Formation and dissociation of tetrahedral nickel(II) and cobalt(II) complexes of 'dithioimidodiphosphato' ligands, measurement of kinetically determined stability constants, and kinetic investigations of the rapid formation of tetrahedral Bis(ligand)copper(II) complexes and their rates of reduction to trinucleair copper(I) species

Article abstract of DOI:10.1002/hlca.200590055

The stopped-flow technique was used to measure the rates of formation and dissociation of tetrahedral [ML₂] complexes (M²⁺ = Ni ²⁺ or Co²⁺) of four bidentate S₂-donor 'dithioimidodiphosphato' ligands L^- (HL = [R¹R ²P(=S)]NH[P(=S)R³R⁴], R¹ to R ⁴ = alkyl) at 25.0° in MeOH/H₂O 95:5 (v/v) solution and in the presence of either MOPS (= 3-(morpholin-4-yl)propane-1-sulfonic acid) or 2,6-lutidine (= 2,6-dimethylpyridine) buffers. The kinetically determined equilibrium formation constants for [ML]⁺ ions (M = Ni or Co) are 10^-5 K = 0.50 ± 0.01 or 1.64 ± 0.071 mol^-1 for L = L³ (R¹ = R² = Me(CH₂) ₂CH(Me), R³ = R⁴ = Me₂CH), 1.27 ± 0.02 or 7.93 ± 0.091 mol^-1 for L = L⁷ (R¹ to R⁴ = Me₂CHCH₂), 0.88 ± 0.04 or 3.84 ± 0.131 mol^-1 for L = L⁸ (R ¹ to R⁴ = Me₂CH), and in case of Ni ²⁺ 1.88 ± 0.041 mol^-1 for L = L⁶ (R ¹ = R³ = Bu, R² = R⁴ = ¹Bu) (see Table 3: for L³ and L⁶-L⁸, see Table 1). Whereas the tetrahedral Ni²⁺ complexes dissociate more slowly than the analogous Co²⁺ species, in all cases, the Co ²⁺ complexes are more stable than those of Ni²⁺ due to their larger formation rate constants (Table 3). Reactions of Cu²⁺ with eight ligands HL (R¹ to R⁴ = alkyl, alkoxy, aryl, and aryloxy) show that formation of intensely colored tetrahedral [Cu ¹¹L₂] species is too fast be measured with the available stopped-flow apparatus (t_1/2 < 2 ms), but the subsequent rates of reduction of [Cu¹¹L₂] to give trinuclear products [Cu ¹₃L₃] are measurable. An X-ray analysis establishes the structure of one of the [Cu₃L₃] complexes, where R¹ = R² = Me₂CHO and R³ = R⁴ = 2-(tert-butyl)phenyl (L = L⁵), and a multiwavelength stopped-flow kinetic experiment establishes the spectrum of a tetrahedral [Cu¹¹L₂] species prior to the reduction reactions. The redox reactions proceed at 25.0° with first-order rate constants in the range 0.285 s^-1 (R¹ to R⁴ = PhO; L = L ¹¹) to 2.58 · 10^-4 s^-1 (R¹ to R⁴ = Me₂CHCH₂; L = L⁷) (Table 4).

Full text of DOI:10.1002/hlca.200590055

782
Helvetica Chimica Acta ± Vol. 88 (2005)
Formation and Dissociation of Tetrahedral Nickel(II) and Cobalt(II)
Complexes of ꢀDithioimidodiphosphatoꢁ Ligands, Measurement of Kinetically
Determined Stability Constants, and Kinetic Investigations of the Rapid
Formation of Tetrahedral Bis(ligand)copper(II) Complexes and Their Rates
of Reduction to Trinuclear Copper(I) Species
by Peter Moore*, William Errington, and Satindra P. Sangha
Department of Chemistry, University of Warwick, Coventry, CV4 7AL, UK
(e-mail: p.moore@warwick.ac.uk)
Â
Dedicated to Professor Andre E. Merbach in honor of his retirement
The stopped-flow technique was used to measure the rates of formation and dissociation of tetrahedral
[ML2] complexes (M2‡ ˆ Ni2‡ or Co2‡) of four bidentate S2-donor ꢀdithioimidodiphosphatoꢁ ligands LÀ (HL ˆ
[R1R2P(ˆS)]NH[P(ˆS)R3R4], R1 to R4 ˆ alkyl) at 25.08 in MeOH/H2O 95 :5 (v/v) solution and in the presence
of either MOPS (ˆ 3-(morpholin-4-yl)propane-1-sulfonic acid) or 2,6-lutidine (ˆ2,6-dimethylpyridine) buffers.
‡
The kinetically determined equilibrium formation constants for [ML] ions (M ˆ Ni or Co) are 10À5 K ˆ 0.50 Æ
0.01 or 1.64 Æ 0.07 l molÀ1 for L ˆ L3 (R1 ˆ R2 ˆ Me(CH2)2CH(Me), R3 ˆ R4 ˆ Me2CH), 1.27 Æ 0.02 or 7.93 Æ
0.09 l molÀ1 for L ˆ L7 (R1 to R4 ˆ Me2CHCH2), 0.88 Æ 0.04 or 3.84 Æ 0.13 l molÀ1 for L ˆ L8 (R1 to R4 ˆ
Me2CH), and in case of Ni2‡ 1.88 Æ 0.04 l molÀ1 for L ˆ L6 (R1 ˆ R3 ˆ Bu, R2 ˆ R4 ˆ tBu) (see Table 3; for L3 and
L6 ± L8, see Table 1). Whereas the tetrahedral Ni2‡ complexes dissociate more slowly than the analogous Co2‡
species, in all cases, the Co2‡ complexes are more stable than those of Ni2‡ due to their larger formation rate
constants (Table 3). Reactions of Cu2‡ with eight ligands HL (R1 to R4 ˆ alkyl, alkoxy, aryl, and aryloxy) show
that formation of intensely colored tetrahedral [CuIIL2] species is too fast be measured with the available
stopped-flow apparatus (t1/2 < 2 ms), but the subsequent rates of reduction of [CuIIL2] to give trinuclear products
[CuI3L3] are measurable. An X-ray analysis establishes the structure of one of the [Cu3L3] complexes, where
R1 ˆ R2 ˆ Me2CHO and R3 ˆ R4 ˆ 2-(tert-butyl)phenyl (L ˆ L5), and a multiwavelength stopped-flow kinetic
experiment establishes the spectrum of a tetrahedral [CuIIL2] species prior to the reduction reactions. The redox
reactions proceed at 25.08 with first-order rate constants in the range 0.285 sÀ1 (R1 to R4 ˆ PhO; L ˆ L11) to 2.58 ´
10À4 sÀ1 (R1 to R4 ˆ Me2CHCH2; L ˆ L7) (Table 4).
Introduction. ± Ligands HL 1 of the type (S ˆ PR2)2NH (R ˆ alkyl, alkoxy, aryl,
aryloxy)1)are versatile chelating S2-donors with the ability to form inorganic (C-free)
six-membered M(SPNPS) chelate rings with both main-group and transition metal ions
M [1 ± 30]. A review of such systems has appeared [19]. Coordination with M2‡ metal
ions usually occurs via the two S-atoms with loss of the ligand NH proton, and this leads
to 2 and neutral bis(ligand)metal(II) complexes [ML2] with either a tetrahedral or
square-planar geometry around the metal ion [19]. With Ru2‡, 5-and 6c-oordinate
structures of the types [Ru(PPh3){(S ˆ PPh2)2N}2] and [Ru(NH3)(PPh3){(S ˆ
PPh2)2N}2] have been reported [26], whilst, with Bi3‡, a neutral [Bi{(S ˆ PPh2)2N}3]
complex with a distorted octahedral geometry has been established [11]. Five-
1
)
For convenience, the general name ꢀdithioimidodiphosphatoꢁ is also used for (S ˆ PR2)2NH with R alkoxy
or aryloxy.
ꢂ 2005 Verlag Helvetica Chimica Acta AG, Zürich

