Homoleptic cobalt and copper phenolate A2[M(OAr)4] compounds: The effect of phenoxide fluorination

Article abstract of DOI:10.1021/ic0493954

Two series of homoleptic phenolate complexes with fluorinated aryloxide ligands A₂[M(OAr)₄] with M = Co²⁺ or Cu ²⁺, OAr^- = (OC₆F₅)^- (OAr^F) or {3,5-OC₆H₃(CF₃) ₂}^- (OAr′), A⁺ = K (18-crown-6) ⁺, Tl⁺, Ph₄P⁺, Et₃HN ⁺, or Me₄N⁺ have been synthesized. Two related complexes with nonfluorinated phenoxide ligands have been synthesized and studied in comparison to the fluorinated aryloxides demonstrating the dramatic structural changes effected by modification of OPh to OAr^F. The compounds {K(18-crown-6)}₂[Cu(OAr^F)₄], 1a; {K(18-crown-6)}₂[Cu(OAr′)₄], 1b; [Tl ₂Cu(OAr^F)₄], 2a; [Tl₂Cu(OAr′) ₄], 2b; (Ph₄P)₂[Cu(OAr^F) ₄], 3; (ⁿBu₄N)₂[Cu(OAr ^F)₄], 4; (HEt₃N)₂[Cu(OAr ^F)₄], 5; {K(18-crown-6)}₂[Cu ₂(μ₂-OC₆H₅)₂(OC ₆H₅)₄], 6; {K(18-crown-6)} ₂[Co(OAr^F)₄], 7a; {K(18-crown-6)} ₂[Co(OAr′)₄], 7b; [Tl₂Co(OAr ^F)₄], 8a; [Tl₂Co(OAr′)₄], 8b; (Me₄N)₂[Co(OAr^F)₄], 9; [Cp ₂Co]₂[Co(OAr′)₄], 10; and {K(18-crown-6)}₂[Co₂(μ₂-OC₆H ₅)₂(OC₆H₅)₄], 11, have been characterized with UV-vis and multinuclear NMR spectroscopy and solution magnetic moment studies. Cyclic voltammetry was used to study 1a, 1b, 7a, and 7b, X-ray crystallography was used to characterize 1b, 3, 4, 5, 6, 7a, 7b; 10, and 11. The related [MX₄]^2- compound (Ph ₄P)₂[Co(OAr^F)₂Cl₂], 12, has also been synthesized and characterized spectroscopically, as well as with conductivity and single-crystal X-ray diffraction. Use of fluorinated aryloxides permits synthesis and isolation of the mononuclear, homoleptic phenolate anions in good yield without oligomerized side products. The reaction conditions that result in homoleptic 1a and 7a with OAr^F upon changing the ligand to OPh result in μ₂-OPh bridging phenoxides and the dimeric complexes 6 and 11. The [M(OAr^F)₄]^2- and [M(OAr′)₄]^2- anions in 1a, 1b, 3, 4, 5, 7a, 7b, 9, and 10 demonstrate that stable, isolable homoleptic phenolate anions do not need to be coordinatively or sterically saturated and can be achieved by increasing the electronegativity of the ligand.

Full text of DOI:10.1021/ic0493954

Inorg. Chem. 2004, 43, 7709
−7725
Homoleptic Cobalt and Copper Phenolate A2[M(OAr)4] Compounds: The
Effect of Phenoxide Fluorination
Marisa C. Buzzeo,† Amber H. Iqbal,† Charli M. Long,† David Millar,† Sonal Patel,† Matthew A. Pellow,†
Sahar A. Saddoughi,† Abigail L. Smenton,† John F. C. Turner,‡ Jay D. Wadhawan,§
,
Richard G. Compton,§ James A. Golen,| Arnold L. Rheingold,| and Linda H. Doerrer*,†
Chemistry Department, Barnard College, 3009 Broadway, New York, New York 10027,
Department of Chemistry, UniVersity of Tennessee, 409 Buehler Hall, KnoxVille, Tennessee 37996,
Physical and Theoretical Chemistry Laboratory, UniVersity of Oxford, South Parks Road,
Oxford, OX1 3QZ, U.K., Department of Chemistry and Biochemistry, UniVersity of
CaliforniasSan Diego, 9500 Gilman DriVe, MC 0358, La Jolla, California 92093, and
Department of Chemistry and Biochemistry, UniVersity of MassachusettssDartmouth,
North Dartmouth, Massachusetts 02747
Received May 8, 2004
+
Two series of homoleptic phenolate complexes with fluorinated aryloxide ligands A2[M(OAr)4] with M
)
Co2 or
-
Cu2+, OAr-
)
(OC6F5)- (OArF) or
{
3,5-OC6H3(CF3)2
}
(OAr′ )
), A+ K (18-crown-6)+, Tl+, Ph4P+, Et3HN+, or
Me4N+ have been synthesized. Two related complexes with nonfluorinated phenoxide ligands have been synthesized
and studied in comparison to the fluorinated aryloxides demonstrating the dramatic structural changes effected by
modification of OPh to OArF. The compounds
[Tl2Cu(OArF)4], 2a; [Tl2Cu(OAr
K(18-crown-6) 2[Cu2( 2-OC6H5)2(OC6H5)4], 6;
[Tl2Co(OArF)4], 8a; [Tl2Co(OAr )4], 8b; (Me4N)2[Co(OArF)4], 9; [Cp2Co]2[Co(OAr
[Co2( 2-OC6H5)2(OC6H5)4], 11, have been characterized with UV vis and multinuclear NMR spectroscopy and solution
{
K(18-crown-6)
)4], 2b; (Ph4P)2[Cu(OArF)4], 3; (nBu4N)2[Cu(OArF)4], 4; (HEt3N)2[Cu(OArF)4], 5;
K(18-crown-6) K(18-crown-6) 2[Co(OAr )4], 7b;
2[Co(OArF)4], 7a;
)4], 10; and K(18-crown-6)
} {K(18-crown-6)}2[Cu(OAr′)4], 1b;
2[Cu(OArF)4], 1a;
′
{
}
µ
{
}
{
}
′
′
′
{
} -
2
µ
−
magnetic moment studies. Cyclic voltammetry was used to study 1a, 1b, 7a, and 7b. X-ray crystallography was
-
used to characterize 1b, 3, 4, 5, 6, 7a, 7b, 10, and 11. The related [MX4]2 compound (Ph4P)2[Co(OArF)2Cl2], 12,
has also been synthesized and characterized spectroscopically, as well as with conductivity and single-crystal
X-ray diffraction. Use of fluorinated aryloxides permits synthesis and isolation of the mononuclear, homoleptic
phenolate anions in good yield without oligomerized side products. The reaction conditions that result in homoleptic
1a and 7a with OArF upon changing the ligand to OPh result in
µ
2-OPh bridging phenoxides and the dimeric
-
-
complexes 6 and 11. The [M(OArF)4]2 and [M(OAr
′
)4]2 anions in 1a, 1b, 3, 4, 5, 7a, 7b, 9, and 10 demonstrate
that stable, isolable homoleptic phenolate anions do not need to be coordinatively or sterically saturated and can
be achieved by increasing the electronegativity of the ligand.
Introduction
often provide archetypal examples of structural and electronic
motifs that are more easily understood because of the
compositional simplicity of the coordination sphere (e.g.,
ferrocene,2 carbonyls,3 dithiolates,4 2,2′-bipyridyl,5 and hy-
drides6). Not all monodentate ligands stabilize metals equally
Homoleptic transition-metal complexes have been an
important part of the inorganic chemical literature since
Werner’s characterization1 of [Co(NH3)6]3+. Such compounds
* Author to whom correspondence should be addressed. E-mail:
ldoerrer@barnard.edu.
(2) Wilkinson, G.; Rosenblum, M.; Whiting, M. C.; Woodward, R. B. J.
Am. Chem. Soc. 1952, 74, 2125-2126.
† Barnard College.
‡ University of Tennessee.
(3) Lupinetti, A. J.; Strauss, S. H.; Frenking, G. Prog. Inorg. Chem. 2001,
49, 1-112.
§ University of Oxford.
| University of CaliforniasSan Diego.
University of MassachusettssDartmouth.
(1) Kauffman, G. B. J. Chem. Educ. 1966, 43, 155-156.
(4) Stiefel, E. I.; Eisenberg, R.; Rosenberg, R. C.; Gray, H. B. J. Am.
Chem. Soc. 1966, 88, 2956-2966.
(5) Constable, E. C. AdV. Inorg. Chem. 1989, 34, 1-63.
10.1021/ic0493954 CCC: $27.50
Published on Web 10/29/2004
© 2004 American Chemical Society
Inorganic Chemistry, Vol. 43, No. 24, 2004 7709

Buzzeo et al.
well, and some have proven quite challenging to impose on
transition metals such as methyl groups in homoleptic alkyl
compounds.7-9 Phenoxide and aryloxide groups have been
resistive also to preparation as homoleptic derivatives.10,11
Compounds that contain a subset of aryloxide ligands are
legion, owing to interest in these ligands as metal-oxide
precursors,12,13 ancilliary ligands in catalysis,14-16 and models
for metal-tyrosine interactions in biology.17-20 Much work
has demonstrated the tendency of the phenoxide oxygen to
bridge metal centers, forming oligomers and polymers that
resist purification and characterization.10
Inclusion of the phenoxide group in a chelate ring, such
as salen-21 or calixarene-based22 ligands is the most common
approach to forming mononuclear species. Early transition
metals in their highest oxidation state with no ligand-field
stabliziation energies favor a coordinatively saturated metal
center, such as the W6+ center in [W(OC6H4-p-Me)6].23 In
the early studies of homoleptic alkoxides, this “preference”
for coordinative saturation of metal centers was observed to
have predictive value in structural determination.24 Bridging
phenoxides also have been avoided by introducing steric bulk
in the 2,6 positions on the phenyl ring, preventing close
contact of the oxygen atom with more than one metal
center.15,25 Examples of this can be found in the s,26 p,27,28
d,29,30 and f 31 blocks of the periodic table. More recently,
fluorinated aryloxides have been investigated, and a few
sterically saturated, homoleptic species have been reported
for Al3+, Nb5+, and Ta5+. These [M(OAr)n]- anions are
coordinatively saturated at the metal center and were prepared
as possible weakly coordinating anions (WCAs) for use in
olefin polymerization catalysis,32,33 but it was observed that
highly electrophilic metal centers (i.e., [Cp2ZrMe]+) could
interact strongly enough with the fluorine atoms in the anions
to reduce requisite olefin binding.
Previously, late metal examples of homoleptic alkoxide
or aryloxide anions analogous to the well-known anions
[MX4]2-, X ) halogen or pseudohalogen, were not well-
known, and their conspicuous paucity was noted in the lit-
erature.11 Herein we report a detailed and thorough structural
study of two families of homoleptic phenolate anions,
[Co(OAr)4]2- and [Cu(OAr)4]2-, that use neither a chelate
ring nor steric encumbrance to prevent oligomerization and
yet are not coordinatively saturated at the metal center.
Experimental Section
General Information. All studies were carried out at room
temperature on a Schlenk line or in a nitrogen- or argon-filled
glovebox. For solvents dried in stills, CH2Cl2 and CD2Cl2 were
refluxed over CaH2. Tetrahydrofuran (THF) and hexane were
refluxed over potassium. CH3CN was predried over P2O5 and
distilled from CaH2. Toluene was refluxed over sodium, and ether
was refluxed over NaK3. All solvents were refluxed under nitro-
gen or argon. Some work also was done with solvents (hexanes,
CH2Cl2, and toluene) dried in a nitrogen-filled MBraun solvent
purification system using Al2O3. Celite was dried overnight in vacuo
while heated to 125 °C with an oil bath. KOC6H3(CF3)2 (KOAr′)
and KOC6H5 (KOPh) were prepared according to our previously
published procedure.34 {nBu4N}2[CuCl4] was prepared by the
literature procedure.35 18-crown-6 was recrystallized from either
hexanes or CH3CN. All other reagents were obtained commercially
and were not purified further. NMR spectra were measured on a
(6) King, R. B. Coord. Chem. ReV. 2000, 200-202, 813-829.
(7) LaPointe, A. M. Inorg. Chim. Acta 2003, 345, 359-362.
(8) Schulzke, C.; Enright, D.; Sugiyama, H.; LeBlanc, G.; Gambarotta,
S.; Yap, G. P. A.; Thompson, L. K.; Wilson, D. R.; Duchateau, R.,
Organometallics 2002, 21, 3810-3816.
(9) Wilkinson, G. Sci. Prog. 1994, 77, 15-27.
(10) Bradley, D. C.; Mehrotra, R. C.; Rothwell, I. P.; Singh, A. Alkoxo
and Aryloxo DeriVatiVes of Metals; Academic Press: New York, 2001.
(11) Cotton, F. A.; Liu, C. Y.; Murillo, C. A.; Wang, X. Inorg. Chem.
2003, 42, 4619-4623.
1
Varian 300-MHz, Bruker 300- or 400-MHz spectrometer. H and
(12) Henderson, K. W.; Mulvey, R. E.; Reinhard, F. B. M.; Clegg, W.;
Horsburgh, L. J. Am. Chem. Soc. 1994, 116, 10777-10778.
(13) Boyle, T. J.; Pedrotty, D. M.; Alam, T. M.; Vick, S. C.; Rodriguez,
M. A. Inorg. Chem. 2000, 39, 5133-5146.
13C chemical shifts were referenced to (CH3)4Si via the resonance
of residual protiosolvent (1H) or the 13C resonance of the solvent.
19F shifts were referenced to external CFCl3. UV-vis data were
recorded with a Varian Cary 50 spectrometer. IR spectra were
recorded on a Perkin-Elmer Spectrum BX instrument with Nujol
mulls between NaCl plates. Solution phase magnetic susceptibilities
were determined by Evans’ method36,37 in d6-acetone with 10%
(Me3Si)2O and the same solution as an internal reference and are
reported after correction using appropriate diamagnetic terms.38,39
Microanalyses were preformed by H. Kolbe Microanalytisches
Laboratorium, Mu¨lheim an der Ruhr, Germany.
(14) Churchill, M. R.; Ziller, J. W.; Freudenberger, J. H.; Schrock, R. R.
Organometallics 1984, 3, 1554-1562.
(15) Rothwell, I. P. Chem. Commun. 1997, 1331-1338.
(16) Darensbourg, D. J.; Wildeson, J. R.; Yarbrough, J. C.; Reibenspies,
J. H. J. Am. Chem. Soc. 2000, 122, 12487-12496.
(17) Koch, S. A.; Millar, M. J. Am. Chem. Soc. 1982, 104, 5255-5257.
(18) Kim, E.; Chufan, E. E.; Kamaraj, K.; Karlin, K. D. Chem. ReV. 2004,
104, 1077-1133.
(19) Lewis, E. A.; Tolman, W. B. Chem. ReV. 2004, 104, 1047-1076.
(20) Mirica, L. M.; Ottenwaelder, X.; Stack, T. D. P. Chem. ReV. 2004,
104, 1013-1045.
(21) Jacobsen, E. N. Acc. Chem. Res. 2000, 33, 421-431.
(22) Roundhill, D. M. Prog. Inorg. Chem. 1995, 43, 533-592.
(23) Clegg, W.; Errington, R. J.; Kraxner, P.; Redshaw, C. J. Chem. Soc.,
Dalton Trans. 1992, 1431-1438.
Synthetic Procedures. K(OArF). This compound has been
reported previously in the literature as a reagent,40-44 but we are
(24) Bradley, D. C.; Mehrotra, R. C.; Rothwell, I. P.; Singh, A. In Alkoxo
and Aryloxo DeriVatiVes of Metals; Academic Press: New York, 2001;
Chapter 2, p 3.
(25) Gardiner, I. M.; Bruck, M. A.; Wexler, P. A.; Wigley, D. E. Inorg.
Chem. 1989, 28, 3688-3695.
(26) Bell, N. A.; Coates, G. E.; Shearer, H. M. M.; Twiss, J. J. Chem.
Soc., Chem. Commun. 1983, 840-841.
(27) Cetinkaya, B.; Gumrukcu, I.; Lappert, M. F.; Atwood, J. L.; Rogers,
R. D.; Zaworotko, M. J. J. Am. Chem. Soc. 1980, 102, 2088-2089.
(28) Hitchcock, P. B.; Lappert, M. F.; Thomas, S. A.; Thorne, A. J.; Carty,
A. J.; Taylor, N. J. J. Organomet. Chem. 1986, 315, 27-44.
(29) Meyer, D.; Osborn, J. A.; Wesolek, M. Polyhedron 1990, 9, 1311-
1315.
(31) McKee, S. D.; Burns, C. J.; Avens, L. R. Inorg. Chem. 1998, 37,
4040-4045.
(32) Sun, Y.; Metz, M. V.; Stern, C. L.; Marks, T. J. Organometallics 2000,
19, 1625-1627.
(33) Metz, M. V.; Sun, Y.; Stern, C. L.; Marks, T. J. Organometallics 2002,
21, 3091-3102.
(34) Buzzeo, M. C.; Zakharov, L. N.; Rheingold, A. L.; Doerrer, L. H. J.
Mol. Struct. 2003, 657, 19-24.
(35) Gill, N. S.; Taylor, F. B. Inorg. Synth. 1967, 9, 136-142.
(36) Evans, D. F. J. Chem. Soc. 1959, 2003-2005.
(37) Sur, S. K. J. Magn. Reson. 1989, 82, 169-173.
(38) O’Connor, J. C. Prog. Inorg. Chem. 1982, 29, 203-283.
(39) Kahn, O. Molecular Magnetism; VCH: New York, 1993.
(40) Pasenok, S. V.; Yagupolskii, Y. L.; Tyrra, W.; Naumann, D. Z. Anorg.
Allg. Chem. 1999, 625, 831-833.
(30) Darensbourg, D. J.; Wildeson, J. R.; Lewis, S. J.; Yarbrough, J. C. J.
Am. Chem. Soc. 2002, 124, 7075-7083.
7710 Inorganic Chemistry, Vol. 43, No. 24, 2004

