Reversible binding of water, methanol, and ethanol to a five-coordinate ruthenium(ii) complex

Article abstract of DOI:10.1039/c3dt32909g

The known green, five-coordinate, square-pyramidal trans-RuCl ₂(P-N)(PPh₃) complex reversibly binds water, MeOH and EtOH in the vacant coordination site in the solid state and in CH₂Cl ₂ solution to give pink adducts (P-N = o-diphenylphosphino-N, N′-dimethylaniline). The adducts are well characterized, including X-ray analysis of the aqua complex, trans-RuCl₂(P-N)(PPh ₃)(H₂O), which crystallizes in two different benzene-solvated forms. Comparison of the structural data with those determined previously for the binding of H₂S, thiols, and H₂, which form cis-RuX₂(P-N)(PPh₃)L products (X = Cl, Br; L = a S-ligand or H₂) reveals the trans-influence trend P > H ₂S ～ thiols > H₂ > Cl ～ Br > H ₂O. Thermodynamic data for the binding of water were estimated in solution by UV-Vis spectroscopy, and ΔH^o data for the aqua and alcohol adducts in the solid state were obtained by differential scanning calorimetry. Inclusion of published data for the S-ligand adducts reveals the thermal stability trend of the solid complexes as MeSH > MeOH > H ₂S > H₂O > EtSH > EtOH.

Full text of DOI:10.1039/c3dt32909g

Dalton
Transactions
View Article Online
View Journal | View Issue
PAPER
Reversible binding of water, methanol, and ethanol to
a five-coordinate ruthenium(II) complex†
Cite this: Dalton Trans., 2013, 42, 4291
Erin S. F. Ma, Brian O. Patrick and Brian R. James*
The known green, five-coordinate, square-pyramidal trans-RuCl2(P–N)(PPh3) complex reversibly binds
water, MeOH and EtOH in the vacant coordination site in the solid state and in CH2Cl2 solution to give
pink adducts (P–N = o-diphenylphosphino-N,N’-dimethylaniline). The adducts are well characterized,
including X-ray analysis of the aqua complex, trans-RuCl2(P–N)(PPh3)(H2O), which crystallizes in two
different benzene-solvated forms. Comparison of the structural data with those determined previously
for the binding of H2S, thiols, and H2, which form cis-RuX2(P–N)(PPh3)L products (X = Cl, Br; L = a S-ligand
or H2) reveals the trans-influence trend P > H2S ∼ thiols > H2 > Cl ∼ Br > H2O. Thermodynamic data for
the binding of water were estimated in solution by UV-Vis spectroscopy, and ΔHo data for the aqua
and alcohol adducts in the solid state were obtained by differential scanning calorimetry. Inclusion of
published data for the S-ligand adducts reveals the thermal stability trend of the solid complexes as
MeSH > MeOH > H2S > H2O > EtSH > EtOH.
Received 4th December 2012,
Accepted 8th January 2013
DOI: 10.1039/c3dt32909g
www.rsc.org/dalton
Introduction
Experimental section
General
We recently published details on the reversible binding in
solution of H2S and alkyl thiols to five-coordinate complexes of
the type trans-RuX2(P–N)(PR3) where P–N = o-diphenylphos-
phino-N,N′-dimethylaniline, X = halide, and R = Ph or p-tolyl
(cf. eqn (1)); thermodynamic data were also presented for for-
mation of some selected H2S, MeSH and EtSH products, in
which the halide ligands are now cis.1 In publications from the
1990s,2 we had mentioned corresponding coordination of
MeOH, EtOH and H2O but the details have been available only
in Ph.D. dissertations.3 This current paper now describes the
alcohol and aqua products, which have trans-chlorides (eqn
(1)). The findings include crystallographic data for the aqua
complexes, and thermodynamic data for reversible binding of
these O-donors, which allow for comparison with data for the
S-donor ligand systems.
Unless stated otherwise, all manipulations were performed
under an oxygen-free Ar or N2 atmosphere at ambient tempera-
tures using standard Schlenk techniques. Commercially avail-
able compounds were supplied by Aldrich or Fisher, and were
used as received unless stated otherwise. Spectral or analytical
grade solvents were refluxed, distilled over appropriate drying
agents,4 and then purged with Ar or N2 before being trans-
ferred into a reaction flask via a cannula. Deuterated solvents,
obtained from Cambridge Isotope Laboratories, were stored
over activated molecular sieves (Fisher 4 Å, 4–8 mesh), and
immediately before use were de-oxygenated (via the freeze–
pump–thaw method), and stored under Ar.
NMR spectra were recorded, unless stated otherwise, at
room temperature (r.t. ∼22 °C) on Varian XL300 (300.0 MHz
1
for H, 121.4 MHz for 31P) or Bruker AMX500 (500.0 MHz for
1H and 202.5 MHz for 31P) instruments. Residual deuterated
solvent protons (relative to external SiMe4) and external
P(OMe)3 (δ 141.0 relative to 85% H3PO4) were used as references
(s = singlet, d = doublet, t = triplet, q = quartet, m = multiplet,
br = broad); J values are reported in hertz (Hz). Samples were
prepared in 5 mm NMR tubes equipped with poly(tetrafluoro-
ethylene), J.-Young valves (Aldrich). Calibrated 1H NMR probes
were used to determine the temperatures used for van’t Hoff
analysis. ATLI Mattson Genesis FTIR and Bomem Michelson
far-IR spectrophotometers were used to record spectra from
500–4000 cm−1 (KBr) and 200–3000 cm−1 (CsI); data are
Department of Chemistry, University of British Columbia, Vancouver,
British Columbia, Canada, V6T 1Z1. E-mail: brj@chem.ubc.ca
†Electronic supplementary information (ESI) available: ORTEP of 2b; TGA of 2a;
31P{1H}- and 1H-NMR spectra of 2a at different temperatures; 31P{1H} and UV-vis
spectral changes on addition of H2O to 1a to form 2a, and data for determi-
nation of thermodynamic data for this reaction; 1H NMR spectrum of 3. CCDC
913686 and 913687. For ESI and crystallographic data in CIF or other electronic
format see DOI: 10.1039/c3dt32909g
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 4291–4298 | 4291

View Article Online
Paper
Dalton Transactions
reported in cm−1. UV-Vis spectra were recorded on a Hewlett (2 × 5 mL), and dried by passing Ar through the sample at r.t.