Helvetica Chimica Acta ± Vol. 88 (2005)
783
coordinate trigonal bipyramidal structures with In3‡ of the type [InCl{(S ˆ Pr2))2N}2]
i
(R ˆ Ph, Pr) are also known [22]. With Sm2‡, Y3‡, and other lanthanide ions, in
addition to S,S-binding, terdentate coordination involving both S-atoms and the
deprotonated N-atom has been found [25].
When coordinated through the two S-atoms, ꢀdithioimidodiphosphatoꢁ ligands L1)
are reminiscent of the well-known pentane-2,4-dionato chelators such as acetylaceto-
nato (acacÀ), but differ in forming nonplanar chelate rings with puckered PÀNÀP
groups. The other significant difference is the presence of four bulky R groups attached
to the P-atoms and flanking the donor S-atoms, compared with only two possible R
groups attached to the carbonyl C-atoms of 1,3-diketones like pentane-2,4-dione
(Hacac). The greater steric hindrance produced by the four bulky groups leads to the
ready formation of neutral tetrahedral bis(ligand)metal(II) complexes with 1, and such
ligands have been developed for the selective hydrometallurgical extraction of some
divalent metal ions. For example, a process for the selective removal of Zn2‡ from Fe3‡
(recovery of zinc from rusty galvanized steel) has been developed [31]. Theoretical
calculations suggest that ligands containing an a-branched alkyl chain at the P-atom are
good extractants for zinc(II), whereas those having a primary C-atom or an O-atom at
the P-atom are poor extractants [20].
Early structural studies concentrated largely on the coordination chemistry of the
tetraphenyl ligand [Ph2P(ˆS)]2NH and to a lesser extent of the other symmetric
tetraalkyl analogues (S ˆ PMe2)2NH and (S ˆ PiPr2)2NH (iPrˆ isopropyl), as well as of
some unsymmetric ligands [1][2][6][9][10][12][13]. A comparison with other
bidentate S2-donors such as 1,1-dithiolato ligands shows that they have a smaller S ´´´
´´´ S separation (ca. 3 Š) compared to (SPNPS)À ligands (ca. 4 Š), and this produces
more strain in the tighter chelate ring. Whereas the 1,1-dithiolato ligands have been
found to form square-planar bis(ligand)nickel(II) chelates, the analogous complexes
with [(S ˆ PR2)N(R2P ˆ S)]À ligands are more frequently found to be tetrahedral (e.g.,
R1 to R4 ˆ Ph, Me, iPr) [11][19][21][24]. This is believed to arise from the steric effects
of the bulky R2P groups which are in close proximity to the large S-donor atoms,
although interestingly, the less-symmetric complex [Ni{(S ˆ PPh2)N(Me2P ˆ S}2] has
been found to have a square-planar geometry [18].
In this study, we investigated ꢀdithioimidodiphosphatoꢁ1) ligands 1 (ˆ HL)
containing four bulky groups R1 to R4 that readily form neutral bis(ligand)metal(II)
complexes with a range of transition and main-group metal ions M (Eqn. 1) in the
presence of excess buffer base (B).
‡
M2‡ ‡ 2HL ‡ 2Bꢀbase† > ‰ML2Š ‡ 2BH
(1)

784
Helvetica Chimica Acta ± Vol. 88 (2005)
For the ligands studied here (Table 1), all of the bis(ligand)metal(II) complexes
have geometries close to tetrahedral at the metal center. X-Ray crystal structures of
[M(L1)2] and [M(L6)2] have been determined for M ˆ Co, Ni, Zn, and Hg, and found to
have distorted tetrahedral geometries; square-planar geometries were established for
the palladium(II) complexes [PdL2] (L ˆ L5 and L6). These crystal structures will be
reported elsewhere [24].
Table 1. Structures of the Ligands 1 (ˆ HL ˆ [R1R2P(ˆS)]NH[P(ˆS)R3R4]) Investigated
Ligand
R1
R2
R3
R4
HL1
HL2
HL3
HL4
HL5
HL6
HL7
HL8
HL9
HL10
HL11
EtO
EtO
cyclohexyl
Me(CH2)2CH(Me)
2,4-Me2C6H3O
iPrO
Bu
Me2CHCH2
iPr
Ph
Ph
Ph
cyclohexyl
iPr
Ph
cyclohexyl
iPr
cyclohexyl
Me(CH2)2CH(Me)
2,4-Me2C6H3O
iPrO
Ph
Ph
2-tBuC6H4O
2-tBuC6H4O
tBu
tBu
Bu
Me2CHCH2
Me2CHCH2
iPr
2-tBuC6H4O
Ph
Me2CHCH2
iPr
2-(MeCH2CH(Me))C6H4O
Ph
iPr
2-tBuC6H4O
Ph
PhO
PhO
PhO
PhO
Although there have been numerous synthetic and structural investigations of the
complexes formed by such ligands, to date there have been no reports of the kinetics of
formation and dissociation of such species. In this study, the stopped-flow technique
was used to measure rates of formation and dissociation of complexes formed by Ni2‡
and Co2‡ ions, and to estimate some equilibrium formation constants (K) for the
mono(ligand) complexes from the relationship K ˆ kf/kd where kf and kd are the rate
constants for formation and dissociation of the metal/ligand 1:1 complexes, respec-
tively.
Kinetic studies were also made of the reactions of Cu2‡ with these ligands. Here the
rates of formation of tetrahedral [Cu(ligand)2] were too fast to be measured at 258 with
the available stopped-flow technique used (t1/2 < 2 ms), but subsequent slower redox
‡
reactions to give trinuclear Cu complexes, [Cu3(ligand)3], were measurable, and these
rate constants are reported together with the UV/VIS spectrum of one of the rapidly
formed tetrahedral [Cu(ligand)2] complexes. The X-ray structure of [Cu3(L5)3] was
also determined and is in good agreement with that found for analogous copper(I)
complexes [13][23]. A distorted tetrahedral CuI4 cluster has recently been obtained
from reactions of [Cu(MeCN)4]X (X ˆ BF4 or PF6] with (E ˆ PPh2)NH (E ˆ S, Se)
[30]. Crystal structures of [M(Ph3P){(S ˆ PPh2)2N}] (M ˆ CuI, AgI, AuI) are also
reported [29].
Experimental. ± General. The ꢀdithioimidodiphosphatoꢁ ligands 1 were obtained from Zeneca Specialties,
with the exception of HL7 which was prepared by reacting sBu2P(ˆS)NH2 with sBu2P(ˆS)Br as detailed below.
Ligands HL3 and HL11 were recrystallized from CHCl3, and all other ligands from EtOH. All solvents and other
reagents were obtained from Aldrich. Purity of the ligands was checked by NMR and MS measurements. Kinetic
and spectroscopic measurements were made by using spectroscopic grade solvents. Metal solvates of formula
[M(solvent)6][BF4]2 (solvent ˆ MeOH, DMSO) were prepared by a published method [32], and solns. were
standardized by EDTA titration [33]. Kinetic measurements were made with an Applied Photophyics-SX18-