Homoleptic Cobalt and Copper Phenolate Compounds
not aware of any characterization data and therefore present the
synthesis and spectroscopy here. A portion of HOArF (6.85 g, 37.4
mmol) was dissolved in 20 mL of THF to which a slurry of KH
(1.5 g, 37.4 mmol) in 20 mL of THF was added slowly over several
minutes with vigorous stirring. The solution was allowed to stir
for 45 min, and the THF was removed in vacuo. The crude material
was recrystallized from THF and pentanes. Recrystallized yield:
7.5 g (90%). 13C{1H} NMR (δ, ppm {(CD3)2CO}): 127.5 (p, d,
(212.6 mg, 0.804 mmol) and 8 mL of CH2Cl2 were added. After
30 min, the deep-green solution was separated from a white
precipitate, presumed to be KBr, via filtration through Celite and
concentrated to dryness in vacuo, and the crude product was
recrystallized from CH2Cl2 and pentanes. The product was iso-
1
lated in 53% recrystallized yield (300 mg). H NMR (δ, ppm,
{(CD3)2CO}): 3.64 (s, 48H, CH2). 13C{1H} NMR (δ, ppm,
{(CD3)2CO}): 70.97 (s, CH2). UV-vis (THF) (λmax, nm (ꢀM, cm-1
M-1)): 265 (3900), 329 (1800), 422 (1010). Anal. Calcd for
C48H48O16F20K2Cu: C, 41.10; H, 3.45. Found: C, 41.18; H, 3.50.
µeff (Evans’ method) ) 1.74µB.
1
1JCF ) 228.5 Hz), 139.63 (m, d, JCF ) 245.8 Hz), 142.32 (o, d,
1JCF ) 229.0 Hz), 147.7 (i, d, 2JCF ) 13.2 Hz). 19F NMR (δ, ppm,
3
{(CD3)2CO}): -173.6 (o-F, m-F, multiplet), -193.9 (p-F, t, JFF
) 18.3 Hz).
{K(18-crown-6)}2[Cu(OAr′)4], 1b. A portion of KOAr′ (350.0
mg, 1.305 mmol) was dissolved in 4 mL of THF and then added
to CuBr2 (72.9 mg, 0.326 mmol) following the same procedure as
for 1a, including the addition of 2 equiv of 18-crown-6 (172.5 mg,
0.653 mmol). The crude product was recrystallized from
CH2Cl2 and pentanes in 48% yield (248 mg). X-ray quality crystals
Tl(OArF). This compound also previously has been reported in
the literature45-50 as a reagent, and we report the synthesis and
spectroscopic data that characterize this material here for the same
reasons as for K(OArF). A 25-g portion of TlOEt (100.23 mmol)
was added via syringe to 18.45 g of HOArF (100.23 mmol) in 500
mL of CH2Cl2 with vigorous stirring. The reaction was allowed to
stir for 5 h after which time a large amount of white precipitate
had formed. The supernatant was pale purple-brown and was
decanted. The product was washed four times with ∼125 mL
portions of hexane and dried in vacuo. Yield: 33.9 g (87.4%).
13C{1H} NMR (δ, ppm, {(CD3)2CO}): 131.77 (p, d, 1JCF ) 239.4
1
were grown from the same solvent mixture. H NMR (δ, ppm,
{(CD3)2CO}): 3.72 (s, 48H, CH2). 13C{1H} NMR (δ, ppm,
{(CD3)2CO}): 71.02 (s, CH2). UV-vis (THF) (λmax, nm (ꢀM,
cm-1 M-1)) 308 (7810), 360 (sh) (1659). Anal. Calcd for
C56H60O16F24K2Cu: C, 42.39; H, 3.81. Found: C, 42.35; H, 3.89.
µeff (Evans’ method) ) 1.79µB.
1
Hz), 139.37 (ipso, s), 139.53 (m, d, JCF ) 239.4 Hz), 142.39 (o,
[Tl2Cu(OArF)4], 2a. A portion of CuBr2 (72.1 mg, 0.3226 mmol)
was dissolved in 6 mL of THF to give an orange-brown solu-
tion. A 3-mL solution of 4 equiv of TlOArF (500.0 mg, 1.291
mmol) was added to the first solution causing instant precipi-
tation of a white solid, assumed to be TlBr. The solution was
allowed to stir overnight, after which time the solid was removed
via filtration, and the filtrate was concentrated to ∼2 mL in vacuo.
The solution volume was increased to 20 mL with pentane and
left at room temperature. Very fine brown needles precipitated
overnight. This material was recrystallized twice more from
THF/pentane at -40 °C, collected by filtration, and dried in
vacuo. Final recrystallized yield 210 mg (54% yield). UV-vis
(THF) (λmax, nm (ꢀM, cm-1 M-1)) 441 (2450). Anal. Calcd for
C24O4F20CuTl2: C, 23.93, F, 31.54. Found: C, 24.10, F, 31.31.
µeff (Evans’ method) ) 1.79µB.
1
d, JCF ) 233.0 Hz). 19F NMR (δ, ppm, {(CD3)2CO}): -159.73
3
(p-F, s), -165.57 (m-F, t, JFF ) 21.2 Hz), -176.88 (o-F, m).
Tl(OAr′). A procedure similar to that for Tl(OArF) was followed.
A 19.73-g portion of TlOEt (79 mmol) was added via syringe to
19.11 g of HOAr′ (83 mmol) in 100 mL of toluene with vigorous
stirring. The reaction was allowed to stir overnight after which
time a large amount of white precipitate had formed. Toluene and
ethanol were removed in vacuo, and the product was redissolved
in warm toluene, filtered, and upon cooling layered with room-
1
temperature hexanes. Yield: 25.6 g (74.7%). H NMR (δ, ppm,
{(CD3)2CO}): 7.004 (p, 1H), 7.288 (o, 2H). 13C{1H} NMR (δ,
3
ppm, {(CD3)2CO}): 107.42 (p, septet, JCF ) 4.1 Hz), 119.86 (o,
2
1
s), 132.69 (m, q, JCF ) 31.78 Hz), 125.31 (CF3, q, JCF ) 271.9
Hz), 168.17 (ipso, s). 19F NMR (δ, ppm, {(CD3)2CO}): -62.20
(CF3, s).
[Tl2Cu(OAr′)4], 2b. A procedure analogous to that for 2a was
followed with CuBr2 (51.5 mg, 0.2307 mmol) and 4 equiv of
TlOArF (400.0 mg, 0.9228 mmol). Recrystallization from THF and
hexanes gave 278 mg (45.6% yield) of dark brown microcrystalline
product. UV-vis (THF) (λmax, nm (ꢀM, cm-1 M-1)): 421 (2500).
Anal. Calcd for C46H40O7.5F24CuTl2 (2b‚3.5 THF): C, 33.67; H,
2.46; F, 27.78. Found: C, 33.54; H, 2.41; F, 27.92. µeff (Evans’
method) ) 1.94µB.
{K(18-crown-6)}2[Cu(OArF)4], 1a. A portion of KOArF (357
mg, 1.609 mmol) was dissolved in 4 mL of THF and then added
to an orange-brown slurry of (89.8 mg, 0.402 mmol) CuBr2 in 4
mL of THF causing immediate formation of a dark green mixture.
The solution was stirred at room temperature overnight after which
time the solution was concentrated to dryness and triturated three
times with 5 mL of hexanes. A 2-equiv portion of 18-crown-6
(Ph4P)2[Cu(OArF)4], 3. A portion of CuBr2 (223.4 mg, 1.00
mmol) was dissolved in 10 mL of THF, and 4 equiv of TlOArF
(1.5 g, 4.00 mmol) in a further 10 mL of THF was added. Formation
of a precipitate, presumably TlBr, was observed instantly. The
solution was stirred overnight and then filtered through Celite to
remove the precipitate. To the resulting orange-brown solution, 2
equiv of Ph4PI (932.6 mg, 2.00 mmol) was added, causing instant
formation of a yellow solid, presumed to be TlI, and a solution
color change to deep green. The solution was allowed to stir for 2
h and filtered again through Celite, and the filtrate was concentrated
to dryness in vacuo. The crude powder was recrystallized three
times from CH2Cl2/pentane mixtures. Yield: 1.07 g (73%). 1H
NMR (δ, ppm, {(CD3)2CO}): 7.838 (br), 8.018 (br). 13C{1H} NMR
(δ, ppm, {(CD3)2CO}): 119.15 (ipso, d, 1JC-P ) 89.6 Hz), 131.51
(o, d, 2JC-P ) 12.7 Hz), 135.83 (m, d, 3JC-P ) 10.1 Hz), 136.49 (s,
para). UV-vis (THF) (λmax, nm (ꢀM, cm-1 M-1)): 330 (sh) (5470),
(41) Seidel, S. W.; Schrock, R. R.; Davis, W. M. Organometallics 1998,
17, 1058-1068.
(42) Schofield, M. H.; Schrock, R. R.; Park, L. Y. Organometallics 1991,
10, 1844-1851.
(43) Schrock, R. R.; Kolodziej, R. M.; Liu, A. H.; Davis, W. M.; Vale, M.
G. J. Am. Chem. Soc. 1990, 112, 4338-4345.
(44) Liu, A. H.; Murray, R. C.; Dewan, J. C.; Santarsiero, B. D.; Schrock,
R. R. J. Am. Chem. Soc. 1987, 109, 4282-4291.
(45) Zanella, P.; Mascellani, N.; Cason, A.; Garon, S.; Rossetto, G.; Carta,
G. Appl. Organomet. Chem. 2001, 15, 717-724.
(46) Curnow, O. J.; Hughes, R. P. J. Am. Chem. Soc. 1992, 114, 5895-
5897.
(47) Danopoulos, A. A.; Wilkinson, G.; Sweet, T. K. N.; Hursthouse, M.
B. Polyhedron 1994, 13, 2899-2905.
(48) Danopoulos, A. A.; Wilkinson, G.; Sweet, T. K. N.; Hursthouse, M.
B. J. Chem. Soc., Dalton Trans. 1994, 1037-1049.
(49) Hughes, R. P.; Zheng, X.; Ostrander, R. L.; Rheingold, A. L.
Organometallics 1994, 13, 1567-1568.
(50) Davidson, J. L.; Holz, B.; Leverd, P. C.; Lindsell, W. E.; Simpson,
N. J. J. Chem. Soc., Dalton Trans. 1994, 3527-3532.
Inorganic Chemistry, Vol. 43, No. 24, 2004 7711