Packard 8452A diode-array spectrophotometer, equipped with for ∼15 min. Yield: 45 mg, 56%. Anal. Calcd C39H39NOCl2-
a thermostated compartment using an anaerobic, 1 cm quartz P2Ru: C, 60.70; H, 5.09; N, 1.82. Found: C, 61.0; H, 5.1; N, 1.8.
cell, joined to a side-arm flask for mixing solutions; data are 31P{1H} NMR (CD2Cl2): δ 77.46 (d, P–N), 47.16 (d, PPh3);
reported as λmax in nm (ε in units of M−1 cm−1). Thermogravi- 2JPP = 36.7. 1H NMR (CD2Cl2): δ 7.9–6.9 (29H, m, Ph), 3.33 (3H,
metric analysis (TGA) was performed on a TA Q50 Instrument: d, OCH3), 3.16 (6H, s, N(CH3)2), 1.33 (1H, q, OH).
solid samples were weighed (10 to 15 mg) into a Pt pan, and
the samples were then heated under N2 (flow rate = 100 cc the method described for
trans-RuCl2(P–N)(PPh3)(EtOH) (4). Attempts to prepare 4 by
were unsuccessful. Several
3
min−1) at a rate of 10 °C min−1 to ∼500 °C. Differential scan- different solvent combinations, including acetone mixtures
ning calorimetry (DSC) data were collected on a TA 910S Instru- with Et2O, EtOH or hexanes, failed to precipitate any solid.
ment, with 2–5 mg samples being heated under N2 (flow rate = When P–N (40.5 mg, 0.13 mmol) in EtOH (2 mL) was added to
40 cc min−1) at a rate of 5 °C min−1 up to 500 °C. Microana- a brown suspension of RuCl2(PPh3)3 (122.8 mg, 0.13 mmol) in
lyses were performed in this department on a Carlo Erba 1106 EtOH (8 mL), and the mixture stirred for 1 week, an orange-
instrument.
pink solution containing a small amount of a light brown pre-
The RuCl2(PR3)3 (R = Ph,5 p-tolyl6), RuCl2(P–N)(PPh3) (1a),2a cipitate formed. This solid (∼20 mg) was collected and washed
and RuCl2(P–N)(P(p-tolyl)3) (1b)2a complexes were prepared by with EtOH (5 mL), but it was insoluble in common solvents
the literature methods, the precursor RuCl3·xH2O being (acetone, CDCl3, C6D6, CD2Cl2). Also, the EtOH was removed
donated by Colonial Metals, Inc.
under vacuum from the pink filtrates, and hexanes (10 mL)
trans-RuCl2(P–N)(PPh3)(H2O) (2a). The complex was pre- was then added to the oily residue. The solvent was again
pared by adding a mixture of H2O (2 mL) and acetone (2 mL) removed and EtOH (2 mL) was added to dissolve the residue;
to a stirred solution of RuCl2(PPh3)3 (200 mg, 0.21 mmol) and stirring for 15 min generated a pink precipitate. Hexanes
P–N (64 mg, 0.21 mmol) in acetone (5 mL) at r.t. The instantly (10 mL) was then added to precipitate more solid, which was
formed orange-pink solution was stirred for 3 h during which collected by filtration, washed with hexanes (5 mL), and drying
time a pink solid precipitated; this was filtered off, washed attempted by using an Ar stream as for 3. Yield: 33 mg, 33%.
with acetone (2 × 5 mL), and dried in vacuo for 24 h. Yield: Anal. Calcd C40H41NOCl2P2Ru: C, 61.15; H, 5.26; N, 1.78.
115 mg, 73%. Anal. Calcd
C
38H37NOCl2P2Ru·(acetone): Found: C, 62.2; H, 5.1; N, 1.9. 31P{1H} NMR (CD2Cl2): δ 79.79
C, 60.37; H, 5.31; N, 1.72. Found: C, 60.37; H, 5.46; N, 1.67. (d, P–N), 46.90 (d, PPh3); JPP = 36.2. 1H NMR (CD2Cl2):
31P{1H} NMR (C6D6): δ 73.52 (d, P–N), 49.30 (d, PPh3); δ 7.9–6.9 (29H, m, Ph), 3.61 (2H, d of q, OCH2), 3.18 (6H, s,
2JPP = 38.0. 1H NMR (C6D6): δ 8.4–7.0 (29H, m, Ph), 3.05 (6H, s, N(CH3)2), 1.40 (1H, t, OH)), 1.16 (3H, t, OCH2CH3).
N(CH3)2), 2.15 (2H, br s, Ru–OH2), 1.55 (6H, s, acetone). UV-Vis
2
X-Ray crystallographic analyses
(see Results and discussion). IR: νOH 3556 s, 3295 s, 1605 s,
νCO 1707 s (acetone). Two different type crystals (2a·2C6H6 and X-ray analyses of 2a·2C6H6 and 2a·1.5C6H6 were carried out at
2a·1.5C6H6) were isolated from evaporation of a saturated C6H6 180 K on a Rigaku/ADSC CCD area detector with graphite
solution of the complex over 24 h (see X-Ray crystallographic monochromated MoKα radiation (0.71069 Å). The crystals
analyses). Complex 2a was also readily prepared in situ by differ in appearance as well as having different unit cells, the
adding one mole equiv. of H2O (0.70 μL, 0.04 mmol) to a green pink crystals (2a·2C6H6) being monoclinic, and the yellow-
CDCl3 solution (0.8 mL) of precursor 1a (29.6 mg, 0.04 mmol); brown crystals (2a·1.5C6H6) being triclinic. Some crystallo-
the NMR data were essentially the same as those of isolated 2a. graphic data for the 2a·2C6H6 structure are: 38 827 total reflec-
trans-RuCl2(P–N)(P(p-tolyl)3)(H2O) (2b). The complex was tions, 10 710 unique (Rint = 0.074), 6767 observed [I > 2σ(I)],
prepared in the same manner as described for 2a, but using R1 = 0.076; wR2 = 0.167; GOF = 1.11; residual density = −1.39 e
RuCl2(P(p-tolyl)3)3 (200 mg, 0.19 mmol) as precursor. Yield: Å−3. Corresponding data for 2a·1.5C6H6 are: 18577, 9156
122 mg, 77%. Anal. Calcd
C41H43NCl2OP2Ru·(acetone): (0.039), 6959, 0.062, 0.098, 1.01, and −0.0.87. Data were pro-
C, 61.61; H, 5.76; N, 1.63. Found: C, 62.0; H, 5.7; N, 1.8. cessed using the d*TREK area detector program,7 and the
31P{1H} NMR (C6D6): δ 63.63 (d, P–N), 45.91 (d, PPh3); structures were solved by direct methods.8 All refinements
2JPP = 38.1. 1H NMR (C6D6): δ 8.2–6.8 (26H, m, Ph), 3.10 (6H, s, were performed using the SHELXL-97 program9 via the WinGX
N(CH3)2), 2.00 (3H, s, p-CH3), 2.15 (2H, br s, Ru–OH2), 1.55 interface.10 For both structures, all non H-atoms were refined
(6H, s, acetone). Complex 2b can also be made in situ by a 1 : 1 anisotropically; the H-atoms of the coordinated H2O were
reaction of H2O with 1b in CDCl3, as noted above for 2a.