Helvetica Chimica Acta ± Vol. 88 (2005)
785
MV kinetic workstation, VIS and IR Spectra: Philips Unicam-PU-87000 and Perkin-Elmer Paragon-1000
spectrophotometers, resp.; in nm and cmÀ1, resp. NMR Spectra: Bruker ACP400 spectrometer; d in ppm rel. to
(
Me4Si (1H and 13C) and H3PO4 31P), J in Hz. Mass spectra: Kratos MS80 instrument; in m/z. Elemental
analyses: (C,H,N): Leeman-Labs CE440 instrument.
Bis(2-methylpropyl)phosphinothioic Bromide (sBu2P(ˆS)Br). Under anh. conditions, a soln. of P(ˆS)Cl3
(101.6 g, 0.60 mol) in dry Et2O (200 ml) was added dropwise with overhead stirring to an ice-cold soln. of
sBuMgCl (210.4 g, 1.8 mol) in Et2O (200 ml). After the addition, the mixture was allowed to warm to r.t. and
then heated under reflux for 2 h. After cooling, the mixture was quenched with aq. 2m HCl (1 l). The aq. layer
extracted with more Et2O (3 Â 200 ml) and the combined Et2O extract dried (MgSO4), and evaporated to leave
a pungent oily residue (sBu4P2S4). The oil was dissolved in CH2Cl2 (400 ml), the soln. cooled to 08, and a soln. of
Br2 (57.5 g, 0.36 mol) in CH2Cl2 (400 ml) added dropwise with stirring to give a pink-brown soln. After warming
to r.t., the solvent was evaporated and the orange liquid product distilled at 120 ± 1228/2.5 Torr): sBu2P(ˆS)Br.
31P{1H}-NMR: 120.5 (s).
P, P-Bis(2-methylpropyl)phosphinothioic Amide (sBu2P(ˆS)NH2). NH3 Gas was bubbled through a soln. of
sBu2P(ˆS)Br (9 g, 0.045 mol) in THF (100 ml) for 1 h. Distilled H2O was added to dissolve any NH4Br and
excess NH3. The aq. layer was extracted with more THF (3 Â 100 ml) and the combined THF extract dried
(MgSO4) and evaporated: sBu2P(ˆS)NH2. Colorless oil. 31P{1H}-NMR: 79.63 (t).
P, P-Bis(2-methylpropyl)-N-[bis(2-methylpropyl)phosphinothioyl]phosphinothioic Amide (HL7 ˆ (S ˆ
PsBu2)2NH). To an ice-cold slurry of NaH (4.5 g of a 60% dispersion in paraffin oil) in THF (200 ml), a soln.
of sBu2P(ˆS)NH2 (8 g, 58.8 mmol) in THF (30 ml) was added under stirring at 08. The mixture was allowed to
s
warm to r.t., then left for 30 min before cooling again to 08. A soln. of Bu2P(ˆS)Br (9 g, 0.045 mol) in THF
(30 ml) was added dropwise with stirring, and the mixture was then heated overnight under reflux. After
cooling, distilled H2O (50 ml) was added cautiously to destroy any excess NaH. Conc. HCl soln. was then added
dropwise until the mixture was acidic. The solvent was evaporated and the residue extracted with Et2O (50 ml).
The Et2O extract was washed with H2O (3 Â 30 ml), dried (MgSO4), and evaporated and the oily residue taken
up in dry hexane (5 ml) and left at 08 to produce white crystals of HL7 (3.73 g). IR: 3215 (n(NH)); 900, 959, 767
n(PNP)); 1296 (d(NH)). 1H-NMR (CDCl3): 1.02 (t, 3J(H,H) ˆ 7.33, 12 H); 1.25 (m, 2J(P,H) ˆ 7.32, 7.32,
3J(H,H) ˆ 1.36, 12 H); 1.39 (m, 4 H); 1.97 (m, 4 H); 2.25 (m, 4 H); 2.81 (t, 2J(P,H) ˆ 7.34, 1 H). 13C-NMR
(CDCl3): 12.45 (m, 4 C); 13.34 (s, 4 C); 23.47 (m, 4 C); 39.93 (m, 4 C). 31P{1H}-NMR (CDCl3): 89.70 (m,
2
‡
2J(P,H) ˆ 12.39, J(P,P) ˆ 19.28). FAB-MS (pos.): 370 (M ).
Tris{P, P-bis[2-(1,1-dimethylethyl)phenyl]-P',P'-bis(1-methylethyl)dithioimidodiphosphato}tricopper(I)
([CuI3(L5)3]). A soln. of [Cu(H2O)6][BF4]2 (21 mg, 0.0896 mmol) in EtOH (5 ml) was added to a soln. of HL5
(0.10 g, 0.179 mmol) in EtOH (5 ml) to give a deep purple soln. This deep purple soln. slowly decolorized to give
a pale pink soln. from which X-ray-quality crystals precipitated after a few hours and were collected by filtration.
‡
Yield 88 %. 31P{1H}-NMR (CDCl3): 37.2 (d); 33.8 (d); (2J(P,P) 45.8. FAB-MS (pos.): 1867 ([Cu3(L5)3] ). Anal.
calc. for C26H41CuNO4P2S2: C 50.3, H 6.65, N 2.25; found: C 50.3, H 6.68, N 2.43.
The X-ray structure of [CuI3(L5)3] is shown in Fig. 1, and selected bond lengths and angles are collected in
Table 2.
X-Ray Crystallography: Data Collection and Processing. The crystal of [CuI3(L5)3] was held at 180(2) K
with an Oxford-Cryosystem-Cryostream cooler. Data were collected with a Siemens Smart-CCD-area-detector
diffractometer. A full hemisphere of reciprocal space was scanned by a combination of three sets of exposures;
each set had a different f angle for the crystal, and each exposure of 10 s covered 0.38 in w. The crystal-to-
detector distance was 5.01 cm. Crystal decay was monitored by repeating the initial frames at the end of the data
collection and analyzing the duplicate reflection; a multi-scan absorption correction was applied by using
SADABS [34]. The structures were solved by direct methods with SHELXTL-PC [35] and refined by full
matrix least-squares on F2 for all data with SHELXL-97 [36]. H-atoms were added at calculated positions and
refined using a riding model. Anisotropic temperature factors were used for all non-H-atoms; H-atoms were
given isotropic temperature factors equal to 1.2 (or 1.5 for Me H-atoms) times the equivalent isotropic
displacement parameter of the atom to which the H-atom is attached. Crystallographic data have been
deposited with the Cambridge Crystallographic Data Centre as supplementary publication no. CCDC-2541862).
2
)
Copies of the data can be obtained free of charge on application to CCDC, 12 Union Road, Cambridge
CB2 1EZ, UK (fax: ‡ 44-1223/336-033; e-mail: deposit@ccdc.cam.ac.uk).