Buzzeo et al.
411 (3630). Anal. Calcd for C72H40O4F20CuP2: C, 58.65; H, 2.73.
Found: C, 58.63; H, 2.75. µeff (Evans’ method) ) 1.69µB.
(nBu4N)2[Cu(OArF)4], 4. A portion of (nBu4N)2[CuCl4] (1.78 g,
0.258 mmol) and 4 equiv of TlOArF (4.00 g, 1.03 mmol) were
dissolved in toluene resulting in precipitation of presumed TlCl
and formation of a green solution. The white precipitate was
removed via filtration, and the solution was placed at -80 °C
overnight. Green crystals were formed, isolated via filtration,
washed with hexanes, and dried under vacuum. Single crystals for
X-ray crystallography were grown slowly from THF/pentane
mixtures. 1H NMR (δ, ppm, {(CD3)2CO}): 1.034 (b, 3H, C-CH3),
1.502 (b, 2H, CH2-CH3), 1.753 (b, 2H, CH2-CH2-CH3), 3.245
(b, 2H, CH2-C-C-C), 13C NMR (δ, ppm, {(CD3)2CO}): 13.359
(C-CH3), 19.930 (CH2-CH3), 24.055 (CH2-CH2-CH3), 59.032
(CH2-C-C-C). UV-vis (THF) (λmax, nm (ꢀM, cm-1 M-1)): 322
(sh) (4810), 414 (3370). Anal. Calcd for C56H72N2O4F20Cu: C,
52.52; H, 5.67; N, 2.19. Found: C, 52.61; H, 5.69; N, 2.09. µeff
(Evans’ method) ) 1.87µB.
powder. A 2-equiv portion of 18-crown-6 (0.900 mmol, 238.0 mg)
was added to 10 mL of CH2Cl2 to give a cobalt blue solution that
was filtered through Celite and concentrated to dryness. The crude
product was recrystallized from CH2Cl2 layered with hexanes to
give crystals that rapidly desolvated upon removal of the super-
natant. Complete removal of the solvent yielded 390 mg of blue
powder (62% yield). X-ray quality crystals were grown from THF
1
and pentanes. H NMR (δ, ppm, {(CD3)2CO}): 3.440 (s, 48H,
CH2). 13C{1H} NMR (δ, ppm, {(CD3)2CO}): 70.72 (s, CH2).
UV-vis (THF) (λmax, nm (ꢀM, cm-1 M-1)): 532 (110), 598 (220),
651 (260). Anal. Calcd for C48H48O16F20K2Co: C, 41.24; H, 3.46.
Found: C, 41.28; H, 3.43. µeff (Evans’ method) ) 4.53µB.
{K(18-crown-6)}2[Co(OAr′)4], 7b. A portion of CoI2 (0.1896
mmol, 59.3 mg) was dissolved in 5 mL of THF to which was added
a solution of 4 equiv of KOAr′ (0.7585 mmol, 208.9 mg) in 5 mL
of THF. Subsequent workup was analogous to that of 7a, including
addition of 2 equiv of 18-crown-6 (0.3792 mmol, 100 mg), removal
of precipitates, and recrystallization from CH2Cl2 and hexanes to
give 161 mg (54.2% yield) of cobalt blue crystalline material. X-ray
(HEt3N)2[Cu(OArF)4], 5. To a portion of CuCl2‚2H2O (830.6
mg, 6.18 mmol) in 25 mL of CH2Cl2, a mixture of 4 equiv of Et3N
(3.44 mL, 24.71 mmol) and 4 equiv of HOArF (4.549 g, 24.71
mmol) in 25 mL of CH2Cl2 was added. N.B. The amine and phenol
must react together prior to introduction of CuII because Et3N readily
deprotonates water from CuCl2‚2H2O and forms Et3HNCl and the
insoluble, bright blue [Cu(OH)2] precipitate. The reaction mixture
was allowed to stir overnight and then added to an equal volume
of water. The ammonium salts were removed, the organic layer
was washed with water twice more, and the combined organic
fractions were dried over MgSO4. The desiccant was removed via
filtration, and the product was recrystallized from THF and hexanes
to give 5.23 g (84.7% yield) of a yellow-green powder. X-ray
1
quality crystals were grown from the same mixture. H NMR (δ,
ppm, {(CD3)2CO}): 3.674 (s, 48H, CH2). 13C{1H} NMR (δ, ppm,
{(CD3)2CO}): 70.96 (s, CH2). UV-vis (THF) (λmax, nm (ꢀM, cm-1
M-1)): 550 (360), 597 (540), 632 (510). Anal. Calcd for
C56H60O16F24K2Co: C, 42.51; H, 3.82. Found: C, 42.58; H, 3.87.
µeff (Evans’ method) ) 5.02µB.
[Tl2Co(OArF)4], 8a. In the drybox 4 equiv of TlOArF (1.292 g,
3.33 mmol) was dissolved in THF and 1 equiv of CoI2 (0.261 g,
0.833 mmol) was added while the solution was stirring. A dark
deep blue solution with yellow precipitate was formed. After the
solution was stirred overnight, it was filtered through Celite, and
the solvent was removed in vacuo. A dark purple solid remained,
which was triturated 3 times with 2 mL of hexane. The product
was dissolved in a minimum amount of toluene, filtered, and layered
with hexane. The mixture was stored at -30 °C, and after 1 day
the product was isolated by filtration, leaving a purple colored solid
in 73% yield. X-ray quality crystals were grown from toluene and
hexanes. UV-vis (toluene) (λmax, nm (ꢀM, M-1 cm-1)): 510 (320),
530 (350), 584.9 (530). UV-vis (THF) (λmax, nm (ꢀM, M-1
cm-1)): 533 (230), 586 (340). Anal. Calcd for C24O4F20CoTl2: C,
24.02; F, 31.67. Found: C, 23.86; F, 31.43. µeff (Evans’ method)
) 4.46µB.
1
quality crystals were grown from the same mixture. H NMR (δ,
ppm, {(CD3)2CO}): 1.386 (br, 12H, CH3), 3.006 (br, 8H, CH2).
13C{1H} NMR (δ, ppm, {(CD3)2CO}): 9.21 (s, CH3), 69.27 (s,
CH2) UV-vis (THF) (λmax, nm (ꢀM, cm-1 M-1)): 264 (9870), 330
(sh) (2380) 432 (1990). Anal. Calcd for C36H32N2O4F20Cu: C,
43.23; H, 3.22; N, 2.80. Found: C, 43.32; H, 3.16; N, 2.77. µeff
(Evans’ method) ) 1.72µB.
{K(18-crown-6)}2[Cu2(µ2-OC6H5)2(OC6H5)4], 6. A 1-equiv
portion of CuBr2 (1.009 mmol, 225.3 mg) was dissolved in 5 mL
of THF to which 3 equiv of KOPh (3.026 mmol, 400.0 mg) in 15
mL of THF was added. The reaction mixture was stirred overnight,
then the resulting brown solution was filtered through Celite to
remove white (presumed) KBr, and the filtrate was concentrated
to dryness in vacuo. After the sample was tritruated three times
with 3 mL of CH2Cl2 and 1 equiv of 18-crown-6 (1.009 mmol,
266.6 mg) was added, the solution was filtered again through
Celite and concentrated to dryness and the orange-brown prod-
uct was recrystallized from CH2Cl2 and hexanes to give 160 mg
of dark brown microcrystalline material (24.6% yield). X-ray
quality crystals were grown from the same mixture. 1H NMR
(δ, ppm, {(CD3)2CO}): 3.660 (s, 24H, CH2). 13C{1H} NMR (δ,
ppm, {(CD3)2CO}): 71.01 (s, CH2) UV-vis (THF) (λmax, nm (ꢀM,
cm-1 M-1)): 281 (13,250), 390 (br) (1370) Anal. Calcd For
C60H78O18K2Cu2: C, 55.75; H, 6.08. Found: C, 55.59; H, 6.15.
µeff (Evans’ method) ) 2.74µB.
[Tl2Co(OAr′)4], 8b. A 4-equiv portion of Tl(OAr′) (3.763 g,
8.68 mmol) was dissolved in THF, and 1 equiv of CoI2 (0.679 g,
2.17 mmol) was added while stirring. A dark blue solution with
yellow precipitate was formed. Workup followed the same proce-
dure as that for 8a. The resulting purple powder was dissolved in
a minimum amount of toluene, filtered, layered with hexane for
recrystallization, and left at -30 °C. A purple crystalline solid was
obtained in 52% yield. Crystals were suitable for X-ray crystal-
1
lography. H NMR (δ, ppm, CD2Cl2): 2.64 (s, 3H, C6H5-CH3),
7.79 (b, 5H, C6H5-CH3) 13C NMR (δ, ppm, CD2Cl2): 22.15(s,
C6H5-CH3), 139.76 (ipso, C6H5-CH3), 129.87 (o, C6H5-CH3),
130.70 (m, C6H5-CH3), 126. 94 (p, C6H5-CH3). 19F NMR (δ, ppm,
CD2Cl2): -73.74 (s, CF3). UV-vis (toluene) (λmax, nm (ꢀM, M-1
cm -1)): 547 (300), 598 (430). UV-vis (THF) (λmax, nm (ꢀM, M-1
cm -1)): 479 (110), 482 (110), 522 (200), 544 (230), 599 (300).
Anal. Calcd for C46H28O4F24Tl2Co (8b‚2 toluene): C, 35.23; H,
1.80; F, 29.07. Found: C, 35.04; H, 1.71; F, 29.22. µeff (Evans’
method) ) 6.06µB.
{K(18-crown-6)}2[Co(OArF)4], 7a. A 1-equiv portion of CoI2
(0.450 mmol, 140.8 mg) and 4 equiv of KOArF (1.80 mmol and
400.0 mg) were combined in 20 mL of THF and allowed to stir
overnight. The solution turned deep cobalt blue, and white
(presumed) KI precipitated. The solid was removed via filtration
through Celite, the solvent was removed in vacuo, and the resultant
blue powder was triturated 3 times with hexane to give a purple
(Me4N)2[Co(OArF)4], 9. In the drybox, 1 equiv of [Tl2Co-
(OArF)4], 8a, (100 mg, 0.8334 mmol) was slurried in CH2Cl2, and
2 equiv of Me4NI (33.5 mg, 0.833 mmol) was added while the
7712 Inorganic Chemistry, Vol. 43, No. 24, 2004