located in a difference map and refined isotropically. All other
trans-RuCl2(P–N)(PPh3)(MeOH) (3). A mixture of MeOH H-atoms were placed in calculated positions. For the crystals
(2 mL) and acetone (1 mL) was purged with Ar and cannula with the 1.5 molecules of C6H6 in the asymmetric unit, one
transferred to a stirred solution of RuCl2(PPh3)3 (100 mg, half-benzene resides on an inversion centre, while within the
0.10 mmol) and P–N (32 mg, 0.10 mmol) in acetone (5 mL), material with two C6H6 molecules, one was disordered and
which had been heated to 50 °C. The instantly formed orange was modeled in three orientations such that their combined
solution was stirred at 20 °C for 24 h, and the volume was then occupancies summed to 1.0. The ORTEP plots and selected
reduced to ∼1 mL, when hexanes (10 mL) was added to pre- bond lengths and angles of the 2a structures are shown in
cipitate a pink solid; this was collected, washed with MeOH Fig. 1 and 2, and Tables 1 and 2, while the full experimental
4292 | Dalton Trans., 2013, 42, 4291–4298
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
Table 2 Selected bond angles (°) for benzene solvated trans-RuCl2(P–N)(PPh3)-
(H2O) (2a·2C6H6 and 2a·1.5C6H6), and trans-RuCl2(P–N)(P(p-tolyl)3)(H2O) (2b),
with estimated standard deviations in parentheses
Bond
2a·2C6H6
2a·1.5C6H6
2b
H(1)–O(1)–H(2)
Ru(1)–O(1)–H(1)
Ru(1)–O(1)–H(2)
Cl(1)–Ru(1)–O(1)
Cl(2)–Ru(1)–O(1)
Cl(1)–Ru(1)–P(1)
Cl(1)–Ru(1)–P(2)
Cl(1)–Ru(1)–N(1)
Cl(1)–Ru(1)–Cl(2)
O(1)–Ru(1)–P(1)
O(1)–Ru(1)–P(2)
O(1)–Ru(1)–N(1)
Cl(2)–Ru(1)–P(1)
Cl(2)–Ru(1)–P(2)
Cl(2)–Ru(1)–N(1)
P(1)–Ru(1)–P(2)
P(1)–Ru(1)–N(1)
P(2)–Ru(1)–N(1)
107(3)
118(6)
114(6)
109(3)
121(3)
115(3)
111(6)
112(4)
105(6)
81.6(1)
82.62(11)
83.67(11)
104.20(5)
86.94(5)
91.09(11)
165.17(5)
168.22(11)
90.87(11)
88.90(16)
88.43(5)
98.89(5)
83.03(11)
98.99(5)
81.46(11)
178.03(12)
80.76(6)
85.26(6)
105.31(3)
86.98(2)
92.27(6)
165.58(2)
169.78(6)
88.69(6)
90.32(8)
88.04(3)
96.26(2)
84.25(6)
99.71(3)
81.33(6)
178.85(6)
82.2(1)
104.30(5)
89.74(5)
90.8(1)
162.91(4)
168.8(1)
91.4(1)
90.3(1)
90.73(4)
96.26(5)
83.7(1)
98.04(5)
80.20(9)
178.24(9)
Fig. 1 ORTEP diagram of trans-RuCl2(P–N)(PPh3)(H2O) (2a·2C6H6) with 50%
probability thermal ellipsoids.
parameters and details of the two structures are given in CIF
format in the ESI.†
Reddish crystals of 2b, deposited over 3 h in the NMR tube
of an in situ synthesis, were analysed crystallographically at r.t.
in 1994 as a non-solvated molecule on a Rigaku AFC6S instru-
ment with CuKα radiation. An ORTEP structure (Fig. S1†) and
crystallographic details (Appendix A, in non-CIF format) are
given in the ESI;† some bond lengths and angles are also given
in Tables 1 and 2.
Results and discussion
The aquo complexes
The coordination chemistry of H2O is ubiquitous, although
after an extensive search we have been unable to find any
report on formation of a neutral, six-coordinate RuII-aqua
complex via reaction of H2O with a five-coordinate precursor,
as exemplified here in eqn (1) by the reactivity of the
RuCl2(P–N)(PR3) complexes (R = Ph, 1a; p-tolyl, 1b); the aqua
product, like the precursor, has trans-chlorides. As mentioned
in the Introduction, H2S (and thiols) similarly coordinate,
but the isolated product with these S-ligands has cis-chlorides
with the H2S trans to a Cl.1 Of note, however, there is a
recent example of reversible addition of H2O to a cationic, five-
coordinate RuCl(PNNP)+ complex to give a mixture of cis- and
trans-products, where the P-atoms (and the N-atoms) occupy
cis-positions within the tetradentate ligand.11
Fig. 2 ORTEP diagram of trans-RuCl2(P–N)(PPh3)(H2O) (2a·1.5C6H6) with 50%
probability thermal ellipsoids.
