Metal complexes with cis α topology from stereoselective quadridentate ligands with amine, pyridine, and quinoline donor groups

Article abstract of DOI:10.1021/ic990475v

Though the principles governing quadridentate topology and metal stereochemistry have been known for some time, the cis α topology has been little exploited in designing catalysts for asymmetric reactions. Investigation of the inorganic chemistry of labile metal cis α complexes was undertaken as a prelude to exploring their potential to serve as catalysts for a variety of different reactions. The synthesis of a series of first row transition metal complexes of quadridentate ligands with ethylenediamine (en) and S-propylenediamine (S-pn) backbones that have been alkylated at nitrogen with either pyridine (py) or quinoline (qn) donor groups as well as with noncoordinating benzyl (Bn) or pentafluorobenzyl (F₅Bn) groups was undertaken. The steric and electronic properties vary throughout the ligand series, en(Bn)py, 1, en(F₅Bn)py, 2, S-pn(F₅Bn)py, 3, and S-pn(F₅Bn)qn, 4. These ligands were reacted with MCl_n salts (n = 2, M = Mn, Fe, Co, Ni, Cu, Zn; n = 3, M = Fe) to generate, in most cases, octahedral complexes with the targeted cis a topology. UV/vis, NMR, IR, cyclic voltammetry (CV), and conductivity analysis are described for the metal compounds. X-ray structural analysis of [Cu{en(F₅Bn)py}Cl]Cl reveals a five coordinate square pyramidal geometry. Single or major diastereomers were obtained for all diamagnetic Zn(II) complexes as well as for Co(III) analogues that were prepared by oxidation of Co(II) species using Br₂ as the oxidant. Electronic differences among ligands are reflected in the oxidation potentials of the respective metal complexes as determined by CV, with fluorinated systems showing greater resistance to oxidation, as expected.

Full text of DOI:10.1021/ic990475v

Inorg. Chem. 1999, 38, 5545-5556
5545
Metal Complexes with Cis r Topology from Stereoselective Quadridentate Ligands with
Amine, Pyridine, and Quinoline Donor Groups
Christina Ng, Michal Sabat, and Cassandra L. Fraser*
Department of Chemistry, University of Virginia, McCormick Road, Charlottesville, Virginia 22901
ReceiVed May 4, 1999
Though the principles governing quadridentate topology and metal stereochemistry have been known for some
time, the cis R topology has been little exploited in designing catalysts for asymmetric reactions. Investigation of
the inorganic chemistry of labile metal cis R complexes was undertaken as a prelude to exploring their potential
to serve as catalysts for a variety of different reactions. The synthesis of a series of first row transition metal
complexes of quadridentate ligands with ethylenediamine (en) and S-propylenediamine (S-pn) backbones that
have been alkylated at nitrogen with either pyridine (py) or quinoline (qn) donor groups as well as with
noncoordinating benzyl (Bn) or pentafluorobenzyl (F5Bn) groups was undertaken. The steric and electronic
properties vary throughout the ligand series, en(Bn)py, 1, en(F5Bn)py, 2, S-pn(F5Bn)py, 3, and S-pn(F5Bn)qn, 4.
These ligands were reacted with MCln salts (n ) 2, M ) Mn, Fe, Co, Ni, Cu, Zn; n ) 3, M ) Fe) to generate,
in most cases, octahedral complexes with the targeted cis R topology. UV/vis, NMR, IR, cyclic voltammetry
(CV), and conductivity analysis are described for the metal compounds. X-ray structural analysis of [Cu{en(F5-
Bn)py}Cl]Cl reveals a five coordinate square pyramidal geometry. Single or major diastereomers were obtained
for all diamagnetic Zn(II) complexes as well as for Co(III) analogues that were prepared by oxidation of Co(II)
species using Br2 as the oxidant. Electronic differences among ligands are reflected in the oxidation potentials of
the respective metal complexes as determined by CV, with fluorinated systems showing greater resistance to
oxidation, as expected.
Introduction
Recently there has been a resurgence of interest in chiral
coordination compounds with nitrogen ligands1-4 for their
potential as asymmetric catalysts in both small molecule and
polymer synthesis. For example, chiral metal salen analogues
have been utilized as asymmetric oxygen1-8 and nitrogen atom
transfer catalysts5,9,10 and as catalysts for the ring opening of
epoxides (Figure 1).11,12 Complexes of oxazoline ligands, such
as pybox, are effective Lewis acid catalysts and promote
cyclopropanation and aziridination reactions.13-21 Osmium-
(1) Ojima, I., Ed. Catalytic Asymmetric Synthesis; VCH: New York, 1993.
(2) Lucet, D.; Le Gall, T.; Mioskowski, C. Angew. Chem., Int. Ed. Engl.
1998, 37, 2580-627.
Figure 1. Examples of coordination compounds with chelating nitrogen
donor ligands used in catalysis.
(3) Togni, A.; Venanzi, A. Angew. Chem., Int. Ed. Engl. 1994, 33, 497-
526.
(4) Wills, M. J. Chem. Soc., Perkin Trans. 1 1998, 3101-20.
(5) Canali, L.; Sherrington, D. C. Chem. Soc. ReV. 1999, 28, 85-93.
(6) Palucki, M.; Finney, N. S.; Pospisil, P. J.; Guler, M. L.; Ishida, T.;
Jacobsen, E. N. J. Am. Chem. Soc. 1998, 120, 948-54 and references
therein.
alkaloid systems have been developed into highly selective
catalysts for dihydroxylation22-24 and aminohydroxylation.25
Interestingly, simple coordination compounds are also playing
key roles as catalysts in contemporary polymer chemistry. Atom
(7) Chang, S.; Lee, N.; Jacobsen, E. N. J. Org. Chem. 1993, 58, 6939-
41.
(8) Brandes, B. D.; Jacobsen, E. N. J. Org. Chem. 1994, 59, 4378-80.
(9) Hansen, K. B.: Leighton, J. L.; Jacobsen, E. N. J. Am. Chem. Soc.
1996, 118, 10924-5.
(16) Cyclopropanation: Doyle, M. P.; Protopopova, M. N. Tetrahedron
1998, 54, 7919-46.
(17) Doyle, M. P.; Forbes, D. C. Chem. ReV. 1998, 98, 911-35.
(18) Harm, A. M.; Knight, J. G.; Stemp, G. Synlett 1996, 677-8.
(19) Denmark, S. E.; Stavenger, R. A.; Faucher, A.-M.; Edwards, J. P. J.
Org. Chem. 1997, 62, 3375-89.
(10) Konsler, R. G.; Karl, J.; Jacobsen, E. N. J. Am. Chem. Soc. 1998,
120, 10780-1.
(11) Du Bois, J.; Tomooka, C. S.; Hong, J.; Carreira, E. M. Acc. Chem.
Res. 1997, 30, 364-72.
(12) Minakata, S.; Ando, T.; Nishimura, M.; Ryu, I.; Komatsu, M. Angew.
Chem., Int. Ed. Engl. 1998, 37, 3392-4.
(13) Ghosh, A. K. Tetrahedron Asymmetry 1998, 9, 1-45.
(14) Lewis acids: Evans, D. A.; Olhava, E. J.; Johnson, J. S.; Janey, J. M.
Angew. Chem., Int. Ed. Engl. 1998, 37, 3372-5 and references therein.
(15) Yao, S.; Johannsen, M.; Hazell, R. G.; Jorgensen, K. A. Angew. Chem.,
Int. Ed. Engl. 1998, 37, 3121-4.
(20) Bedekar, A. V.; Koroleva, E. B.; Anderson, P. G. J. Org. Chem. 1997,
62, 2518-26.
(21) Aziridination: Evans, D. A.; Faul, M. M.; Bilodeau, M. T.; Anderson,
B. A.; Barnes, D. M. J. Am. Chem. Soc. 1993, 115, 5328-9.
(22) Becker, H.; Sharpless, K. B. Angew. Chem., Int. Ed. Engl. 1996, 35,
448-51.
(23) Kolb, H. C.; VanNieuwenhze, M. S.; Sharpless, K. B. Chem. ReV.
1994, 94, 2483-547.
10.1021/ic990475v CCC: $18.00 © 1999 American Chemical Society
Published on Web 11/09/1999