Homoleptic Cobalt and Copper Phenolate Compounds
Table 1. Summary of X-ray Crystallographic Data for Copper Complexes
1b
3
4
5
6
formula
C56H60CuF24K2O16
1586.78
P1h
C72H40CuF20O4P2
1474.52
P21/c
12.0060(12)
18.7195(18)
14.9767(15)
C56H72CuF20N2O4
1280.70
Pbca
15.0601(9)
17.5377(11)
22.8468(15)
C36H32CuF20N2O4
1000.18
P21/c
12.5488(7)
12.4550(7)
14.0038(8)
C61H80Cl2Cu2K2O18
1377.43
P21/n
15.205(5)
21.230(8)
20.040(7)
-
formula weight
space group
a, Å
b, Å
c, Å
12.3827(13)
13.0390(14)
13.2356(14)
100.929(2)
112.815(2)
111.913(2)
1682.1(3)
1, 0.5
R, deg
â, deg
108.055(2)
114.282(1)
91.224(7)
γ, deg
V, Å3
3200.2(5)
2, 0.5
6034.3(7)
4, 0.5
1995.10(19)
2, 0.5
6468(4)
4, 1
Z, Z′
crystal color, habit
F(calc), g cm-3
µ(Mo KR), cm-1
temp, K
green, block
1.566
5.78
green, block
1.530
5.01
green, block
1.410
4.69
green, block
1.665
6.83
black, block
1.415
9.37
218(2)
293(2)
293
218(2)
218(2)
reflections measured
reflections ind
R(F), %a
10 718
7145 (Rint ) 0.0200)
10.50
15 445
5015 (Rint ) 0.0617)
5.20
36 069
7277 (Rint ) 0.0313)
5.05
14 444
4751 (Rint ) 0.0242)
3.47
38 465
11384 (Rint ) 0.0368)
4.98
R(wF2),%b
28.85
9.63
15.09
9.58
12.95
a
b
2
2
2
2
2
R ) ∑||Fo| - |Fc||/∑|Fo|. R(ωF2) ) {∑[ω(Fo - Fc )2]/∑[ω(Fo )2]}1/2; ω ) 1/[σ2(Fo ) + (aP)2 + bP], P ) [2Fc + max(Fo,0)]/3.
solution was stirring. A dark blue solution with yellow precipitate
was formed. After the solution was stirred overnight, a dark brown-
purple precipitate had formed around the yellow precipitate. THF
was added to dissolve the dark brown-purple precipitate. After the
addition of THF, the solution turned purple and was filtered through
Celite, and the solvent was removed in vacuo. A purple solid
remained, which was triturated three times with 2 mL of hexane.
The product was dissolved in a minimum amount of THF, filtered,
layered with hexane for recrystallization, and stored at -30 °C.
After 1 day, the product was collected via filtration, leaving a light
purple solid in 82% yield. 1H NMR (δ, ppm, {(CD3)2CO}): 2.481
(s, 12H, NCH3). 13C{1H} NMR (δ, ppm, {(CD3)2CO}): 69.29 (s,
CH3). UV-vis (CH2Cl2) (λmax, nm (ꢀM, M-1 cm -1)): 539 (200),
592 (320). Anal. Calcd for C32H24N2O4F20Co: C, 40.91; H, 2.57;
N 2.98. Found: C, 40.89; H, 2.50; N 2.91. µeff (Evans’ method) )
4.56µB.
each, to give the crude product as a blue powder. Subsequently, 2
equiv of 18-crown-6 (150 mg, 0.57 mmol) were added to the crude
product in 15 mL of CH2Cl2 and stirred for 1 h. The subsequent
blue solution was filtered through Celite, concentrated to dryness,
and recrystallized from a slow diffusion of hexanes into CH2Cl2
affording blue crystals with a blocklike habit (60 mg, 55% yield).
1H NMR (δ, ppm, {(CD3)2CO}): 3.787 (s, 24H, CH2). 13C{1H}
NMR (δ, ppm, {(CD3)2CO}): 70.95 (s, CH2). UV-vis (THF) (λmax
,
nm (ꢀM, M-1 cm-1)): 496 (260), 598 (440), 659 (540). (Anal. Calcd
For C60H78O18K2Co2: C, 56.15; H, 6.13. Found: C, 56.22; H, 6.18.
µeff (Evans’ method) ) 5.68µB.
(Ph4P)2[CoCl2(OArF)2], 12. A portion of CoCl2 (83.8 mg, 0.6453
mmol) was reacted with 4 equiv of TlOArF (1.00 g, 2.581 mmol)
in 25 mL of toluene and allowed to stir overnight. A cobalt blue
color evolved slowly with the formation of some white precipitate.
The reaction was filtered through Celite the following day to remove
the precipitate, and the blue solution was concentrated to dryness.
A 2-equiv portion of Ph4PCl (484 mg, 1.291 mmol) and 20 mL of
CH2Cl2 were added to the crude product. Formation of white solid
was observed instantly, and the reaction was allowed to stir
overnight. The presumed TlCl precipitate was again removed
through Celite, and the bright blue oily product was recrystallized
from CH2Cl2 and hexanes. Crystalline material required slow
growth over several days from room temperature solutions, and
173 mg (22.3% yield) was eventually obtained. 1H NMR (δ, ppm,
{(CD3)2CO}): 7.559 (br, CH), 7.590 (br, CH), 7.632 (br, CH), 7.802
(br, CH). 13C{1H}NMR (δ, ppm, {(CD3)2CO}): 118.61 (i, d, JCP
[Cp2Co]2[Co(OAr′)4], 10. A portion of Cp2Co (300 mg, 1.586
mmol) was dissolved in 6 mL of CH2Cl2 to give a dark red-brown
solution. A syringe was used to add 1.33 equiv of HOAr′ (322 µL,
2.115 mmol) with no obvious color changes. The solution was
sealed in an ampule and heated to reflux with an oil bath. A dark
green color was observable within 2 h, and the heating was allowed
to continue overnight. Upon completion, the solvents were re-
moved in vacuo, and the oily green product was washed with
pentanes and dried in vacuo. The oil was recrystallized from tol-
uene and hexanes. (N. B. Metathesis reactions using 2:1 ratios of
[Cp2Co]PF6 and [Co(OAr′)4]2- salts in CH3CN/THF showed colors
consistent with 10, but recrystallization attempts reformed the
) 89.9 Hz), 131.52 (o, d, JCP ) 12.8 Hz), 136.08 (m, d, JCP
)
1
10.1 Hz), 136.51 (p, s). UV-vis (CH2Cl2) (λmax, nm (ꢀM, cm-1
M-1)): 350 (310), 635 (270), 674 (270). Anal. Calcd for
C60H40O2F10P2Cl2Co: C, 61.35; H, 3.43. Found: C, 61.22; H, 3.37.
µeff (Evans’ method) ) 4.45µB.
starting material mixture.) H NMR (δ, ppm, {(CD3)2CO}): 6.59
(v. br). 13C{1H} NMR (δ, ppm, {(CD3)2CO}): 87.55 (HC Cp). 19
F
NMR (δ, ppm, {(CD3)2CO}): -66.80 (CF3). UV-vis (THF) (λmax
,
nm (ꢀM, M-1 cm -1)): 315 (17,160), 615 (510), 662 (470). Anal.
Calcd for C52H32O4F24Co3: C, 46.14; H, 2.38. Found: C, 46.16;
H, 2.12. µeff (Evans’ method) ) 4.67µB.
X-ray Crystallography. A summary of crystal data collection
and refinement parameters for all compounds are given in Tables
1 and 2. X-ray diffraction data were collected on either a Bruker
SMART51 APEX CCD (7b, 12) or a Siemens P4 diffractometer
equipped with a SMART51 Apex CCD detector (1b, 3, 4, 5, 6, 7a,
10, 11). Space groups and lattice symmetries were determined
through inspection of systematic absences and intensity statistics
in 1b. All data sets were corrected for absorption using the program
{K(18-crown-6)}2[Co2(µ2-OC6H5)2(OC6H5)4], 11. A portion of
CoI2 (88.7 mg, 0.29 mmol) was dissolved in 6 mL of THF to give
a blue-green solution. To this solution, 3 equiv of KOPh (120 mg,
0.91 mmol) was added, giving an intense blue color. The reaction
mixture was stirred for 3 h at room temperature after which time
there appeared some slight turbidity. The reaction mixture was
filtered through Celite to remove (presumed) KI with the addition
of 4 mL of additional THF. The filtrate was evaporated to dryness
under reduced pressure and triturated twice with 5 mL hexanes
(51) SMART: Software for the CCD Detector System; Bruker AXS:
Madison, WI, 2000.
Inorganic Chemistry, Vol. 43, No. 24, 2004 7713

Buzzeo et al.
Table 2. Summary of X-ray Crystallographic Data for Cobalt Complexes
7a
7b
10
11
12
formula
C56H66CoF20K2O18
1540.19
C2/c
22.0161(10)
16.0513(7)
19.1081(9)
C56H60CoF24K2O16
1582.17
C2/c
23.3983(13)
12.7964(6)
24.0070(13)
C52H32Co3F24O4
1353.57
C2/c
24.202(3)
13.9911(17)
17.1697(19)
C31H41Cl2CoKO9
726.57
P21/n
13.487(4)
15.672(4)
16.844(4)
C60H40Cl2CoF10O2P2
1174.69
P21/c
19.7866(13)
14.2273(10)
18.6950(13)
formula weight
space group
a, Å
b, Å
c, Å
R, deg
â, deg
101.730(1)
109.109(2)
114.678(2)
105.970(6)
95.751(1)
γ, deg
V, Å3
6611.5(5)
4, 0.5
6792.0(6)
4, 0.5
5282.9(11)
4
3422.9(15)
4, 0.5
5236.3(6)
4, 1
Z, Z′
crystal color, habit
F(calc), g cm-3
µ(Mo KR), cm-1
temp, K
blue, block
1.547
5.09
blue, block
1.547
5.03
green, block
1.702
10.61
green, block
1.410
8.29
blue, block
1.490
5.71
218(2)
150(2)
293(2)
218(2)
100(2)
reflections measured
reflection ind
R(F), %a
21 480
6485 (Rint ) 0.0358)
6.17
21 302
7978 (Rint ) 0.0288)
8.36
17 151
5957 (Rint ) 0.0435)
9.67
18 579
5378 (Rint ) 0.0888)
6.80
33 218
12328 (Rint ) 0.0379)
5.27
R(wF2),%b
15.55
21.21
29.37
15.09
13.86
a R ) ∑||Fo| - |Fc||/∑|Fo|. b R(ωF2) ) {∑[ω(Fo - Fc )2]/∑[ω(Fo )2]}1/2; ω ) 1/[σ2(Fo ) + (aP)2 + bP], P ) [2Fc + max(Fo,0)]/3.
2
2
2
2
2
SADABS.52 The structures were solved using direct methods or
the Patterson function and difference map techniques and were
refined by full-matrix least-squares procedures on F2. All non-
hydrogen atoms were refined with anisotropic displacement pa-
rameters. The H atoms were treated as idealized contributions except
5 where they were located in the Fourier difference map and refined
with isotropic thermal parameters. One CF3 group in 10, all the
CF3 groups in 7b, one terminal CH3 group in 4, and the O(6) atom
in 7a are disordered over two positions. Highly disordered CH2Cl2
solvent molecules in 6 were verified by the SQUEEZE program.53
Corrections of the X-ray data with 155 electrons/cell were close to
the required value of 168 electrons/cell. All software and sources
of scattering factors are contained in the SHELXTL program
package.54
Electrochemistry. Cyclic voltammetry experiments were carried
out in CH2Cl2 with 1.0 M (nBu4N)PF6 (recrystallized three times
from CH2Cl2) with analyte concentrations of approximately 2.0 mM.
Experiments were performed in a nitrogen-filled drybox, using a
three-electrode cell with all leads connected through binding posts
to an external EG&G model 273a potentiostat (Princeton Applied
Research) run by a personal computer with M270/250 Research
Electrochemistry software, version 4.30. A platinum (2 mm
diameter) working electrode, Ag wire quasi-reference electrode, and
Pt counter electrode were used. The working electrode was cleaned
between experiments using 1 mm Al2O3 slurry (Buehler), care-
fully dried, and returned to the drybox. Scan rates were in the
range 20-500 mV/s. Solutions of [K{18-crown-6}(OArF)] and
[K{18-crown-6}(OAr′)] were prepared by adding equimolar amounts
of the appropriate KOAr salt and 18-crown-6.
phenoxide, OC6H3(CF3)2- (OAr′). We anticipated that fluo-
rination of the aryl rings, which should result in decreased
electron density at oxygen, would reduce the tendency toward
bridging and formation of oligomeric or polymeric materials.
Other electronic effects may also play a role. (E.g., for OArF,
there is substantial participation of the peripheral F atoms
in the π system of the ring, whereas for OAr′ no such effects
are formally present.) The inductive influence of fluorination
of some or all of the carbon atoms in the two phenols is
reflected in the pH values for both materials, which are 8.42
for HOArF 55 and 8.26 for HOAr′,55 compared to 9.78 for
phenol55 and 10.20 for 3,5-dimethylphenol.56 Fluorination
also reduces the likelihood of unwanted C-H activation and
alters the possibility of ligand reduction or oxidation reactions
(such as preventing oxidation to the quinone) in comparison
to the perprotio-congeners.
Syntheses. The tetrakisphenolate anion [M(OAr)4]2- is
formed readily by simple metathesis reactions as shown in
Figures 1 and 2. The potassium phenoxides are reacted in
1
THF with /4 equiv of CuBr2. Addition of 2 equiv of 18-
crown-6 in CH2Cl2 and separation of residual KBr give
the compounds {K(18-crown-6)}2[Cu(OArF)4], 1a, and
{K(18-crown-6)}2[Cu(OAr′)4], 1b, in good yield. Syntheses
of the [Tl2Cu(OAr)4] complexes proceed under different con-
ditions. Unlike the potassium phenoxides, the thallium
compounds TlOArF and TlOAr′ are quite soluble in toluene.
The synthetic result is that CuBr2 reacts readily in toluene
with 4 equiv of TlOAr to form [Tl2Cu(OArF)4], 2a, and
[Tl2Cu(OAr′)4], 2b. Analytically pure 2a was recrystallized
from toluene and hexane to give a fine brown powder, and
2b was recrystallized with THF solvation. The thallium
cations can be metathesized easily from 2a and 2b to form
a variety of other A2[Cu(OAr)4] compounds, including the
Ph4P+ derivative (Ph4P)2[Cu(OArF)4], 3, via exchange with
Ph4PI, demonstrating the robust nature of the phenolate anion
Results and Discussion
In our research group, we were interested in preparing
homoleptic aryloxide anions that were neither coordinatively
nor sterically saturated at the metal center. We chose two
highly fluorinated aryloxides for this purpose, pentafluo-
rophenoxide, OC6F5 (OArF), and bis-3,5-trifluoromethyl-
-
(52) Sheldrick, G. M. SADABS: Area Detector Absorption Correction;
University of Go¨ttingen: Go¨ttingen, Germany, 2001.
(53) Van der Sluis, P.; Spek, A. L. Acta Crystallogr. 1990, A46, 194-
201.
(54) Sheldrick, G. M. SHELXTL: Program Library for Structure Solution
and Molecular Graphics; Bruker AXS: Madison, WI, 2000.
(55) Abraham, M. H.; Duce, P. P.; Morris, J. J.; Taylor, P. J. J. Chem.
Soc., Faraday Trans. 1 1987, 83, 2867-81.
(56) The Chemistry of the Hydroxyl Group; Patai, S., Ed.; Interscience:
London, New York, 1971; Part 1.
7714 Inorganic Chemistry, Vol. 43, No. 24, 2004