786
Helvetica Chimica Acta ± Vol. 88 (2005)
Fig. 1. a) Structure of [CuI3(L5)3] showing the atomic numbering (H-atoms omitted for clarity). b) Structure of
[CuI3(L5)3] showing the atomic numbering (H-, O-, and C-atoms omitted)

Helvetica Chimica Acta ± Vol. 88 (2005)
787
Table 2. Selected Bond Lengths and Angles in [CuI3(L5)3]a). Arbitrary numbering.
Bond lengths [Š]
Bond angle [8]
Cu(1)ÀS(2)
Cu(1)ÀS(1)
Cu(1)ÀS(1)#1
Cu(1)ÀCu(1)#1
Cu(1)ÀCu(1)#2
S(1)ÀP(1)
2.216(2)
2.264(2)
2.272(2)
2.9785(16)
2.9785(16)
2.018(3)
2.272(2)
1.975(2)
1.505(8)
1.549(6)
1.553(6)
1.578(4)
1.582(6)
1.593(4)
S(2)ÀCu(1)ÀS(1)
128.92(8)
122.81(8)
108.16(11)
115.98(6)
98.91(6)
48.83(6)
119.46(5)
49.08(6)
98.71(6)
60.0
S(2)ÀCu(1)ÀS(1)#1
S(1)ÀCu(1)ÀS(1)#1
S(2)ÀCu(1)ÀCu(1)#1
S(1)ÀCu(1)ÀCu(1)#1
S(1)#1ÀCu(1)ÀCu(1)#1
S(2)ÀCu(1)ÀCu(1)#2
S(1)ÀCu(1)ÀCu(1)#2
S(1)#1ÀCu(1)ÀCu(1)#2
Cu(1)#1ÀCu(1)ÀCu(1)#2
P(1)ÀS(1)ÀCu(1)
S(1)ÀCu(1)#2
S(2)ÀP(2)
P(1)ÀO(11)
P(1)ÀN(1)
P(1)ÀO(12)
P(2)ÀO(22)
P(2)ÀN(1)#1
P(2)ÀO(21)
105.30(10)
102.44(10)
82.09(7)
103.71(9)
P(1)ÀS(1)ÀCu(1)#2
Cu(1)ÀS(1)ÀCu(1)#2
P(2)ÀS(2)ÀCu(1)
a #1 means A; #2 means B.
Results. ± Studies of the Formation of Nickel(II) and Cobalt(II) Complexes. To
drive the equilibrium of Eqn. 1 to the right, an excess of ligand (HL) and an excess of
buffer base (B) was used. Studies were made either in 100% MeOH in the presence of
2,6-dimethylpyridine (2,6-lutidine) as buffer, or in MeOH/H2O 95 :5 (v/v; for solubility
reasons) in the presence of 3-(morpholin-4-yl)propane-1-sulfonic acid (MOPS) as
buffer. These two buffers were chosen since they do not form complexes with the metal
ions, and because they have pKa values around 6 ± 7 where the ꢀdithioimidodiphos-
phatesꢁ are not deprotonated before coordination to the metal ions. Jobꢁs method of
continuous variations [37] confirmed the formation of a ligand/metal 2 :1 complex, and
the molar extinction coefficients of both the mono-and bis(ligand)nickel(II)
complexes (e > 150 l molÀ1 cmÀ1) were consistent with the formation of tetrahedral
rather than octahedral complexes. Typical spectra of the d ± d bands obtained in MeOH
for the ligand/nickel(II) 1:1 and 2 :1 complexes recorded for ligand HL3 in the
presence of either excess Ni2‡ or excess HL3, respectively, are shown in Fig. 2. The
spectra are as expected if both the mono-and bisl-igand species have a tetrahedral
geometry, based on the relatively large molar extinction coefficients compared with
those for octahedral complexes that normally have e values for the d ± d bands in the
range of 10 ± 20 l molÀ1 cmÀ1 [38]. The d ± d bands are at longer wavelength for
‡
[Ni(L3)(MeOH)2] compared with [Ni(L3)2] as expected for the greater ligand-field
strength of the latter.
For reactions with Ni2‡, initial kinetic studies were investigated in MeOH solution
in the presence of 2,6-lutidine buffer. Stopped-flow experiments were carried out by
using excess ligand (5 ´ 10À4 m) over [Ni2‡] (5 ´ 10À5 m), and with [2,6-lutidine] (0.020m)
in large excess. The ionic strength was controlled by addition of [Bu4N][BF4] (0.010m).
Reactions were monitored at 25.08 and 380 nm and revealed a single exponential
kinetic trace. The observed pseudo-first-order rate constants were interpreted as that
associated with the formation of the tetrahedral mono(ligand)nickel(II) complexes,
‡
‡
[NiL(MeOH)2] , with rapid conversion of [NiL(MeOH)2] to [NiL2] (Eqns. 2 and 3).

