvent. Ethanol was added,
and the solution was cooled overnight to yield a microcrystalline dark
green precipitate. The mixture was filtered under vacuum and the
product was washed exhaustively with ethanol (150-300 mL), dried
overnight at room temperature, and finally dried at 40 °C for 8 h.
Yield: 6.40 g (35%). Anal. (Cu2C84H84Cl4N6O21) C, H, N, Cu: calcd,
56.60, 4.75, 4.72, 7.13; found, 56.16, 4.94, 4.86, 7.52.
produce dinuclear ([CuII (RCOO)4L2]) and mononuclear ([CuII-
2
(RCOO)2L2]) complexes, where R ) aryl or phenyl group, and
L has included a variety of ligands such as water, caffeine,
imidazole, diethylamine, 3-picoline, N,N-dimethylformamide,
dimethyl sulfoxide, 3-pyridylmethanol, and pyridine.22,23,25-31
The characterization of Cu(II) complexes of IndoH is
important because of their increasing use as veterinary phar-
maceutics and their potential for use as potent human anti-
inflammatory drugs. Here we report on the synthesis, the
electronic, zero-field EPR, and X-band EPR spectroscopies, the
magnetic properties, and the X-ray powder and crystal diffrac-
tion characteristics of Cu(II)-Indo and its solvent adducts with
DMF, DMA, NMP, and water. These studies are important to
establish the structures of the active components of the
pharmaceutical preparations.
Diaquatetrakis-µ-(O,O ′-Indo)dicopper(II) sesquihydrate, [Cu2-
(Indo)4(OH2)2]‚1.5H2O. IndoH (4 g, 0.01 mol) in ethanol (50 mL) was
added to Cu(II) acetate monohydrate (1 g, 0.005 mol) in ethanol (50
mL), and a pale green Cu complex began to precipitate within minutes
of mixing. While the resultant solution was being shaken vigorously,
ethanol (150 mL) was added: then the mixture was left to stand at
room temperature overnight. The mixture was filtered under vacuum
and dried overnight at room temperature. Yield: 3.9 g (78%). Anal.
(Cu2C76H69Cl4N4O20.5) C, H, N, Cu: calcd, 56.44, 4.18, 3.46, 7.86;
found, 56.30, 4.26. 3.62, 8.10.
Experimental Section
Synthesis. All of the complexes are green crystalline or powdered
solids soluble in DMF, tetrahydrofuran (THF), and acetonitrile and
insoluble in water. IndoH was of pharmaceutical grade (Sigma
Pharmaceuticals). All of the other chemicals were of analytical grade
(Aldrich or Sigma) and were used without further purification.
Bis(N,N ′-dimethylformamide)tetrakis-µ-(O,O ′-Indo)dicopper-
(II) N,N ′-Dimethylformamide Dihydrate, [Cu2(Indo)4(DMF)2]‚
DMF‚2H2O. To a warm (∼50 °C) DMF (20 mL) solution of IndoH
Diaquatetrakis-µ-(O,O ′-Indo)dicopper(II), [Cu2(Indo)4(H2O)2].
To a sodium hydroxide solution (0.1 M, 100 mL) was added IndoH
(14.3, 0.04 mol), and the resulting solution was filtered to remove
undissolved IndoH. Copper(II) chloride dihydrate (1 g, 0.006 mol) in
water (25 mL) was added and the solution was heated to 50 °C, during
which time the Cu complex separated as a pale green powder. The
mixture was filtered under vacuum, and the green product was washed
exhaustively with ethanol and water. Yield: 11.9 g (78%). Anal.
(Cu2C76H64Cl4N4O18) C, H, N, Cu: calcd, 57.40, 4.06, 3.52, 7.99; found,
57.65. 4.06, 3.98, 7.90.
(18) Fenner, H. Semin. Arthritis Rheum. 1997, 26, 28-33.
(19) Goodman, L. S.; Gilman, A. The Pharmacological Basis of Thera-
peutics, 5th ed.; Macmillan Publishing Co., Inc.: New York, 1975.
(20) Reynolds, J. E. F. Martindale. The Extra Pharmacopoeia, 13th ed.;
The Pharmaceutical Press: London, 1993.
(21) Auer, D. E. Copper, Inflammation and Copper Containing Anti-
inflammatory Drugs in the Horse. Ph.D. Thesis, University of
Queensland, Brisbane, 1987; pp 1-273.
(22) Abuhijleh, A. L.; Bogas, E.; Le Guenniou, G. Inorg. Chim. Acta 1992,
195, 67-71.
Bis(N-methyl-2-pyrrolidone)tetrakis-µ-(O,O ′-Indo)dicopper-
(II) Dihydrate, [Cu2(Indo)4(NMP)2]‚2H2O. The N-methylpyrrolidone
complex was prepared by the method described for [Cu2(Indo)4(DMF)2]
except that N-methylpyrrolidone was used as the solvent. Yield: 12.08
g (66%). Anal. (Cu2C86H82Cl4N6O20) C, H, N, Cu: calcd, 57.75, 4.62,
4.70, 7.11; found: 57.76, 4.54, 4.76, 7.85.
Physical Measurements. Metal analyses were determined using a
Varian AA-800 air-acetylene flame atomic absorption spectropho-
tometer. The C, H, N elemental analyses of the powdered samples were
performed by the Department of Chemical Engineering, University of
Sydney. The C, H, N elemental analysis of the [Cu2(Indo)4(DMF)2]‚
1.6(DMF) crystal was performed by the National Analytical Labora-
tories (NAL).
(23) Abuhijleh, A. L. J. Inorg. Biochem. 1994, 55, 255-262.
(24) Dendrinou-Samara, C.; Kessissoglou, D. P.; Manoussakis, G. E.;
Mentzafos, D.; Terzis, A. J. Chem. Soc., Dalton Trans. 1990, 959-
965.
(25) Dendrinou-Samara, C.; Jannakoudakis, P. D.; Kessissoglou, D. P.;
Manoussakis, G. E.; Mentzafos, D.; Terzis, A. J. Chem. Soc., Dalton
Trans. 1992, 3259-3264.
(26) Regtop, H. L.; Biffin, J. R. Preparation of Divalent Metal Salts of
Indomethacin. US Patent 5310936; Biochemical Veterinary Research
Pty. Ltd., Australia, 1994.
(27) Underhill, A. E.; Bougourd, S. A.; Flugge, M. L.; Gale, S. E.; Gomm,
P. S. J. Inorg. Biochem. 1993, 52, 139-144.
(28) Bhirud, R. G.; Srivastave, T. S. Inorg. Chim. Acta 1990, 173, 121-
125.