5546 Inorganic Chemistry, Vol. 38, No. 24, 1999
Ng et al.
transfer radical polymerization (ATRP) of styrene and acrylate
monomers26 utilizes Cu(I) complexes with bipyridine,26-28
chelating amine,29-31 and Schiff base ligands,32-34 whereas metal
complexes with diimine ligands figure prominently in recent
advances in a polyolefin synthesis, with many showing high
catalytic activity.35-37 Chiral variants of some of these systems
have been employed in an attempt to control polymer tacticity.38
Though systematic catalyst tuning can be difficult for
asymmetric complexes possessing C1 symmetry, numerous
derivatives of the dissymmetric C2-symmetric quadridentate
salen1-8 and bi- or tridentate oxazoline-based systems13-21 are
known. Upon coordination, these ligands typically adopt planar
topologies in which carbon stereocenters on the ligand frame-
work generate a chiral array at the reactive center. Though much
success has been achieved with these planar catalysts, com-
paratively little has been reported about the reactivity of
nonplanar chiral chelates despite the fact that methods for
controlling linear quadridentate chelate topology and metal
absolute configuration have been known for some time. Many
studies have been performed using nitrogen donor ligands,39,40
with mixed pyridylamine chelates figuring prominently.41
Structural features and physical characterization of many inert
and labile metal pyridylamine systems have been investigated.42-49
This family of ligands has been exploited in bimetallic enzyme
Figure 2. Quadridentate chelate topologies.
models50-54 and as catalysts for redox processes55-58 and olefin
polymerization.59
By the appropriate selection of metal chelate ring sizes, donor
groups, and chiral centers on the ligand backbone, it is possible
to generate ligands that result in major, if not single diastere-
omers upon coordination to metal ions. Since five-membered
chelate rings typically possess bite angles of less than 90°,
cumulative angle strain is minimized when chelate rings are
positioned out of plane with respect to each other.60 Others have
pointed out that “B strain,” namely unfavorable in-plane
nonbonding interactions between R-hydrogens on pyridyl
residues, also disfavor the planar orientation.43 Thus, quadri-
dentate chelates with three consecutive five-membered metal
chelate rings preferentially adopt cis topologies, either cis R or
cis â, over trans structures (Figure 2). For quadridentates in
which the two internal donor groups are secondary amines, cis
â complexes or mixtures of cis R and cis â isomers are typically
obtained.41,43,50,61-64 However, when these internal donor groups
are tertiary amines50 or sulfides,65-67 cis R structures have been
observed. It is assumed that apical binding is energetically
preferable to the cis â in-plane chelation, since the terminal
(24) Norrby P.-O.; Becker, H.; Sharpless, K. B. J. Am. Chem. Soc. 1996,
118, 35-42.
(25) Bruncko, M.; Schlingloff, G.; Sharpless, K. B. Angew. Chem., Int.
Ed. Engl. 1997, 36, 1483-6.
(26) Patten, T. E.; Matyjaszewski, K. AdV. Mater. 1998, 10, 901-15.
(27) Matyjaszewski, K.; Patten, T. E.; Xia, J. J. Am. Chem. Soc. 1997,
119, 9, 674-80.
(28) Matyjaszewski, K.; Wang, J.-L.; Grimaud, T.; Shipp, D. A. Macro-
molecules 1998, 31, 1527-34.
(29) Xia, J.; Matyjaszewski, K. Macromolecules 1999, 32, 2434-7.
(30) Xia, J.; Gaynor, S. G.; Matyjaszewski, K. Macromolecules 1998, 31,
5958-9.
(31) Xia, J.; Matyjaszewski, K. Macromolecules 1997, 30, 7697-700.
(32) Haddleton, D. M.; Crossman, M. C.; Dana, B. H.; Duncalf, D. J.;
Heming, A. M.; Kukulj, D.; Shooter, A. J. Macromolecules 1999, 32,
2110-9.
(33) Haddleton, D. M.; Kukulj, D.; Duncalf, D. J.; Heming, A. M.; Shooter,
A. J. Macromolecules 1998, 31, 5201-5.
(34) Haddleton, D. M.; Jasieczek, C. B.; Hannon, M. J.; Shooter, A. J.
Macromolecules 1997, 30, 2190-3.
(35) Small, B. L.; Brookhart, M. Macromolecules 1999, 32, 2120-30.
(36) Small, B. L.; Brookhart, M. J. Am. Chem. Soc. 1998, 120, 7143-4.
(37) Small, B. L.; Brookhart, M.; Bennett, A. M. A. J. Am. Chem. Soc.
1998, 120, 4049-50.
(38) Haddleton, D. M.; Duncalf, D. J.; Kukulj, D.; Heming, A. M.; Shooter,
A. J.; Clark, A. J. J. Mater. Chem. 1998, 8, 1525-32.
(39) Sargeson, A. M.; Searle, G. H. Inorg. Chem. 1967, 6, 787.
(40) ComprehensiVe Coordination Chemistry; Wilkinson, G., Ed.; Pergamon
Press: Oxford, U.K., 1987; Vol. 2, Chapter 13.1, pp 50-6 and
references therein.
(53) Glerup, J.; Goodson, P. A.; Hazell, A.; Hazell, R.; Hodgson, D. J.;
McKenzie, C. J.; Michelsen, K.; Rychlewska, U.; Toftlund, H. Inorg.
Chem. 1994, 33, 4105-11.
(54) Branca, M.; Pispisa, B.; Aurisicchio, C. J. Chem. Soc., Dalton Trans.
1976, 1543-6.
(55) Melnyk, A. C.; Kjeldahl, N. K.; Redina, A. R.; Busch, D. H. J. Am.
Chem. Soc. 1979, 101, 3232-40.
(56) Cairns, C. J.; Heckman, R. A.; Melnyk, A. C.; Davis, W. M.; Busch,
D. H. J. Chem. Soc., Dalton Trans. 1987, 2505-10.
(57) Leising, R. A.; Kim, J.; Perez, M. A.; Que, L., Jr. J. Am. Chem. Soc.
1993, 115, 9524-30.
(58) Rabion, A.; Chen, S.; Wang, J.; Buchanan, R. M.; Seris, J.-L.; Fish,
R. H. J. Am. Chem. Soc. 1995, 117, 12356-7.
(59) Rieger, B.; Abu-Surrah, A. S.; Fawzi, R.; Steiman, M. J. Organomet.
Chem. 1995, 497, 73-9.
(41) Aldrich-Wright, J. R.; Vagg, R. S.; Williams, P. A. Coord. Chem.
ReV. 1997, 166, 361-89.
(42) Goodwin, H. A.; Lions, F. J. Am. Chem. Soc. 1960, 82, 5013-23.
(43) Gibson, J. G.; McKenzie, E. D. J. Chem. Soc. A 1971, 1666-83 and
references therein.
(60) For a discussion of this issue in relation to the trien ligand, refer to
the following: Bosnich, B.; Gillard, R. D.; McKenzie, E. D.; Webb,
G. A. J. Chem. Soc. A 1966, 1331-9 and references therein.
(61) Fenton, R. R.; Vagg, R. S.; Williams, P. A. Inorg. Chim. Acta 1988,
148, 37-44.
(44) Bosnich, B. Proc. R. Soc. London Ser. A 1967, 88.
(45) Bosnich, B.; Kneen, W. R. Inorg. Chem. 1970, 9, 2191-4.
(46) Fenton, R. R.; Stephens, F. S.; Vagg, R. S.; Williams, P. A. Inorg.
Chim. Acta 1992, 201, 157-64.
(62) Chambers, J. A.; Mulqi, M. W.; Williams, P. A.; Vagg, R. S. Inorg.
Chim. Acta 1984, 81, 55-60.
(63) Chambers, J. A.; Goodwin, T. J.; Mulqi, M. W.; Williams, P. A.; Vagg,
R. S. Inorg. Chim. Acta 1984, 88, 193-9.
(47) Fenton, R. R.; Stephens, F. S.; Vagg, R. S.; Williams, P. A. Inorg.
Chim. Acta 1991, 182, 67-75.
(64) Heinrichs, M. A.; Hodgson, D. J.; Michelsen, K.; Pedersen, E. Inorg.
Chem. 1984, 23, 3174-80.
(48) Fenton, R. R.; Stephens, F. S.; Vagg, R. S.; Williams, P. A. Inorg.
Chim. Acta 1992, 197, 233-42.
(65) Bosnich, B.; Kneen, W. R.; Phillip, A. T. Inorg. Chem. 1969, 8, 2567-
74.
(49) Michelsen, K.; Nielsen, K. M. Acta Chem. Scand. A 1980, 34, 755-
64.
(66) Bernal, I.; Cetrullo, J.; Worrell, J. H.; Li, T. Polyhedron 1994, 13,
463-8.
(67) Toscano, P. J.; Fordon, K. J.; Engelhardt, L. M.; Skelton, B. W.; White,
A. H.; Marzilli, P. A. Polyhedron 1990, 9, 1079.
(68) Dwyer, F. P.; Garvan, F. L. J. Am. Chem. Soc. 1959, 81, 2955-7.
(69) Dwyer, F. P.; MacDermott, T. E. J. Am. Chem. Soc. 1963, 85, 2916-
9.
(50) Glerup, J.; Goodson, P. A.; Hodgson, D. J.; Michelsen, K. Inorg. Chem.
1995, 34, 6255-64.
(51) Glerup, J.; Michelsen, K.; Arulsamy, N.; Hodgson, D. J. Inorg. Chim.
Acta 1998, 274, 155-166.
(52) Hazell, R.; Jensen, K. B.; McKenzie, C. J.; Toftlund, H. J. Chem.
Soc., Dalton Trans. 1995, 707-17.

Metal Complexes with Cis R Topology
Inorganic Chemistry, Vol. 38, No. 24, 1999 5547
Figure 3. Comparison of cis R and ansa metallocene complexes.
Figure 5. Quadridentate pyridylamine ligands.
as catalysts for a variety of reactions. The first goal was to
identify convenient and versatile synthetic routes to ligand
targets. Ethylenediamine (en) and the chiral S-propylenediamine
(S-pn) were employed as the backbones for comparison.46,74
Ligands based on the bulkier R,R-diaminocyclohexane and their
complexation chemistry are the subject of a future report.75 For
the S-pn system, ligands with pyridyl, 3, and quinolyl, 4, donor
groups were prepared to vary the steric bulk in the terminal
donor positions of the quadridentate. To determine whether
subtle differences in the electronic nature of substituents might
influence physical properties, both -CH2C6H5, 1,76 and
-CH2C6F5, 2, analogues of the achiral en backbone ligand were
synthesized. Benzyl groups were also chosen for their steric
bulk and their potential to improve solubility in organic solvents.
Since many reactions are promoted by Lewis acidic metal
centers, electron withdrawing fluorine substituents were intro-
duced to generate more electron deficient complexes. Moreover,
since pyridine and quinoline groups are π acidic, these donors
are also expected to enhance Lewis acidity. Depending upon
the particular metal ion, the steric features of the ligand, and
the donor strength of the counterions, six, five, or four coordinate
structures could be obtained upon reaction of quadridentates
with divalent metal halide salts, MX2. Ultimately, complexes
prepared from metal halides may be further activated to
coordinate Lewis basic substrates by exchanging the halides for
triflates, OTf-, hexafluoroantimonates, SbF6-, hexafluorophos-
phates, PF6-, tetraphenylborates, BPh4-, or other weakly
coordinating counterions via metathesis with the respective silver
salts. The synthesis of these ligands and many of their first-
row transition metal complexes, as well as structural and
physical characterization, are discussed below.
Figure 4. Topological and conformational chirality in quadridentate
chelates.
donor group and the substituent on N (or lone pair on S) are
oriented anti to each other in the cis R arrangement.47 It is
interesting to compare cis R complexes and ansa metallo-
cenes,70-73 which are similar in topology. They differ in that
strapped metallocene ligands bind “face on” via cyclopentadi-
enyl (Cp) and its analogues, with substituents projecting out
and away from the reaction wedge for many derivatives. In
contrast, the terminal donor groups on cis R pyridylamine
ligands bind “edge on” and R groups point more directly at
substrate binding sites (Figure 3).
While these principles may be exploited to impose a
nonplanar C2-symmetric cis R topology, other features must be
incorporated to make the ligand stereoselective and, in some
cases, even stereospecific.41 In many chelating ligands, the
introduction of one or more chiral centers on the internal chelate
ring backbone has been sufficient to ensure that a single
diastereomer is obtained. For example, when the sexadentate
chelate S-propylenediamine tetraacetic acid (S-PDTA) binds to
metal ions, the methyl substituent assumes a pseudoequatorial
position on the puckered five-membered chelate ring. This
preference fixes the conformational chirality of the metal
propylenediamine ring as δ, which in turn, controls the way in
which the terminal donor groups wrap to give exclusively the
∆ absolute configuration about the metal center.68,69 A similar
kind of diastereoselection has been observed in amine-based
quadridentate chelates prepared from enantiomerically pure
chiral diamines (Figure 4).41
Experimental Section
General Considerations. All reagents and solvents were used as
received from commercial sources (Aldrich, Acros, Strem) unless
otherwise indicated. Pyridinecarboxaldehyde was distilled under vacuum
prior to use. THF was dried and purified on alumina columns.77 Cyclic
voltammetric measurements were made either with a Bioanalytical
Systems, Inc., model CV-27 or a Bioanalytical Systems model CV-
Taking these concepts and precedent with related systems
into account, we have prepared a series of quadridentate ligands
(Figure 5) and their first-row transition metal complexes. This
study was undertaken as a prelude to screening these complexes
(74) Goodwin, T. J.; Vagg, R. S.; Williams, P. A. J. Proc. R. Soc. New
South Wales 1984, 117, 1-6.
(75) Ng, C.; Savage, S. A.; Derringer, D.; Sabat, M.; Campana, C.; Fraser,
C. L. Manuscript in preparation.
(70) Hoveyda, A. H.; Morken, J. P. Angew. Chem., Int. Ed. Engl. 1996,
35, 1263-84.
(71) Alt, H. G.; Samuel, E. Chem. Soc. ReV. 1998, 27, 323-9.
(72) Petasis, N. A.; Hu, Y. H. Curr. Org. Chem. 1997, 1, 249-86.
(73) Soga, K.; Shiono, T. Prog. Polym. Sci. 1997, 22, 1503-46.
(76) Reference to the prior synthesis of en(Bn)py appeared in the course
of our work with this ligand, 1:41 Aldrich-Wright, J. R. Ph.D. Thesis,
Macquarie University, 1993.
(77) Pangborn, A. B.; Giardello, M. A.; Grubbs, R. H.; Rosen, R. K.;
Timmers, F. J. Organometallics 1996, 15, 1518-20.