788
Helvetica Chimica Acta ± Vol. 88 (2005)
‡
Fig. 2. VIS Spectra for the d ± d bands of tetrahedral [Ni(L3)(MeOH)2] (lower trace) and [Ni(L3)2] (upper
trace) in MeOH containing 2,6-lutidine buffer (0.1m). Upper trace: [HL3] ˆ 5 ´ 10À3
m
and
[[Ni(MeOH)6][BF4]2] ˆ 5 ´ 10À5 m; lower trace: [HL3] ˆ 5 ´ 10À4 m and [[Ni(MeOH)6][BF4]2] ˆ 5 ´ 10À2 m, with
allowance made for the absorbance of excess Ni2‡
‡
‡
[Ni(MeOH)6]2‡ ‡ HL ‡ B (buffer) ! [NiL(MeOH)2] ‡ 4 MeOH ‡ BH
(2)
(3)
‡
‡
[NiL(MeOH)2] ‡ HL ‡ B ! [NiL2] ‡ 2 MeOH ‡ BH (rapid)
The pseudo-first-order rate constants (kobs) were found to increase with the
concentration of the 2,6-lutidine buffer [B] as shown in Fig. 3 for the reactions with
ligand HL3. The data plotted in Fig. 3 were fitted to Eqn. 4 by using nonlinear least-
squares analysis to give the continuous curve shown in Fig. 3.
a ‡ b‰BŠ
kobs
ˆ
(4)
1 ‡ K‰BŠ
K is the equilibrium constant for the rapid H-bonding of HL with B (K ˆ [B ´´´ HL]/
[B][HL]), and a and b are constants. Scheme 1 fits the observed variation of kobs
with [B].
Scheme 1

Helvetica Chimica Acta ± Vol. 88 (2005)
789
Fig. 3. Plot of the pseudo-first-order rate constants (kobs) vs. 2,6-lutidine buffer concentration at 258 for the
reaction of Ni2‡ (5 ´ 10À5 m) with HL3 (10À3 m) at ionic strength controlled by [Bu4N][BF4] (0.1m). The
continuous curve was calculated by using a nonlinear least-squares fit to Eqn. 4
ka and kb are the rate constants for reactions of Ni2‡ with non-H-bonded and H-
bonded buffer± ligand species, respectively. These two reactions, that each involve
‡
formation of [Ni(L)] , are the rate-determining steps for the overall reactions to give
[Ni(L)2]. The rate law associated with Scheme 1 is outlined in Scheme 2.
Scheme 2
d‰NiL2Š
dt
1 d
2 dt
ˆ ka‰Ni2‡Š‰HLŠ ‡ kb‰Ni2‡Š‰B Á Á Á HLŠ ˆ À
ꢀ‰HLŠ ‡ ‰B Á Á Á HLŠ†
hence;
1 d
2 dt
‰Ni2‡Š‰HLŠfka ‡ kbK‰BŠg ˆ À
‰HLŠ ꢀ1 ‡ K‰BŠ†
and in the presence of excess HL and B,
d‰HLŠ
2‰Ni2‡Š‰HLŠꢀka ‡ kbK‰BŠ†
À
ˆ
ˆ kobs‰Ni2‡
Š
dt
1 ‡ K‰BŠ
where;
2‰HLŠꢀka ‡ kbK‰BŠ†
kobs
ˆ
1 ‡ K‰BŠ
in agreement with the observed Eqn. 4 where
a ˆ 2ka‰HLŠ; b ˆ 2kbK‰HLŠ and b=a ˆ kbK=ka
With 2,6-lutidine as buffer, it is possible to determine an approximate equilibrium
constant for the buffer± ligand interaction in MeOH (for L3, K ˆ (42.5 Æ 8.5) l molÀ1],
but the rate constants determined this way were too inaccurate. Therefore, an

790
Helvetica Chimica Acta ± Vol. 88 (2005)
alternative buffer 3-(morpholin-4-yl)propane-1-sulfonic acid (MOPS) was selected to
allow a more-accurate estimation of the formation rate constants. MOPS has poorer H-
bonding properties compared with 2,6-lutidine, and at the concentration used does not
affect the observed rate constants.
Kinetic Studies of the Complex Formations in the Presence of the Buffer MOPS.
MOPS has a pKa of 7.2 in H2O but is not soluble enough in MeOH; thus investigations
were carried out in MeOH/H2O 95 :5 (v/v). Spectrophotometric titrations showed that
the equilibrium of Eqn. 1 is driven to the right in the presence of excess [MOPS]
(0.02m). Kinetic studies were carried out at 25.08 under pseudo-first-order conditions
by using the stopped-flow method with [ligand] in large excess ((1 ± 15) ´ 10À3 m) over
[metal ion] (5 ´ 10À5 m). Ligand solutions were freshly prepared prior to each set of
measurements to avoid any possible ligand decomposition. Equimolar amounts of
MOPS were added to the metal ion and the ligand solutions before mixing, and rate
data were collected at 380 nm. A single pseudo-first-order rate process was observed,
and plots of the observed pseudo-first order rate constants, kobs/sÀ1 vs. [ligand] were
linear with kobs close to zero at zero [ligand]. A typical plot for the reaction of Ni2‡ with
HL3 is shown in Fig. 4, and from such plots, the second-order rate constants for complex
formation were obtained from the slopes of the lines by using weighted linear least-
squares analysis (Table 3). For these calculations, the observed rate constants (kobs
)
were weighted by the inverse square of their standard deviations. Values of kobs were
the average of at least 3 measurements at each ligand concentration. Studies of the rates
of the formation reactions with ligands HL3, HL6, HL7, HL8 (all carrying aliphatic
substituents) were all possible in MeOH/H2O 95 :5 by the stopped-flow method, but
studies of all of the ligands substituted with alkoxy, aryl, and aryloxy groups were
outside the range of this technique. For the latter, this is attributed to the smaller
Fig. 4. Plot of the pseudo-first-order rate constants (kobs) at 258 vs. [HL3] for the reaction with Ni2‡ (5 ´ 10À5 m) in
MeOH/H2O 95 :5 (v/v) containing MOPS buffer (2 ´ 10À2 m)