Homoleptic Cobalt and Copper Phenolate Compounds
Figure 1. Syntheses of homoleptic copper phenolate anions [Cu(OAr)4]2- and related compounds.
under these conditions. The related compound (nBu4N)2[Cu-
(OArF)4], 4, was prepared from (nBu4N)2[CuCl4]35 and
TlOArF.
conditions, with no care to exclude water or oxygen. Control
experiments with HOPh led to intractable, insoluble oily
brown materials. These intractable mixtures could be the
results of phenol polymerization57 and are under further
investigation.
The analogous reactions were studied with KOPh and
TlOPh. Reaction of 4 equiv of TlOPh with 1 equiv of
CuBr2 in THF yielded only insoluble material, suggesting
oligomers or polymers, whereas 2a and 2b are significantly
soluble not only in THF but also in toluene. Reaction of
KOPh and addition of 18-crown-6 with a ligand-metal ratio
of 4:1 yielded the dinuclear complex {K(18-crown-6)}2-
[Cu2(µ-OPh)2(OPh)4], 6, and residual KOPh. Compound 6
(as well as 11) was subsequently synthesized from a 1:3
ratio of Cu(II) (Co(II) for 11) and KOArF (Experimental Sec-
tion). These control experiments demonstrate that substantial
fluorination on the phenyl ring allows the formation of
mononuclear compounds under conditions in which nonflu-
orinated phenoxide ligands yield dimeric or insoluble species.
The cobalt derivatives can be synthesized in a similar
manner. Reaction of CoI2 with an appropriate stoichio-
metric quantity of KOArF and 18-crown-6, as for 1a and
1b, resulted in facile formation of {K(18-crown-6)}2[Co-
(OArF)4], 7a, and {K(18-crown-6)}2[Co(OAr′)4], 7b, in good
yield. The related thallium [Tl2Cu(OAr)4] complexes, 8a and
8b, were prepared in toluene, analogous to 2a and 2b, as
shown in Figure 2. Reaction of CoCl2 under similar condi-
tions yields some TlCl precipitate and a cobalt-blue product
but also unreacted TlOArF. Treatment of the cobalt product
of these conditions with Ph4PCl and removal of TlCl led to
isolation of (Ph4P)2[CoCl2(OArF)2], 12, presumably through
an intermediate with the stiochiometry {Tl2CoCl2(OArF)2}.
Complete substitution at cobalt required CoI2 as a starting
The [Cu(OArF)4]2- anion can be formed also by the
deprotonation of the fluorinated phenols with amines to form
the ammonium salt (Et3HN)2[Cu(OArF)4], 5. The reaction
shown in Figure 1 gave good yields of 5 under aerobic
(57) Aromi, G.; Gamez, P.; Kooijman, H.; Spek, A. L.; Driessen, W. L.;
Reedijk, J. Eur. J. Inorg. Chem. 2003, 1394-1400.
Inorganic Chemistry, Vol. 43, No. 24, 2004 7715

Buzzeo et al.
Figure 2. Syntheses of homoleptic cobalt phenolate anions [Co(OAr)4]2- and related compounds.
material which is quite soluble in THF, unlike CoCl2. In the
reaction between TlOArF and CoCl2, either the driving force
of TlCl lattice formation is not enough to completely breakup
the CoCl2 lattice or the reaction kinetics are too slow to be
synthetically useful. In contrast, Co-I bond cleavage and
TlI precipitation are complete within hours of the reaction
between TlOArF and CoI2. Metathesis of 8a with Me4NI
resulted in elimination of yellow TlI and formation of
(Me4N)2[Co(OArF)4], 9, analogous to the synthesis of 3, vide
supra.
With regard to metal chlorides as starting materials, CuCl2
works just as well as CuBr2 in the preparation of [Cu(OAr)4]2-
species. Reaction of CoCl2 with either KOAr or TlOAr,
however, led to residual aryloxide and incomplete substitu-
tion as demonstrated with the preparation of 12. Because
use of CoI2 solved this problem, we suspect that bridging in
CoX2 is less strong with the heavier halides but have not
made a detailed study of this phenomenon.
The cobalticenium compound [Cp2Co]2[Co(OAr′)4], 10,
was discovered as a byproduct in the original preparation of
[Cp2Co][Ar′O‚‚‚H‚‚‚OAr′]34 and was synthesized again by
the direct route reported in the Experimental Section.
Cobaltocene (3 equiv) is reacted with 4 equiv of HOAr′,
resulting in good yields of 10. The reaction stoichiometry
suggests concomitant formation of 2 equiv of cyclopenta-
diene from the protonation of two Cp ligands on the same
cobalt center, which subsequently acquires four phenolate
[Cp2Co]+ and 2 mol of H2 is produced. A 1H NMR study of
the reaction (Figure S1) is consistent with this stoichiometry,
although no H2 was detected. Oxidation of [Cp2Co] to
[Cp2Co]+ with Brønsted acids is well precedented.58,59 The
compound is initially separated as a blue-green oil and is
crystallized only over a period of weeks.
These A2[M(OAr)4] complexes are the first homoleptic
phenolate anions of late transition metal ions to be structur-
ally and spectroscopically characterized. A previous report
of some [M(OArF)4]2- anions provided spectroscopic char-
acterization60 and proposed metal coordination61 geometries,
some of which are shown to be incorrect (vide infra). An
earlier report62 exists of the compounds Li2Co(OPh)4‚4THF,
Na2Co(OPh)4‚5THF, and K2Co(OPh)4‚0.75THF but no struc-
tural characterization was performed, and substantial color
changes from blue to red were observed upon loss of THF,
suggesting metal coordination environment changes from
tetrahedral to octahedral, vide infra. No data were pro-
vided to distinguish M2[Co(OPh)4] structures from [MOPh]/
[MCo(OPh)3] mixtures. The Li2Co(OPh)4 structures that are
blue in the solid state become red (assumed to be octahedral)
when dissolved in THF.63 This behavior is in distinct contrast
to compound 11 reported herein. Magnetic data also sug-
(58) Wilkinson, G.; Pauson, P. L.; Cotton, F. A. J. Am. Chem. Soc. 1954,
76, 1970-1974.
(59) Wilkinson, G.; Cotton, F. A.; Birmingham, J. M. J. Inorg. Nucl. Chem.
1956, 2, 95-113.
(60) Hollebone, B. R.; Nyholm, R. S. J. Chem. Soc. A 1971, 332-337.
(61) Hollebone, B. R. J. Chem. Soc. A 1971, 481-486.
(62) Ibrahim, A. I.; Gaube, W.; Kalies, W.; Witt, B. J. Prakt. Chem. 1991,
333, 397-406.
(63) El-Behairy, M.; Ibrahim, A. I.; Emara, A. A. A.; Tawfic, T. A. J.
Chem. Res., Synop. 1996, 22.
3[Cp2Co] + 4HOAr′ f [Cp2Co]2[Co(OAr′)4] + 2CpH + 2H2
ligands. The other two [Cp2Co] equivalents are oxidized to
7716 Inorganic Chemistry, Vol. 43, No. 24, 2004

Homoleptic Cobalt and Copper Phenolate Compounds
selves. The O-Cu-O angles are also slightly different from
90°, with the longer sides of the rectangle subtending wider
angles. In 3, the phenyl rings are staggered with the shortest
distance of 2.988(6) Å between the two ortho atoms, C1 and
C7. There are no close contacts in 3 or 4 with either of the
n
structurally unexceptional Ph4P+ or Bu4N+ cations. The
average Cu-O distances in 3 and 4 are 1.919(8) and
1.928(4) Å, respectively, and the average O-C distances and
Cu-O-C angles are 1.302(5) and 1.299(4) Å and 124.0(9)
and 125.7(2)° in 3 and 4, respectively. A search of the
Cambridge Structural Database (CSD)65 for Cu-phenoxide
structures in which the phenoxide ligand is neither bridging
nor part of a chelate ring yielded 29 structures and 34
crystallographically independent Cu-O-Ar linkages. Among
these compounds the average parameters are 1.93(15) Å for
the Cu-O distances, 1.32(2) Å for the O-C distances, and
128(7)° for the Cu-O-C angles. Thus, the average Cu-O
and O-C bond lengths in compounds 3 and 4 (see Table 3)
are comparable with the average values for these bond
lengths.
Single-crystal X-ray diffraction studies revealed similar
geometries for the anions of [K{18-crown-6}]2[Cu(OArF)4],
1a, and [K{18-crown-6}]2[Cu(OAr′)4], 1b. The structure of
only 1b, Figure 4, is reported here in detail because disorder
in the cation of several crystals of 1a resulted in a structure
of unpublishable quality, though the connectivity of the
structural units present in the crystal were consistent with
the chemical formulation.66 In 1b, the copper atom has an
approximately square-planar geometry in the center of four
oxygen donors that form a rectangle with sides 2.662(5) and
2.797(5) Å, similar to the analogous distances in 3 and 4.
The greater steric bulk of the OAr′ anion does not permit
the adjacent phenoxides to come into such close contact as
with OArF, however. The C(11)‚‚‚C(21) distance in 1b is
4.279(7) Å. The shorter sides of the rectangle are each
loosely bridged by a {K(18-crown-6)}+ cation between
aryloxide oxygen atoms. These interactions are weak as
demonstrated by the long K-O(aryloxide) average distance
of 2.76(5) Å, by an absence of shortening in the O‚‚‚O
distance in the anion and no significant lengthening of the
Cu-O bonds in contrast to that observed in 6 (vide infra).
The average K‚‚‚O (noncrown) distance from 68 examples
in the CSD of {K(18-crown-6)}+ cations with two additional
oxygen atom donors is 2.8(2) Å. In 1b, the unique Cu-O
distances are 1.917(3) and 1.944(4) Å, and the Cu-O-C
angles have a range of 127.7(3) to 130.2(3)°. These
parameters are not significantly different from the purely
monodentate aryloxides in 3 and 4 described above, dem-
onstrating that the potassium atoms exert little perturbation
on the anion geometry. The data for 1a show a structure
quite similar to 1b, an approximately square-planar Cu(II)
center and long K-O (aryloxide) contacts on two sides of
the rectangle. The {K(18-crown-6)}+ cations in 1a bridge
the short edges of the rectangle away from the OArF rings,
analogous to the ammonium cations in 5, vide infra. Previous
Figure 3. ORTEP diagrams of (Ph4P)2[Cu(OArF)4], 3, (top) and anion
alone (bottom). The copper atom sits on an inversion center. Hydrogen
atoms have been omitted for clarity. Selected distances and angles:
Cu(1)-O(1) ) 1.912(3) Å, Cu(1)-O(2) ) 1.926(3) Å, Cu-O-Cavg
124.0(8)°. Ellipsoids are shown at the 50% probability level.
)
gested that these compounds did not contain [M(OPh)4]2-
cobalt centers.
A study of several [Al(ORF)4]- “superweak” anions of
Lewis acid catalysts64 demonstrated a linear relationship
between the pKa of the fluoro alcohol and the Lewis basicity
of the aluminate anion. Analogously, the lower pKa values
of HOArF and HOAr′ versus HOPh and HOC6H3(CH3)2 seem
to reduce the basicity of the [M(OAr)4]2- anions and thus
prevent the bridging behavior and favor mononuclear anions
whose structures are described in detail below.
Structural Characterization. First we discuss the struc-
tural features of the [M(OAr)4]2- anions with the completely
uncoordinating cations and then describe structures with
cations showing some associative interaction. Crystallo-
graphic data collection parameters for the copper compounds
are summarized in Table 1 and those for the cobalt com-
pounds, in Table 2. Selected distances and angles are
collected in Tables 3 (Cu) and 4 (Co).
Two different salts of the uncoordinated [Cu(OArF)4]2-
anion, (Ph4P)2[Cu(OArF)4], 3, and (nBu4N)2[Cu(OArF)4], 4,
have been crystallographically characterized. The cation and
anion of 3 are shown in Figure 3, and those of 4 are shown
in Figure S2. Both copper centers in 3 and 4 exhibit the same
approximately square planar geometry. Each copper center
is coplanar with the four oxygen atoms that lie at the corners
of a rectangle having sides of 2.644(5) and 2.780(5) Å in
the case of 3 and of 2.653(3) and 2.797(3) Å in 4. The longer
side of the rectangle, the O1‚‚‚O2 vector in Figure 3, is the
side on which the perfluorophenyl groups arrange them-
(64) Ivanova, S. M.; Nolan, B. G.; Kobayashi, Y.; Miller, S. M.; Anderson,
(65) Allen, F. H. Acta Crystallogr. 2002, B58, 380-388.
(66) Pnnm, a ) 9.2001 Å, b ) 17.6783 Å, c) 19.2815 Å.
O. P.; Strauss, S. H. Chem.sEur. J. 2001, 7, 503-510.
Inorganic Chemistry, Vol. 43, No. 24, 2004 7717