5548 Inorganic Chemistry, Vol. 38, No. 24, 1999
Ng et al.
50W instrument on dichloromethane solutions that contained 0.1 M
tetra-n-butylammonium hexafluorophosphate (TBAH) as the supporting
electrolyte. A glassy carbon electrode and a Pt-wire electrode were
utilized. E1/2 values, determined as (Ep,a + Ep,c)/2, were referenced to
the aqueous Ag/AgCl electrode at room temperature and are uncorrected
for junction potentials. Under our experimental conditions the ferro-
cenium/ferrocene couple is at E1/2 ) 0.47 V vs Ag/AgCl. Voltammo-
grams shown in Figure 8 were referenced vs a nonaqueous (CH3CN)
Ag/AgCl electrode. Under these conditions the ferrocene couple is
observed at E1/2 ) 0.24 V. Conductivity measurements were performed
at room temperature on 1 mM acetonitrile solutions using a YSI model
35 conductance meter. Infrared spectra were recorded as mineral oil
(Nujol) mulls supported on NaCl plates in the region 4000-600 cm-1
using a Nicolet Impact 400 Fourier transform spectrometer. Electronic
absorption spectra were recorded on Hewlett-Packard diode array
spectrophotometers, either model HP8452A (200-800 nm) or model
(400 mL) of the HCl salt (21.33 g, 0.153 mol) through a column of
strongly basic Amberlite 402 anion-exchange resin (225 g, 17%
capacity): 10.84 g, Yield: 69% by NMR integrations. The resulting
ester was reacted with NH3 in a saturated MeOH solution to form the
amide, which was subsequently reduced with BH3‚THF and worked
up as described by Miller et al.82 to produce S-propylenediamine as its
HCl salt. S-propylenediamine‚2HCl (0.980 g, 6.66 mmol) was crushed
into a powder and then was suspended in CH3CN (50 mL). Triethyl-
amine (18 mL, 0.13 mol) and 2-pyridinecarboxaldehyde (1.3 mL, 0.014
mol) were added, and the reaction was allowed to stir for ∼15 h at 25
°C. After this time, H2O (30 mL) was added and the aqueous layer
was extracted with CH2Cl2 (5 × 50 mL). Combined CH2Cl2 layers
were dried over Na2SO4, filtered, and concentrated in vacuo to give
the crude Schiff base product as a yellow orange oil: 1.50 g, 89%. 1H
NMR (CDCl3, 300 MHz): δ 1.39 (d, J ) 6.1 Hz, 3H), 3.89 (m, 3H),
7.28 (m, 2H), 7.70 (m, 2H), 7.97 (m, 2H), 8.36 (s, 1H), 8.40 (s, 1H),
8.61 (m, 2H). The resulting diimine was reduced with NaBH4 in MeOH
by the general procedure described above for 1: 1.14 g, 75%.83 1H
NMR (CDCl3, 300 MHz): δ 1.11 (d, J ) 6.1 Hz, 3H), 2.71-2.56 (br
m, 2H), 2.84 (m, 1H), 4.01-3.83 (br m, 4H), 7.14 (m, 2H), 7.33 (d, J
) 7.3 Hz, 2H), 7.62 (m, 2H), 8.49 (d, J ) 4.9 Hz, 2H). Alkylation of
the resulting diamine was effected by the method described for 1 using
R-bromo-2,3,4,5,6-pentafluorotoluene in place of benzyl bromide. In
the final workup, after the aqueous layer was made basic, the alkylated
product was extracted with Et2O instead of EtOAc. The crude ligand,
3, was obtained as a viscous brown oil which solidified upon
standing: 2.37 g, 87%. Typically the crude ligand 3 was not purified
prior to reaction with metal chloride salts. However further purification
may be effected by dissolving the crude ligand in Et2O and adding
hexanes to precipitate a viscous brown impurity. After filtration through
Celite and concentration, the resulting residue may be recrystallized
1
HP8453 (200-1100 nm). H and 13C NMR spectra were recorded at
300 and 75 MHz, respectively, on either a GE QE-300 or a GN-300
spectrometer. Elemental microanalyses were performed on a Perkin-
Elmer model 2400 Series II CHNS/O analyzer or were determined by
Atlantic Microlab, Inc., Norcross, GA.
Ligand Syntheses. 1,6-Diphenyl-2,5-bis(2-methylpyridyl)-2,5-di-
azahexane, en(Bn)py (1). Ethylenediamine (2.1 mL, 0.031 mol) was
stirred with 2-pyridinecarboxaldehyde (6 mL, 0.063 mol) in CH2Cl2
(40 mL) containing molecular sieves for 2 h. The reaction was then
filtered and concentrated to give the Schiff base as a pale yellow oil:
7.10 g, 95%.78,79 The diimine (7.10 g, 0.030 mol) was dissolved in
MeOH (50 mL), NaBH4 (2.48 g, 0.066 mol) was added, and the reaction
was stirred for 30 min at room temperature. Aqueous HCl (3 M, 24
mL) was added slowly with stirring, and then the MeOH was removed
via rotary evaporation. Aqueous NaOH (20%) was added until the
solution reached pH > 10. The product was extracted with CH2Cl2 (4
× 50 mL), was washed with H2O (2 × 50 mL) and brine (50 mL), and
then was dried over Na2SO4. Following filtration and concentration,
the diamine was obtained as a yellow oil: 6.0 g, 83%.42,80 Alkylation
was effected by dissolving the diamine (0.742 g, 3.06 mmol) in dry
THF (50 mL) and deprotonating with NaH (0.476 g, 19.8 mmol). The
mixture was cooled to 0 °C and was stirred for ∼20 min prior to
addition of benzyl bromide (1.1 mL, 9.3 mmol). The reaction was
allowed to warm to room temperature slowly and then was stirred at
that temperature for an additional 12 h. Water (15 mL) and then HCl
(3 M) were added until the solution reached pH ) 2. The acidic mixture
was concentrated in vacuo to remove THF and then was washed with
Et2O (3 × 20 mL). The aqueous solution was made basic (pH > 10)
through addition of NaOH (10%), was saturated with NaCl, and then
was extracted with EtOAc (4 × 20 mL). The solution was dried over
Na2SO4, filtered, and concentrated to yield the product, 1, as an off-
white solid: 1.22 g, 80%. The ligand may be further purified by silica
1
from Et2O/hexanes to yield 3 as a beige powder. H NMR (CDCl3,
300 MHz): δ 1.07 (d, J ) 6.6 Hz, 3H), 2.43 (m, 1H), 2.73 (m, 1H),
2.98 (q, J ) 6.6 Hz, 1H), 3.70 (m, 8H), 7.11 (m, 2H), 7.39 (d, J ) 7.8
Hz, 2H), 7.59 (m, 2H), 8.42 (d, J ) 4.2 Hz, 1H), 8.47 (d, J ) 5.1 Hz,
1H).
3S-Methyl-1,6-bis(pentafluorophenyl)-2,5-bis(2-methylquinolyl)-
2,5-diazahexane, S-pn(F5Bn)qn (4). Crushed S-pn‚2HCl was sus-
pended in CH3CN (54 mL), and then Et3N (18 mL) and 2-quinoline
carboxyaldehyde (2.22 g, 14.0 mmol) were added. The reaction was
stirred for 15 h at 25 °C. H2O (20 mL) was added, and the solution
was extracted with CH2Cl2 (4 × 40 mL). Combined organic layers
were washed with H2O (2 × 40 mL) and with saturated brine (2 × 40
mL) and then were dried over Na2SO4. After filtration and concentration
via rotovap, the crude Schiff base product was obtained as a brownish
solid. The crude solid was dissolved in a minimal amount of CH2Cl2,
and hexanes were added until a dark brown residue precipitated. The
remaining supernatant was filtered through Celite and then was
concentrated in vacuo just to the point when crystallization commenced.
After being chilled at 0 °C for 2 h, the pale yellow crystalline solid
was collected by filtration, was washed with a minimal amount of
hexanes, and then was dried in vacuo: 1.48 g, 62%. 1H NMR (CDCl3,
300 MHz): δ 1.45 (d, J ) 5.5 Hz, 3H), 4.00 (m, 3H), 7.55 (m, 2H),
7.72 (m, 2H), 7.82 (d, J ) 7.9 Hz, 2H), 8.08 (s, 1H), 8.11 (s, 1H),
8.16 (m, 4H), 8.57 (s, 1H), 8.61 (s, 1H). The diimine was reduced
with NaBH4 in MeOH by the general procedure described for 1 to give
the crude diamine as a tan oil in essentially quantitative yield. 1H NMR
(CDCl3, 300 MHz): δ 1.17 (d, J ) 6.1 Hz, 3H), 3.02-2.68 (br m,
3H), 4.24-4.05 (br m, 4H), 7.52-7.47 (m, 4H), 7.66 (m, 2H), 7.77
(d, J ) 8.6 Hz, 2H), 8.02 (d, J ) 3.6 Hz, 2H), 8.09 (d, J ) 3.6 Hz,
2H). The diamine was subsequently alkylated with R-bromo-2,3,4,5,6-
pentafluorotoluene in THF in the presence of NaH by the standard
procedure described for 1. In the final workup, after the aqueous layer
was made basic, the alkylated product was extracted with Et2O instead
of EtOAc. The product, 4, was obtained as a brittle yellow solid: 2.19
1
gel flash chromatography with EtOAc. H NMR (CDCl3, 300 MHz):
δ 2.68 (s, 4H), 3.57 (s, 4H), 3.70 (s, 4H), 7.11 (m, 2H), 7.25 (m, 10H),
7.45 (d, J ) 8.1 Hz, 2H), 7.57 (m, 2H), 8.48 (d, J ) 5.0, 2H).
1,6-Bis-(pentafluorophenyl)-2,5-bis(2-methylpyridyl)-2,5-diaza-
hexane, en(F5Bn)py (2). The ligand 2 was prepared by the method
described for 1 using R-bromo-2,3,4,5,6-pentafluorotoluene instead of
benzyl bromide. In the final work up, after the aqueous layer was made
basic, the alkylated product was extracted with CH2Cl2 instead of
EtOAc. The ligand, 2, was obtained as a white crystalline solid: 2.44
1
g, 87%. H NMR (CDCl3, 300 MHz): δ 2.66 (s, 4H), 3.74 (s, 4H),
3.76 (s, 4H), 7.14 (m, 2H), 7.34 (d, J ) 7.9 Hz, 2H), 7.60 (m, 2H),
8.48 (d, J ) 4.3 Hz, 2H).
3S-Methyl-1,6-bis(pentafluorophenyl)-2,5-bis(2-methylpyridyl)-
2,5-diazahexane, S-pn(F5Bn)py (3). L-Alanine methyl ester hydro-
chloride was converted to the free base by passing a MeOH solution
(78) For other preparations of the diimine see: DeVos, D. E.; Feijen, E. J.
P.; Schoonheydt, R. A.; Jacobs, P. A. J. Am. Chem. Soc. 1994, 116,
4746-52.
(79) Busch, D. H.; Bailer, J. C., Jr. J. Am. Chem. Soc. 1956, 78, 1137.
(80) For previous preparations and uses of this pyridyldiamine see ref 43
and: Newkome, G. R.; Frere, Y. A.; Fronczek, F. R.; Gupta, V. K.
Inorg. Chem. 1985, 24, 1001-6.
(81) Basak, A. K.; Martell, A. E. Inorg. Chem. 1988, 27, 1948-55.
(82) Miller, D. D.; Hsu, F.-L.; Ruffolo, R. R., Jr.; Patil, P. N. J. Med. Chem.
1976, 19, 1382-4.
(83) For another preparation of â-pnHpy diamine see ref 49 and the
following: McCollum, D. G.; Fraser, C.; Ostrander, R.; Rheingold,
A. L.; Bosnich, B. Inorg. Chem. 1994, 33, 2383-92.