Room temperature magnetic moments (µeff) were measured on a
Sherwood Scientific magnetic susceptibility balance. Variable temper-
ature magnetic data were collected in the range 5-300 K using a
Quantum Design MPMS SQUID magnetometer. The samples were
cooled in zero field. In all cases the raw data were corrected for the
diamagnetic sample holder. The strength of the antiferromagnetic
coupling constant, J, defined as
(29) Valach, F.; Tokarcik, M.; Kubinex, P.; Melnik, M.; Macaskova, L.
Polyhedron 1997, 16, 1461-1464.
(30) Melnik, M.; Potocnak, I.; Macaskova, L.; Miklos, D.; Holloway, C.
E. Polyhedron 1996, 15, 2159-2164.
(31) Reimann, G. W.; Kokoszka, G. F.; Gordon, G. Inorg. Chem. 1965, 4,
1082-1084.
Hex ) -2JSˆ1‚Sˆ2
(1)

1738 Inorganic Chemistry, Vol. 38, No. 8, 1999
was obtained by fitting the data to the Bleaney-Bowers equation,32
3 exp(-2J/kT)
Weder et al.
eters at 21 °C were determined by a least-squares fit to the setting
parameters of 25 independent reflections.
Nâ2gj2
3kT
4x
Intensity data were collected in the range 1° < θ < 21° using an
ω-0.67θ scan. The scan widths and horizontal counter apertures
employed were (1.30 + 0.35 tan θ)° and (2.30 + 0.5 tan θ) mm. Data
reduction and application of Lorentz, polarization, and decomposition
(6.4%) corrections were carried out using the Enraf-Nonius Structure
Determination Package.38 Absorption corrections were not applied. Of
the 4572 independent nonzero reflections collected, 2744 with I > 2.5σ-
(I) were considered observed and used in the calculations.
The structure was solved by direct methods using SHELXS-76,39
and the solution was extended by difference Fourier methods. Hydrogen
atoms were included at calculated sites (C-H 0.97 Å) with group
isotropic thermal parameters, and all other atoms, with the exception
of minor contributors to disordered species (DMF molecules), were
refined anistropically. Blocked-matrix least-squares refinements of an
overall scale factor and positional and thermal parameters converged
with all shifts <0.4 σ. Maximum excursions in a final difference map
were +0.33 and -0.25 e Å-3. Scattering factors and anomalous
dispersion terms used for Cu (neutral Cu) were taken from International
Tables for X-Ray Crystallography,40 and all others used were those
supplied by SHELX-76.39 All calculations were carried out using
SHELX-76,39 and plots were drawn using ORTEP.41
Solution Stability. The stability of [Cu2(Indo)4(DMF)2] in aqueous
micelle solutions of the biological buffers HEPES (N-2-hydroxyeth-
ylpiperazine-N ′-2-ethanesulfonic acid) and TES (N-tris(hydroxymeth-
yl)methyl-2-aminoethanesulfonic acid) was determined by dissolving
∼1-2 mg mL-1 of the complex in 100 mM buffer with 40 mM sodium
dodecyl sulfate (SDS) in Milli-Q water adjusted to pH ∼7.4 and
measuring the EPR and UV-vis spectra of the resulting solutions. In
addition, the stability of [Cu2(Indo)4(DMF)2] in an aqueous veterinary
manufacturing solution comprising 48 mg mL-1 of the complex
dissolved in micelles of Span 80 plus tetraglycol was also determined.
The % Cu(II) monomer concentrations were calculated by determining
the double integral of the Cu(II) monomer spectra of the unknown
samples and comparing to a CuCl2 calibration curve. Glycerin 20%
was added to the biological buffer and CuCl2 calibration samples to
produce vitrified samples suitable for EPR spectroscopy. The UV-vis
and EPR spectra of the [Cu2(Indo)4(DMF)2] HEPES and TES buffers
were measured over a period of 3 days.
øA )
(1 - x)
+ NR +
(2)
[
]
T
3 exp(-2J/kT) + 1
where gj is the average g value for the dimer, øA represents the magnetic
susceptibility per Cu, x is the fraction of monomer, â is the Bohr
magneton, N is the Avogadro constant, and NR is a temperature
independent contribution to the susceptibility. In the fitting procedure,
all of the experimentally observed susceptibilities were equally weighted
and final fits were obtained by minimizing the function in eq 3.
E )
[(øobsd - øcalcd)T]2
(3)
∑
Solid state IR (400-2000 cm-1) and far-IR (75-500 cm-1) spectra
were acquired in pressed disks of KBr and polyethylene, respectively,
on a Bruker IFS 66v FTIR spectrophotometer at a resolution of 1.0
cm-1
.
Diffuse reflectance solid state and solution (DMF) UV-vis electronic
spectra were recorded from 200 to 1200 nm and 400-900 nm,
respectively, using a Varian Cary 5E UV-vis-near-IR spectropho-
tometer. Spectra of the solid samples of the complexes were acquired
in a KBr matrix. The spectra of the DMF solutions were acquired in
1-cm quartz cells in the double-beam acquisition mode.
Zero-field EPR spectra (ZFEPR) of powdered samples were recorded
using a resonant reflection zero-field EPR spectrometer similar to that
described elsewhere.33-35 The zero-field splitting parameters, D, E, and
A|, were obtained from eq 4,33,35
2
HS ) D(Sz2 - S(S + 1)/3) + E(Sx2 - Sy ) + A|SzIz
(4)
with the axial direction being along the Cu-Cu vector. As the
perpendicular hyperfine constituents were unresolved, they are not
included in the equation. In the analysis it was assumed that the
antiferromagnetic coupling of the two Cu(II) sites was sufficiently strong
that the singlet-triplet separation was significantly greater than D.
Low-temperature X-band powder EPR spectra of the solid [Cu2-
(Indo)2L2] samples were measured at X-band frequencies (∼9.5 GHz)
using a Bruker EMX EPR spectrometer equipped with a standard
ER4120 X-band cavity, gaussmeter, frequency counter, BVT2000
variable temperature unit, and Oxford Instruments E900 continuous
flow cryostat. The values of g| and g were determined from the
resulting X-band spectra using the values of D and E obtained from
the ZFEPR studies. The powder EPR spectra are described by the spin
Hamiltonian,
Results
Syntheses. The complexes [Cu2(Indo)4(OH2)2]‚1.5H2O and
[Cu2(Indo)4L2] (L ) DMF, DMA, NMP) were prepared by
heating a 2:1 molar ratio of IndoH and Cu(II) acetate mono-
hydrate (or 6.7:1 molar ratio of IndoH/Cu(II) chloride dihydrate
for [Cu2(Indo)4(OH2)2]) in the appropriate solvent. The dimer
was precipitated by the addition of excess ethanol. A dimeric
species was also prepared using MeCN as the solvent, but on
precipitation with EtOH, microanalysis demonstrated that the
resulting complex was the water adduct, [Cu2(Indo)4(OH2)2].