Table 1 Selected bond lengths (Å) for benzene solvated trans-RuCl2(P–N)-
(PPh3)(H2O) (2a·2C6H6 and 2a·1.5C6H6), and trans-RuCl2(P–N)(P(p-tolyl)3)(H2O) (2b),
with estimated standard deviations in parentheses
Bond
2a·2C6H6
2a·1.5C6H6
2b
Ru(1)–O(1)
Ru(1)–P(1)
Ru(1)–P(2)
Ru(1)–N(1)
Ru(1)–Cl(1)
Ru(1)–Cl(2)
O(1)–H(1)
O(1)–H(2)
H(1)⋯Cl(2)
2.232(4)
2.191(2)
2.2333(8)
2.3091(7)
2.311(2)
2.4311(7)
2.3951(7)
0.863(10)
0.866(10)
2.31(3)
2.252(4)
2.220(1)
2.284(1)
2.326(4)
2.385(1)
2.418(1)
0.69(6)
2.2305(14)
2.3143(14)
2.312(4)
2.3957(13)
2.4195(13)
0.870(10)
0.870(10)
2.27(7)
ð1Þ
The air-sensitive, pink complexes (2a and 2b) can be pre-
0.96(6)
2.46(7)
pared by stirring 1a and 1b in a 4 : 1 mixture of acetone and
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 4291–4298 | 4293

View Article Online
Paper
Dalton Transactions
H2O under Ar at r.t., the products being isolated as the sol-
As noted above, the trans-RuCl2(P–N)(PR3) precursors bind
vated trans-RuCl2(P–N)(PR3)(H2O)·(acetone) complexes (R = Ph, H2S, MeSH and EtSH to form cis-RuCl2(P–N)(PR3)L products
p-tolyl). More conveniently, the required unsaturated, green (L = the S-donor),1 and the P(1) of the P–N ligand, like L, is
precursor species (1a/1b) can be formed in situ from P–N and also trans to a Cl-atom. In the aqua species, P(1) is trans to
RuCl2(PR3)3, as described in the Experimental section. Heating H2O, and the Ru–P(1) bonds in 2a·2C6H6, 2a·1.5C6H6 and 2b
2a/2b in vacuo at 80 °C regenerates 1a/1b. The loss of H2O was (2.231, 2.233, and 2.220 Å, respectively) are shorter than the
confirmed by the TGA of 2a (Fig. S2†), where a weight loss of average value of 2.27 Å for the S-containing complexes, indicat-
11% between 80 to 110 °C, prior to thermal decomposition, ing that Cl has a stronger trans-influence than H2O toward a
agrees well with the theoretical combined 9% weight of phosphine P-atom; this agrees with ab initio calculations,18
1
acetone and H2O present. Exposure of 1a and 1b to a moist, and JPtP NMR data for trans-[Pt(H2O)(CH3)(diphos)]+ and
oxygen-free atmosphere reversibly gives within minutes the trans-[PtCl(CH3)(diphos)].19 The correlation between 31P NMR
aqua species, with formation of 2b (<3 min) being noticeably data and trans-influence will be further discussed below.
faster (via colour change) than 2a (>15 min). The faster rate
with 1b may be related to particle size, but the X-ray structure pattern with δPA of the P–N ligand (eqn (1)) being at lower field
The NMR data in C6D6 for the aqua-adducts: an AX 31P{1H}
of 1b shows no agostic interaction between the Ru and an (by ∼17–25 ppm) than δPX with 2J ∼ 38 Hz, and singlets at δH
∼
o-phenyl H-atom from the P(p-tolyl)3 ligand,2a and thus the 3.50–3.00 in the 1H spectrum for equivalent NMe2 groups,1 are
species does have an easily accessible, vacant coordination consistent with the solid state structures. The resolution and
site; the implication is that 1a might show such an interaction, shifts of the NMR resonances for the aqua-adducts are,
but unfortunately many attempts to grow crystals of this however, dependent on the solvent, temperature and concen-
complex were unsuccessful.
tration of added H2O. Because of the rapid equilibrium (eqn
The structures of 2a in both solvate forms and 2b (Fig. 1, 2 (1)), some of the resonances of 2a and 1a are broadened on
and S1;† Tables 1 and 2) reveal pseudo-octahedral geometry at the NMR-timescale. The sharp doublets at δPX ∼ 48 for 2a are
the Ru with trans-chlorides, with the H2O being trans to the little affected but, for example on dissolution of 2a in CD2Cl2
P-atom of the P–N ligand; the H-atoms of the H2O were isotro- (Fig. S3†), the composite broadened PA signal changes from
pically refined in the structures. The non-disordered crystals of ∼80 to 62 ppm on going from 25 to −80 °C: at 25 °C, 1a is
2a·1.5C6H6 revealed a 2.77(3) Å distance between the aqua H(2) favoured (δPA 80.1). At −50 °C, the concentrations of 1a and 2a
and benzene C(40), suggesting an OH/π-benzene ring inter- must be similar as the resonances coalesce into the base line,
action. This possibly results in the observed shorter Ru–O and at −80 °C, 2a dominates. Such coalescence resulting from
bond in 2a·1.5C6H6 than in 2a·2C6H6 by ∼0.04 Å. The value is the trans-effect of a P-atom on a H2O ligand has been noted
intermediate between those of a relatively weakly bound aqua previously within the weakly bonded aqua complexes, trans,
ligand [e.g., 2.215 Å in cis-[RuCl(H2O)(PNNP)]+ mentioned mer-[MCl2(H2O)(PMe2Ph)3][ClO4] (M = Rh21 or Ir22). Table S1†
above,11 2.218 Å in trans-[Ru(H2O)(PEt3)2(trpy)]2+ (trpy = 2,2′,2′′- gives the 31P{1H} data for 1a and isolated samples of 2a at r.t.