Metal Complexes with Cis R Topology
Inorganic Chemistry, Vol. 38, No. 24, 1999 5549
1
g, 77%. H NMR (CDCl3, 300 MHz): δ 1.12 (d, J ) 6.9 Hz, 3H),
[Cu{en(F5Bn)py}Cl2]‚0.5CH3OH. The Cu(II) compound was ob-
tained as a mint green microcrystalline solid after slow evaporation of
CH2Cl2 from a CH2Cl2/MeOH solution of the complex. Yield: 93%.
ΛM ) 54 Ω-1 mol-1 cm2. Anal. Calcd for C28.5H22N4O0.5F10Cl2Cu: C,
45.46; H, 2.95; N, 7.44. Found: C, 45.54; H, 2.99; N, 7.13.
[Zn{en(F5Bn)py}Cl2]‚CH3OH. This complex was obtained as white
needles. Yield: 84%. ΛM ) 5.1 Ω-1 mol-1 cm2. Anal. Calcd for
C29H24N4OF10Cl2Zn: C, 45.14; H, 3.14; N, 7.27. Found: C, 45.42; H,
3.01; N, 7.14. 1H NMR (CDCl3, 300 MHz): δ 2.23, 2.49 (system AB,
JAB ) 11 Hz, 4H), 3.36 (d, J ) 15.4 Hz, 2H), 3.48 (m, 2H), 4.83 (d,
J ) 14.6 Hz, 2H), 5.05 (d, J ) 14.3 Hz, 2H), 7.30 (d, J ) 7.7 Hz,
2H), 7.47 (t, 2H), 7.87 (t, 2H), 9.75 (d, J ) 4.6 Hz, 2H). 13C NMR
(CDCl3, 75 MHz): δ 42.4, 45.5, 57.3, 123.7, 124.0, 135.4, 139.2, 142.6,
143.6, 147.0, 149.5, 153.6.
2.51 (m, 1H), 2.84 (m, 1H), 3.08 (m, 1H), 3.78 (m, 8H), 7.49 (m, 4H),
7.68 (m, 2H), 7.76 (m, 2H), 7.99 (m, 4H).
Preparation of Metal Complexes. Metal complexes were prepared
by reaction of alcohol (EtOH or MeOH) solutions of the appropriate
metal chloride salt with an alcohol solution of the ligand. A typical
reaction scale is as follows: ligand (0.15 mmol); metal chloride salt
(0.16 mmol); alcohol (3 mL total). For the less soluble ligands, 2 and
4, a minimal amount of methylene chloride was sometimes added to
facilitate dissolution. In cases where the complexes precipitated from
the reaction solution, they were collected by filtration. Otherwise the
solutions were concentrated in vacuo and the residues were purified
by recrystallization. Specific purification procedures and deviations from
this standard method are indicated below for the respective compounds.
Metal Complexes of en(Bn)py (1). [Mn{en(Bn)py}Cl2]‚0.5CH2Cl2.
The reaction solution was concentrated in vacuo. The resulting residue
was recrystallized from CH2Cl2/Et2O to give the Mn(II) complex as a
white microcrystalline solid. Yield: 74%. ΛM ) 0.20 Ω-1 mol-1 cm2.
Anal. Calcd for C28.5H31N4Cl3Mn: C, 57.93; H, 5.29; N, 9.48. Found:
C, 57.61; H, 5.68; N, 9.30.
Metal complexes of S-pn(F5Bn)py (3). [Mn{S-pn(F5Bn)py}Cl2].
The Mn(II) complex precipitated from the EtOH reaction solution as a
white microcrystalline solid. Yield: 85%. ΛM ) 2.0 Ω-1 mol-1 cm2.
Anal. Calcd for C29H22N4F10Cl2Mn: C, 46.92; H, 2.99; N, 7.55.
Found: C, 46.69; H, 3.28; N, 7.35.
[Fe{S-pn(F5Bn)py}Cl2]‚0.5CH2Cl2. The EtOH reaction mixture was
concentrated in vacuo. The resulting residue was dissolved in CH2Cl2
and precipitated from hexanes to produce a canary yellow powdery
solid. Yield: 78%. ΛM ) 0.5 Ω-1 mol-1 cm2. Anal. Calcd for
[Fe{en(Bn)py}Cl2]‚0.5CH2Cl2. The CH2Cl2/EtOH reaction mixture
was concentrated in vacuo. The resultant yellow residue was recrystal-
lized from CH2Cl2/hexanes to give canary yellow needles. Yield: 82%.
ΛM ) 0.13 Ω-1 mol-1 cm2. Anal. Calcd for C28.5H31N4Cl3Fe: C, 57.84;
H, 5.28; N, 9.47. Found: C, 58.23; H, 5.44; N, 9.43.
C29.5H23N4F10Cl3Fe: C, 45.10; H, 2.95; N, 7.13. Found: C, 45.17; H,
3.26; N, 7.14.
[Co{en(Bn)py}Cl2]‚0.5CH2Cl2. The Co(II) compound was recrys-
tallized by vapor diffusion of Et2O into a CH2Cl2/Et2O solution of the
[Co{S-pn(F5Bn)py}Cl2]. The Co(II) complex precipitated from the
EtOH reaction solution as a lavender microcrystalline solid. Yield: 82%.
ΛM ) 7.0 Ω-1 mol-1 cm2. Anal. Calcd for C29H22N4F10Cl2Co: C, 46.67;
H, 2.97; N, 7.51. Found: C, 46.28; H, 3.33; N, 7.16.
complex to yield a purple microcrystalline solid. Yield: 85%. ΛM
)
5.6 Ω-1 mol-1 cm2. Anal. Calcd for C28.5H31N4Cl3Co: C, 57.54; H,
5.25; N, 9.42. Found: C, 57.14; H, 5.60; N, 9.32.
[Ni{S-pn(F5Bn)py}Cl2]. The Ni(II) complex precipitated from the
EtOH reaction solution as a mint green microcrystalline solid. Yield:
34%. ΛM ) 30 Ω-1 mol-1 cm2. Anal. Calcd for C29H22N4F10Cl2Ni: C,
46.69; H, 2.97; N, 7.51. Found: C, 46.53; H, 3.31; N, 7.33.
[Cu{S-pn(F5Bn)py}Cl2]‚0.5CH2Cl2. The Cu(II) compound precipi-
tated as a lime green powder from a CH2Cl2/Et2O solution. Yield: 71%.
ΛM ) 92 Ω-1 mol-1 cm2. Anal. Calcd for C29.5H23N4F10Cl3Cu: C,
44.66; H, 2.92; N, 7.06. Found: C, 44.54; H, 3.27; N, 7.30.
[Zn{S-pn(F5Bn)py}Cl2]. The Zn(II) complex precipitated from the
EtOH reaction solution as a white microcrystalline solid. Yield: 67%.
ΛM ) 36 Ω-1 mol-1 cm2. Anal. Calcd for C29H22N4F10Cl2Zn: C, 46.27;
H, 2.95; N, 7.44. Found: 46.13; H, 3.24; N, 7.24. Major isomer: 1H
NMR (CDCl3, 300 MHz) δ 0.66 (d, J ) 7.3 Hz, 3 H), 2.12 (d, J )
11.6 Hz, 1H), 2.47 (m, 1H), 2.87 (m, 1H), 3.34 (d, J ) 6.1 Hz, 1H),
3.39 (d, J ) 5.5 Hz, 1H), 3.45 (d, J ) 14.0 Hz, 1H), 4.05 (d, J ) 14.0,
1H), 4.76 (d, J ) 16.5 Hz, 1H), 4.98 (m, 2H), 5.21 (d, J ) 14.0 Hz,
1H), 7.30 (d, J ) 7.9 Hz, 1H), 7.37 (d, J ) 7.3 Hz, 1H), 7.46 (m, 2H),
7.89 (m, 2H), 9.72 (m, 2H); 13C NMR (CDCl3, 75 MHz) δ 11.7, 39.5,
43.0, 50.8, 51.3, 55.8, 57.2, 123.6, 123.8, 124.1, 139.1, 149.6, 149.7,
153.4, 154.4.
[Zn{en(Bn)py}Cl2]. The Zn(II) compound was recrystallized by
vapor diffusion of Et2O into a CH2Cl2/Et2O solution to yield a pale
yellow microcrystalline solid. Yield: 66%. ΛM ) 1.7 Ω-1 mol-1 cm2.
1H NMR (CDCl3, 300 MHz): δ 2.34, 2.57 (system AB, JAB ) 4.4 Hz,
4H), 3.18, 4.96 (system AB, JAB ) 14.4 Hz, 4H), 3.42, 4.64 (system
AB, JAB ) 13.7, 4H), 7.12 (br s, 4H), 7.22 (d, J ) 6.39, 2H), 7.35 (m,
6H), 7.41 (t, 2H), 7.81 (t, 2H), 9.73 (s, 2H). 13C NMR (CDCl3, 75
MHz): δ 43.8, 54.4, 58.3, 123.3, 123.6, 127.9, 128.1, 131.0, 131.9,
138.5, 149.6, 154.1.
Metal Complexes of en(F5Bn)py (2). [Mn{en(F5Bn)py}Cl2]‚
0.5CH3CH2OH. The Mn(II) complex was prepared from MnCl2 and
the tetra-HCl salt of the ligand, 2, in EtOH solution. White needles
precipitated from the EtOH solution. Yield: 97%. ΛM ) 1.4 Ω-1 mol-1
cm2. Anal. Calcd for C29H33N4O0.5F10Cl2Mn: C, 46.36; H, 3.09; N,
7.46. Found: C, 46.27; H, 3.04; N, 7.26.
[Fe{en(F5Bn)py}Cl2]PF6. The Fe(III) complex precipitated from
EtOH solution upon addition of excess NH4PF6 (3 equiv). The resulting
yellow residue was recrystallized from CH3CN/EtOH to yield a canary
yellow microcrystalline solid. Yield: 62%. ΛM ) 97 Ω-1 mol-1 cm2.
Anal. Calcd for C28H20N4PF16Cl2Fe: C, 38.50; H, 2.31; N, 6.41.
Found: C, 38.42; H, 2.46; N, 6.16.
[Fe{en(F5Bn)py}Cl2]‚CH2Cl2. The EtOH reaction mixture was
concentrated in vacuo, and the resulting crude Fe(II) compound was
recrystallized from CH2Cl2/hexanes to give powdery canary yellow
needles. Yield: 92%. ΛM ) 0.8 Ω-1 mol-1 cm2. Anal. Calcd for
C29H22N4F10Cl4Fe: C, 42.78; H, 2.72; N, 6.88. Found: C, 42.45; H,
2.92; N, 6.65.
Metal Complexes of S-pn(F5Bn)qn (4). [Co{S-pn(F5Bn)qn}Cl2]‚
0.5CH2Cl2. The Co(II) complex was obtained as a periwinkle micro-
crystalline solid after precipitation from CH2Cl2/Et2O. Yield: 70%. ΛM
) 43 Ω-1 mol-1 cm2. Anal. Calcd for C37.5H27N4F10Cl3Co: C, 50.67;
H, 3.06; N, 6.30. Found: C, 50.78; H, 3.45; N, 6.23.
[Zn{S-pn(F5Bn)qn}Cl2]. The Zn(II) compound was obtained as a
beige solid upon concentration of the EtOH/CH2Cl2 solution. Yield:
46%. ΛM ) 29 Ω-1 mol-1 cm2. Major isomer: 1H NMR (CDCl3, 300
MHz) δ 1.09 (d, J ) 5.5 Hz, 3H), 2.99 (d, J ) 15.3 Hz, 1H), 3.65 (m,
1H), 3.87 (m, 1H), 4.00 (d, J ) 19.5 Hz, 1H), 4.09 (d, J ) 15.9 Hz,
1H), 4.26 (d, J ) 9.16 Hz, 1H), 4.32 (d, J ) 12.8 Hz, 1H), 4.54 (d, J
) 15.3 Hz, 1H), 4.79 (d, J ) 15.3 Hz, 1H), 5.22 (m, 2H), 6.96 (m,
1H), 7.31 (m, 2H), 7.73 (m, 4H), 7.90 (d, J ) 7.9 Hz, 1H), 8.03 (m,
1H), 8.25 (d, J ) 7.9 Hz, 1H), 8.34 (d, J ) 7.9 Hz, 1H), 9.69 (d, J )
6.7 Hz, 1H); 13C NMR (CDCl3, 75 MHz) δ 13.2, 42.4, 46.2, 49.8,
55.8, 56.2, 59.6, 104.6, 108.8, 121.2, 122.7, 126.2, 127.0, 127.3, 128.0,
128.3, 129.7, 130.3, 135.6, 138.9, 139.6, 139.9, 140.9, 143.6, 144.0,
145.3, 147.2, 157.6, 159.7.
[Co{en(F5Bn)py}Cl2]‚0.5CH3CH2OH.0.5CH2Cl2. The Co(II) com-
plex was obtained as lavender needles by concentration of the EtOH
reaction mixture, followed by recrystallization of the resulting residue
by evaporation of CH2Cl2 from a CH2Cl2/EtOH solution. Yield: 90%.
ΛM ) 7.6 Ω-1 mol-1 cm2. Anal. Calcd for C29.5H24N4O0.5F10Cl3Co:
C, 44.10; H, 3.03; N, 7.02. Found: C, 44.33; H, 3.00; N, 7.03. (Note:
1
The H NMR spectrum obtained after oxidation of this complex to
Co(III) confirms the presence of these associated solvents. This
spectrum is provided as part of the Supporting Information.)
[Ni{en(F5Bn)py}Cl2]‚0.5CH3CH2OH. The Ni(II) complex was
prepared from NiCl2 and the tetra-HCl salt of the ligand. It precipitated
as mint green needles from the EtOH reaction medium. Yield: 79%.
ΛM ) 22 Ω-1 mol-1 cm2. Anal. Calcd for C29H23N4O0.5F10Cl2Ni: C,
45.76; H, 2.96; N, 7.49. Found: C, 46.05; H, 2.79; N, 7.48.
Oxidation of Co(II) Complexes. Co(II) complexes were oxidized
with H2O2 according to the procedure described by Fenton et al.46
Oxidations with Br2 were performed by the procedure described below
1
for [Co{en(Bn)py}Cl2], and then H NMR spectra were recorded. If