Crystal Structure of the [Cu2(Indo)4(DMF)2]‚1.6(DMF)
Complex. The crystal data and data collection parameters for
[Cu2(Indo)4(DMF)2]‚1.6(DMF) are given in Table 1 and selected
bond lengths and angles in Table 2. The atomic numbering
system and ORTEP diagram are given in Figures 1 and 2,
respectively. Tables of non-hydrogen coordinates and isotropic
thermal parameters, bond distances, bond angles, anisotropic
thermal parameters, hydrogen atom positions, and isotropic
thermal parameters for [Cu2(Indo)4(DMF)2]‚1.6(DMF) and a
2
H ) âBgS + D(Sx2 - 2â) + E(Sx2 - Sy )
(5)
where B, the external magnetic field, g, the anisotropic Lande splitting
tensor, S, the total spin vector, and D and E, the zero-field splitting
parameters, were derived directly from the observed zero-field EPR
spectrum. The values of g were obtained using the EPRPOW and
SIMER simulation programs.36,37
Powder X-ray diffraction patterns were collected at room temperature
using Cu KR radiation on a SIEMENS D5000 diffractometer with
divergence and antiscatter slits of 1 mm, and receiver and detector slits
of 0.2 and 0.6 mm, respectively. The data were collected over the range
2.0-35.0° in steps of 0.04° in 2θ, and a count time per step of 15.0 s.
The X-ray crystallographic data for [Cu2(Indo)4(DMF)2]‚1.6(DMF)
were measured on an Enraf-Nonius CAD4F four-circle diffractometer
employing graphite-monochromated Mo KR radiation. Lattice param-
(32) Bleaney, B.; Bowers, K. Proc. R. Soc. London, Ser. A 1952, 451-
465.
(33) Delfs, C. D.; Bramley, R. Chem. Phys. Lett. 1997, 264, 333-337.
(34) Delfs, C. D.; Bramley, R. J. Chem. Phys. 1997, 107, 8840-8847.
(35) Bramley, R.; Strach, S. Chem. ReV. 1983, 83, 49-82.
(36) MacLachlan, D. EPRPOW Computer Program for Simulation of EPR
Spectra; University of Sydney: Sydney, 1996.
(38) Enraf-Nonius Structure Determination Package; Enraf-Nonius: Delft,
Holland, 1985.
(39) Sheldrick, G. M. SHELX-76: A Program for X-ray Crystal Structure
Determination; University of Cambridge: Cambridge, 1976.
(40) Cromer, D. T.; Waber, J. T. International Tables for X-Ray Crystal-
lography; The Kynoch Press: Birmingham, England, 1974; Vol. 4.
(41) Johnson, C. K. ORTEP. A Thermal Ellipsoid Plotting Program; Oak
Ridge National Laboratories: Oak Ridge, TN, 1976.
(37) MacLachlan, D. SIMER Computer Program for Rapid Simulation of
EPR Spectra; University of Sydney: Sydney, 1996.

Anti-Inflammatory Cu(II)-Indomethacin Complexes
Inorganic Chemistry, Vol. 38, No. 8, 1999 1739
Table 1. Crystal and Data Collection Parameters for
[Cu2(Indo)4(DMF)2]‚1.6(DMF)
formula
Mr
cryst color
cryst dimens (mm)
cryst syst
space group
a (Å)
b (Å)
c (Å)
R (deg)
â (deg)
γ (deg)
V (Å3)
Cu2C86.8H85.2Cl4N7.6O19.6
1817.2
green
0.10 × 0.15 × 0.15
triclinic
P1h
10.848(3)
13.336(6)
16.457(4)
104.67(3)
100.94(2)
107.16(3)
2109(2)
1
Z
D
calc (g cm-3
)
1.437
radiation (monochromated incident
0.7107
beam) λ(Mo KR) (Å)
µ(Mo KR) (cm-1
)
6.88
0.046
0.048
R1a
wR2b
a R1 ) ∑(||Fo| - |Fc||)/∑|Fo|. b wR2 ) ∑(w1/2||Fo| - |Fc||)/∑w1/2|Fo|,
w ) 0.75/(σ2(Fo) + 0.0010Fo ).
2
Table 2. Selected Interatomic Distances (Å) and Angles (deg) in
[Cu2(Indo)4(DMF)2]‚1.6(DMF)a
Figure 1. Atomic numbering system for the complex in the crystal
structure of [Cu2(Indo)4(DMF)2]‚1.6DMF.
Cu1-Cu1*
Cu-O1
2.630(1)
1.960(4)
1.947(4)
Cu1*-O5
Cu1-O6
Cu1-O9
1.967(4)
1.963(5)
2.143(5)
Cu1*-O2
O6-Cu1-O1
O9-Cu1-O1
O9-Cu1-O6
Cu1-O6-C21
O9-Cu1-O5*
C21-O5-Cu1*
Cu1-O1-C1
O1-C1-O2
88.9(2)
93.9(2)
94.5(2)
122.5(5)
97.2(2)
122.4(5)
123.2(5)
123.9(6)
Cu1*-O2-C1
O6-Cu1-Cu1*
O5*-Cu1-Cu1*
O6-Cu1-O2*
O2*-Cu1-O5*
O5*-Cu1-O1
O6-C21-O5
124.5(5)
84.2(1)
84.1(1)
89.1(2)
89.8(2)
89.8(2)
126.9(7)
a Atoms are labeled as depicted in Figure 1. Asterisks (*) depict
atom labeling assuming an inverted point of symmetry centered midway
along the Cu-Cu vector. Standard deviations, in parentheses, occur in
the last significant figure.
figure presenting the atomic numbering system for Indo in the
crystal structure of [Cu2(Indo)4(DMF)2]‚1.6(DMF) are given in
the Supporting Information.
The structure consists of two Cu atoms linked by four Indo
groups in a fashion similar to that of [Cu2(Indo)4(DMSO)2]42
and the dinuclear structure of [Cu2(CH3COO)4(H2O)2].43-46
A
DMF molecule is bound trans to the Cu-Cu vector on each of
the Cu atoms, and a further two DMF molecules are loosely
held in the lattice. These lattice DMF sites were highly mobile,
and their sites were only partially occupied. During the data
collection, some reductions in the intensities of the standard
reflections were noted, and these are consistent with slow loss
of these molecules of crystallization. The fused and conjugated
five- and six-membered rings are close to being coplanar
(deviations less than 0.04 Å), and the Cu dimer makes no close
contacts with other molecules in the lattice. The structure of
Figure 2. ORTEP diagram of the complex in the crystal structure of
[Cu2(Indo)4(DMF)2]‚1.6 DMF.
the [Cu2(Indo)4(DMF)2]‚1.6(DMF) complex is similar to that
reported for the DMSO complex,42 although the latter structure
was of a much lower precision.