terpyridine)12a and 2.203 Å in [Ru(η6-p-MeC6H4 Pr)(H2O)L]+ in various solvents. Fig. S4† clearly shows that increasing
i
(HL
= ] and a more [H2O] is required to fully form 2a in d6-acetone where, in the
S-(α-methylbenzyl)-salicylaldimine)12b
strongly bound one [e.g. 2.15 Å in [RuH(H2O)(CO)2(PPh3)2]+,13 absence of H2O, 1a exists as the acetone adduct:1 with increas-
2.127 Å in [Ru(η6-C6H6)(H2O)3]2+,14 2.141 and 2.115 Å in Ru- ing [H2O], the broadened doublet PA signal at δ 70.5 gradually
(H2O)2(η1(O):η2(C,C′)–OCOCH2CHvCHCH3)2,15 2.122 Å in [Ru- becomes the sharp doublet of 2a at ≥300 equiv. of H2O, while
(H2O)6]2+,16 and 2.158 and 2.095 Å in [Ru(cod)(H2O)4]2+ 17]; the sharp PX doublet changes little. The H NMR spectra of 2a
1
there are several other examples within cationic RuII systems.11 in various solvents are consistent with the 31P{1H} data,
The shorter Ru–O bonds in 2a·1.5C6H6 and 2a·2C6H6 relative although the distinction between the resonances of 2a and 1a
to the one in 2b perhaps contributes to a greater interaction is not as obvious. For example, in CD2Cl2 solutions of 2a, the
between H(1)⋯Cl(2) in the 2a species. Of note, the O–H bonds NMe2 signals of 2a and 1a overlap as seen in Fig. S5;† of note,
are up to ∼0.25 Å shorter than the 0.956 Å value in free H2O, when the solution is cooled from 25 to −80 °C, the Ru–OH2
while the approximate H–O–H angles are somewhat greater signal moves downfield from δ 2.18 to 3.42, whereas the
than that of free H2O (105°); the weak OH⋯benzene and NMe2 signals shift upfield from δ 3.20 to 2.85. In cis-
OH⋯Cl interactions almost certainly play a role. In the H2S RuCl2(P– N)(PR3)L complexes, where L is H2S, MeSH, EtSH or
1,2a,3b
analogue of 2a, isolated as an acetone solvate in which H2
the two Me groups are inequivalent as established by
H-bonding to a chloride is also present,20 the H–S–H angle crystallographic and 1H data.
1
102(2)° is larger than that of free H2S (92.5°), and the Ru–S
The 31P{1H} and H NMR spectra for the 1a/2a equilibrium
bond length (2.35 Å) is as expected longer than the Ru–O contrast with those of the corresponding reversible binding of
bond in the aqua species. The non-linear Cl–Ru–Cl angles S-ligands to give cis-products, where under similar conditions
(∼165 Å) in the 2a and 2b structures must result from the both 1a and the H2S- and RSH-adducts, for example, are
bending of the chlorides towards the aqua ligand due to readily distinguished by 1H NMR data, and equilibria and
H-bonding interactions, which are thought more generally thermodynamic data for the reversible binding could be obtained
(even in non-chloro-containing species) to stabilize RuII-aqua by these data.1 The aqua system is clearly much more labile on
complexes.11
the NMR-timescale compared to the H2S system; however,
4294 | Dalton Trans., 2013, 42, 4291–4298
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
Table 3 Comparison of 31P{1H} NMR data and Ru–P bond lengths (in Å)
Complex (in CDCl3 at 20 °C)a
δPA
Ru–PA
δPX
Ru–PX
2JPP (Hz)
t-RuCl2(P–N)(P(p-tolyl)3) (1b)b
t-RuCl2(P–N)(P(p-tolyl)3)(H2O) (2b)c
t-RuCl2(P–N)(PPh3)(H2O) (2a·2C6H6)
(2a·1.5C6H6)
81.46
71.80
68.50
2.170(1)
2.220(1)
47.64
47.62
47.70
2.290(1)
2.284(1)
37.15
38.12
37.76
2.2305(14)
2.2333(8)
2.2560(4)
2.2712(6)
2.262(1)
2.2802(8)
2.2743(4)
2.2884(7)
2.3143(14)
2.3091(7)
2.3040(3)
2.3110(7)
2.301(1)2
2.3109(8)
2.3110(5)
2.3098(6)
c-RuCl2(P–N)(P(p-tolyl)3)(H2S)d (5)
c-RuCl2(P–N)(PPh3)(H2S)e (6)
c-RuBr2(P–N)(PPh3)(H2S)e (7)
c-RuCl2(P–N)(PPh3)(MeSH)f (8)
c-RuCl2(P–N)(PPh3)(EtSH)f (9)
c-RuCl2(P–N)(PPh3)(H2)g (10)
51.91
50.60
53.41
51.43
50.97
49.30
42.58
44.48
44.36
42.37
42.48
45.48
30.41
30.23
29.20
29.87
30.05
26.83
a t = trans; c = cis. b Ref. 2a. c Ref 3a. d Ref. 2b. e Ref. 20. f Ref. 1. g Ref. 3b.
equilibria data for the aquo system could be determined by
UV-Vis spectroscopy, a faster timescale technique (see below).
The rapid and reversible coordination of H2O must result
from the trans-effect of the PA-atom on the H2O ligand. Conver-
sely, the trans-influence of the H2O must weaken the Ru–PA
bond relative to its strength in 1a. The trans-influence of the
ligand trans to PA is demonstrated by 31P{1H} NMR data for the
RuCl2(P–N)(PR3)L complexes (Table 3). For example, the ligand
trans to PA is Cl in cis-RuCl2(PA–N)(PXR3)L, and trans to H2O in
trans-RuCl2(PA–N)(PXR3)(H2O); in both species (and 1a), the
N-atom is trans to PPh3 and the δPX value of ∼45 ppm is rela-
tively insensitive to L or the orientation of the Cl-atoms
(Table 3). The negligible cis-influence of ligands on phos-
1
phines is also well established by δP and JPtP values for PtII-
phosphine systems.23 The δPA values, however, are dependent
on the ligand trans to PA: a more downfield PA signal corres- Fig. 3 Relationship between Ru–PA bond length (Å) and δPA (in CDCl3) for the
complexes listed in Table 3.
