A study of syn-anti isomerism of six-coordinate ruthenium(II) complexes containing PhP(CH2CH2CH2PCy2) 2 (Cyttp) ligand

Article abstract of DOI:10.1016/j.molcata.2004.07.029

A study is reported on synthesis and ligand substitution/modification reactions of the syn and anti geometric isomers of the six-coordinate ruthenium(II) complexes [mer-Ru(κ²-O₂CX)(CO)(Cyttp)] ⁿ (n = 0, X = O (2); n = +1, X = Me (3), Ph (4), OMe (5), OEt (6); Cyttp = PhP(CH₂CH₂CH₂PCy₂) ₂) and cis-mer-RuX₂(CO)(Cyttp) (X = I (7), Cl (8)) (syn (s) and anti (a) refer to the orientation of the Ph group on the central P atom of Cyttp). Reaction of cis-mer-Ru(BF₄)₂(CO)(Cyttp) (1) in freshly prepared aqueous acetone solution with each of sodium carbonate, sodium acetate, and sodium benzoate affords 2a, 3a(BF₄), and 4a(BF ₄), respectively. The syn isomer of 2a, 2s, was synthesized by prolonged treatment of solid mer-Ru(CO)₂(Cyttp) with gaseous O ₂. Complexes 2a and 2s do not interconvert even on heating in toluene or methanol at reflux temperature for 24 h. Alkylation of 2a and 2s with 1 equiv. of [Me₃O]BF₄ or [Et₃O]PF₆ affords the methylcarbonato (5a(BF₄) and 5s(BF₄), respectively) and ethylcarbonato (6a(PF₆) and 6s(PF₆), respectively) complexes. Use of >2 equiv. of [Me₃O]BF₄ still furnishes 5s(BF₄) from 2s, but yields 1 instead of 5a(BF ₄) from 2a. The foregoing reactions represent, to our knowledge, the first example of different reactivity of syn and anti isomers of metal Cyttp complexes. Prolonged treatment of 2a with an excess of MeI at room temperature results in the formation of 7a. A parallel reaction of 2s with MeI to yield 7s requires heating; at ambient temperature, 5s(I) is obtained instead. Each of the complexes 2, 3(BF₄)-5(BF₄), and 6(PF₆) is converted to 8 with retention of the syn or anti configuration by the action of hydrochloric acid. It is concluded from this and previous studies that complexes stable to ligand dissociation do not undergo syn-anti isomerization, in contrast to those that contain weakly coordinated ligands. ¹³C{ ¹H} and ³¹P{¹H} NMR spectroscopic generalizations were developed that enable syn-anti assignment to be made for this class of complexes. The structures of 2a (as 2a·CH ₂Cl₂·2H₂O), 2s (as 2s·(3/4)MeC(O) Me·2H₂O), and 5s (as 5s(BF₄)·C ₆H₆) were determined by single crystal X-ray diffraction analysis. The syn and anti isomers of [mer-Ru(κ²-O ₂CX)(CO)(Cyttp)]ⁿ (n = 0, X = O; n = +1, X = Me, Ph, OMe, OEt) and cis-mer-RuX₂(CO)(Cyttp) (X = I or Cl) were synthesized and found to be configurationally stable with respect to isomerization and upon ligand substitution. This stability contrasts with that of previously studied related ruthenium(II) Cyttp complexes containing weakly coordinating ligands.