(42) Weser, U.; Sellinger, K. H.; Lengfelder, E.; Werner, W.; Strahle, J.
Biochim. Biophys. Acta 1980, 631, 232-245.
(43) van Niekerk, J. N.; Schoening, F. R. L. Acta Crystallogr. 1953, 6,
227-232.
The individual Cu atoms have a Jahn-Teller-distorted,
octahedral geometry, with four short Cu-ORCOO (1.960(4)-
1.967(4) Å) bond lengths and a long solvent Cu-O (2.143(5)
Å) bond length. The two copper atoms in each dimeric unit are
separated by a distance of 2.630(1) Å. This distance is similar
to that found in [Cu2(Indo)4(DMSO)2]42 (2.632(6) Å) and other
(44) de Meester, P.; Fletcher, S. R.; Skapski, A. C. J. Chem. Soc., Dalton
Trans. 1973, 2575-2578.
(45) Brown, G. M.; Chidambaram, R. Acta Crystallogr., Sect. B 1973, B29,
2393-2403.
(46) Wilkinson, G. ComprehensiVe Coordination Chemistry. The Synthesis,
Reaction, Properties & Application of Coordination Compounds;
Pergamon Press: Oxford, 1987; Vol. 5, pp 634-774.

1740 Inorganic Chemistry, Vol. 38, No. 8, 1999
Weder et al.
Table 3. Lattice Parameters and Selected Details of Refinements of the Powder Diffraction Patterns of [Cu2(Indo)4(L)2] Complexes
params
[Cu2(Indo)4(DMF)2]
[Cu2(Indo)4(OH2)2]‚1.5H2O
[Cu2(Indo)4(NMP)2]
[Cu2(Indo)4(DMA)2]
a (Å)
b (Å)
c (Å)
R (deg)
â (deg)
γ (deg)
10.850(7)
13.360(7)
16.56(1)
105.8(1)
101.1(1)
106.9(1)
triclinic
11.17(3)
14.84(4)
16.26(7)
103.1(3)
104.8 (3)
100.0(3)
triclinic
22.104(6)
25.941(9)
15.375(7)
91.80(3)
107.17(3)
108.48(4)
triclinic
11.10(2)
12.98(2)
15.97(12)
100.7(3)
101.2(3)
109.7(2)
triclinic
complex, there appears to be a doubling of both the a and b
cell dimensions, while c remains unchanged, when compared
to the DMF complex. No clear relationship, however, is
observed between the a, b, and c cell dimensions for the NMP
complex when compared to the unit cells of the other complexes.
UV-Vis and IR Spectroscopies. The diffuse reflectance
UV-vis spectra of (a) IndoH, (b) [Cu2(Indo)4(DMF)2], and (c)
Cu(II) acetate monohydrate in a KBr matrix and solid state IR
spectra of (a) [Cu2(Indo)4(DMF)2] and (b) IndoH in a KBr
matrix are provided as Supporting Information. The solid and
solution state electronic spectra of all of the [Cu2(Indo)4L2]
complexes show a broad absorption band (band I) in the visible
region around 700 nm (Table 4), which is assigned to a dxy,yz
f dx -y transition.25,31,48 There is a slight shoulder around 790
nm, which is typical of a Jahn-Teller-distorted Cu(II). This is
similar to that reported for [Cu2(Indo)4(DMSO)2].42 The solid
state spectra all display a shoulder at ∼330 nm (band II). The
origin of band II remains unclear but has been assigned to a
charge-transfer absorption.49 Band II is believed to be indicative
of a dimeric complex23,30,50 even though a number of monomeric
and polymeric Cu(II) compounds exhibit this band.49,51,52 All
of the complexes, however, display bands I and II in the usual
range for Cu(II) compounds in a square-pyramidal CuO4O
environment.46,53 None of the solid state spectra display a
carboxy-copper(II) charge-transfer absorption in the 240-280-
nm range (band III),54 because IndoH has a strong absorption
in the 200-400-nm range. Consequently, band II of the Cu
complexes partially overlaps with this band and band III is
concealed. The diffuse reflectance spectra of the [Cu2(Indo)4L2]
complexes do, however, show a small shoulder at 300-330 nm
(Table 4) and possibly a second small absorption band at 240-
280 nm. The shoulder at 300-330 nm is similar to those
displayed in Cu(II) acetate monohydrate, Cu(II) cyanoacetate,
and Cu(II) cyanoacetate monohydrate55 and may be due to band
II. The assignment of the shoulder near 260 nm (band III) is
less certain.
2
2
Figure 3. X-ray powder diffraction patterns of (a) [Cu2(Indo)4(OH2)2],
(b) [Cu2(Indo)4(OH2)2]‚1.5H2O, (c) [Cu2(Indo)4(NMP)2], (d) [Cu2(Indo)4-
(DMA)2], and (e) [Cu2(Indo)4(DMF)2].
Cu carboxylate dimers23-26 and only slightly longer than that
found in Cu metal (2.56 Å).47
X-ray Powder Diffraction Patterns. The powder patterns
of all of the complexes were indexed as having a triclinic cell
with lattice parameters given in Table 3. X-ray powder
diffraction patterns for [Cu2(Indo)4(DMF)2], [Cu2(Indo)4-
(DMA)2], [Cu2(Indo)4(NMP)2], [Cu2(Indo)4(OH2)2]‚1.5H2O and
[Cu2(Indo)4(OH2)2] are shown in Figure 3.
With the exception of the [Cu2(Indo)4(OH2)2] complex, the
microcrystalline samples were reasonably crystalline and gave
well-resolved X-ray diffraction patterns over the range 2° <
2θ < 35°. The small number of well-resolved lines observed
in the diffraction pattern of [Cu2(Indo)4(OH2)2] prevented
indexing of this pattern. X-ray diffraction patterns for the other
complexes could only be indexed on the basis of a triclinic cell.
The cell dimensions of the DMF complex are similar to those
obtained from the single-crystal measurements; the small
differences probably arose from different DMF/H2O solvents
of crystallization in the two samples. The cell dimensions of
the DMF complex are also similar to those of the DMSO
complex.42 Figure 3 clearly shows that the [Cu2(Indo)4(OH2)2]
and [Cu2(Indo)4(OH2)2]‚1.5H2O complexes are isostructural, and
the DMF and DMA complexes are isostructural. It is clear that
these two pairs of compounds are not isostructural with each
other, as is evident from the strong 002 reflection near 7.5° in
the DMF and DMA complexes.