Helvetica Chimica Acta ± Vol. 88 (2005)
791
Table 3. Comparison of Rate Constants at 25.08 for the Formation (kf) and Dissociation (kd) of Tetrahedral
Mono(ligand)metal(II) Complexes of Ni2‡ and Co2‡ with Four ꢀDithioimidodiphosphateꢁ Ligands in MeOH/
H2O 95 : 5 (v/v), and the Kinetically Determined Equilibrium Formation Constants, K ˆ kf/kd
Ligand
Ni2‡
Co2‡
kf [l molÀ1 sÀ1
]
104 kd [sÀ1
]
10À5K [l molÀ1
]
kf [l molÀ1 sÀ1
]
104 kd [sÀ1
]
10À5K [l molÀ1
1.64 Æ 0.07
]
HL3
HL6
HL7
HL8
38.1 Æ 1.0
119.4 Æ 2.7
55.3 Æ 0.6
43.5 Æ 2.0
7.55 Æ 0.01
6.34 Æ 0.05
4.35 Æ 0.06
4.92 Æ 0.02
0.50 Æ 0.01
1.88 Æ 0.04
1.27 Æ 0.02
0.88 Æ 0.04
731 Æ 33
3113 Æ 41
1023 Æ 37
867 Æ 29
44.67 Æ 0.01
12.90 Æ 0.01
22.55 Æ 0.09
7.93 Æ 0.09
3.84 Æ 0.13
equilibrium formation constants, resulting in large rate constants for the back
(dissociation) reactions.
The ligands HL2, HL10, and HL11 are too insoluble for studies in MeOH/H2O, so
studies were attempted in CHCl3/DMSO 95 :5 (v/v). The DMSO was used to allow
sufficient solubility of the [Ni(DMSO)6][BF4]2 salt used. The buffer MOPS does not
dissolve in this solvent mixture, so the sterically hindered base 1,4-diazabicy-
clo[2.2.2]octane (DABCO) was chosen instead. For HL2, the reaction rates in
CHCl3/DMSO were measured at 258 in the presence of excess ligand (2.5 ± 12.5 ´ 10À4 m)
over [Ni2‡] (5 ´ 10À5 m) in the presence of DABCO (3 ´ 10À3 m) buffer and [Bu4N][BF4]
(10À2 m) for ionic strength control. For HL2 that carries cyclohexyl substituents, the
second-order rate constant in CHCl3/DMSO 95 :5 (v/v) was found to be 4391 Æ 265 l
molÀ1 sÀ1 at 258. This was obtained from the slope of the plot of kobs vs. [HL3]. Attempts
to study HL10 and HL11 that carry aromatic substituents were again unsuccessful due to
too rapid conversions for measurements by the available stopped-flow method.
Reactions of [Co(MeOH)6][BF4]2 with ligands HL3, HL6, HL7, and HL8 were
studied in MeOH/H2O 95 :5 (v/v) at 365 nm under the same conditions as those used
for the nickel(II) measurements. As expected, the reactions were more rapid than for
nickel(II), in line with the greater solvent-exchange rate for cobalt(II). The rate
constants are collected in Table 3.
Kinetic Studies of the Complex Dissociation Reactions. Reversal of the equilibrium
of Eqn. 1 is possible by addition of an excess of 2,6-lutidinium tetrafluoroborate.
Solutions of largely the mono(ligand)metal(II) complexes were prepared by mixing
excess M2‡ ion (0.020m; M ˆ Ni or Co) with ligand (10À3 m) in MeOH/H2O 95 :5 (v/v)
containing 2,6-lutidine (10À2 m) and [Bu4N][BF4] (0.1m). A second series of solutions
of 2,6-lutidinium tetrafluoroborate (0.01 ± 0.10m) were prepared in the same solvent
mixture, and the ionic strength was maintained at 0.10 by addition of the requisite
amount of [Bu4N][BF4]. The dissociation reactions were monitored by mixing the
complex and 2,6-lutidinium solutions at 258 and following the absorbance decreases at
either 380 or 365 nm for the nickel(II) and cobalt(II) complexes, respectively. A single
first-order reaction was observed in each case, and the observed rate constants were
found to increase only slightly as the concentration of the 2,6-lutidinium ion was
increased. The dissociation rate constants in Table 3 were obtained from plots of kobs vs.
[2,6-lutidinium ion] extrapolated to [2,6-lutidinium ion] ˆ 0.
Kinetic, Equilibrium, and Structural Studies of the Reactions of ꢀDithioimidodi-
phosphatesꢁ with Cu2‡ Ion. Previous studies of the reactions of ꢀdithioimidodiphos-

792
Helvetica Chimica Acta ± Vol. 88 (2005)
phatesꢁ with copper(II) have been shown to involve rapid formation of intensely
colored bis(ligand)copper(II) complexes followed by reduction of these intermediates
to colorless copper(I) complexes of varying stoichiometry. In the presence of
triphenylphosphine, a mononuclear complex [Cu(Ph3P){(S ˆ PPh2)2N}] was isolated
and characterized by means of an X-ray crystal structure [15]. An early attempt to
prepare a copper(II) complex led to the isolation of a copper(I) product [3], but a later
study involving the reaction of K[(S ˆ PPh2)2N] with CuCl2 at À 788 led to the isolation
of the dark blue copper(II) complex [Cu{(S ˆ PPh2)2N}2]. At 258 in CH2Cl2 solution,
this blue product was reported to be reduced to a CuI product with a first-order rate
constant of ca. 2 ´ 10À3 minÀ1 [7]. In further studies of this reaction, the CuI product was
isolated and characterized by a crystal structure as [Cu4L3][CuCl2] ´ CCl4 (HL ˆ (S ˆ
PPh2)2NH) that dissociated to [CuI3L3] plus the oxygenated ligand
(HOPPh2)N(PPh2 ˆ S) and elemental sulfur [8]. Involvement of moisture in these
reactions was indicated by our study of analogous reactions of Cu2‡ with HL5 in EtOH
and MeCN. At room temperature, [Cu(H2O)6][BF4]2 and 2 mol-equiv. of HL5 in EtOH
‡
immediately formed a deep purple solution which was reduced to a Cu product that
precipitated as white [CuI3(L5)3]. This was characterized by 1H-, 13C-, and 31P-NMR, an
elemental analysis, a FAB mass spectrum, and an X-ray crystal structure (Fig. 1). When
[Cu(MeCN)6][BF4]2 was treated with HL5 under dry Ar in dry MeCN, with either a
Cu(II)/HL5 1:1 or 1:2 stoichiometry, the purple solution was again obtained which
persisted for several hours. However, upon exposure of this purple solution to moisture
from the air, a solid precipitate of [CuI3(L5)3] was obtained which was examined by 31P-
NMR. Besides the complex, it contained mainly free ligand; but also a new resonance
appeared at d 1.31, in line with some formation of the oxygenated ligand
[(Me2CHO)2P ˆ O]NH[(2-tBuC6H4O)2P ˆ S] (HL'). A yellow residue of sulfur was
also found. By comparison of these observations with those of related studies [8], the
overall reaction of Eqn. 5 was deduced to have occurred.
‡
6 Cu2‡ ‡ 9 HL5 ‡ 3 H2O ! 2[Cu3(L5)3] ‡ 3 S ‡ 3 HL' ‡ 12 H
(5)
Stopped-flow kinetic studies were carried out with unbuffered solutions. When HL3
(5 ´ 10À4 m) was reacted with [Cu(MeOH)6][BF4]2 (5´ 10À5 m) [26] at 258 in MeOH, the
last stages of the very rapid formation of highly colored [Cu(L3)2] was just detected at
380 nm. This reaction was followed by a much slower fall in absorbance as the reduction
to colorless [Cu3(L3)3] occurs. The rate constant for the formation of the bis(ligand)-
copper(II) intermediate could not be measured reliably, but the spectrum of [Cu(L3)2]
was obtained 2 ms after mixing by means of a variable-wavelength stopped-flow
experiment, conducted over the wavelength range 320 ± 650 nm (Fig. 5).
Two intense absorption bands were found at lmax (e/l molÀ1 cmÀ1) 580 (4990) and 390
(4480) nm. These two bands are similar to those recorded for the tetrahedral [Cu{(S ˆ
PPh2)2N}2] in CH2Cl2, i.e., lmax at 575 (3610) and 407 (4100) nm. This latter complex was
isolated at À 788, and an approximate spectrum obtained immediately after dissolving
in CH2Cl2 [7]. Further stopped-flow studies of the reactions of [Cu(MeOH)6]2‡ (5 ´
10À5 m) with variable amounts of HL3 (5 to 20 ´ 10À5 m) showed that the maximum
absorbance for [Cu(L3)2], recorded by the stopped-flow technique 2 ms after mixing,