Full text of DOI:10.1016/j.molcata.2004.07.029

Journal of Molecular Catalysis A: Chemical 224 (2004) 133–144
A study of syn–anti isomerism of six-coordinate ruthenium(II)
complexes containing PhP(CH2CH2CH2PCy2)2 (Cyttp) ligandଝ
Patrick W. Blosser, Judith C. Gallucci1, Andrew Wojcicki∗
Department of Chemistry, The Ohio State University, 100 W. 18th Avenue, Columbus, OH 43210-1185, USA
Received 5 April 2004; received in revised form 22 July 2004; accepted 26 July 2004
Dedicated to Professor Jo´zef J. Zio´łkowski on the occasion of his 70th birthday.
Abstract
A study is reported on synthesis and ligand substitution/modification reactions of the syn and anti geometric isomers of the six-
coordinate ruthenium(II) complexes [mer-Ru(␬2-O2CX)(CO)(Cyttp)]n (n = 0, X = O (2); n = +1, X = Me (3), Ph (4), OMe (5), OEt (6);
Cyttp = PhP(CH2CH2CH2PCy2)2) and cis-mer-RuX2(CO)(Cyttp) (X = I (7), Cl (8)) (syn (s) and anti (a) refer to the orientation of the Ph
group on the central P atom of Cyttp). Reaction of cis-mer-Ru(BF4)2(CO)(Cyttp) (1) in freshly prepared aqueous acetone solution with
each of sodium carbonate, sodium acetate, and sodium benzoate affords 2a, 3a(BF4), and 4a(BF4), respectively. The syn isomer of 2a, 2s,
was synthesized by prolonged treatment of solid mer-Ru(CO)2(Cyttp) with gaseous O2. Complexes 2a and 2s do not interconvert even on
heating in toluene or methanol at reflux temperature for 24 h. Alkylation of 2a and 2s with 1 equiv. of [Me3O]BF4 or [Et3O]PF6 affords the
methylcarbonato (5a(BF4) and 5s(BF4), respectively) and ethylcarbonato (6a(PF6) and 6s(PF6), respectively) complexes. Use of >2 equiv. of
[Me3O]BF4 still furnishes 5s(BF4) from 2s, but yields 1 instead of 5a(BF4) from 2a. The foregoing reactions represent, to our knowledge,
the first example of different reactivity of syn and anti isomers of metal Cyttp complexes. Prolonged treatment of 2a with an excess of MeI
at room temperature results in the formation of 7a. A parallel reaction of 2s with MeI to yield 7s requires heating; at ambient temperature,
5s(I) is obtained instead. Each of the complexes 2, 3(BF4)–5(BF4), and 6(PF6) is converted to 8 with retention of the syn or anti configuration
by the action of hydrochloric acid. It is concluded from this and previous studies that complexes stable to ligand dissociation do not undergo
syn–anti isomerization, in contrast to those that contain weakly coordinated ligands. 13C{1H} and 31P{1H} NMR spectroscopic generalizations
were developed that enable syn–anti assignment to be made for this class of complexes. The structures of 2a (as 2a·CH2Cl2·2H2O), 2s (as
2s·(3/4)MeC(O)Me·2H2O), and 5s (as 5s(BF4)·C6H6) were determined by single crystal X-ray diffraction analysis.
© 2004 Elsevier B.V. All rights reserved.
Keywords: Ruthenium complexes; Triphosphine complexes; Carbonato complexes; syn–anti isomerism; Ligand substitution; Crystal structures
1. Introduction
ଝ
Crystallographic data for the structural analysis have been de-
posited with the Cambridge Crystallographic Data Centre, CCDC
nos. 228319, 227599, and 227406 for compounds 2a·CH2Cl2·2H2O,
2s·(3/4)MeC(O)Me·2H2O, and 5s(BF4)·C6H6, respectively. Copies of the
data can be obtained free of charge from The Director, CCDC, 12
Union Road, Cambridge CB2 1EZ, UK (Fax: +44-1223-336-033; e-mail:
ma∗ilto:deposit@ccdc.cam.ac.uk or http://www.ccdc.cam.ac.uk/).
Corresponding author. Tel.: +1 614 292 4750; fax: +1 614 292 1685.
E-mail addresses: gallucci.1@osu.edu (J.C. Gallucci),
Transition-metal complexes with polydentate phosphine
ligands have attracted considerable attention in the past few
decades largely because of their diverse structures and cat-
alytic properties [1,2]. Such ligands provide an important
advantage over monodentate phosphine ligands with respect
to greater control of the coordination number, stoichiometry,
and stereochemistry of their complexes.
wojcicki@chemistry.ohio-state.edu (A. Wojcicki).
Linear polydentate phosphines of the type R2P
((CH2)nP)m(CH2)nPR2 have been extensively utilized in
1
To whom inquiries concerning the X-ray crystallographic studies should
be addressed. Fax: +1 614 292 1685.
1381-1169/$ – see front matter © 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcata.2004.07.029