ponds to a greater trans-influence of the trans ligand because
this is determined by the ability of this ligand to deshield PA.24
This results from the efficacy of the ligand to compete for the
Table 4 Ru–Cl bond lengths (Å) for cis- and trans-RuCl2(P–N)(PR3)L
metal orbital’s s-character,25 which also reflects the σ-donating
1
ability of the ligand, as demonstrated by JMP data for M = RhI
Complex
Ru–ClA
Ru–ClB
and PtII systems.24b,25b,26 A larger J value reflects stronger
σ-bonds and indicates a weaker influence of the trans
ligand.25b,27
trans-RuCl2(P–N)(PR3)L
R = p-tolyl, L = vacant (1b) 2.387(1) 2.379(1)
R = Ph, L = H2O (2a·2C6H6) 2.3957(13) 2.4195(13)
R = Ph, L = H2O (2a·1.5C6H6) 2.3951(7) 2.4311(7)
For the cis- and trans-RuX2(P–N)(PR3)L complexes (X = Cl,
Br), the Ru–PX bond lengths are in the 2.28–2.31 Å range
(Table 3), whereas there is an inverse dependence of δPA on
the Ru–PA length (Table 3, Fig. 3), a trend also noted for
R = p-tolyl, L = H2O (2b)
2.385(1) 2.418(1)
cis-RuCl2(P–N)(PR3)L
R = p-tolyl, L = H2S (5)
R = Ph, L = H2S (6)
R = Ph, L = MeSH (8)
R = Ph, L = EtSH (9)
R = Ph, L = η2-H2 (10)
29
related RuII-complexes containing PPh328 and Ph2P(CH2)4PPh2
2.429(3) 2.469(4)
2.4238(6) 2.4721(5)
2.4241(7) 2.4472(7)
2.4204(6) 2.4674(5)
2.4090(6) 2.4543(7)
ligands. Remarkably, the slopes and intercepts of all three
plots (cf. Fig. 3) are essentially the same, about −3.0 × 10−3
Å ppm−1, and ∼2.43 Å, respectively. From the Fig. 3 plot, the
order of decreasing trans-influence is Cl ∼ Br > H2O, consistent
with halides being better σ- and π-donors than H2O. Of note,
2JPP values for the trans complexes (37–38 Hz) are larger than
those of the cis complexes (27–30 Hz). Table 4 lists the Ru–Cl PA-atom of the P–N ligand has a greater trans-influence than Cl
bond distances for the trans- and cis-complexes: the average as shown by the relatively long Ru–ClB bonds for the cis-
trans Ru–Cl bond distances and the Ru–ClA bond distances in species, consistent with the same trend already established for
the cis-species imply that S-ligands have a stronger trans-influ- some PtII and RhI complexes.24b,25b Overall, assuming that cis-
ence than Cl. Further, the average Ru–ClA value of 2.424 Å for effects are negligible, the above observations suggest a trans-
the S-ligand cis-species perhaps implies a slightly greater trans- influence order of PA > H2S ∼ thiols > H2 > Cl ∼ Br > H2O.
influence for the S-ligands than for H2 (Ru–ClA = 2.409 Å). The The trend offers plausible mechanistic insight. Formation of
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 4291–4298 | 4295

View Article Online
Paper
Dalton Transactions
trans-2a and -2b involves binding of the H2O at the vacant posi- (at 25 °C, based on K = 36 1 M−1) for the coordination of H2O
tion of the square pyramidal precursors (eqn (1)), whereas to 1a (see Appendix B); the exothermicity and negative entropy
such coordination of the S-ligands and H2, which give the cis- are consistent with an equilibrium such as eqn (1).1 Compari-
adducts, is likely disfavoured because of the mutual trans- son of the K = 36 M−1 value with corresponding values
influences of PA and the S-ligands. Rearrangement of 2a/2b to obtained for the S-ligand systems (in C6D6)1 indicates that the
a trigonal bipyramidal intermediate with a Cl trans to PA (and formation of RuCl2(P–N)(PPh3)L is favoured in the order:
PX, N and Cl in the trigonal plane), and subsequent attack at L = MeSH > EtSH ∼ H2S > H2O, with K decreasing from 296 to
the equatorial position between PX and N, would give the 36 M−1. Kinetic studies on ligand substitution were attempted,
favoured cis-product. Such
a
rearrangement has been for example, by exposing solutions of 2a (in acetone containing
suggested previously, following dissociation of H2O from trans, >1.0 M H2O, or in CH2Cl2 solution containing 0.13 M H2O) to
mer-[MCl2(H2O)(PMe2Ph)3]+ (M = Rh21 or Ir30), although routes 1 atm H2S. However, the solution changed ‘instantaneously’
involving initial dissociation of Cl− are feasible.
from pink to the yellow colour of the H2S adduct,1 and the
Equilibria data for the H2O-binding to 1a were obtained by reaction rates were too rapid to be measured.
UV-Vis spectroscopy. The spectral changes on addition of H2O
to 1a in CH2Cl2 (and in C6H6) to form 2a (Fig. 4 and S6†) reveal
three isosbestic points.31 Related changes are observed when
acetone and THF solutions of 1a are used, although there are
differences in the isosbestic regions (Fig. S7 and S8†) that arise
because these solvents compete with H2O for the vacant site.
The equilibrium constant K for H2O-binding in CH2Cl2 was
estimated using eqn (2), the concentrations of 1a and 2a being
determined from the absorbance at 678 nm (Fig. 4).
The alcohol complexes
Syntheses of the trans-RuCl2(P–N)(PPh3)(ROH) adducts
(3, R = Me; and 4, R = Et) were complicated as trace moisture
led to formation of the aqua adduct 2a, but the pink complex
3 of good elemental analysis was eventually isolated in 56%
yield from a mixture of RuCl2(PPh3)3 and P–N in a vigorously
dried 2 : 5 blend of MeOH and acetone. The 31P{1H} NMR spec-
trum of isolated 3 in CD2Cl2 was similar to that of 2a
(cf. Fig. S3†), and again implies an equilibrium with 1a but, in
the presence of 50 equiv. of MeOH, both PA and PX signals
were seen as doublets at δ ∼77 and ∼47, respectively, with
2JPP = 36.7 Hz. The 1H NMR data (Fig. S10) are similar to those
of 2a, and are consistent with the trans structure. Isolation of
analytically pure 4 was not achieved by mixing RuCl2(PPh3)3
and P–N over long periods in EtOH-containing, mixed sol-
vents. An uncharacterized, insoluble, brown solid was first
readily isolated and, after a lengthy work-up procedure, a pink
compound was subsequently isolated in 33% yield from the fil-
trate. NMR spectra in CD2Cl2 of this pink solid (e.g. Fig. 5)
showed the absence of impurities and are completely consist-
ent with the EtOH adduct, but the C-analysis is 1.0% high,
almost certainly due to contamination by hexanes used for
washing; attempts to achieve more effective drying were foiled
by accompanying loss of the EtOH, noted visibly by colour
changes. Rapid reversible coordination of the alcohol in solu-
tion is apparent from the 31P{1H} NMR spectrum, and the vari-
able temperature NMR data for 4 again resemble those of 2a.
logf½2aꢀ=½1aꢀg ¼ log K þ log½H2Oꢀ
ð2Þ
A K value of 36 1 M−1 at 25 °C was estimated from the inter-
cept of the log–log plot (Fig. S9†), using data up to the solubi-
lity limit of H2O in CH2Cl2 (∼0.13 M)32 at this temperature, the
value being based on repeat experiments (Appendix B); the
1.06 slope of the plot is consistent with a 1 : 1 equilibrium.