Buzzeo et al.
Table 3. Selected Interatomic Distances, Bond Lengths, and Angles for
Copper Compoundsa
compound
distances
(Å)
angles
(deg)
1b
Cu1-O1
Cu1-O2
O1‚‚‚O2
1.917(3)
1.944(4)
2.797(5)
O1-Cu1-O2
O1-Cu1-O2_2
Cu1-O1-C11
Cu1-O2-C21
92.83(15)
87.17(15)
127.7(3)
O1‚‚‚O2_2 2.662(5)
130.2(3)
O1-C11
O2-C21
1.315(6)
1.306(6)
C11‚‚‚C21 4.279(7)
K1-O1
K1-O2
2.793(4)
2.718(4)
3
Cu1-O1
Cu1-O2
O1‚‚‚O2
1.912(3)
1.926(3)
2.780(5)
O1-Cu1-O2
O1-Cu1-O2_2
Cu1-O1-C1
Cu1-O2-C7
92.83(12)
87.17(12)
123.2(3)
O1‚‚‚O2_2 2.644(5)
124.8(2)
O1-C1
O2-C7
C1‚‚‚C7
1.305(5)
1.298(5)
2.988(6)
4
5
Cu1-O1
Cu1-O2
O1‚‚‚O2
1.9243(16) O1-Cu1-O2
1.9311(15) O1-Cu1-O2_5
86.98(7)
93.02(7)
125.73(14)
125.65(13)
2.797(3)
Cu1-O1-C1
Cu1-O2-C7
Figure 4. ORTEP diagram of the [Cu(OAr′)4]2- anion and K+ cations
from 1b. The copper atom sits on an inversion center. Crown ether,
hydrogen, and fluorine atoms have been omitted for clarity. Selected
distances and angles: Cu(1)-O(1) ) 1.917(3) Å, Cu(1)-O(2) ) 1.944(4)
Å, Cu-O-Cavg ) 129(1)°. Ellipsoids are shown at the 50% probability
level.
O1‚‚‚O2_5 2.653(3)
O1-C1
O2-C7
C1‚‚‚C7
Cu1-O1
Cu1-O2
O1‚‚‚O2
1.298(3)
1.300(3)
2.985(4)
1.9195(10) O1-Cu1-O2
1.9290(11) O1-Cu1-O2_3
95.27(5)
84.73(5)
125.13(9)
122.83(9)
2.593(2)
Cu1-O1-C1
Cu1-O2-C7
O1‚‚‚O2_3 2.844(2)
O1-C1
O2-C7
1.314(2)
1.313(2)
C1‚‚‚C7_3 2.951(2)
O1‚‚‚N1
O2‚‚‚N1
2.950(2)
2.882(2)
6
Cu1-O1
Cu1-O2
Cu1-O3
Cu1-O4
Cu2-O1
Cu2-O2
Cu2-O5
Cu2-O6
1.9473(19) O1-Cu1-O2
1.9527(19) O2-Cu1-O3
1.9046(19) O3-Cu1-O4
1.8823(19) O4-Cu1-O1
1.9577(19) O2-Cu1-O4
1.9608(19) O1-Cu1-O3
1.8835(19) O1-Cu2-O2
1.8990(19) O2-Cu2-O5
77.73(8)
100.62(8)
89.35(8)
99.22(9)
158.62(9)
159.83(9)
77.30(8)
99.87(9)
89.55(9)
Cu1‚‚‚Cu2 3.0483(9)
O5-Cu2-O6
O6-Cu2-O1
O1-Cu2-O5
O2-Cu2-O6
Cu1-O1-Cu2
Cu1-O2-Cu2
Cu1-O1-C1
Cu1-O2-C7
Cu1-O3-C13
Cu1-O4-C19
Cu2-O1-C1
Cu2-O2-C7
Cu2-O5-C25
Cu2-O6-C31
O1‚‚‚O4
O2‚‚‚O3
O1‚‚‚O6
O2‚‚‚O5
O1‚‚‚O2
O3‚‚‚O4
O5‚‚‚O6
O1-C1
O2-C7
O3-C13
O4-C19
O5-C25
O6-C31
K1-O3
K1-O4
K2-O5
K2-O6
2.917(3)
2.968(2)
2.947(3)
2.943(2)
2.447(2)
2.663(3)
2.664(3)
1.358(4)
1.355(3)
1.326(3)
1.322(4)
1.318(4)
1.324(4)
2.743(2)
2.715(2)
2.792(2)
2.722(2)
99.64(9)
159.85(9)
160.04(9)
102.64(9)
102.32(9)
122.22(17)
122.93(15)
120.52(17)
124.93(18)
122.36(16)
122.57(16)
122.40(18)
120.47(17)
Figure 5. ORTEP diagram of (Et3NH)2[Cu(OArF)4], 5, showing hydrogen
bonding between cation and anion. The copper atom sits on an inver-
sion center. Hydrogen atoms are omitted for clarity. Selected distances:
Cu(1)-O(1) ) 1.9195(10) Å, Cu(1)-O(2) ) 1.9290(11) Å, O(1)‚‚‚O(2)
) 2.593(2) Å, O(1)‚‚‚N(1) ) 2.950(2) Å, O(2)‚‚‚N(1) 2.882(2) Å. Ellipsoids
are shown at the 50% probability level.
a Numbers in parentheses are estimated deviations of the last significant
figure.
ously, except for the bifurcating hydrogen bonds between
the Et3HN+ cations and the aryloxide oxygen atoms (Table
3) in which the average O-N distance is 2.916(4) Å, quite
normal for such bonds.71 The bridging proton was not located
spectroscopic characterization of (Et4N)2[Cu(OArF)4] was
used to propose a tetrahedral geometry.61
The structures of the four thallium salts, 2a,67 2b,68 8a,69
and 8b,70 have all been determined, have significantly more
complicated bimetallic structures with bridging phenoxide
ligands, and will be discussed in a separate publication.
Contacts between the cation and anion are also observed
in (Et3HN)2[Cu(OArF)4], 5. The structure in 5, shown in
Figure 5, is generally similar to the three discussed previ-
(67) P1h, a ) 14.858 Å, b ) 20.1452 Å, c ) 28.5113 Å, R ) 96.734°, â
) 94.611°, γ ) 93.648°.
(68) Pbcn, a ) 15.3095 Å, b ) 30.0215 Å, c ) 16.7395 Å.
(69) P1h, a ) 11.7969 Å, b ) 13.9222 Å, c ) 25.2309 Å, R ) 96.740°, â
) 102.643°, γ ) 96.954°.
(70) P1h, a ) 11.8345 Å, b ) 13.0593 Å, c ) 15.9152 Å, R ) 100.657°,
â ) 96.325°, γ ) 113.222°.
(71) Jeffrey, G. A. An Introduction to Hydrogen Bonding; Oxford University
Press: New York, 1997.
7718 Inorganic Chemistry, Vol. 43, No. 24, 2004

Homoleptic Cobalt and Copper Phenolate Compounds
CuO4 coordination and the two terminal Cu atoms have
approximately C2V CuO3 coordination. Steric repulsions be-
tween bridging OtBu groups are likely responsible for the
substantial distortion from square-planar geometry of the
internal copper centers.73 The ends of 6, like 1a and 1b,
are loosely coordinated to the two {K(18-crown-6)}+ cat-
ions with an average distance from K-O of 2.74(3) (Å)
(Table 3).
The terminal Cu-O(Ar) distances reported for 6 in Table
3 are well within the range of reported bond lengths for
acyclic linkages in the literature (34 crystallographically
independent described above) that have an average of 1.93-
(15) Å. The average Cu-O(terminal) distance is 1.892(11)
Å, slightly shorter than the related distances in 1b, 3, 4, and
5, having the fluorinated aryloxides described above. These
data suggest a correlation between the basicity of the ligand
(measured by the pKa of the parent alcohol) and the length
of the M-O bond. The average µ2-OAr distances (Å) in 6,
1.955(6), are longer than those from the nonbridging phe-
noxides, as expected, and again consistent with literature
range of 1.99(9). The average of Cu-O-C angles (deg) in
the literature is 128(7) compared to 122.3(1.4) in 6 from
Table 3. The Cu-O and O-C distances in 6 are also
unexceptional and within the literature range. Again, there
is no correlation between Cu-O-C angles and Cu-O
distances, consistent with what has been observed for other
transition-metal monodentate aryloxides.72
Figure 7 shows an ORTEP of [Cp2Co]2[Co(OAr′)4], 10,
in which the anion has no bonding interactions with the
cations. The Co-O4 geometry is distorted, having O‚‚‚O
contacts that range from 2.973(7) to 3.286(7) Å (Table 4)
and proportionately wider angles for the corresponding
O-Co-O angles, 99.3(3)° and 115.4(2)°, respectively.
These distortions are reminiscent of the well-studied D2d
[CoCl4]2- anions.74 In a CSD survey of 129 [CoCl4]2- anions
(coordination number of cobalt ) 4, of each chlorine ) 1),
the average minimum and maximum Co-Cl distances (Å)
were 3.60(9) and 3.82(8) with an average |max - min| value
of 0.18(16). The [Co(OAr′)4]2- anion thus exhibits the same
distortion seen in [CoCl4]2-.
The cobalt compounds 7a and 7b, analogues of 1a and
1b, are shown in Figures 8 and S3, respectively. Both cobalt
centers have approximately D2d symmetry, like 10 above,
rather than Td, with O-Co-O angles (deg) in the range
97.02(10) to 133.5(2) for 7a and 96.10(12) to 131.0(2) for
7b. A Td geometry was proposed for (Ph4P)2[Co(OArF)4].61
The corresponding minimum and maximum O‚‚‚O contacts
are 2.898(4) to 3.482(4) Å and 2.892(4) to 3.345(4) Å,
respectively. In each anion, two {K(18-crown-6)}+ cations
weakly bridge the O1‚‚‚O2 sides of the anion. The distant
coordination of the potassium atom to the aryloxide oxygen
atoms may be responsible for the slight distortion of 7a and
7b from the geometry observed in 10. The K‚‚‚O contacts
in 7a and 7b, listed in Table 4, are long and weak, analogous
Figure 6. ORTEP diagram of the anion of {K(18-crown-6)}2[Cu2(µ-OPh)2-
(OPh)4], 6. Crown ether and hydrogen atoms are omitted for clarity. Selected
distances and angles: Cu(1)-O(1) ) 1.9473(19) Å, Cu(1)-O(3) ) 1.9046-
(19) Å, Cu(1)‚‚‚Cu(2) ) 3.0483(9) Å, Cu(1)-O(1)-Cu(2) ) 77.73(8)°.
Ellipsoids are shown at the 50% probability level.
in the difference map but calculated. Again, the Cu atom is
in the center of a rectangle with Cu-O distances close to
1.92 Å and edges differing slightly in length. The Et3HN+
ions bridge the short sides of the rectangle which are shorter
by ∼0.25 Å, and the angles at Cu in the Cu-O4 plane are
close to 85° and 95°. In 1b, 3, and 4 the average difference
in side length was ∼0.15 Å and the angles 3° removed from
90°, suggesting a small distortion of the anion by the
hydrogen bonding cations. The Cu-O bond lengths in 5 are
not longer than the other three phenolate anions however;
thus the Et3HN+ cations, such as the {K(18-crown-6)}+
cations, are not strongly bound to the aryloxide groups.
The small variations among Cu-O-C angles are detailed
in Table 3 and show no correspondence with Cu-O bond
lengths. Although the angles are significantly wider than a
canonical tetrahedral angle, several other groups have
established that no correlation exists between larger
M-O-C(Ar) angles and shorter M-O distances that might
suggest increased M-O bond order.72 The CSD search
detailed above demonstrates that angles more obtuse than
109.5° are quite common in such linkages.
In the dinuclear compound 6, shown in Figure 6, each
Cu(II) center is significantly distorted from square-planar
geometry; the dihedral angle between the O(1)-Cu(1)-O(2)
and O(3)-Cu(1)-O(4) planes is 30.19°, and the dihedral
angle between the O(1)-Cu(2)-O(2) and O(5)-Cu(2)-O(6)
planes is 28.94°. There are 20 compounds in the CSD with
coordination pattern of {O2Cu(µ2-OR)2CuO2}, the majority
of which have two square-planar Cu(II) centers. Only one
compound has copper centers more distorted toward tetra-
hedral than 6, namely, the linear compound [Cu4(µ2-OtBu)6-
{O(CF3)3}2] in which the central two copper atoms have D2d
(73) Purdy, A. P.; George, C. F.; Brewer, G. A. Inorg. Chem. 1992, 31,
2633-2638.
(74) Cotton, F. A.; Wilkinson, G.; Murillo, C. A.; Bochmann, M. AdVanced
Inorganic Chemistry; John Wiley & Sons: New York, 1999.
(72) Bradley, D. C.; Mehrotra, R. C.; Rothwell, I. P.; Singh, A. In Alkoxo
and Aryloxo DeriVatiVes of Metals; Academic Press: New York, 2001;
Chapter 6, p 671.
Inorganic Chemistry, Vol. 43, No. 24, 2004 7719