Since the intensity of band I can be appreciably higher for
dinuclear carboxylate complexes22,23,25,28,56-58 than for mono-
nuclear complexes, it has been suggested that this is diagnos-
tic.22,23 However, the molar absorptivities for the monomeric
(48) Kato, M.; Muto, Y. Coord. Chem. ReV. 1988, 92, 45-83.
(49) Catterick, J.; Thornton, P. AdV. Inorg. Chem. Radiochem. 1977, 291-
362.
(50) Dubicki, L.; Martin, R. L. Inorg. Chem. 1966, 5, 2203-2209.
(51) Hathaway, B. J.; Billing, D. E. Coord. Chem. ReV. 1970, 5, 143-
207.
(52) Yawney, D. B. W.; Moreland, J. A.; Doedens, R. J. J. Am. Chem.
Soc. 1973, 95, 1164-1170.
(53) Agterberg, F. P. W.; Provo Kluit, H. A. J.; Driessen, W. L.; Reedijk,
J.; Oevering, J.; Buijs, W.; Veldman, N.; Lakin, M. T.; Spek, A. L.
Inorg. Chim. Acta 1998, 267, 183-192.
The diffraction pattern of the NMP complex is unique when
compared to those observed for either the DMF/DMA or the
H2O/[Cu2(Indo)4(OH2)2]‚1.5H2O complexes. For the NMP
(54) Graddon, D. P. J. Inorg. Nucl. Chem. 1961, 17, 222-231.
(55) Wasson, J. R.; Shyr, C.-I.; Trapp, C. Inorg. Chem. 1968, 7, 469-
474.
(56) Ahmed, I. Y.; Abuhijleh, A. L. Inorg. Chim. Acta 1982, 61, 241-
246.
(47) Aylward, G. H.; Findlay, T. J. V. SI Chemical Data, 2nd ed.; John
(57) Abuhijleh, A. L.; Woods, C. Inorg. Chim. Acta 1993, 209, 187-193.
(58) Kokot, E.; Martin, R. L. Inorg. Chem. 1964, 3, 1306-1312.
Wiley & Sons Australasia Pty. Ltd.: New York, 1971.

Anti-Inflammatory Cu(II)-Indomethacin Complexes
Inorganic Chemistry, Vol. 38, No. 8, 1999 1741
Figure 4. Plots of øM (O, observed; s, calculated) and µ ([, observed; s, calculated) versus T for (a) [Cu2(Indo)4(OH2)2], (b) [Cu2(Indo)4-
(DMF)2], and (c) [Cu2(Indo)4(OH2)2]‚1.5H2O.
Table 4. Room Temperature Magnetic Moments, SQUID Variable Temperature Magnetic Data, and Electronic Absorption Spectral Data for
the [Cu2(Indo)4(L)2] Complexes
a,b
µeff
fractionb
ν
asym(COO)
(cm-1
νCu-O
(cm-1
)
λ
max (nm)c
(ꢀmax (M-1 cm-1))
b
compd
(T ) 300 K) g valueb
J (cm-1
)
NR (cgs)b monomer
)
[Cu2(Indo)4(DMF)2]
[Cu2(Indo)4(OH2)2]‚1.5H2O
[Cu2(Indo)4(DMA)2]
[Cu2(Indo)4(NMP)2]
[Cu2(Indo)4(OH2)2]
1.42
1.40
1.52
1.53
1.51
1.53d
2.10
2.06
2.16
2.25
2.17
2.1
-152.5(1) 0.000 06
-149.8(1) 0.000 06
-142.3(1) 0.000 06
-142.3(1) 0.000 06
0.006
0.002
0.09
0.01
0.065
1618 s
1630 s
1599 s
1622 s
1630 s
1618 s
363 s, 321 s 709 s (417), 326 sh
385 s, 331 s 708 s (460), 327 sh
382 s, 319 s 709 s (395), 329 sh
388 s, 340 s 709 s (441), 319 sh
386 s, 332 s 709 s (450), 324 sh
376 s, 331 s 703 s (191), 355 sh
-144(1)
-142(1)
0.000 06
[Cu2(CH3COO)4(OH2)2]
0.000 06b
a Calculated using monomeric formula weight. b Derived from variable temperature magnetic susceptibility data. c s ) strong; sh ) shoulder.
d Taken from ref 68.
pyridine analogue of the Cu complexes of the NSAIDs naprosyn,
ꢀdmf ) 301 M-1 cm-1, and ibuprofen, ꢀdmf ) 263 M-1 cm-1 25
are similar to that of a dimeric DMSO Cu(II) complex of
ibuprofen, ꢀdmf ) 398 M-1 cm-1 25
Furthermore, the reported
value of the molar absorptivity for the dimeric DMSO Cu(II)
complex of ibuprofen (ꢀ ) 178 M-1 cm-1 24
is approximately
half that for the dimeric [Cu2(Indo)4(L)2] complexes under
investigation (ꢀ ≈ 349 M-1 cm-1). Thus it is clear that the
intensity of band I cannot be used to definitively determine the
presence of monomeric or dimeric units.
UV-vis and IR spectroscopies, together with elemental
analysis, of the [Cu2(Indo)4L2] complexes are suggestive, rather
than definitive, of a dimeric Jahn-Teller-distorted Cu site. Thus,
magnetic susceptibility and EPR measurements were undertaken
to determine both the nature and extent of any Cu-Cu
interactions in the complexes.
Magnetic Susceptibility. The room temperature magnetic
moments per Cu (µeff ) 1.40-1.56 µB, Table 4) are similar to
those observed for other dinuclear Cu(II) compounds of the type
[Cu2(RCOO)4L2]65,66 and are lower than the d9 spin-only
magnetic moment, µeff ) 1.73 µB. This observation is consistent
with antiferromagnetic exchange between the two Cu(II)
ions.32,67 The variable temperature magnetic results for [Cu2-
(Indo)4(OH2)2], [Cu2(Indo)4(DMF)2], and [Cu2(Indo)4(OH2)2]‚
,
.
)
In principle, the mode of coordination of the carboxylate to
the central Cu ion can be determined from the value of ∆ (∆ )
νasym - νsym)
59-61 in the IR spectra. This criterion could not be
used in the present [Cu2(Indo)4L2] complexes, since the car-
boxylate stretching frequencies are obscured by other bands
(Supporting Information). The Cu-O stretching frequencies
(Table 4) are comparable to those reported for Cu(II) acetate
type dimers.62-64
(62) Faniran, J. A.; Patel, K. S. J. Inorg. Nucl. Chem. 1974, 36, 2261-
2263.