5550 Inorganic Chemistry, Vol. 38, No. 24, 1999
Scheme 1. Quadridentate Ligand Synthesis
Ng et al.
the following manner: diamine(alkyl group)terminal donor
group. The diamine backbone is indicated first (en or S-pn),
followed by the noncoordinating alkyl group given in paren-
theses (Bn or F5Bn) and then the terminal donor group (py )
pyridine; qn ) quinoline). The Schiff base precursor to ligands
1 and 2 was prepared by stirring ethylenediamine with 2 equiv
of pyridinecarboxaldehyde in CH2Cl2 containing molecular
sieves. Filtration and concentration yielded the diimine products
in high yield. While this approach works well for en backbone
ligands, since S-pn is isolated as its di-HCl salt via an
asymmetric synthesis starting from L-alanine methyl ester
hydrochloride,82 analogous Schiff base syntheses using S-pn to
make ligands 3 and 4 required the prior liberation of the free
base. This can be achieved using either concentrated basic
solutions or a basic ion exchange column. Since these methods
for generating the S-pn free base are somewhat tedious and they
result in significant losses of product, a direct route was devised
that involved the reaction of S-pn‚2HCl with the desired
aldehyde in the presence of excess base, i.e., triethylamine, in
acetonitrile solution. After workup, the desired S-pn Schiff bases
were obtained in good yield.
The Schiff bases were then reduced with sodium borohydride
in methanol to yield quadridentates with secondary diamines
at the internal donor positions.86,87 This approach is more
convenient than other reported methods using diborane83 or
reductions with Zn/HOAc42 or Pd/C.88 Reduction products were
obtained relatively cleanly and in high yield. The final step of
ligand synthesis involves alkylation of the secondary nitrogen
centers to produce tertiary amines. Though reductive amination
using formaldehyde and sodium cyanoborohydride works well
for methylation,47 this route is not effective for larger, bulkier
aldehydes. Instead it was discovered that reaction of the 2°
diamines with NaH in THF, followed by addition of benzyl
bromide, C6H5CH2Br, or R-bromo-2,3,4,5,6-pentafluorotoluene,
C6F5CH2Br, resulted in the efficient formation of the desired
alkylated products.89 Most quadridentate ligands were used as
prepared for making complexes. In certain cases, they were
purified by chromatography or by conversion to their tetra-HCl
salts followed by recrystallization.
Synthesis and Characterization of Metal Complexes. Metal
complexes were prepared by combining alcohol solutions of
the free ligands with those of the appropriate metal chloride,
MCln (n ) 2, M ) Mn, Fe, Co, Ni, Cu, Zn; n ) 3, M ) Fe).
In a few cases (M ) Mn and Ni), following literature precedent,
the tetra-HCl salt of the en(F5Bn)py ligand was utilized in place
of the free base. Since the S-pn(F5Bn)qn ligand, 4, showed very
limited solubility in methanol or ethanol, it was dissolved in a
warm CH2Cl2/EtOH solvent mixture prior to addition of ethanol
solutions of the metal salts. Immediate color changes indicative
of complexation were observed in all cases. For certain ligand-
metal combinations, namely for S-pn(F5Bn)py with M ) Co,
Mn, Ni, and Zn and en(F5Bn)py for M ) Mn and Ni, complexes
precipitated directly from the alcohol reaction medium. Soluble
complexes were isolated by concentration in vacuo followed
by crystallization using either vapor diffusion or layering
techniques with the indicated two solvent systems. Crude
necessary, the solutions were warmed to dissolve the complexes prior
to addition of bromine. Chemical shifts attributable to the major isomers
evident in the 1H NMR spectra of the bromine reactions are tabulated
below. Note: In all cases, minor products characterized by resonances
at ∼9.88-9.95 ppm are present. It is estimated that these comprise up
to 15-20% of the reaction mixture as determined by integration.
[Co{en(Bn)py}Cl2] Oxidation. [Co{en(Bn)py}Cl2] (8.7 mg, 0.015
mmol) and CDCl3 (0.6 mL) were combined in an NMR tube. Bromine
1
(0.050 mL, 0.97 mmol) was added, and a H NMR spectrum of the
resultant red-brown solution was recorded. 1H NMR (CDCl3, 300
MHz): δ 2.85 (s, 4H), 3.35 (m, 4H), 4.48 (m, 2H), 5.13 (m, 2H), 7.34
(m, 4H), 7.55 (m, 6H), 7.62 (m, 2H), 7.72 (m, 2H), 8.09 (m, 2H), 9.64
(d, J ) 5.8 Hz, 2H).
[Co{en(F5Bn)py}Cl2] Oxidation. 1H NMR (CDCl3, 300 MHz): δ
2.97 (s, 4H), 3.42 (d, J ) 14.2 Hz, 2H), 3.93 (d, J ) 15.0 Hz, 2H),
4.72 (d, J ) 15.4 Hz, 2H), 5.21 (d, J ) 15.4 Hz, 2H), 7.76 (m, 4H),
8.17 (m, 2H), 9.68 (d, J ) 5.8 Hz, 2H). This spectrum is provided as
part of the Supporting Information.
1
[Co{S-pn(F5Bn)py}Cl2] Oxidation. H NMR (CDCl3, 300 MHz):
δ 1.27 (m, 3H), 2.34 (m, 1H), 2.62 (d, J ) 14.2 Hz, 1H), 3.32-2.96
(br m, 2H), 3.49 (m, 1H), 4.02 (d, J ) 16.6 Hz, 1H), 5.01-4.67 (br m,
3H), 5.12 (d, J ) 13.9 Hz, 1H), 5.45 (d, J ) 16.2 Hz, 1H), 7.94-7.73
(br m, 4H), 8.20 (m, 2H), 9.64 (m, 2H).
X-ray Structure Determination. A thin green plate of dimensions
0.32 × 0.11 × 0.48 mm was used for all X-ray experiments. The data
collection was carried out on a Rigaku AFC6S diffractometer at -120
°C using Mo KR radiation (λ ) 0.710 69 Å). Unit cell dimensions
were determined by applying the setting angles of 25 high-angle
reflections. Intensities of three standard reflections were monitored
during the data collection showing no significant variance. The
intensities were corrected for absorption by using ψ scans of several
reflections. The transmission factors ranged from 0.76 to 1.00. The
structure was solved by direct methods (SIR92).84 Calculations were
performed on a Silicon Graphics Indigo 2 Extreme computer by
employing the teXsan 1.7 software.85 Full-matrix least-squares refine-
ment with anisotropic thermal displacement parameters for the Cu, Cl,
and F atoms yielded a final R of 0.057 (Rw ) 0.078). An inspection of
a difference Fourier map indicated the presence of three weak peaks
corresponding to a partially populated ethanol molecule. The non-
hydrogen atoms of this molecule were refined isotropically with the
occupancy of 0.5. The final difference map was essentially featureless
with the highest peak of 0.59 e/Å3.
Results and Discussion
Ligand Synthesis. Ligands 1-4 were prepared from the
diamines by condensation with aldehydes, followed by reduction
of the resulting imines to secondary amines, and subsequent
alkylation at the secondary nitrogen centers to give the desired
quadridentate chelates (Scheme 1). Ligands are designated in
(86) For other examples of the use of NaBH4 as the reductant to make
ligands of this type, see: Toftlund, H.; Pedersen, E.; Yde-Andersen,
S. Acta Chem. Scand. A 1984, 38, 693-7.
(87) Branca, M.; Checconi, P.; Pispisa, B. J. Chem. Soc., Dalton Trans.
1976, 481-8 and references therein.
(84) Altomare, A.; Burla, M. C.; Camalli, G.; Cascarano, G.; Giacovazzo,
C.; Guagliardi, A.; Polidori, G. J. Appl. Crystallogr. 1994, 27, 435.
(85) teXsan 1.7. Crystal Structure Analysis Package; Molecular Structure
Corp.: The Woodlands, TX, 1992.
(88) Gruenwedel, D. W. Inorg. Chem. 1968, 3, 495-501.
(89) For examples of alternate preparations of related ligands see refs 41,
59, and Arulsamy, N.; Glerup, J.; Hazell, A.; Hodgson, D. J.;
McKenzie, C.; Tofflund, H. Inorg. Chem. 1994, 33, 3023-5.