Helvetica Chimica Acta ± Vol. 88 (2005)
793
Fig. 5. UV/VIS Spectrum of [Cu(L3)2] Recorded between 340 and 650 nm by means of a stopped-flow
multiwavelength experiment, with absorbances measured 2 ms after reacting [Cu(MeOH)6][BF4] (5 ´ 10À5 m)
with excess HL3 (5 ´ 10À4 m) in MeOH at 258
occurred when the metal/ligand ratio was 1:2, thus confirming the formation of the
bis(ligand)copper(II) species.
The rate constants for the reduction of [CuL2] species to give [Cux L3] (L ˆ L11, L1,
3
L4, L5, L9, L3, L8, and L7 in order of decreasing rate) were measured keeping ligand/
copper(II) ꢁ 10 :1. A single first-order rate process was observed in each case, and the
rate constants at 258 are collected in Table 4. The rates did not vary as the excess ligand
concentration was increased by a factor of four.
Table 4. Rate Constants at 258 for the Formation of [CuI3L3] from [CuIIL2]. Standard deviations in parentheses.
L
L11
L1
L4
L5
L9
L3
L8
L7
105 k/sÀ1
2849(7)
1110(4)
154(4)
21.0(0.1)
12.9(0.1)
5.00(0.03)
4.52(0.04)
2.58(0.03)
Discussion. ± Of the ꢀdithioimidodiphosphateꢁ ligands investigated (Table 1), only
those with electron-donating alkyl substituents, as opposed to the electron-withdrawing
alkoxy, aryl, or aryloxy substituents, on the P-atom react slowly enough for measure-
ments of rates of formation and dissociation of their metal complexes with Ni2‡ and
Co2‡ ions by means of the available stopped-flow technique (t1/2 > 2 ms). Reactions of
Ni2‡ and Co2‡ ions with HL3, HL6, HL7, and HL8 were measured and found to involve
‡
formation of tetrahedral [ML(solvent)2] ions (M ˆ Co, Ni) in the rate-determining
‡
step, followed by rapid conversion of [ML(solvent)2] to the tetrahedral [ML2]
products. This behavior is understandable since rates of solvent exchange usually
control the rates of complex formation (the Eigen ± Wilkins mechanism [39]), and
solvent exchange with octahedral [M(solvent)6]2‡ ions are much slower than with
‡
‡
tetrahedral [ML(solvent)2] ions [39]. The tetrahedral geometries of [ML(solvent)2]

794
Helvetica Chimica Acta ± Vol. 88 (2005)
and [ML2] complexes was established from their VIS spectra (e.g., Fig. 2), and from X-
ray structures of some of the [ML2] complexes [24].
From the data in Table 3, it can be seen that the rates of complex formation for Ni2‡
are more than an order of magnitude slower than for Co2‡, which is in line with the
relative rates of solvent exchange for the [M(solvent)6]2‡ ions [40]. Rates of
‡
dissociation of the [NiL(solvent)2] ions are also slower than those of corresponding
‡
[CoL(solvent)2] ions, by factors of ca. 3 ± 6. However, the smaller variation in values of
the dissociation rate constants, kd(Ni2‡)/kd(Co2‡), compared with the greater variation
in values of the formation rate constants, kf(Ni2‡)/kf(Co2‡), results in stability constants
(K ˆ kf/kd) for the Co2‡ complexes being larger than those of Ni2‡. For octahedral metal
complexes, stability constants usually vary according to the Irving ± Williams sequence
Mn2‡ < Fe2‡ < Co2‡ < Ni2‡ < Cu2‡ > Zn2‡, which is in line with calculated variations in
ligand-field stabilization energies (LFSE), with allowance being made for the greater
LFSE for Cu2‡ due to Jahn ± Teller distortion. For the tetrahedral complexes of the M2‡
ions studied here, the trend KNi < KCo contrasts with that found for octahedral
complexes where KNi > KCo, but is consistent with calculations of LFSE for tetrahedral
systems. The e4t2 electronic configuration of Co2‡ ions in a tetrahedral ligand field
3
results in greater stability than for Ni2‡ due to a calculated LFSE of (À6/5)Dt ‡ 2P for
Co2‡ (P ˆ pairing energy Dt ˆ tetrahedral-ligand-field splitting parameter), compared
with the e4t24 configuration of Ni2‡ that gives a smaller calculated LFSE of (À4/5)Dt ‡
3P. Therefore, ligands that force a tetrahedral geometry on first-row d-block M2‡ ions
are expected to be the most stable for Co2‡, in line with the results in Table 3.
The hydrometallurgical use of ꢀdithioimidodiphosphatesꢁ to extract divalent metal
ions is aided by the formation of neutral [ML2] complexes that extract from aqueous
media into organic solvents like kerosene, but a complication arises due to the redox
reactions observed with Cu2‡ ions. Although formation of [CuL2] species is very rapid,
‡
‡
reduction to Cu species follows to give multinuclear products such as [Cu4L3] [8] and
[Cu3L3] (Fig. 1) [17] and destruction of some of the extractant ligand. Therefore,
effective use of such ligands in hydrometallurgy requires the careful removal of
oxidizing metal ions like Cu2‡ beforehand. The rates of the reduction reactions to give
[Cu3L3] (Table 4) vary by at least a factor of 103, with those ligands carrying elcctron-
withdrawing substituents like L1 and L11 being much more reactive than those with
electron-donating substituents like L3, L7, and L8.
We thank Zeneca Specialties for provision of most of the ligands used in this study and for financial support,
and the EPSRC for provision of stopped-flow, X-ray, and NMR facilities.
REFERENCES
[1] A. Schmidpeter, R. Boehm, H. Groeger, Angew. Chem. 1964, 76, 860; A. Schmidpeter, H. Groeger, Z.
Anorg. Allg. Chem. 1966 345, 106; A. Schmidpeter, J. Ebeling, Chem. Ber. 1968 101, 815.
[2] R. A. Shaw, E. H. M. Ibrahim, Angew. Chem., Int. Ed. 1967, 6, 556.
[3] M. R. Churchill, J. Cooke, J. Wormald, A. Davison, E. S. Switkes, J. Am. Chem. Soc. 1969, 91, 6518; M. R.
Churchill, J. Wormald, J. Chem. Soc., Chem. Commun. 1970, 703; M. R. Churchill, J. Cooke, J. P. Fennessey,
J. Wormald, Inorg. Chem. 1971, 10, 1031; M. R. Churchill, J. Wormald, Inorg. Chem. 1971, 10, 1778.
[4] A. Davison, E. S. Switkes, Inorg. Chem. 1971, 10, 837.
[5] O. Siiman, H. B. Gray, Inorg. Chem. 1971, 10, 1778.
[6] H. Steinberger, W. Kuchen, Z. Naturforsch. B 1974, 29, 611.