(63) Mathey, Y.; Greig, D. R.; Shriver, D. F. Inorg. Chem. 1982, 21, 3409-
3413.
(64) Fanisan, J. A.; Patel, K. S. J. Inorg. Nucl. Chem. 1979, 41, 1495-
1496.
(65) Casanova, J.; Alzuet, G.; LaTorre, J.; Borras, J. Inorg. Chem. 1997,
36, 2052-2058.
(66) Doedens, R. J. Prog. Inorg. Chem. 1976, 21, 209-231.
(67) Figgis, B. N.; Martin, R. L. J. Chem. Soc. 1956, 3837-3846.
(59) Deacon, G. B.; Phillips, R. J. Coord. Chem. ReV. 1980, 33, 227-250.
(60) Nakamoto, K. Infrared Spectra of Inorganic and Organometallic
Compounds; Wiley: New York, 1963.
(61) Tyagi, A. S.; Srivastava, C. P. J. Indian Chem. Soc. 1981, 58, 284-
286.

1742 Inorganic Chemistry, Vol. 38, No. 8, 1999
Weder et al.
Table 5. X-Band and Zero-Field EPR Data for the [Cu2(Indo)4(L)2] Complexes
D (GHz)
A| (MHz)
b
b
b
sample
T (K)
g valuesa (110 K)
(10-2 cm-1
)
coordination
E (GHz) (cm-1
)
(104 cm-1
)
[Cu2(Indo)4(H2O)2]d
90 g| 2.35, g 2.05, geff 2.15c 9.975(1) (33.273(3))
CuO5
CuO5
CuO5
CuO5
CuO5
0.028(2) (9.3(6) × 10-4
0.051(2) (1.70(6) × 10-3) 230.0(5) (76.7(2))
0.085(1) (2.84(3) × 10-3) 219.2(5) (73.1(2))
0.296e (0.01 ( 0.005)f
)
)
230.5(2) (76.88(7))
229.0(5) (75.4(2))
[Cu2(Indo)4(OH2)]‚1.5H2Od 120 g| 2.35, g 2.05, geff 2.15c 10.003(1) (33.366(3))
0.017(3) (5.7(9) × 10-4
[Cu2(Indo)4(DMF)2]d
[Cu2(Indo)4(NMP)2]d
[Cu2(CH3COO)4(OH2)2]
90 g| 2.36, g 2.06, geff 2.16c 10.217(1) (34.080(3))
145 g| 2.36, g 2.06, geff 2.16c 10.317(1) (34.414(3))
g| 2.42, g 2.08, geff 2.19f 10.08(2)e (33.623(6))
a g value derived from X-band simulations; for an anisotropic system: g| ) gz, and g ) gxy. b D, E, and A| derived from ZFEPR simulations.
c geff ) (1/3)(g| + 2g ). d Sample contained traces of monomer. e Taken from ref 35. f Taken from ref 32.
1.5H2O are plotted in Figure 4, panels a, b, and c, respectively,
along with the lines of best fit, and the calculated values of J
are given in Table 4 for all of the [Cu2(Indo)4(L)2] complexes.
The magnetic moments of the Cu-Indo complexes decrease
from around 1.50 µB at 300 K to approximately 0.2 µB at 4.5
K, as a consequence of depopulation of the excited triplet (S )
1) state.58,67,68 The susceptibilities display a maximum at
approximately 300 K and decrease to a minimum value near
50 K. Below this, there is a rapid increase in magnetic
susceptibility as the temperature is lowered further to 4.5 K.
The behavior below 50 K is a consequence of a small amount
(e1%, Table 4) of a monomeric Cu(II) ion.
The temperature dependence of the magnetic properties of
[Cu2(Indo)4(OH2)2] was significantly different. The magnetic
moment decreased with decreasing temperature from 1.51 µB
at 300 K to 0.63 µB at 50 K in a manner similar to that observed
for the other dimers. At temperatures below 50 K, the value of
µeff initially increases with decreasing temperature, reaching a
maximum of 1.5 µB at 10 K, and then it decreases again to
∼0.88 µB at 4.5 K. This increase in magnetic moment with
decreasing temperature from 50 to 10 K is unusual and is
suggestive either of interdimer ferromagnetic coupling, perhaps
via some hydrogen-bonding interaction involving the water
molecules, or of a magnetic impurity.
Figure 5. Experimental (solid line) and calculated (dashed line) ZFEPR
spectra of [Cu2(Indo)4(OH2)2], T ) 90 K, B( ) 2 mT.
Zero-Field EPR Spectroscopy. ZFEPR spectra were mea-
sured for all of the complexes at 90, 120, and 145 K to determine
values for D and E, prior to simulation of the X-band EPR
spectra. Despite the atypical magnetic susceptibilities of [Cu2-
(Indo)4(OH2)2] below 50 K, the magnetic properties for all of
the complexes studied were similar at the temperatures relevant
to the EPR studies. The results of the least-squares refinements
for the values of D, E, g, and A| are set out in Table 5.
As illustrated for [Cu2(Indo)4(OH2)2] in Figure 5, the com-
plexes exhibit 7-line ZFEPR spectra centered near 10 GHz. In
all cases, the inner peaks were of comparable intensity. The D
and g parameters are typical of those for octahedral Cu(II)
2
2
carboxylate complexes where the unpaired electron is in a dx -y
orbital.69,70 The observed D values, 0.333 and 0.344 cm-1, are
comparable to those found for dinuclear Cu compounds with
exclusively oxygen donors (D ) 0.34 cm-1).71,72 The best-fit
E values are all very small, in keeping with the strong axial
symmetry observed in related X-band spectra.
X-Band EPR Powder Spectra. Typical powder spectra for
the [Cu2(Indo)4L2] complexes are shown in Figure 6, and the
best-fit parameters are summarized in Table 5. The line shape
and peak positions of the spectra were not substantially
Figure 6. [Cu2(Indo)4(DMF)2] X-band powder EPR spectra: (a)
simulation of 110 K spectrum; (b) 295 K; (c) 110 K.
influenced by the axial ligands. All spectra exhibited features
typical of those for dimeric complexes with axial symmetry and
66,69,73
2
2
g| > g
and a dx -y (or dxy) ground state.46 The spectra
are better resolved than that of [Cu2(Indo)4(DMSO)2] described
by Weser,42 and in all cases they exhibit, at room temperature,
a broad resonance at geff ≈ 2.1 (H ≈ 4720 G), with weak
(68) Hadjikostas, C. C.; Katsoulos, G. A.; Sigalas, M. P.; Tsipis, C. A.;
Mrozinski, J. Inorg. Chim. Acta 1990, 167, 165-169.
(69) Melnik, M. Coord. Chem. ReV. 1981, 36, 1-44.