A more approximate K value of ∼10 M−1 was estimated from
1
the H NMR spectra of 2a in CD2Cl2 at 25 °C. UV-Vis analysis
in C6H6 gave a K value of ∼28 M−1 (Appendix B), reasonable
agreement with the 36 M−1 value, implying perhaps that
CH2Cl2 and C6H6 are likely both non-coordinating toward 1a
(or have similar, weak binding properties). K values in CH2Cl2
were measured from 10 to 38 °C, but reproducible values at
the extreme temperatures could not be obtained. Nevertheless,
ΔHo, ΔSo and ΔGo values were estimated to be, respectively,
−50 20 kJ mol−1, −140 40 kJ mol−1, and −8.9 0.2 kJ mol−1
Fig. 4 Spectral changes observed upon addition of H2O to RuCl2(P–N)(PPh3) (1a)
(1.04 × 10−3 M) in CH2Cl2 at 25 °C. Added [H2O] = (a) 0.0, (b) 0.0056, (c) 0.0111,
(d) 0.0333, (e) 0.0500, (f) 0.0666, (g) 0.0999, (h) 0.1110 M.
Fig. 5 1H NMR spectrum (300 MHz) of trans-RuCl2(P–N)(PPh3)(EtOH) (4) in
CD2Cl2 at r.t.
4296 | Dalton Trans., 2013, 42, 4291–4298
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
reversibly binds N2 and N2O to generate cis-species, but the
coordination is weaker than with the oxygen- and sulfur-donor
ligands.2a,39
Conclusions
The ability of square pyramidal complexes of the type trans-
RuCl2(P–N)(PR3), where R is Ph or p-tolyl, to bind small mole-
cules is extended to include water, MeOH and EtOH. The study
of formation of the trans-products, together with previously
reported findings for the binding of H2, N2, N2O, H2S, MeSH
and EtSH, which all form cis-products, uniquely illustrates for-
mation of six-coordinate products from a five-coordinate pre-
cursor with such a range of small molecules. Crystallographic
and thermodynamic data (in both solution and the solid state)
for the reversible equilibria systems allow for creation of
trends describing trans-influence and insight into the bond
strengths of the small molecule ligands. The five-coordinate
precursor also reacts with NH3, CO and acetylenes but these
systems lead to more than a single product, chemistry that will
be described elsewhere.
Fig. 6 DSC curves for trans-RuCl2(P–N)(PPh3)L complexes. Greater error in ΔHo
for 4 results from the inability of obtaining it analytically pure (see text).
The alcohol adducts are isolated only under absolutely an-
hydrous conditions and, even in the solid state under 1 atm
Ar, the complexes lose the solvent to regenerate the green, five-
coordinate 1a.
Stabilization of octahedral RuII–phosphine complexes by
the presence of MeOH and EtOH ligands has been known for
decades, for example. as in RuCl2(EtOH)(PMe2Ph)3,33 [RuH-
(PPh3)2(H2O)2(MeOH)]+,34 [Ru(Y)Cl2(MeOH)(PPh3)2] (Y = CO or
CS),35 [RuH(PMe2Ph)4(MeOH)]+,36 and [RuH(dppe)2(EtOH)]+,36
and our findings now extend this to a P–N coordinated
species. A labile alcohol ligand can play a key role in catalysis,
as exemplified within a Ru-BINAP-MeOH species used for
homogeneous asymmetric hydrogenation.37
Acknowledgements
We thank the Natural Sciences and Engineering Council of
Canada for funding and Colonial Metals Inc. for a loan of
RuCl3·xH2O.
References
Differential scanning calorimetry (DSC) data
The ΔHo values for dissociation of the H2O, MeOH and EtOH
from the respective complexes cis-2a, -3 and -4 to give 1a were
obtained from DSC experiments (Fig. 6). Taken together with
our earlier DSC data for dissociation of the S-ligands from the
trans complexes,1 ΔHo is seen to decrease in the order: MeSH
(94 kJ mol−1) > MeOH (90) > H2S (85) > H2O (75) > EtSH (64) >
EtOH (36). Thus, in the solid state, the S-ligands, which are
trans to Cl, require higher dissociation energy than the corres-
ponding O-ligands, which are trans to PA, a site of strong trans-
influence (see above). The MeSH1 and MeOH adducts are
noticeably more thermally stable than the other complexes,
perhaps because these Me-containing molecules are of the
most compatible size and electronic structure to occupy the
vacant site. The ΔHo value of −75 4 kJ mol−1 for coordination
of H2O to 1a in the solid state compares with the more ques-
tionable value of −50 20 kJ mol−1 estimated in CH2Cl2 solu-
tion (see above); if there is any real difference in these values,
weak bonding of the CH2Cl2 could be a factor, since RuII–
CH2Cl2 complexes are known.38 Finally, DSC data for 2b, the
1 E. S. F. Ma, S. J. Rettig, B. O. Patrick and B. R. James, Inorg.
Chem., 2012, 51, 5427.
2 (a) D. C. Mudalige, S. J. Rettig, B. R. James and
W. R. Cullen, J. Chem. Soc., Chem. Commun., 1993, 830;
(b) D. C. Mudalige, E. S. F. Ma, S. J. Rettig, B. R. James and
W. R. Cullen, Inorg. Chem., 1997, 36, 5426.
3 (a) D. C. Mudalige, PhD thesis, The University of British
Columbia, Vancouver, British Columbia, Canada, 1994;
(b) E. S. F. Ma, PhD thesis, The University of British
Columbia, Vancouver, British Columbia, Canada, 1999.
4 D. D. Perrin, W. L. F. Armarego and D. R. Perrin, Purifi-
cation of Laboratory Chemicals, Pergamon, Oxford, 2nd edn,
1980.
5 P. S. Hallman, T. A. Stephenson and G. Wilkinson, Inorg.
Synth., 1970, 12, 237.
6 P. W. Armit, W. J. Sime, T. A. Stephenson and L. Scott,
J. Organomet. Chem., 1978, 161, 391.
7 d*TREK. Area Detector Software, version 4.13, Molecular
Structure Corp, The Woodlands, TX, 1996–1998.