Metal Complexes with Cis R Topology
Inorganic Chemistry, Vol. 38, No. 24, 1999 5551
Table 1. UV/Vis Spectral Data (250-800 nm Range) for Ligands
products were typically redissolved in either CH2Cl2 or CH3-
CN and were precipitated using ethanol, diethyl ether, or, more
rarely, hexanes. When it proved difficult to obtain crystalline
materials, complexes were sometimes isolated as analytically
pure powders by rapid precipitation from similar solvent
combinations. Details pertaining to specific metal complexes
are provided in the Experimental Section. In general it was our
experience that complexes of Mn, Fe(II), Co, and Ni tended to
be easier to isolate as crystalline or powdery solids than
complexes of Zn, Cu, and Fe(III) throughout the ligand series.
This was especially true for Cu and Fe(II) complexes of the
more bulky quinoline bearing ligand, 4, which appeared to be
unstable. Certain iron complexes of related ligands are unstable
as well.59
and Complexes in CH3CN Solution
compd
en(Bn)py
en(F5Bn)py
S-pn(F5Bn)py
S-pn(F5Bn)qn
wavelength (nm)a
262 (6300), 270 sh, 300 sh
262 (8500), 270 sh, 300 sh
262 (7800), 270 sh, 300 sh
263 (9000), 303 (6000), 308 sh,
316 (7300)
Mn{en(Bn)py}Cl2
262 (8400), 269 sh
262 (3400), 270 sh, 320 sh
264 (7200), 270 sh
258 (10 000), 413 (1100)
260 (8400), 338 (1900), 402 (1200)
260 (7900), 341 (1600), 398 (1200)
256 (19 000), 265 sh, 300 sh, 376 (420)
258 (7100), 270 sh, 300 sh, 510 sh,
536 (24), 660 sh
Mn{en(F5Bn)py}Cl2
Mn{S-pn(F5Bn)py}Cl2
Fe{en(Bn)py}Cl2
Fe{en(F5Bn)py}Cl2
Fe{S-pn(F5Bn)py}Cl2
[Fe{en(F5Bn)py}Cl2]PF6
Co{en(Bn)py}Cl2
It should be noted that it was not possible to prepare the entire
series of metal complexes for each of the ligands. Often
analytical data for isolated materials did not correlate well with
predicted structures. Especially puzzling was the fact that though
many complexes of en(Bn)py, 1, yielded beautifully crystalline
materials, it was difficult to determine the compositions of these
samples based upon CHN analytical data. This could be due to
partial dissociation of halide ligands and the presence of
fractional amounts of solvent(s) associated with complexes in
the solid state, to solvento species, or to the presence of minor
impurities that proved difficult to remove. Though we were
unable to obtain complexes of satisfactory purity with certain
ligand and metal combinations after screening a variety of
different solvent systems and purification techniques, it may
still be possible to achieve pure and stable quadridentate
compounds if different metal salts (e.g. other halides, acetates)
and solvents are employed. Only those samples for which
elemental analyses correlate well with predicted products are
included in the Experimental Section. Since diamagnetic Zn-
(II) complexes provide valuable information about metal
Co{en(F5Bn)py}Cl2
Co{S-pn(F5Bn)py}Cl2
Co{S-pn(F5Bn)qn}Cl2
260 (6400), 270 sh, 300 sh, 510 sh,
540 (33), 640 sh
260 (8300), 270 sh, 300 sh, 536 (38),
660 sh
292 sh, 297 sh, 305 (7400), 317 (7600),
588 (210), 633 sh, 666 sh,
685 (240)
258 (9400), 270 sh, 398 sh, 638 (16)
258 (16 000), 270 sh, 400 sh, 625 (23)
262 (11 000), 290 (4500), 794 (97)
261 (20 000), 289 sh, 460 (75),
794 (260)
Ni{en(F5Bn)py}Cl2
Ni{S-pn(F5Bn)py}Cl2
Cu{en(F5Bn)py}Cl2
Cu{S-pn(F5Bn)py}Cl2
Zn{en(F5Bn)py}Cl2
Zn{S-pn(F5Bn)py}Cl2
264 (6000), 270 sh
264 (6300), 270 sh
a ꢀmax values in parentheses (M-1 cm-1); sh ) shoulder.
complexes of all the ligands. Structural differences among Zn-
(II) complexes correlate with the degree of steric bulk sur-
rounding the metal center. Our work with ligands 1-4 and
others of this type75 has revealed that all three features of the
ligandsthe diamine backbone, the pendant alkyl substituent,
and the apical donor groupsplay a role.
1
complex structure, H NMR spectral data in CDCl3 solution
are tabulated for all Zn complexes, regardless of analytical
purity. Likewise, 1H NMR spectral data for Co(III) complexes
generated in situ from Co(II) precursors are also provided. Metal
complexes were routinely characterized by conductivity mea-
surements, electronic absorption and IR spectroscopy, and cyclic
voltammetry.
Zinc complexes of ligands with ethylenediamine backbones
and pyridyl donor groups, en(Bn)py, 1, and en(F5Bn)py, 2, adopt
C2-symmetric structures in CDCl3 solution as evidenced by their
1
NMR spectra. The H and 13C spectra for [Zn{en(F5Bn)py}-
Cl2] in CDCl3 are shown in Figure 6. This, combined with the
fact that these complexes are nonconducting even in acetonitrile
solution, is consistent with octahedral, cis R structures in which
both chlorides are coordinated. The Zn(II) complex of the S-pn
backbone ligand, S-pn(F5Bn)py, 3, is also present as a major
diastereomer in CDCl3 solution. However, other species are also
evident in trace amounts (∼5%). Since 3 bears a methyl group
on the internal chelate ring backbone, the major isomer adopts
a cis R topology but the complex is not C2 symmetric. Most of
the resonances in the aromatic region of the 1H and 13C spectra
are relatively insensitive to the methyl group on the diamine
backbone. Differences are greater in benzylic and aliphatic
regions; here all proton and carbon signals for [Zn{S-pn(F5-
Bn)py}Cl2] are unique. Aside from expected differences arising
from the unsymmetrical S-pn backbone, Zn complexes of 1-3
give rise to very similar spectra. This suggests that they all
possess the same cis R topology for the sole or major isomers.
In addition to varying the diamine backbone, it was also of
interest to us to see how differences in steric demand at the
terminal donor positions of the quadridentate ligand influence
metal geometry. Thus, a Zn(II) complex of S-pn(F5Bn)qn, 4,
was prepared for comparison. The 1H NMR of the Zn complex
of S-pn(F5Bn)qn in chloroform solution suggests that one major
isomer (>95%) is formed. However, the spectrum is complex,
and all protons exhibit unique resonances. It is also interesting
The FT-IR spectra of the ligands and complexes all contain
major bands at approximately 1655, 1525, 1503, and 1124 cm-1.
Free ligands 1-3 also exhibit pyridyl ring vibrations at ∼1585-
1590 cm-1. These shift to higher energies upon complexation
(∼1601-1611 cm-1). By comparison, for the free quinoline
ligand, 4, an absorption appears at 1596 cm-1 and this shifts
very little upon complexation (e.g. Co complex: 1603 cm-1),
which may correlate with weaker binding to this bulkier ligand.
Complexes that precipitate with associated alcohol solvent
typically exhibit OH stretches in the ∼3400-3050 cm-1 range.
The Fe(III) complex of 2 shows a peak at 843 cm-1, charac-
teristic of the PF6- counterion. These observations are consistent
with what has been observed previously for related complexes.43
Additional details about the structure and properties of quadri-
dentate metal complexes are provided below and are organized
according to metal ion in reverse order from Zn to Mn.
Zinc Complexes. All Zn(II) complexes are white or pale
yellow in color with UV/vis spectra exhibiting bands in the UV
1
region attributable to ligand absorptions (Table 1). H and 13C
NMR spectra, combined with conductivity data, reveal consider-
able information about the structure of Zn(II) complexes of the
quadridentate ligand series. In chloroform solution, either single
or major diastereomers (>95%) were obtained for Zn(II)

5552 Inorganic Chemistry, Vol. 38, No. 24, 1999
Ng et al.
Figure 7. ORTEP drawing and labeling scheme for [Cu{en(F5Bn)-
py}Cl2]. Ellipsoids are at 30% probability. Hydrogens are omitted for
clarity.
ligands (qn vs py), may force a cis â topology in Zn complexes
of the S-pn(F5Bn)qn ligand.
Copper Complexes. Copper complexes of the quadridentate
ligands were typically isolated as green solids. Structural
variation throughout the series of ligands for Cu(II) complexes
is indicated by differences in color and conductivity in solution
and by geometry in the solid state. In acetonitrile solution,
complexes that are nonconducting or only slightly conducting
(i.e. chlorides are coordinated in the inner sphere) tend to be
green in color. Those in which chlorides are dissociated in a
portion of the material are often blue-green or teal in color,
whereas complexes from which the chloride ligands are removed
using silver salts (e.g. Ag(SbF6)) generally form blue solutions
in both polar (CH3CN, MeOH) and noncoordinating (CH2Cl2)
solvents.42,75,90-93 Likewise, complexes become more blue in
color as the electronic and steric features of the ligand are varied.
For example, acetonitrile solutions of Cu(II) complexes of en-
(Bn)py are green and nonconducting and those of the more
electron deficient en(F5Bn)py and S-pn(F5Bn)py ligands are teal
and slightly conducting, whereas a related ligand, R,R-cn(F5-
Bn)py with the bulkier 1R,2R-diaminocyclohexane backbone,75
is a 1:1 electrolyte and blue in color. These observations
correlate well with predicted trends in the energy of d-d bands
as the geometry at the Cu(II) center changes for a given ligand
set75 and may be compared to Cu complexes of related
ligands.43,94,95 Transitions are typically lower in energy (green)
for distorted octahedral structures whereas square planar ge-
ometries typically give rise to higher energy transitions (blue).
As is common for Cu(II) complexes, multiple Cu(II) species
are likely present in acetonitrile solution as evidenced by
complex UV/vis spectra and conductivities between values that
would be expected for 1:1 electrolytes and nonconducting
samples. This flexibility within the coordination sphere of Cu
complexes has previously been described as the “plasticity
effect”.96 Moreover, as was seen in the solid state for [Cu{en-
(F5Bn)py}Cl2], it is also possible that both chlorides are bound
Figure 6. 1H (CD3CN) and 13C (CDCl3) NMR spectra of Zn{en(F5-
Bn)py}Cl2‚MeOH.
1
to note that the H NMR spectrum of this qn complex shares
considerable overlap in the benzylic and aliphatic regions with
a minor isomer evident in the spectrum of [Zn{S-pn(F5Bn)py}-
Cl2]. Both complexes exhibit resonances at ∼9.7 ppm which,
in our experience, seems to correlate with six coordinate
structures. These observations, and comparison with data for
cis â Co(III) complexes and 13C spectra of analogous bimetallic
Zn complexes with cis R and cis â topologies,41,43,50,61-64
strongly point to a cis â assignment for the major isomer of
[Zn{S-pn(F5Bn)qn}Cl2].50 It should be noted, however, that both
a nonideal cis-R structure, as was suggested by Rieger et al.59
for a ferrous complex of a related but less bulky en(Me)qn
ligand, and a five coordinate structure with one donor group
dissociated, analogous to that seen for [Cu{en(F5Bn)py}Cl2] in
the solid state (Figure 7), are also consistent with the observed
conductivity and NMR data.
Unlike the Zn(II) complexes of the en backbone ligands, those
1
of S-pn ligands 3 and 4 give rise to more complex H NMR
spectra upon dissolution in CD3CN. New species emerge that
are characterized both by the appearance of resonances in the
8-9 ppm and 3.5-4.6 ppm regions and the diminution of the
downfield resonance at ∼9.7 ppm. Moreover, these complexes
conduct to a small extent (ΛM ∼30 Ω-1 mol-1 cm2) in CH3CN
solution. These observations suggest that a chloride ligand
dissociates and either a five coordinate structure or a chloro-
solvento species comprises a fraction of these samples. In
summary, changes as subtle as the addition of a methyl
substituent to the ethylenediamine backbone lead to observable
differences in product distributions for Zn complexes of this
family of quadridentate ligands. Whereas en-based ligands form
single isomers, the S-pn systems show greater structural diversity
in both polar and nonpolar solvents. This, combined with
additional steric bulk at the terminal donor positions of these
(90) Hathaway, B. J. J. Chem. Soc., Dalton Trans. 1972, 1196-99.
(91) Hathaway, B. J.; Billing, D. E. Coord. Chem. ReV. 1970, 5, 143-
207.
(92) ComprehensiVe Coordination Chemistry; Wilkinson, G., Ed.; Pergamon
Press: Oxford, U.K., 1987; Vol. 5; Chapter 53, pp 533-774.
(93) Also see: Amundsen, A. R.; Whelan, J.; Bosnich, B. J. Am. Chem.
Soc. 1977, 99, 6730-9.
(94) Nikles, D. E.; Powers, M. J.; Urbach, F. L. Inorg. Chem. 1983, 22,
3210-7.
(95) Sakurai, T.; Kimura, M.; Nakahara, A. Bull. Chem. Soc. Jpn. 1981,
54, 2976-8.