Buzzeo et al.
Table 4. Selected Interatomic Distances, Bond Lengths, and Angles for
Cobalt Compoundsa
compound
distances
(Å)
angles
(deg)
7a
Co1-O1
Co1-O2
O1‚‚‚O2
O1‚‚‚O1_2
O1‚‚‚O2_2
O2‚‚‚O2_2
O1-C1
1.937(2)
1.932(3)
2.898(4)
3.482(4)
3.131(4)
3.362(4)
1.310(4)
1.305(5)
2.784(2)
2.811(3)
O1-Co1-O2
O1-Co1-O1_1
O2-Co1_O1
O2-Co1-O2_1
Co1-O1-C1
Co1-O2-C7
97.02(10)
128.00(15)
108.04(12)
121.0(2)
133.5(2)
125.9(2)
O2-C7
K1-O1
K1-O2
7b
Co1-O1
Co1-O2
O1‚‚‚O2
O1‚‚‚O1_7
O1‚‚‚O2_7
O2‚‚‚O2_7
O1-C1
1.947(3)
1.941(3)
2.892(4)
3.342(4)
3.345(4)
3.186(4)
1.309(4)
1.307(5)
2.824(3)
2.761(3)
O1-Co1-O2
O1-Co1-O1_7
O2-Co1-O1_7
O2-Co1-O2_7
Co1-O1-C1
Co1-O2-C9
96.10(12)
118.29(16)
118.69(12)
110.27(18)
131.0(2)
128.5(2)
O2-C9
K1-O1
K1-O2
10
11
Co1-O1
Co1-O2
O1‚‚‚O2
1.951(5)
1.937(5)
3.286(7)
2.973(7)
3.099(7)
3.260(7)
1.325(8)
1.313(9)
O1-Co1-O2
O1-Co1-O1_2
O1-Co1-O2_2
O2-Co1-O2_2
Co1-O1-C1
Co1-O2-C7
115.4(2)
99.3(3)
105.7(2)
114.5(3)
125.0(4)
135.6(5)
O1‚‚‚O1_5
O1‚‚‚O2_5
O2‚‚‚O2_5
O1-C1
O2-C7
Co1-O1
Co1-O2
Co1-O3
Co1-O1_2
1.962(3)
1.911(4)
1.901(3)
2.000(3)
O1-Co1-O2
O2-Co1-O3
O3-Co1-O1_2
O1-Co1-O1_2
O2-Co1-O1_2
O1-Co1-O3
Co1-O1-Co1_2
Co1-O1-C1
Co1-O2-C7
Co1-O3-C13
117.76(15)
114.12(16)
104.80(15)
80.33(15)
117.75(15)
117.00(16)
99.67(15)
130.6(3)
Co1‚‚‚Co1_2 3.028(2)
O1‚‚‚O2
O2‚‚‚O3
O1‚‚‚O3
O1‚‚‚O1_2
O2‚‚‚O1_2
O3‚‚‚O1_2
O1-C1
3.316(5)
3.199(5)
3.294(5)
2.556(5)
3.348(5)
3.091(5)
1.341(6)
1.318(6)
1.315(6)
3.302(6)
3.458(7)
3.240(6)
3.251(7)
3.493(5)
Figure 7. ORTEP diagrams of [Cp2Co]2[Co(OAr′)4], 10 (top), with
hydrogen and fluorine atoms omitted for clarity and anion of 10 (bottom)
with only hydrogen atoms removed. Co(1) sits on a 2-fold rotation axis.
Selected distances and angles: Co(1)-O(1) ) 1.951(5) Å, Co(1)-O(2) )
1.937(5) Å, O(1)-Co(1)-O(2) ) 115.4(2)°. Ellipsoids are shown at the
50% probability level.
126.1(3)
128.6(4)
O2-C7
O3-C13
K1-C10
K1-C15
K1-C16
K1-C17
K1-C18
to 1a and 1b. Like the copper cases described above, the
Co-O bonds show no significant lengthening because of
potassium interaction, and the [Co(OAr)4]2- anions contain
effectively monodentate phenolates.
Similarly, the {Ph4P}+ cations in 12 do not coordinate to
the [Co(OArF)2Cl2]2- anion and the cobalt center is closest
to tetrahedral of 7a, 7b, 10, and 12, based on the angles
(deg) 99.69(9)-117.12(8). The longer Co-Cl versus Co-O
bond lengths result in a disparity between O‚‚‚O and
Cl‚‚‚Cl contact lengths that does not correspond to the narrow
range of angles at cobalt. The anion has C2V symmetry
because of the two different ligands but is otherwise quite
similar to that in the other three pseudotetrahedral cobalt
anions.
12
Co1-O1
Co1-O2
Co1-Cl1
Co1-Cl2
O1‚‚‚O2
O1‚‚‚Cl1
O1‚‚‚Cl2
O2‚‚‚Cl1
O2‚‚‚Cl2
Cl1‚‚‚Cl2
O1-C1
1.962(2)
1.969(2)
2.2782(8) O1-Co1-Cl2
2.2845(8) O2-Co1-Cl1
3.005(3)
3.353(2)
3.627(2)
3.579(2)
3.470(2)
3.774(2)
1.308(3)
1.280(4)
O1-Co1-O2
O1-Co1-Cl1
99.69(9)
104.28(6)
117.12(8)
114.65(7)
109.12(7)
111.62(3)
120.66(18)
131.30(19)
O2-Co1-Cl2
Cl1-Co1-Cl2
Co1-O1-C1
Co1-O2-C7
O2-C7
a Numbers in parentheses are estimated deviations of the last significant
figure.
The range of Co-O distances for monodentate aryloxides
in all four structurally characterized fluoroaryloxide cobalt
compounds, see Table 4, is narrow. These distances are
unexceptional compared to 17 known distances in the CSD
for nonbridging, acyclic Co-O(aryl) linkages which average
1.92(7) Å. The Co-O-C angles and O-C distances are
unexceptional, and again there is no quantitative or qualitative
correlation between Co-O-C angles and Co-O distances
for the monodentate ligands.
The dimeric structure of 11, Figure 9, like that of 6, shows
the effects of lack of fluorination on the aryloxide binding
motif. In 11, the two cobalt centers each are bridged by two
phenoxide ligands and have two monodentate terminal
phenoxides. The data in Table 4 show bond lengths longer
by ∼0.5 Å for the bridging versus the terminal ligands. Again
like 6, the terminal Co-O distances in 11 are shorter than
7720 Inorganic Chemistry, Vol. 43, No. 24, 2004

Homoleptic Cobalt and Copper Phenolate Compounds
Figure 9. ORTEP diagram of {K(18-crown-6)}2[Co2(µ-OPh)2(OPh)4], 11.
The two cobalt atoms are related by a center of inversion. Crown ether
molecules, hydrogen atoms, and one molecule of lattice CH2Cl2 have
been omitted for clarity. Selected distances and angles: Co(1)-O(1) )
1.962(3) Å, Co(1)-O(2) ) 1.911(4) Å, Co(1)-O(3) ) 1.901(3) Å,
Co(1)‚‚‚Co(1_3) ) 3.028(2) Å, Co(1)-O(1)-Co(1_3) ) 99.67(15)°.
Ellipsoids are shown at the 50% probability level.
Figure 10. ORTEP diagram of (Ph4P)2[CoCl2(OArF)2], 12. Hydrogen
atoms have been omitted for clarity. Selected distances and angles:
Co(1)-O(1) ) 1.962(2) Å, Co(1)-O(2) ) 1.969(2) Å, Co(1)-Cl(1) )
2.2782(5) Å, Co(1)-Cl(2) ) 2.2845(8) Å, O(1)-Co(1)-O(2) ) 99.69(9)°,
Cl(1)-Co(1)-Cl(2) ) 111.62(3)°. Ellipsoids are shown at the 50%
probability level.
analogue 6, the phenoxide rings are perpendicular to the
Cu2O2 core. The dihedral angle between the C2-C1-C6
plane and the Co1-O1-Co1_3 plane is only 5.57°. Only
two other structures with bridging but nonchelating phenox-
ides have been crystallographically characterized, one with
Klaui ligands and pentacoordinate cobalt atoms75 and one
with distorted tetrahedral cobalt centers, (Et4N)2[Co2Cl4-
(µ-4-OC6H4CH3)2].76,77 In the latter, the analogous dihedral
angle is 8.3°. In most aspects, the dinuclear compound 11
has a similar structure to 6 except that the cobalt atoms
are approximately tetrahedral with a dihedral angle at co-
balt of 84.11°. There are no close contacts between the
{K(18-crown-6)}+ cations and the phenolate oxygen atoms,
however. One axial site of the potassium is coordinated by
Figure 8. ORTEP diagrams of {K(18-crown-6)}2[Co(OArF)4], 7a. The
cobalt atom sits on a 2-fold rotation axis. Crown ether molecules and one
unbound THF molecule in the lattice have been removed for clarity. Selected
distances and angles: Co(1)-O(1) ) 1.937(2) Å, Co(1)-O(2) ) 1.932(3)
Å, O(1)-Co(1)-O(2) ) 97.02(10)°. Ellipsoids are shown at the 50%
probability level.
those in 7a, 7b, and 10 because of the greater basicity of
the OPh versus fluorinated OAr ligands. The symmetry at
the crystallographically equivalent cobalt centers is C2V
because the O1‚‚‚O1_3 contact (Å) is 2.556(5) compared to
3.199(5) between O2 and O3, and the subtended angles (deg)
are 80.33(15) and 114.12(16), respectively. The most inter-
esting structural feature is the virtual coplanarity of both
bridging phenoxides and the Co2O2 core. In the copper
(75) Ma, D.-q.; Hikichi, S.; Akita, M.; Moro-oka, Y. Dalton 2000, 1123-
1134.
(76) Coucouvanis, D.; Greiwe, K.; Salifoglou, A.; Challen, P.; Simopoulos,
A.; Kostikas, A. Inorg. Chem. 1988, 27, 593-594.
(77) Emori, S.; Nakashima, M.; Mori, W. Bull. Chem. Soc. Jpn. 2000, 73,
81-84.
Inorganic Chemistry, Vol. 43, No. 24, 2004 7721

Buzzeo et al.
an η4-phenoxide (C15-C18) and the antipodal site is
occupied by an η1-phenoxide (C10). The K-C distances (Å)
listed in Table 4 are consistent with 182 similar contacts in
the CSD, having an average distance of 3.25(16) and the
range 2.90-3.93.
Information on the solution structure of some of these
compounds was obtained from conductivity studies. The
conductivities of six CH2Cl2 solutions each of 12, (Ph4P)2-
[CoCl2(OArF)2], and 8a, [Tl2(Co(OArF)4], were measured at
varying concentrations. The conductivity of 8a is insensitive
to changes in concentration, demonstrating that the thallium
cations are covalently bound to the phenolate anion. In
contrast, the conductivity of solutions of 12 increases linearly
with the increases in concentration, demonstrating separated
cations and anions. Data are plotted in Figure S4.
IR spectra are not particularly informative as they are
dominated by strong absorptions for CdC stretches from the
aryloxide rings. Spectra for compounds 3, 6, 9, and 11 are
presented in Figure S5. In the region of ν(C-O)M bands
(900-1150 cm-1), all four complexes exhibit a very weak
absorption around 1110 cm-1. The mononuclear compounds
(3 and 9) exhibit slightly stronger stretches at 972 and 1007
and 988 and 1011 cm-1, respectively whereas the dinuclear
species (6 and 11) exhibit weak bands at 1585 and 1583
cm-1, respectively. M-O bands for first row metals typically
are below 600 cm-1.24
Figure 11. Visible spectra of [Cu(OAr)4]2- complexes {K(18-crown-6)}2-
[Cu(OArF)4], 1a, {K(18-crown-6)}2[Cu(OAr′)4], 1b, (Ph4P)2[Cu(OArF)4],
3, (nBu4N)2[Cu(OArF)4], 4, and (HEt3N)2[Cu(OArF)4], 5.
2
2
2
B1g(dx -y ) . B2g(dxy) > Eg(dxz, dyz) > Alg(dz ). The solid
samples of 1a, 1b, 3, 4, and 5 are bright green and retain
the color in THF or CH2Cl2 solutions. The visible spectra
for 1a, 1b, 3, 4, and 5 are compared in Figure 11. For
comparison, the visible spectrum of (nBu4N)2[CuCl4] has
absorptions at 340, 400, and 450 nm.78 All the complexes
with [Cu(OArF)4]2- anions have strong absorbances in the
UV region, suggesting that some ligand-to-metal charge-
transfer bands overlap with intraligand π-π* transitions. On
the low energy side of these broad absorbances, there are
shoulders near 330 nm that we assign as B1g r Eg transitions
and broad maxima near 425 nm that are assigned as B1g
r
Electronic Structure. The [CuCl4]2- anion is well-studied
and provides an initial model for discussion of the
[Cu(OAr)4]2- anions described here. Early crystallographic
work78 established a wide range of geometries from distorted
tetrahedral (D2d) to square planar (D4h) for [CuCl4]2-. The
anion is also found in bridged forms such as [Cu2Cl6]2-,
[Cu4Cl10]2-, [Cu4Cl12]4-, and [Cu5Cl14]4-, making spectro-
scopic “signatures” inadequate proof of discrete [CuX4]2-
anions. The [Cu(OAr)4]2- compounds reported herein are
in distinct contrast because they can be routinely prepared
as monomeric units, and thus detailed studies of their
electronic and magnetic properties were undertaken. In a
recent CSD search, 217 structures were found with coordi-
nates for an isolated [CuCl4]2- anion (each chlorine coor-
dination number ) 1), and only 36 of them have maximum
Cl-Cu-Cl angles of greater than 175°, thus square planar
anions are less common but not infrequent. Among these
anions, 21 have a standard deviation in the short Cl‚‚‚Cl
distances of more than 0.01 Å, indicative of symmetry
reduction to D2h geometry.
B2g transitions. The lowest energy transition of all the
phenolate complexes is blue-shifted compared to that of the
[CuCl4]2- species, consistent with larger splitting parameters
for oxygen donor atoms versus chlorine. The effect of
different cations on λmax is minimal with absorptions for 3
and 4 at 411 and 414 nm, respectively. The spectrum of 5
with {Et3NH}+ cations has a λmax value shifted slightly to
432 nm. The [Cu(OAr′)4]2- spectrum in 1b differs from the
other four, which all contain OArF, in the absence of distinct
maxima near 330 or 420 nm but a maximum at 308 nm that
is relatively sharp. Further differences between [M(OArF)4]2-
and [M(OAr′)4]2- spectra are discussed below.
Similarly we reference our discussion of the [Co(OAr)4]2-
electronic structure to the literature on [CoCl4]2- anion.
Extensive structural work has determined a flattened-
tetrahedron D2d structure for [CoCl4]2- anions. In the same
CSD search quoted above (129 hits), the average maximum
Cl-Co-Cl angle is 133(4)°. The electronic spectra have
been described in terms of a tetrahedral ligand field,81 despite
this noticeable distortion. Absorptions in UV-vis spectra
of [CoX4]2- anions have been assigned to three band groups
Our crystallographic work (vide supra) has shown that the
[Cu(OAr)4]2- anions are planar with a slight distortion from
D4h to D2h symmetry. We base our spectroscopic assignments
on D4h symmetry because this distortion is not very large.
Work by Solomon and others79,80 on square planar [CuCl4]2-
has demonstrated that in the ground state the d-orbital
manifold in D4h symmetry has decreasing energies as
4
4
ν1 (4T2 r A2 far UV), ν2 (4T1(F) r A2, approximately
425-830 nm), and ν3 (4T1(P) r 4A2, approximately 1430-
2220 nm).82
In our compounds with [Co(OAr)4]2- anions, there is no
O-Co-O angle greater than 128°, and most angles are less
than 120°, having approximately tetrahedral structures so we
(78) Smith, D. W. Coord. Chem. ReV. 1976, 21, 93-158.
(79) McDonald, R. G.; Hitchman, M. A. Inorg. Chem. 1986, 25, 3273-
81.
(80) Desjardins, S. R.; Penfield, K. W.; Cohen, S. L.; Musselman, R. L.;
Solomon, E. I. J. Am. Chem. Soc. 1983, 105, 4590-4603.
(81) Cotton, F. A.; Goodgame, D. M. L.; Goodgame, M. J. Am. Chem.
Soc. 1961, 83, 4690-4699.
(82) Cotton, F. A.; Goodgame, M. J. Am. Chem. Soc. 1961, 83, 1777-
1780.
7722 Inorganic Chemistry, Vol. 43, No. 24, 2004