8 A. Altomare, M. C. Burla, M. Camalli, G. L. Cascarano,
C. Giacovazzo, A. Guagliardi, A. G. G. Moliterni, G. Polidori
and R. Spagna, J. Appl. Crystallogr., 1999, 32, 115.
9 G. M. Sheldrick, Acta Crystallogr., Sect. A: Found. Crystal-
logr., 2008, 64, 112.
p-tolyl analogue of 2a, give a value of 62
2 kJ mol−1 for
removal of the H2O ligand (Fig. S11†), the number being
consistent with the longer Ru–O bond in 2b (2.252 Å) than
in those determined in the structures of the benzene solvates
of 2a (2.229 and 2.189 Å). It should be noted that 1a also
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 4291–4298 | 4297

View Article Online
Paper
Dalton Transactions
10 WinGX-V1.80.05; L. J. Farrugia, J. Appl. Crystallogr., 1999, 26 K. D. Tau and D. W. Meek, Inorg. Chem., 1979, 18, 3574.
32, 837. 27 R. L. Keiter and J. G. Verkade, Inorg. Chem., 1969, 8, 2115.
11 C. Schotes, M. Ranocchiari and A. Mezzetti, Organo- 28 P. G. Jessop, S. J. Rettig, C.-L. Lee and B. R. James, Inorg.
metallics, 2011, 30, 3596. Chem., 1991, 30, 4617.
12 (a) H. J. Lawson, T. S. Janik, M. R. Churchill and 29 (a) K. S. MacFarlane, A. M. Joshi, S. J. Rettig and
K. J. Takeuchi, Inorg. Chim. Acta, 1990, 174, 197;
(b) S. K. Mandal and A. R. Chakravarty, Inorg. Chem., 1993,
32, 3851.
B. R. James, Inorg. Chem., 1996, 35, 7304; (b) S. L. Queiroz,
A. A. Batista, G. Oliva, M. T. Do, P. Gambardella,
R. H. A. Santos, K. S. MacFarlane, S. J. Rettig and
B. R. James, Inorg. Chim. Acta, 1998, 267, 209.
13 S. M. Boniface, G. R. Clark, T. J. Collins and W. R. Roper,
J. Organomet. Chem., 1981, 206, 109.
14 M. S. Röthlisberger, W. Hummel, P.-A. Pittet, H.-B. Bürgi,
A. Ludi and A. E. Merbach, Inorg. Chem., 1988, 27, 1358.
15 D. V. McGrath and R. H. Grubbs, J. Am. Chem. Soc., 1991,
113, 3611.
16 P. Bernhard, H.-B. Bürgi, J. Hauser, H. Lehmann and
A. Ludi, Inorg. Chem., 1982, 21, 3936.
17 U. Kölle, G. Flunkert, R. Görissen, M. U. Schmidt and
U. Englert, Angew. Chem., Int. Ed. Engl., 1992, 31, 440.
18 H. Basch, M. Krauss, W. J. Stevens and D. Cohen, Inorg.
Chem., 1985, 24, 3313.
30 A. J. Deeming, S. Doherty, J. E. Marshall, J. L. Powell and
A. M. Senior, J. Chem. Soc., Dalton Trans., 1993, 1093.
31 The isosbestic points in Fig. 4 are not quite as sharp as
those usually seen on the HP 8452 instrument, but the
excellent log : log plot (Fig. S9†) from the Fig. 4 data has
the correct slope, and the K value is well determined. The
isosbestics seen in Fig. S6–S8† for spectra in C6H6, acetone,
and THF, respectively, are sharper.
32 IUPAC Solubility Data Series, Vol. 60, Halogenated Methanes
with Water, ed. A. L. Horváth and F. W. Getzen, Oxford
University Press, Oxford, 1995, p. 153.
19 S. Shinoda, Y. Koie and Y. Saito, Bull. Chem. Soc. Jpn., 1986, 33 J. Chatt, G. J. Leigh and R. J. Paske, J. Chem. Soc. A, 1969,
59, 2938. 854.
20 E. S. Ma, S. J. Rettig and B. R. James, Chem. Commun., 34 R. J. Young and G. Wilkinson, J. Chem. Soc., Dalton Trans.,
1999, 2463. 1976, 719.
21 A. J. Deeming and G. P. Proud, Inorg. Chim. Acta, 1985, 100, 35 P. W. Armit, W. J. Sime and T. A. Stephenson, J. Chem. Soc.,
223. Dalton Trans., 1976, 2121.
22 A. J. Deeming, G. P. Proud, H. M. Dawes and 36 T. V. Ashworth and E. Singleton, J. Chem. Soc., Chem.
M. B. Hursthouse, J. Chem. Soc., Dalton Trans., 1986, 2545. Commun., 1976, 706.
23 (a) J. Malito and E. C. Alyea, Transition Met. Chem., 1992, 37 C.-C. Chen, T.-T. Huang, C.-W. Lin, R. Cao, A. S. C. Chan
17, 481; (b) G. K. Anderson and R. Kumar, J. Chem. Res. (S), and W. T. Wong, Inorg. Chim. Acta, 1998, 270, 247.
1998, 48.G. K. Anderson and R. Kumar, J. Chem. Res. (M), 38 (a) D. Huang, J. C. Huffman, J. C. Bollinger, O. Eisenstein
1988, 432.
and K. G. Caulton, J. Am. Chem. Soc., 1997, 119, 7398;
(b) J. Zhang, K. A. Barakat, T. R. Cundari, T. B. Gunnoe,
P. D. Boyle, J. L. Petersen and C. S. Day, Inorg. Chem., 2005,
44, 8379.
24 (a) E. Grimley and D. W. Meek, Inorg. Chem., 1986, 25,
2049; (b) P. Brüggeller, Inorg. Chem., 1987, 26, 4125.
25 (a) A. Pidcock, R. E. Richards and L. M. Venanzi, J. Chem.
Soc. A, 1966, 1707; (b) J. J. Gambaro, W. H. Hohman and 39 C. B. Pamplin, E. S. F. Ma, N. Safari, S. J. Rettig and
D. W. Meek, Inorg. Chem., 1989, 28, 4154.
B. R. James, J. Am. Chem. Soc., 2001, 123, 8596.
4298 | Dalton Trans., 2013, 42, 4291–4298
This journal is © The Royal Society of Chemistry 2013