Metal Complexes with Cis R Topology
Inorganic Chemistry, Vol. 38, No. 24, 1999 5553
Table 2. Crystallographic Data for Cu{en(F5Bn)py}Cl2
Table 4. Comparison of UV/Vis/NIR Spectral Data (250-1100 nm
Range) for Nickel, Cobalt, and Iron Complexes of en(F5Bn)py and
S-pn(F5Bn)py Ligands in Methylene Chloride Solution
chem formula
fw
space group
a, Å
b, Å
c, Å
â, deg
V, Å3
Z
C29H23N4O0.5F10Cl2Cu
759.96
P21/a (No. 14)
15.509(6)
10.936(4)
17.491(5)
96.53(3)
2947(1)
4
compd
wavelength (nm)
Fe{en(F5Bn)py}Cl2
Fe{S-pn(F5Bn)py}Cl2
Co{en(F5Bn)py}Cl2
Co{S-pn(F5Bn)py}Cl2
Co{S-pn(F5Bn)qn}Cl2
336 (1740), 407 (1420)a
335 (604), 406 (527)a
518 sh, 542 (37), 635 sh, 1086 (5)
521 sh, 543 (37), 633 sh, 1084 (5)
300 sh, 307 (8400), 320 (8200),
540 sh, 601 (95), 613 (97),
635 (110), 669 (100), 809 sh
402 (38), 642 (20), 1078 (11)
401 (41), 646 (21), 1083 (8)
T, °C
λ, Å
-120
0.710 69
1.71
Ni{en(F5Bn)py}Cl2
Ni{S-pn(F5Bn)py}Cl2
F
calcd, g cm-3
µ(Mo KR), cm-1
10.17
0.057
0.078
R(Fo)a
a Long tail across visible region.
a
Rw
a R ) (∑|Fo| - |Fc|)/∑|Fo|; Rw ) [(∑w(|Fo| - |Fc|)2/∑w(Fo)2].
shoulder on a much stronger UV band in the spectra of all Ni
complexes. Energies of absorptions for the Ni(II) complex of
weaker field fluorinated ligands 2 and 3 appear at slightly lower
energies than those for [Ni{en(Bn)py}Cl2]. Satisfactory analyti-
cal data were not obtained for the green crystalline [Ni{en-
(Bn)py}Cl2] complex; however, spectra were recorded for
comparison. The following d-d transitions were observed: CH3-
CN, 394 sh, 629 nm; CH2Cl2, 393 sh, 631, 1062 nm. Though
the complex [Ni{en(Bn)py}Cl2] is nonconducting in acetonitrile
solution (ΛM ) 1.3 Ω-1 mol-1 cm2), Ni complexes of
fluorinated ligands, 2 and 3, are slightly conducting in aceto-
nitrile solutions (ΛM ) 22 and 33 Ω-1 mol-1 cm2, respectively).
This suggests that chloride ligands dissociate from the Ni center
in a portion of the sample in this polar solvent. A five coordinate
Ni(II) species or six coordinate solvento-chloro or aquo-chloro
species are all consistent with the observed conductivity data.
No dramatic changes are observed in acetonitrile spectra as
compared with those in methylene chloride that might be
attributed to these new species that are formed. UV/vis spectra
of Ni(II) complexes of related but stronger field secondary
diamine ligands such as en(H)py with a variety of different
ancillary ligands and counterions have been previously recorded
for solids and in a variety of different solvents.43 For these
complexes, many of which are likely to be cis R or even dimeric
structures, d-d bands shift toward higher energies as expected
(580 and 1000 nm for en(H)py vs 630 and 1080 nm for
complexes of 2 and 3).
Cobalt Complexes. Cobalt complexes of ligands 1-3 were
isolated as lavender microcrystalline solids. Electronic absorp-
tion spectra in acetonitrile solution reveal absorptions at
approximately 510, 538, and 650 nm indicative of six coordinate
high-spin d7 Co(II) centers for ligands 1-3 (Table 1).101
Additional bands are present in the near-IR region at ∼1085
nm (CH2Cl2) which are consistent with this assignment (Table
4). Conductivity data also support this assertion, as all complexes
are essentially nonconducting (ΛM < 10 Ω-1 mol-1 cm2 in CH3-
CN). By comparison, the Co(II) complex of the more bulky
S-pn(F5Bn)qn ligand, 4, is slightly more bluish in color in the
solid state. This is consistent with what has been seen previously
for Co(II) complex with a related sterically bulky quinoline
ligand.59 Higher intensity bands at 588 and 685 nm (ꢀ ) 210
and 240 L mol-1 cm-1, respectively) in the UV/vis spectrum
(CH3CN) of this quinoline complex along with a molar
conductivity, ΛM ) 43 Ω-1 mol-1 cm2, suggest that chloride
ligands are dissociated for a fraction of the sample. Most likely,
mixtures of six and five coordinate species are present in
Table 3. Selected Bond Lengths (Å) and Bond Angles (deg) for
Cu{En(F5Bn)py}Cl2
(a) Bond Lengths
Cu-Cl(1)
Cu-N(1)
Cu-N(3)
2.519(3)
2.018(8)
2.425(7)
Cu-Cl(2)
Cu-N(2)
Cu-N(4)
2.369(3)
2.267(8)
2.034(8)
(b) Bond Angles
Cl(1)-Cu-Cl(2)
Cl(1)-Cu-N(2)
Cl(2)-Cu-N(1)
N(1)-Cu-N(4)
104.5(1)
89.2(2)
95.6(2)
171.6(3)
Cl(1)-Cu-N(1)
Cl(1)-Cu-N(4)
Cl(2)-Cu-N(2)
N(2)-Cu-N(4)
89.4(2)
95.5(2)
164.8(2)
95.3(3)
to give nonconducting solutions but that a nitrogen donor group
may be dissociated to produce five coordinate structures overall.
The ORTEP diagram of the green Cu(II) complex, [Cu{en-
(F5Bn)py}Cl2], is shown in Figure 7. Crystallographic param-
eters are provided in Table 2, and selected bond lengths and
bond angles are given in Table 3. The especially long Cu-
N(3) distance of 2.425(7) Å suggests that a five coordinate
square pyramidal structure with an N3Cl2 donor set (an N(1)-
N(2)-N(4)-Cl(2) base and a Cl(1) apical group) best describes
the geometry of this compound in the solid state. Goodwin and
Lions also proposed dissociation of a chelate donor group from
Cu(II) centers to explain their observations with a related
quadridentate ligand.42 This structure may be contrasted with
[Cu{en(H)py}Cl]+, a five-coordinate Cu(II) complex with an
N4Cl donor set,97 and also with a six coordinate Cu(II) complex
of tn(H)py (tn ) trimethylenediamine), a 5-6-5 chelate ring
system with planar chelate coordination (i.e. N4Cl2 donor
set).98,99 These solid-state examples illustrate that minor varia-
tions in the ligand can result in considerable structural differ-
ences in Cu complex geometry.96 Though crystallographic
analysis of metal complexes of chiral backbone ligands could
provide valuable insights into the topology and stereospecificity
of these systems, it proved difficult for us to grow crystals
suitable for analysis using the S-pn backbone ligands, 3 and 4.
Nickel Complexes. Nickel complexes were isolated as light
green paramagnetic solids. All show very similar UV/vis spectra
in CH3CN and CH2Cl2 solutions with bands at ∼630 and ∼1080
nm typical for octahedral systems (Tables 1 and 4).100 It should
be noted that an additional band is present at ∼400 nm as a
(96) Gazo, J.; Bersuker, I. B.; Garaj, J.; Kabesova, M.; Kohout, J.;
Langfelderova, H.; Melnyk, M.; Serator, M.; Valach, F. Coord. Chem.
ReV. 1976, 19, 253-97.
(97) Bailey, N. A.; McKenzie, E. D.; Worthington, J. M. J. Chem. Soc.,
Dalton Trans. 1973, 1227-31.
(98) Pajunen, A.; Pajunen, S. Acta Crystallogr. 1986, C42, 53-6.
(99) McKenzie, E. D.; Stephens, F. S. Inorg. Chim. Acta 1980, 42, 1-10.
(100) ComprehensiVe Coordination Chemistry; Wilkinson, G., Ed.; Per-
gamon Press: New York, 1987; Vol. 5, Chapter 50, pp 86-89.
(101) Lever, A. P. B. Inorganic Electronic Spectroscopy, 2nd ed.;
Elsevier: New York, 1984; pp 480-505.
(102) Banci, L.; Bencini, A.; Benelli, C.; Gatteschi, D.; Zanchini, C. Struct.
Bonding 1982, 52, 77ff.

5554 Inorganic Chemistry, Vol. 38, No. 24, 1999
Ng et al.
solution.102 Weak features present in the UV/vis spectrum of
this complex in acetonitrile solution are more pronounced in
the spectrum run in methylene chloride solution (Tables 1 and
4).
Though some investigators have used this sensitivity to oxidation
to synthetic advantage, no such Fe complexes were isolated with
the quinoline ligand, 4. In contrast, the Fe(II) complexes of 1-3
showed no evidence of sensitivity to oxidation in solution or in
the solid state. This is consistent with the observations of Hazell
et al., who have noted that quadridentates with tertiary amine
donors help to stabilize the lower Fe(II) oxidation state whereas
ligands with stronger field secondary amines in the internal
donor positions are sometimes prone to oxidative dehydroge-
nation, thus forming purple products.105 For en(F5Bn)py, an Fe-
(III) complex was also readily prepared by reaction of the ligand
with FeCl3 followed by addition of NH4PF6 to precipitate the
product, [Fe{en(F5Bn)py}Cl2]PF6, as canary yellow plates. This
complex is a 1:1 electrolyte in acetonitrile solution, consistent
with an octahedral geometry at the metal center. The electronic
absorption spectrum of this complex is similar to that observed
for a related Fe(III) complex, [Fe(en(Me)py)Cl2]ClO4.52 Like-
wise, the cis â Fe(III) complex [Fe{tn(H)py}Cl2]ClO4 (tn )
trimethylenediamine) prepared by Busch and co-workers was
isolated as yellow prisms.106 The other ligands were not
amenable to Fe(III) complex synthesis; brown solids were often
obtained. Though no structural information was obtained for
these paramagnetic species, Fe(II) and Fe(III) complexes of
related quadridentate ligands with pyridyl or quinoline and 3°
donors all possess cis R structures in both monometallic and
bimetallic complexes.51,59,107 Complexes previously prepared by
others have found application as bioinorganic models51,52,107 and
catalysts,59,108 and they have been used in studies of spin
equilibria in diisothiocyanate complexes.86,87
Manganese Complexes. Reaction of MnCl2 with quadriden-
tate ligands 1-4 produced off-white paramagnetic solids in high
yield. White Mn(II) complexes of analogous ligands have been
described previously as intermediates in the synthesis of
bimetallic complexes53 including many mixed valence species
intended to model biological systems or for studying magnetic
interactions between bridging ligands.109-112 However, no
characterization of these intermediates was provided. Related
bimetallic Mn complexes have been utilized as oxidation
catalysts.108 As expected for high-spin d5 systems, electronic
absorption spectra of manganese complexes show ligand bands
but no absorptions attributable to d-d transitions. A magnetic
moment of 6.1 µB was determined for [Mn{en(Bn)py}Cl2] using
the Evans NMR method. This is somewhat higher than the spin
only value of 5.9 µB calculated for this d5 system.
Though Co(II) complexes are paramagnetic, their oxidation
to diamagnetic Co(III) systems allows for structural character-
ization by 1H NMR and for comparison of these quadridentate
ligand systems with analogous Co(III) systems.44-48 A variety
of different methods have been reported for the generation of
Co(III) complexes with these types of quadridentate chelates,
reaction with hydrogen peroxide in ethanol solution being the
most common of these.46 Unfortunately, our attempts to employ
this standard preparation with ligands 1-4 met with little
success. Reaction of CoCl2, en(Bn)py, and hydrogen peroxide
yielded a blue-green solid, as expected by analogy with related
systems; however, it did not precipitate readily from EtOH
solution and was isolated by concentration instead. Inspection
of products by 1H NMR revealed chemical shifts consistent with
what has been previously reported for analogous C2 symmetric
cis R systems46 though resonances were broad. Hydrogen
peroxide reactions with Co(II) chloride and the S-pn(F5Bn)qn
ligand, 4, also exhibited the expected color changes; a green
solid was obtained after partial concentration and precipitation
with additional EtOH. However, as with the en(Bn)py ligand,
1, only broad resonances were observed, indicative of the
presence of paramagnetic impurities. The Co(II) complex of
en(F5Bn)py, 2, turned brown upon reaction with H2O2, and
reaction of Co(II) with S-pn(F5Bn)py and H2O2 resulted in the
precipitation of the unreacted lavender Co(II) complex.
Lack of success with the standard methods prompted us to
explore other common oxidants.103 Reactions with Ag(OTf) did
not generate significant quantities of the desired Co(III)
products; however, reaction of Co(II) complexes with the
stronger oxidant, bromine, did yield diamagnetic Co(III)
products in all cases except for with the bulky ligand S-pn(F5-
1
Bn)qn, 4. The H NMR spectrum of the oxidation product of
[Co{en(F5Bn)py}Cl2] is provided as Supporting Information.
It clearly reveals that a C2-symmetric cis R isomer was obtained.
Cobalt complexes of other ligands also formed major isomers
under these reaction conditions; however, all spectra bear
evidence of at least one other minor product as well (<20% of
the total sample as determined by integration). It should be noted
that Br2 may not be an innocent oxidant in these reactions. The
bromide ion formed could exchange with the chloro compounds
to generate new mixed-halide complexes. Moreover, it has been
reported previously that dichlorocobalt(III) complexes with
analogous quadridentate ligands are highly susceptible to
hydrolysis; they rapidly form chloro-aquo species that are 2:1
electrolytes.46,104 These various factors complicate product
analysis and make it difficult to determine whether minor
isomers are ancillary ligand byproducts, topological isomeric
impurities, or some combination of the two.
Electrochemistry. Cyclic voltammetry is a useful technique
for assessing the purity of samples and for probing the electronic
features of metal complexes. Methylene chloride solutions of
the complexes and ligands were investigated using a glassy
carbon working electrode, an aqueous silver/silver chloride
reference electrode, and tetra-n-butylammonium hexafluoro-
phosphate (TBAH) as the supporting electrolyte. Data are
collected in Table 5. Free diamine ligands undergo oxidation
Iron Complexes. Iron(II) complexes of en(F5Bn)py, S-pn-
(F5Bn)py, and en(Bn)py were isolated as canary yellow solids.
Iron complexes of the bulky quinoline ligand, 4, were not
amenable to synthesis by our standard preparations. Solutions
turned orange and then dark brown upon reaction of the ligand
with ferrous chloride, suggesting ready oxidation to Fe(III)
products. Others have reported that the iron complex of a related
nonfluorinated ligand, namely en(Bn)qn, was also susceptible
to degradation in the presence of water and other solvents.59
(105) See compound 8 found in the Experimental Section of ref 52.
(106) Alcock, N. Acta Crystallogr., Sect. C 1997, C53, 1385-7.
(107) Arulsamy, N.; Goodson, P. A.; Hodgson, D. J.; Glerup, J.; Michelsen,
K. Inorg. Chim. Acta 1994, 216, 21-29.
(108) Tetard, D.; Verlhac, J. J. Mol. Catal. A: Chem. 1996, 113, 223-30.
(109) Goodson, P. A.; Glerup, J.; Hodgson, D. J.; Michelsen, K.; Weihe,
H. Inorg. Chem. 1991, 30, 4909-14 and references therein.
(110) Arulsamy, N.; Glerup, J.; Hazell, A.; Hodgson, D. J.; McKenzie, C.
J.; Toftlund, H. Inorg. Chem. 1994, 33, 3023-4.
(111) Manchanda, R.; Thorp, H. H.; Brudvig, G. W.; Crabtree, R. H. Inorg.
Chem. 1991, 30, 494-7.
(112) For a related monometallic Mn system with a N4O2 donor set see:
Neves, A.; Erthal, S. M. D.; Vencato, I.; Ceccato, A. S.; Mascarenhas,
Y. P.; Nascimento, O. R.; Horner, M.; Batista, A. A. Inorg. Chem.
1992, 31, 4749-55.
(103) Connelly, N. G.; Geiger, W. E. Chem. ReV. 1996, 96, 877.
(104) Also see: Henderson, R. A.; Tobe, M. L. Inorg. Chem. 1977, 16,
2576-83.