(70) Smith, T. D.; Pilbrow, J. R. Coord. Chem. ReV. 1974, 13, 173-278.
(71) Goodgame, D. M. L.; Nishida, Y.; Winpenny, R. E. P. Bull. Chem.
Soc. Jpn. 1986, 59, 344-346.
2
features at ∼5980 (Hz2) and 500 G (Hz1) due to the spin-triplet
state of the dimeric complex and a small resonance at 3300 G
due to a Cu(II) monomer impurity. On cooling to 110 K, the
(72) Price, J. H.; Pilbrow, J. R.; Murray, K. S.; Smith, T. D. J. Chem. Soc.
A 1970, 968-976.
(73) Mabbs, F. E. Chem. Soc. ReV. 1993, 22, 313-324.

Anti-Inflammatory Cu(II)-Indomethacin Complexes
Inorganic Chemistry, Vol. 38, No. 8, 1999 1743
broad resonance at geff ) 2.1 decreases in intensity, with a
corresponding increase in amplitude of the features due to the
triplet state. At 110 K, the monomer feature becomes more
evident and the powder spectrum exhibits a distinctive absorp-
tion at 3300 G. Calculated values of the monomeric fraction
for the complexes studied, as derived from the variable
temperature magnetic susceptibility data, are given in Table 4.
The average values for g for the complexes studied are in
good agreement with those obtained from variable temperature
magnetic susceptibility measurements. In addition, both the g|
and g values for the [Cu2(Indo)4L2] complexes are similar to
those reported for other dinuclear Cu(II) carboxylate ad-
important clues as to the features of the drugs that may be
important in their pharmacology. The DMF complex has no
significant intermolecular contacts in the unit cell. This hydro-
phobic molecule is contained in micelles in some pharmaceutical
preparations and biological buffer solutions, and a hydrophobic
environment of the micelles would help prevent the decomposi-
tion of the dinuclear species. EPR and UV-vis spectroscopies
of the micelle and buffer solutions of the DMF complex confirm
that the complex remains intact in the dinuclear form; however,
lack of sensitivity of these techniques to the nature of the axial
ligand does not allow changes in the axial ligation to be
determined. The hydrophobic nature of these complexes is also
likely to be important in allowing the transport of the drug, in
an intact form, through the biological membranes of the
gastrointestinal tract and hence, minimization of the adverse
biological effects of free IndoH in the gastrointestinal tract.
Copper-Copper Interactions in the Dinuclear Complexes.
The large negative J values obtained from the variable temper-
ature magnetic susceptibility studies (Table 4) show that the
[Cu2(Indo)4(L)2] complexes have a singlet ground state with a
triplet excited state, which is 284-304 cm-1 (2J) higher in
energy. The observed values of J are comparable to those of
other dinuclear Cu(II) complexes,49,68 including Cu(II) acetate
monohydrate (J ) -142 cm-1).49 Dimeric Cu(II) arylcarboxy-
ducts,66,69,74 where g| ≈ 2.3, g ≈ 2.08, and D ≈ 0.34 cm-1 75
,
and for Cu(II)-Indo and its water76 and DMSO adducts.42
Previous workers have proposed that the value of g| is
indicative of an axial environment, with oxygen donors given
a lower value of g| compared to nitrogen donors.28 The finding
of equivalent values of g| for the complexes studied is, therefore,
suggestive of equivalent symmetry around the Cu ion and similar
ligand fields due to the various adducts.
Solution Studies. When [Cu2(Indo)4(DMF)2] was dissolved
as a micellar solution in HEPES and TES, the UV-vis spectra
remained characteristic of the Cu(II) dimer over a 3 day period,
with little change in either the Cu absorbance at 709 nm or the
indomethacin absorbance at 325 nm. In addition, EPR spec-
troscopy indicated that <8% of Cu(II) monomer was present.
Likewise, the Span 80 plus tetraglycol micellar solution
indicated <1% of Cu(II) monomer present.
lates have a stronger coupling, J values around -150 cm-1
,
compared to polymeric materials, J ≈ 50 cm-1 66,68,81
,
further
confirming the dimeric nature of the present complexes.
While the fine details remain obscure, it is clear that dinuclear
compounds with either a CuO4N or CuO5 coordination have
noticeably different average values, of -2J ) 348 and 316
cm-1, respectively.74 This observation is consistent with the
results for [Cu2(Indo)4(H2O)2], [Cu2(Indo)4(NMP)2], and [Cu2-
Discussion
Four Cu(II) dimers of Indo have been prepared and character-
ized. The analytical data and physical measurements are
consistent with the dimeric structures of the [Cu2(Indo)4L2]
complexes.
(Indo)4(DMF)2], where the average value of -2J is 294 cm-1 82
.
The strength of the antiferromagnetic coupling was independent
of the nature of the axial ligand. Conflicting evidence exists in
the literature concerning the influence of the nature of the
terminal ligand on the value of J. Some workers have proposed
that the value of J is only weakly dependent on the nature of
the axial donor ligand and is not a simple function of the base
strength of the bridging carboxylate ligand,66 whereas others
have concluded that the strength of the antiferromagnetic
interaction tends to increase as either the axial ligand, L, or the
carboxylate constituent becomes a stronger electron donor,83 and
the pKa value of L increases.84,85 In addition, larger J values
have been attributed to a weaker σ donation by the axial ligand,
with the suggestion that when there is a corresponding increase
in the ligand field of the four carboxylato oxygen atoms around
the metal ion, larger splitting of the d-d energy levels and a
blue shift of band I result.48,50 It is, however, well established
that the magnitude of the magnetic coupling is sensitive to the
Cu-O-C-O-Cu geometry,69 and the similarity in J and λmax
among the complexes suggests that they all retain a similar
Cu2O8 core.53,74
Structures of Cu Indomethacin in Relation to Pharma-
ceutical Activity. The [Cu2(Indo)4(L)2] complexes act as
effective anti-inflammatory drugs in dogs without leading to
gastrointestinal bleeding and death.26,77 This indicates that Indo
is not released from Cu2+ during ingestion via the gastrointes-
tinal tract, since IndoH itself cannot be used as an anti-
inflammatory drug for dogs because of these fatal side effects.78-80
To understand the pharmacology of these drugs and, hence,
their potential for use as human pharmaceutical agents, it is
important to understand the properties of the complexes in terms
of their stabilities and structures. In particular, it was important
to assess whether the terminal ligands had any effects on the
structures and properties of the complexes, since a variety of
pharmaceutical preparations have been used in which different
solvent molecules are coordinated to the Cu2+ center, in addition
to Indo.