Helvetica Chimica Acta ± Vol. 88 (2005)
795
[7] R. D. Bereman, F. T. Wang, J. Najdzionek, D. M. Braitsch, J. Am. Chem. Soc. 1976, 98, 7266.
[8] C. P. Huber, M. L. Post, O. Siiman, Acta Crystallogr., Sect. B 1978, 34, 2629; O. Siiman, J. Vetuskey, Inorg.
Chem. 1980, 19, 1672.
[9] L. Meznik, A. Marecek, Z. Chem. 1981, 8, 294.
[10] H. Noth, Z. Naturforsch., B 1982, 37, 1491.
[11] D. J. Williams, C. O. Quicksall, K. M. Barkigia, Inorg. Chem. 1982, 21, 2097.
[12] S. Husebye, K. Maartmann-Möe, Acta Chem. Scand., A 1983, 37, 439.
[13] P. B. Hitchcock, J. F. Nixon, I. Silaghi-Dumitrescu, I. Haidue, Inorg. Chim. Acta 1985, 96, 77.
[14] P. Bhattacharyya, J. Novosad, J. R. Phillips A. M. Z. Slawin, D. J. Williams, J. D. Woollins, J. Chem. Soc.,
Dalton Trans. 1995, 1607; J. R. Phillips, A. M. Z.. Slawin, A. J. P. White, D. J. Williams, J. D. Woollins, J.
Chem. Soc., Dalton Trans. 1995, 2467.
Â
[15] C. Sivestru, R. Rösler, I. Haiduc, R. Cea-Olivares, G. Espinosa-Perez, Inorg. Chem. 1995, 34, 3352; I.
Haiduc, R. Cea-Olivares, R. A. Toscano, C. Sivestru, Polyhedron 1995, 14, 1067.
[16] D. C. Cupertino, R. W. Keyte, A. M. Z.. Slawin, D. J. Williams, J. D. Woollins, Inorg. Chem. 1996, 35, 2695;
D. C. Cupertino, R. W. Keyte, A. M. Z.. Slawin, J. D. Woollins, D. J. Williams, Polyhedron 1996, 15, 4441; P.
Bhattacharyya, A. M. Z.. Slawin, M. B. Smith, J. D. Woollins, Inorg. Chem. 1996, 35, 3675; D. J. Williams,
A. M. Z.. Slawin, J. R. Phillips, J. D. Woollins, Polyhedron 1996, 15, 3175.
[17] D. J. Birdsall, D. Cupertino, J. D. Woollins, A. M. Z.. Slawin, Phosphorus, Sulfur Silicon Relat. Elem. 1997,
124/125, 465.
Â
[18] C. Silvestru, R. Rösler, J. E. Drake, J. Yang, G. Espinosa-Perez, I. Haiduc, J. Chem. Soc., Dalton Trans.
1998, 73.
[19] T. Q. Ly, J. D. Woollins, Coord. Chem. Rev. 1998, 176, 451.
[20] J. O. Morley, M. H. Charlton, J. Phys. Chem. A 1998, 102, 6871.
[21] D. C. Cupertino, R. W. Keyte, A. M. Z.. Slawin, J. D. Woollins, Polyhedron 1999, 18, 311; D. C. Cupertino,
R. W. Keyte, A. M. Z.. Slawin, J. D. Woollins, Polyhedron 1998, 17, 4219; D. C. Cupertino, R. W. Keyte,
A. M. Z.. Slawin, J. D. Woollins, Polyhedron 1999, 18, 707.
[22] L. M. Gilby, B. Piggott, Polyhedron 1999, 18, 1077; K. Darwin, L. M. Gilby, P. R. Hodge, B. Piggott,
Polyhedron 1999, 18, 3729.
[23] D. J. Birdsall, A. M. Z.. Slawin, J. D. Woollins, Inorg. Chem. 1999, 38, 4152; D. J. Birdsall, A. M. Z.. Slawin,
J. D. Woollins, Polyhedron 2001, 20, 125.
[24] S. P. Sangha, Ph.D. Thesis, University of Warwick, 1999.
[25] C. G. Pernin, J. A. Ibers, Inorg. Chem. 2000, 39, 1216; C. G. Pernin, J. A. Ibers, Inorg. Chem. 2000, 39, 1222.
[26] Q.-F. Zhang, H. Zheng, W.-Y. Wong, W.-T. Wong, W.-H. Leung, Inorg. Chem. 2000, 39, 5255.
[27] G. L. Abbati, M. C. Aragoni, M. Arca, F. A. Devillanova, A. Francesco, A. C. Fabretti, A. Garau, F. Isaia,
V. Lippolis , G. Verani, J. Chem. Soc., Dalton Trans. 2001, 1105.
[28] E. Simon-Manso, M. Valderrama, D. Boys, Inorg. Chem. 2001, 40, 3647.
[29] W. C. Herndon, I. Haiduc, J. Mol. Struct. 2001, 598, 127.
Â
[30] H. Liu, N. A. G. Bandeira, M. J. Calhorda, M. G. B. Drew, V. Felix, J. Novosad, F. Fabrizi de Biani, P.
Zanello, J. Organomet. Chem. 2004, 689, 2808.
[31] R. F. Dalton, D. C. Cupertino, J. Campbell, GB Pat. No. 2317385, 1998.
[32] A. D. van Ingen Schenau, W. L. Groenevald, J. Reedijk, Recl. Trav. Chim. Pays-Bas 1972, 91, 88.
[33] A. I. Vogel, ꢀTextbook of Quantitative Analysisꢁ, 4th edn., Longman, London, 1978, p. 199.
[34] G. M. Sheldrick, ꢀSADABS, Emprirical Absorption Corrections Programꢁ, University of Göttingen,
Germany, 1996.
[35] Siemens, ꢀSHELXTL-PC Version 5.0 Reference Manualꢁ, Siemens Industrial Automation, Inc., Analytical
Instrumentation, Madison, WI, 1994.
[36] G. M. Sheldrick, ꢀSHELLXL97 Program for Crystal Structure Refinementꢁ, University of Göttingen,
Göttingen, Germany 1997.
[37] W. C. Vosburgh, G. R. Cooper, J. Am. Chem. Soc. 1941, 63, 437.
[38] S. F. A. Kettle, ꢀPhysical Inorganic Chemistry: a Coordination Chemistry Approachꢁ, Spektrum Academic
Publ., 1996, p. 175.
[39] R. G. Wilkins, ꢀKinetics, Mechanisms of Reactions of Transition Metal Complexesꢁ, 2nd edn., VCH,
Weinheim, 1991.
[40] R. G. Pearson, P. Ellgen, Inorg. Chem. 1967, 6, 1379.
Received November 11, 2004