Homoleptic Cobalt and Copper Phenolate Compounds
Figure 13. Visible spectra of [M2(µ-OPh)2(OPh)4]2- compounds
{K(18-crown-6)}2[Cu2(µ2-OC6H5)2(OC6H5)4], 6, and {K(18-crown-6)}2[Co2-
(µ2-OC6H5)2(OC6H5)4], 11.
Figure 12. Visible spectra of [CoX2(OAr)2]2- complexes {K(18-crown-
6)}2[Co(OArF)4], 7a, {K(18-crown-6)}2[Co(OAr′)4], 7b, (Me4N)2[Co-
(OArF)4], 9, [Cp2Co]2[Co(OAr′)4], 10, and (Ph4P)2[Co(OArF)2Cl2], 12.
Table 5. Summary of Magnetic Moment Data
compound
compound formulation
µ
eff (µB)
1a
1b
2a
2b
3
4
5
6
7a
7b
8a
8b
9
10
11
12
{K(18-crown-6)}2[Cu(OArF)4]
{K(18-crown-6)}2[Cu(OAr′)4]
Tl2[Cu(OArF)4]
1.74
1.79
1.79
1.94
1.59
1.87
1.72
2.74
4.53
5.02
4.46
6.06
4.56
4.67
5.68
4.45
also assign the spectra based on Td symmetry. A comparison
of the visible spectra of 7a, 7b, 9, 10, and 12 in THF is
presented in Figure 12. We assign these visible transitions
to the ν2 manifold of 4T1(F) r 4A2 transitions. The [CoCl4]2-
anion in (nBu4N)2[CoCl4] exhibits three maxima at λmax (ꢀ)
) 634 (450), 667 (630), and 699 nm (750 M-1 cm-1).81 The
phenolate spectra for 7a, 7b, and 9 in Figure 12 also show
three maxima in approximately the same region and the
magnitudes of the extinction coefficients for all five com-
plexes are consistent with d-d transitions. The strong, broad
absorbance less than 400 nm in 10 is due to the thoroughly
studied yellow [Cp2Co]+ cations. The presence of both the
yellow cobalticenium cations and blue cobaltate anions
accounts for the observed blue-green color of the crystalline
complex 10 and its solutions.
The spectra in Figure 12 qualitatively fall into three
groups, those with only OArF ligands (7a and 9, in red), those
with only OAr′ ligands (7b and 10, in blue), and a mixed
OArF-Cl derivative 12 (in black). Interestingly, the extinc-
tion coefficients for 7b and 10 are greater by almost a factor
of 2 versus 7a, 9, and 12. This observation together with
the comments about 1b above suggests that the d-d
transitions are more allowed (greater ꢀ values) with the OAr′
ligand than with OArF.
The spectrum for the less symmetrical [CoCl2(OArF)2]2-
anion in 12 shows two not three maxima, as does the
spectrum of 10. The anion in 12 has C2V symmetry in which
the T1 state is expected to split into A2, B1, and B2 terms,
but other C2V [CoX2Y2]2- anions have spectra not substan-
tially different from their [CoX4]2- cousins.83 The larger λmax
at 674 nm for 12 compared to the [Co(OAr)4]2- containing
species is consistent with smaller splitting parameters
expected for Cl donor atoms compared to oxygen donor
atoms. Similarly, all the absorbances in Figure 12 are blue-
shifted relative to those of [CoCl4]2-, consistent with the
comparison of the spectra in Figure 11 and that of the
[CuCl4]2- anion. The UV-vis spectra of many [M(cate-
cholate)2]2- complexes have been reported,84 and those of
Tl2[Cu(OAr′)4]
(Ph4P)2[Cu(OArF)4]
(nBu4N)2[Cu(OArF)4]
(HEt3N)2[Cu(OArF)4]
{K(18-crown-6)}2[Cu2(OPh)6]
{K(18-crown-6)}2[Co(OArF)4]
{K(18-crown-6)}2[Co(OAr′)4]
[Tl2Co(OArF)4]
{Tl2[Co(OAr′)4]‚toluene}2
(Me4N)2[Co(OArF)4]
[Cp2Co]2[Co(OAr′)4]
{K(18-crown-6)}2[Co2(OPh)6]
(Ph4P)2[Co(OArF)2Cl2]
the Co(II) and Cu(II) species are broadly similar to the
spectra reported herein for [M(OAr)4]2- anions. To date, no
structural data for bis-catecholate complexes [M(cate-
cholate)2]2- have been reported.
The visible-region electronic spectra of the two dinuclear
[M2(µ-OPh)2(OPh)4]2- compounds are shown in Figure 13.
The copper complex 6 is brown in solution and the solid
state, in contrast to the green mononuclear derivatives, and
shows no clear maxima in the visible region but a broad
shoulder around 390 nm. The dinuclear cobalt derivative 11
is bright blue as a solid and in solution and shows cor-
respondingly blue-shifted absorbances compared to those of
the dark cobalt-blue mononuclear compounds. There are
three maxima at 496, 598, and 659 nm for the formally C2V
metal centers, but the general pattern matches those of the
more tetrahedral CoO4 centers described above.
Magnetism. Solution measurements of magnetically dilute
samples by the Evans’ method36,37 yielded the magnetic
susceptibilities listed in Table 5. The compounds 1-5 that
contain [Cu(OAr)4]2- anions have moments > 1.73µB (spin-
only moment for one unpaired electron) and in the range
1.74-1.94µB. Magnetic susceptibilities for compounds with
78
[CuCl4]2- anions are ∼1.90µB, with quenching expected
of the orbital angular momentum in D4h symmetry (no T
states).85 Dimeric 6 has a significantly larger value at 2.74µB
(1.37µB /Cu) that indicates some coupling between the two
(83) Cotton, F. A.; Goodgame, D. M. L.; Goodgame, M.; Sacco, A. J.
Am. Chem. Soc. 1961, 83, 4157-4161.
(84) Sever, M. J.; Wilker, J. J. Dalton Trans. 2004, 1061-1072.
Inorganic Chemistry, Vol. 43, No. 24, 2004 7723

Buzzeo et al.
copper centers.39 Detailed magnetic studies of 6 and 11 and
other polynuclear phenolates (2a, 2b, and 8b) will be reported
in a future publication.
phenolate group and not from the metal center. No reduction
voltammetry was observed. Turning to the Co species, similar
observations were made; no reduction voltammetry and
oxidative waves correspond to ligand electrolysis. The only
point of difference being that for [Co(OAr′)4]2- the wave
was diffusional as observed for the related copper compound
1b. Although the Co and Cu species show no reversible
electrochemical oxidation or reduction behavior, which
would suggest possible formation of formally [MIII(OAr)4]-
or [MI(OAr)4]3- species, preliminary chemical studies have
shown that the Cu2+ can be oxidized by strong oxidants such
as Ce4+ and O2.89 Representative cyclic voltammograms for
all six compounds are provided in the Supporting Informa-
tion, Figures S6-S11.
The [Co(OAr)4]2- family of anions show values that are
noticeably higher than the spin-only value of 3.87µB for three
unpaired electrons, namely, in the range 4.45-5.02µB, and
such behavior has been observed previously for several
[CoX4]2- compounds.81-83,86,87 The susceptibility of 8b is
noticeably higher than that of the other compounds, but as
previously mentioned the single-crystal X-ray diffraction
studies of the [Tl2M(OAr)4] family reveal them to contain
aryloxide ligands that bridge the Tl+ and M2+ metal centers
or two M2+ centers. This large magnetic susceptibility in
solution suggests that a structure with bridging aryloxide
ligands and coupling between cobalt centers persists in
solution.
Conclusions
The syntheses of two families of late transition metal
homoleptic phenolate anions with fluorinated aryloxides of
the form [M(OAr)4]2- are reported with thorough spectro-
scopic and structural characterization. By use of fluorinated
aryloxide ligands, the electron density on the ligating oxygen
atoms seems to be sufficiently reduced so that bridging
between transition-metal centers can be prevented. Reactions
with OPh under the same conditions led to dimeric anions
of the form [M2(µ2-OC6H5)2(OC6H5)4]2-. The compounds
herein provide several examples of monodentate aryloxide
coordination without the use of chelating or cyclic ligands.
The structural comparisons between the copper anions in
[Cu(OArF)4]2- and [Cu2(µ2-OC6H5)2(OC6H5)4]2- and between
the cobalt anions in [Co(OArF)4]2- and [Co2(µ2-OC6H5)2-
(OC6H5)4]2- demonstrate that by changing the electronega-
tivity of the ligand the degree of oligomerization can be
affected. After decades of research in metal-aryloxide
chemistry and struggles with bridging ligands, the structures
of these seemingly simple [M(OAr)4]2- anions are only now
being reported for the first time.
An exciting target for these systems is the chemistry of
higher oxidation states that we are actively pursuing at this
time. Oxygen atoms are often the ligand of choice for
stabilizing high oxidation states (e.g., [OsO4], [MnO4]-) and
oxygen-donor ligands can also serve in this capacity. The
organic moieties attached to oxygen can often be attacked
intra- or intermolecularly by the resultant high oxidation state,
and a great deal of work has gone into control or prevention
of such C-H activation.90 Fluorination of the aryloxide rings
not only makes them resistant to C-H activation and
subsequent decomposition but increases their electron-
withdrawing capacity. These readily accessible [M(OAr)4]2-
units open the door to a plethora of new compounds as well
as electronic structure and reactivity studies that were
previously unavailable.
The magnetic susceptibility of dinuclear 11, like that of
6, suggests a coupling interaction between the cobalt centers.
Two related anions [Cl2Co2(µ2-OPh)2Cl2]2- and [Cl2Co2-
(µ2-O-p-tolyl)2Cl2]2- have been characterized in the solid
state both structurally76,77 and magnetically.77 The latter
studies show small antiferromagnetic interactions in the solid
state, consistent with the solution susceptibilities reported
herein. Both compounds with bridging phenoxide ligands,
6 and 11, show higher susceptibilities than either predicted
spin-only moments for mononuclear compounds or the values
observed for the magnetically isolated mononuclear com-
pounds.
Electrochemistry. The four transition metal complexes
{K(18-crown-6)}2[Cu(OArF)4], 1a, {K(18-crown-6)}2-
[Cu(OAr′)4], 1b, {K(18-crown-6)}2[Co(OArF)4], 7a, and
{K(18-crown-6)}2[Co(OAr′)4], 7b, were studied by cyclic
voltammetry as were the related compounds [K{18-crown-
6}(OArF)] and [K{18-crown-6}(OAr′)]. Control studies of
the two potassium phenolates indicated that both ligands
show no reversible or irreversible reduction features within
the solvent window to -2.5 V (versus Ag). Under oxidation,
the ligands are similarly stable out to +2.0 V (versus Ag).
Beyond this point, both phenoxides show chemically ir-
reversible oxidation consistent with formation of a phenoxyl
radical, which undergoes followup chemistry. The behavior
of the pentafluorophenoxyl radical in aqueous solution has
been studied.88
The Cu species show irreversible oxidations at potentials
near 2.50 V with [Cu(OAr′)4]2- and 2.0 V for [Cu(OArF)4]2-
versus Ag. In the former case the parent material appears to
show surface activity (peak current proportional to scan rate),
suggesting the adsorption of the Cu(OAr′) ion/ion pair,
whereas the latter shows diffusional behavior (peak current
proportional to square root of scan rate). In both cases, the
oxidative potentials are close to those of the ligands and
suggest that the oxidation removes electrons from the
Acknowledgment. This work was supported by startup
funds from Barnard College, a Camille and Henry Dreyfus
Faculty Startup Grant (SU-99-065), a Research Corporation
(85) Figgis, B. N.; Hitchman, M. A. Ligand Field Theory and its
Applications; Wiley-VCH: New York, 2000.
(86) Holm, R. H.; Cotton, F. A. J. Chem. Phys. 1959, 31, 788-792.
(87) Holm, R. H.; Cotton, F. A. J. Chem. Phys. 1960, 32, 1168-1172.
(88) Shoute, L. C. T.; Mittal, J. P.; Neta, P. J. Phys. Chem. 1996, 100,
3016-19.
(89) Iqbal, A.; Saddoughi, S.; Turner, J. F. C.; Doerrer, L. H. Unpublished
results.
(90) Collins, T. J. Acc. Chem. Res. 1994, 27, 279-285.
7724 Inorganic Chemistry, Vol. 43, No. 24, 2004

Homoleptic Cobalt and Copper Phenolate Compounds
CCSA Award (CC5398), an NSF-CAREER award (to
L.H.D.; CHE-0134817), and the 2000 Moissan Undergradu-
ate Research Fellowship in Chemistry from the Fluorine
Division of the ACS to M.C.B.. The Barnard College NMR
facility is supported by the National Science Foundation
(DUE-9952633). L.H.D. thanks C. C. Cummins for hospital-
ity during a sabbatical when some of these data were
collected on the MIT DCIF NMR spectrometer funded
through NSF Grant CHE-9808061. We thank J. R. Norton
for the loan of a YSI Conductivity Bridge. We also thank
G. Parkin and D. Churchill for collecting the raw X-ray data
for 1a, 1b, and 10 and thank L. N. Zakharov for data
collection and structure solution for 2a, 2b, 4, 5, 7a-8b,
and 12, modeling the CF3 disorders in 1b and 10, all CIF
file preparations, and thoughtful discussions.
Supporting Information Available: 1H NMR data for the
preparation of 10, structures 4 and 7b, the conductivity data, IR
spectra, cyclic voltammograms, and X-ray crystallographic data
in CIF format. This material is available free of charge at
http://pubs.acs.org.
IC0493954
Inorganic Chemistry, Vol. 43, No. 24, 2004 7725