The results of the spectroscopic and magnetic measurements
establish that a dinuclear Cu acetate-type structure is maintained
for all of the complexes. The X-ray structure of the DMF
complex not only confirmed this but also provided some
(81) Herring, F. G.; Landa, B.; Thomson, R. C.; Schewdtfeger, C. F. J.
Chem. Soc. A. 1971, 528-535.
(82) The unusual magnetic properties are not due to trapped oxygen. This
was confirmed by low-temperature magnetism studies, where O2 has
characteristic behavior. Furthermore, the EPR studies do not show
any triplet signals due to O2. Although several attempts have been
made at growing crystals of the aqua complex, we are yet to obtain
crystals suitable for XRD analysis.
(74) Melnik, M. Coord. Chem. ReV. 1982, 42, 259-293.
(75) Bencini, A.; Gatteschi, D. EPR of Exchange Coupled Systems;
Springer-Verlag: Berlin, 1990.
(76) David, L.; Cosar, O.; Chis, V.; Negoescu, A.; Vlasin, I. Appl. Magn.
Reson. 1994, 6, 521-528.
(77) IVS Annual; MIMS Publishing: Crows Nest, NSW, Sydney, 1997;
pp 145, 276.
(78) Adams, H. R. Veterinary Pharamacology and Therapeutics, 7th ed.;
Iowa State University Press: Ames, IA, 1995; pp 443-444.
(79) Ewing, G. O. J. Am. Vet. Med. Assoc. 1972, 161, 1665-1668.
(80) Menguy, R.; Desbaillets, L. Am. J. Dig. Dis. 1967, 12, 862-866.
(83) Jotham, R. W.; Kettle, S. F. A.; Marks, J. A. J. Chem. Soc., Dalton
Trans. 1972, 428-438.
(84) Marsh, W. E.; Carlisle, G. O.; Hanson, M. V. J. Mol. Struct. 1977,
40, 153-156.
(85) Marsh, W. E.; Carlisle, G. O.; Hanson, M. V. J. Inorg. Nucl. Chem.
1977, 39, 1839-1841.

1744 Inorganic Chemistry, Vol. 38, No. 8, 1999
Weder et al.
The only measurable difference found among all the com-
plexes is in the value of D, as measured by ZFEPR, with the
DMF and NMP ligands contributing to a greater extent to the
exchange between the ground state of one Cu and the excited
state of the other Cu, compared to the water ligand.
The magnitude of the axial splitting, D, is a function of two
terms:70,72,86,87 a dipolar contribution Ddip, which is related to
the Cu-Cu distance, and the exchange contribution, Dexch. The
latter term is more critical, for complexes with weaker-field
donors that have excited states nearer in energy to the ground
state, so yielding large values of Dexch and hence D. Conversely,
for a strong-field donor, the energy gap to the excited state is
larger, leading to a smaller value for Dexch and, hence, D.
pharmaceutical activities of the Cu-Indo drugs with different
axial ligands.26,88
In addition, the other important aspect of the magnetic and
EPR measurements is that they are extremely sensitive to
mononuclear Cu2+ at low temperatures as a result of the
diamagnetic ground states of the dinuclear species. Thus, both
techniques are valuable in quality control of pharmaceutical
preparations, since even small amounts of mononuclear Cu2+
can be quantified readily by such techniques. EPR spectroscopy,
however, is a more convenient technique for this application.
The results reported here establish that the concentrations of
such impurities are very low (<1%) in the complexes used for
pharmaceutical preparations.
Assuming that the Cu-Cu separation and, hence, Ddip are
comparable for the various complexes under investigation, it
would appear that the [Cu2(Indo)4(DMF)2] and [Cu2(Indo)4-
(NMP)2] complexes have a weaker ligand-field contribution
from their respective axial ligands (as indicated by a higher value
for D) when compared to those of the water complexes. These
very small differences in the value of D do not influence either
the UV-vis spectra (the energies of band I and band II are not
sensitive indicators to small differences in bonding) or the
strength of the antiferromagnetic exchange (J).
Acknowledgment. We acknowledge the support of this work
by the University of Sydney Major Equipment fund and an
Australian Research Council (ARC) RIEFP grant for the EPR
spectrometers, an ARC Large grant to Monash University
(K.S.M.), an ARC SPIRT grant (P.A.L., T.W.H., B.J.K., J.R.B.,
H.L.R.), and Biomedical Veterinary Research (BVR) for the
funding of a postgraduate scholarship for J.E.W. We also thank
Dr. Bradley Collins and Mr. Alan MacKenzie for useful
suggestions in the course of the IR study and Prowski refine-
ments, respectively.
The important conclusion that flows from the electronic and
magnetic studies of the complexes is their relative insensitivity
to changes in the axial ligands. Since the axial ligands are likely
to exchange during drug metabolism, these results show that
the physical properties and, hence, reactivities and efficacies
of the drug are likely to be insensitive to the nature of the axial
ligands. This is consistent with the lack of variation in the
Supporting Information Available: Tables of non-hydrogen
positional parameters, bond distances, bond angles, anisotropic thermal
parameters, and hydrogen atom positions and isotropic thermal
parameters for [Cu2(Indo)4(DMF)2], tables of observed and calculated
variable temperature magnetic susceptibility and magnetic moment
values per Cu ion in emu and µB units, respectively, for [Cu2(Indo)4-
(DMF)2], [Cu2(Indo)4(NMP)2], [Cu2(Indo)4(OH2)2]‚1.5H2O and [Cu2-
(Indo)4(OH2)2], figure of the diffuse reflectance UV-vis spectra of (a)
IndoH, (b) [Cu2(Indo)4(DMF)2], and (c) Cu(II) acetate monohydrate in
a KBr matrix, figure of the solid state IR spectra of (a) [Cu2(Indo)4-
(DMF)2] and (b) IndoH in a KBr matrix, and figures of the atomic
numbering system for Indo attached to C21 and C1* in the crystal
structure of [Cu2(Indo)4(DMF)2]. This material is available free of
charge via the Internet at http://pubs.acs.org.
(86) Mabbs, F. E.; Collison, D. Electron Paramagnetic Resonance of d
Transition Metal Compounds; Elsevier: Amsterdam, 1992.
(87) Blake, A. J.; Grant, C. M.; McInnes, E. J. L.; Mabbs, F. E.; Milne, P.
E. Y.; Parsons, S.; Rawson, J. M.; Winpenny, R. E. P. J. Chem. Soc.,
Dalton Trans. 1996, 21, 4077-4082.
(88) Regtop, H. L.; Biffin, J. R. Divalent Metal Complexes of Indomethacin,
Compositions and Medical Methods of Use Thereof. US Patent
5466824; Biochemical Veterinary Research Pty. Ltd., Australia, 1995.
IC981100X