Metal Complexes with Cis R Topology
Table 5. Cyclic Voltammetric Dataa
Inorganic Chemistry, Vol. 38, No. 24, 1999 5555
compd
en(Bn)py
en(F5Bn)py
S-pn(F5Bn)py
S-pn(F5Bn)qn
Mn{en(Bn)py}Cl2
Mn{en(F5Bn)py}Cl2
Mn{S-pn(F5Bn)py}Cl2
Fe{en(Bn)py}Cl2
Fe{en(F5Bn)py}Cl2
Fe{en(Bn)py}Cl2
Fe{en(F5Bn)py}Cl2
Ep,a (V) Ep,c (V) E1/2 (V) ∆Epb (mV)
+1.05
+1.08
+1.28 -1.24
+1.06
+0.95 +0.86
+1.08 +1.00
+1.09 +1.00
+0.40 +0.32
+0.48 +0.41
+0.91
+1.04
+1.04
+0.36
+0.44
90
90
90
85
75
202
112
70
c
+0.18 -0.019 +0.082
+0.26 +0.15
c
+0.20
+0.42
+0.49
[Fe{en(F5Bn)py}Cl2]PF6 +0.46 +0.39
Fe{S-pn(F5Bn)py}Cl2
Co{en(Bn)py}Cl2
+0.54 +0.45
+0.77 +0.12d
+0.86 +0.26d
+0.89 +0.26d
95
Co{en(F5Bn)py}Cl2
Co{S-pn(F5Bn)py}Cl2
Co{S-pn(F5Bn)qn}Cl2
Ni{en(F5Bn)py}Cl2
Ni{S-pn(F5Bn)py}Cl2
Cu{en(F5Bn)py}Cl2
Cu{S-pn(F5Bn)py}Cl2
Zn{en(F5Bn)py} Cl2
Zn{S-pn(F5Bn)py}Cl2
featureless
+1.40 +1.34
+1.37
+1.41
-0.02
65
97
160
+1.46 +1.36
+0.06 -0.11
+1.25 -0.09e
featureless
+1.28
a Instrumentation: BAS model CV-27 electrochemical analyzer.
Working electrode: glassy carbon. Reference electrode: aqueous Ag/
AgCl. Concentration: 0.5 mM. Supporting electrolyte: 0.1 M TBAH
in CH2Cl2. Scan rate: 200 mV/s. E1/2 for the ferrocenium/ferrocene
couple is at +0.47 V vs Ag/AgCl under these experimental conditions.
b ∆Ep is defined here as Ep,a - Ep,c c As in (a) with the following
.
exceptions: Instrumentation: BAS model CV-50W electrochemical
analyzer. Reference electrode: Ag/AgCl in CH3CN. Scan rate: 100
mV/s. E1/2 for the ferrocenium/ferrocene couple is at +0.23 V vs Ag/
AgCl under these experimental conditions. d Product wave. e Additional
minor features present at +0.56 (anode) and +0.52 and +0.68 (cathode).
Figure 8. Comparison of the cyclic voltammograms for (a) [Fe{en-
(Bn)py}Cl2] and (b) [Fe{en(F5Bn)py}Cl2].
are more difficult to oxidize than analogous complexes of the
nonfluorinated en(Bn)py ligand, 1, is also important. The
expected trends are clearly evident in E1/2 values for Mn and
Fe complexes of ligands 1 and 2 that are identical except for
perfluorination of the phenyl rings in 2. For Mn, the fluorinated
ligand is harder to oxidize by ∼130 mV whereas ∆E1/2 ) 80
mV for the two Fe complexes in this series. Voltammograms
of [Fe{en(Bn)py}Cl2] and [Fe{en(F5Bn)py}Cl2] that were
recorded versus a nonaqueous Ag/AgCl reference electrode are
compared in Figure 8.
Steric features of the ligands also influence how easily metal
centers are oxidized. Few differences are observed between
complexes of en(F5Bn)py, 2, and S-pn(F5Bn)py, 3; however,
complexes of the bulky quinoline ligand may exhibit rather
different electrochemistry, as is dramatically demonstrated for
cobalt complexes. Complexes of the structurally similar ligands
1-3 undergo irreversible oxidations at 0.77-0.89 V; however,
the voltammogram recorded for the Co complex of 4 is
featureless throughout. The bulky quinoline ligand is less able
to accommodate the contracted Co(III) oxidation state, perhaps
due to unfavorable steric interactions between apical quinoline
groups and equatorial chloro ligands that are undoubtedly
necessary for stabilizing the octahedral geometry preferred by
Co(III). Thus, [Co{S-pn(F5Bn)qn}Cl2] resists oxidation, both
chemically, using common oxidants such as Br2 or H2O2, and
electrochemically.
at high positive potentials (1.05-1.28 V), and for the S-pn(F5-
Bn)py ligand, 3, a reduction is also observed. It was of interest
to us to determine which metal complexes exhibit reversible
oxidations, as this could be important for certain catalytic
reactions. For manganese, iron, and certain Cu complexes,
quasireversible waves were observed. Manganese(II-III) couples
appear at E1/2 ∼ 0.91-1.04 V, whereas oxidation of Fe(II) to
Fe(III) is more facile under similar experimental conditions (E1/2
∼ 0.36-0.49 V). While the CV of the Cu(II) complex of the
bulky S-pn(F5Bn)qn ligand, 4, contains a quasireversible oxida-
tion, other copper complexes are not as well behaved. A series
of weak oxidations and reductions are evident at different
potentials, including ones at -0.43 V (vs ferrocene) indicative
of free copper. This may be contrasted with the findings of
Urbach and co-workers on related copper complexes, albeit with
different solvent systems.94 For Ni(II) complexes, weak qua-
sireversible processes are evident at high potentials (∼1.4 V).
The Ni(III) oxidation state is typically highly unstable; often
Ni(II)/Ni(III) oxidations occur in the range of +1 to +1.5
V.113-116 As is expected for a d10 system, Zn complexes are
essentially featureless over the potential range under investiga-
tion.
For catalytic systems, it is desirable to be able to tune the
electron richness of the metal center. Thus, determining whether,
and to what extent, complexes of the fluorinated systems 2-4
Conclusion
These investigations further demonstrate how readily the steric
and electronic features of the amino-diimine family of ligands
and their complexes may be varied. This, combined with the
fact that appropriately designed ligands form C2-symmetric cis
R topologies and single (or major) diastereomers upon com-
plexation to metal ions should make them well suited for
(113) ComprehensiVe Coordination Chemistry; Wilkinson, G., Ed.; Per-
gamon Press: New York, 1987; Vol. 5, Chapter 50, p 287 ff.
(114) Freire, C.; de Castro, B. Polyhedron 1998, 17, 4227-35.
(115) Ward, M. S.; Shepherd, R. E. Inorg. Chim. Acta 1999, 286, 197-
206.
(116) Bhattacharyya, S.; Weakley, T. J. R.; Chaudhury, M. Inorg. Chem.
1999, 38, 633-8.

5556 Inorganic Chemistry, Vol. 38, No. 24, 1999
Ng et al.
applications in asymmetric catalysis. While metals such as Fe,
Co, and Mn seem to adopt similar geometries throughout most
of the ligand series, other metals such as Zn, Cu, and, to some
extent, Ni are far more sensitive to the steric and electronic
features of the ligand and can assume six, five, or even four
coordinate structures. These differences could be exploited in
reaction optimization. Moreover, varying the ancillary ligands
from halides to other species, or replacing some or both of them
with noncoordinating counterions, opens up even more pos-
sibilities for tailoring reactivity. Once complexes are proven
viable catalysts for specific transformations, considerably more
work is merited on these systems. Devising alternate routes to
complexes that were difficult to access by standard preparations
reported herein and further varying the steric and electronic
features of these ligands, as well as more detailed characteriza-
tion, are all worth pursuing. Recently we have utilized Cu
complexes of this family of ligands as catalysts for the controlled
polymerization of methyl methacrylate.117 We have also seen
moderate enantioselectivities using Cu complexes of chiral
quadridentate ligands as catalysts for the Mukaiyama aldol
reaction.75 Rieger has described their potential to serve as
catalysts for olefin polymerization.59 Yet these few examples
barely begin to explore the potential of cis R complexes as
catalysts. Numerous other reactions including oxidations, po-
lymerizations, and Lewis acid-catalyzed transformations are ripe
for exploration.
Acknowledgment. We gratefully acknowledge Dupont for
a Young Professor Grant, the Alfred P. Sloan Foundation for a
Research Fellowship, and the University of Virginia for support
for this research. Dr. Daniel Derringer, Janet Cho, Scott A.
Savage, and Laura E. Johnston are thanked for their various
contributions to this project. C.L.F. also expresses her gratitude
to Prof. B. Bosnich, who acquainted her with principles of
inorganic stereochemistry and other factors important in chiral
catalyst design.
1
Supporting Information Available: A H NMR spectrum (300
MHz, CDCl3) for [Co{en(F5Bn)py}Cl2]Br‚1/2CH3CH2OH‚1/2CH2Cl2 and
IR spectral data for ligands and complexes measured as Nujol mulls
(2 pages). An X-ray crystallographic file in CIF format for [Cu{en-
(F5Bn)py}Cl2]. This material is available free of charge via the Internet
at http://pubs.acs.org.
(117) Ng, C.; Johnson, R. M.; Samson, C. M. C.; Fraser, C. L. Unpublished
manuscript.
IC990475V