134
P.W. Blosser et al. / Journal of Molecular Catalysis A: Chemical 224 (2004) 133–144
the above context [1,3,4]. A representative member of this
class, the tridentate PhP(CH2CH2CH2PCy2)2 (Cyttp, Cy = c-
C6H11), was prepared by Meek and coworkers [5], who also
synthesized a number of its metal complexes and investi-
gated their chemistry and possible catalytic applications [3].
Ruthenium complexes of Cyttp, e.g., RuH2(␩2-H2)(Cyttp)
and RuHCl(Cyttp), show reactivities toward alkenes [6],
alkynes [7,8], and CO2-like species [9] that furnish infor-
mation of relevance to catalytic processes. Later studies in
our group resulted in the synthesis of RuX2(CO)(Cyttp)
(X = BF4 or OSO2CF3) with two weakly coordinated an-
ions; these complexes serve as useful precursors of a va-
riety of ruthenium(II) complexes [10]. A rare example of
a hydrido-oxo metal complex, [ReH2(O)(Cyttp)]+, which
transfers both hydrogen and oxygen to some substrates and
catalyzes hydration of nitriles to amides, was also reported
[11,12].
Among the tridentate phosphines investigated, Cyttp
shows exceptional propensity for meridional coordination
around the metal. The trimethylene linkage between phos-
phorus atoms of Cyttp provides an ideal “chelate-bite” angle
of ca. 90◦ for a larger transition metal such as ruthenium(II).
As a result of this feature and of the steric demands of the ter-
minal CH2PCy2 donors, meridional structures are strongly
favored. The consequence is greater control of the coordina-
tion geometry in metal complexes as compared to monoden-
tate phosphine analogues.
species that result from different orientations of the NH hy-
drogen of dien [16,17].
There have been few studies on syn–anti isomerism of
six-coordinate metal Cyttp complexes. Both isomers were
synthesized for each of cis-mer-IrH2Cl(Cyttp) [18,19],[cis-
mer-IrH2(MeCN)(Cyttp)]+ [19], cis-mer-RuH2(P(OMe)3)
(Cyttp) [20], cis-mer-RuH2(P(OPh)3)(Cyttp) [20], and mer-
RuCl(␩3-PhCH C C CPh)(Cyttp)[7,8], someasasyn–anti
mixture that was not separated. Characterization of at least
one isomer in each case was effected by 1H NMR nOe
spectroscopy or X-ray diffraction techniques. Several other
complexes of this general type were obtained as a single
isomer and characterized similarly as syn or anti [8,10,
19,20].
In this paper, we report further investigations into syn–anti
isomerism of ruthenium(II) Cyttp complexes. In particular,
we address questions concerning reactivity of such com-
plexes in isomerization and ligand substitution/modification
processes. Both syn and anti isomers of mer-Ru(␬2-O2CO)
(CO)(Cyttp) (2) were synthesized and completely character-
ized, andtheirstereochemical behaviorinthe aforementioned
types of transformation was investigated.
2. Experimental
2.1. General procedures and measurements
In Cyttp complexes that are six-coordinate and contain
different trans ligands (cis to all three P donors) as, for ex-
ample, in cis-mer-ML2Lꢀ(Cyttp), the Ph group on the central
phosphorus atom can point to either one of them (L or Lꢀ),
thus giving rise to syn and anti geometric isomers (cf. I and
II).2
Reactions and manipulations of air-sensitive compounds
were conducted under an atmosphere of dry argon by use
of standard procedures [21]. Solvents were dried [22], dis-
tilled under argon, and degassed before use. Elemental anal-
yses were carried out by M-H-W Laboratories, Phoenix, AZ.
IR, NMR (1H, 13C, and 31P), and mass spectra (FAB MS)
wereobtainedaspreviouslydescribed[23,24]. Alllistedmass
peaks are those of ions containing 102Ru. Relative peak in-
tensities are given with the assignments.
2.2. Materials
Reagents were procured from various commercial
sources and used as received. The complexes cis-mer-anti-
RuI2(CO)(Cyttp) (7a), cis-mer-anti-RuCl2(CO)(Cyttp) (8a),
and cis-mer-Ru(BF4)2(CO)(Cyttp) (1) were prepared as re-
ported earlier [10]. (Cyttp = PhP(CH2CH2CH2PCy2)2; mer
refers to Cyttp, and cis refers to I2, Cl2, etc.; the ref-
erence center for syn and anti is Ph on central P, cf.
Section 1.)
Even when these trans ligands are identical, in some cases
their inequivalence can be detected by NMR spectroscopy
[12]. This type of syn–anti isomerism has been observed also
for meridional ttp (ttp = PhP(CH2CH2CH2PPh2)2) and re-
lated tridentate phosphine complexes [14,15]. Furthermore,
it is not limited to phosphines. Particularly noteworthy are
recent reports of synthesis and characterization of all pos-
sible geometric isomers of the amine complexes [Co(dien)
(ibn)Cl]2+ (dien = diethylenetriamine, ibn = 1,2-diamino-2-
methylpropane) and [Co(dien)(ampy)Cl]2+ (ampy = 2-ami-
nomethylpyridine), including the meridional syn and anti
2.3. Synthesis of carbonato, carboxylato, and
alkylcarbonato complexes of Ru(CO)(Cyttp)
2.3.1. Preparation of
mer-syn-Ru(␬2-O2CO)(CO)(Cyttp) (2s)
2
The designation syn and anti is made in accordance with the
The procedure employed is a slight modification of that
reported in the literature [25]. A powdery sample of known
Cahn–Ingold–Prelog priority rules [13] as applied to the two ligands that
are mutually trans.

P.W. Blosser et al. / Journal of Molecular Catalysis A: Chemical 224 (2004) 133–144
135
2.3.4. Preparation of [mer-anti-Ru(␬2-O2CPh)(CO)
(Cyttp)]BF4 (4a(BF4))
[25]mer-Ru(CO)2(Cyttp)(0.562 g, 0.755 mmol)inaSchlenk
flask was exposed to a stream of O2 gas for 10 days, dur-
ing which time a gradual color change from pale yellow to
slightly gray was observed. The solid was wetted with Et2O
(30 ml), collected on a filter frit, and washed with benzene
(10 ml)toremoveanyremaining mer-Ru(CO)2(Cyttp)and/or
possible 2a co-product. (Note: Whereas the solubility of 2s
in benzene is only ca. 0.025 g/10 ml, that of 2a is greater than
0.25 g/10 ml.) Rinsing with Et2O (20 ml) and drying under
vacuum afforded 2s as a white solid in 86% yield (0.506 g).
Selected spectroscopic data. 13C{1H} NMR (CD2Cl2): δ
The product was prepared from cis-mer-Ru(BF4)2(CO)
(Cyttp) (1) (0.202 g, 0.227 mmol) and sodium benzoate
(0.055 g, 0.38 mmol) by a procedure analogous to that for
3a(BF4). Yield of 4a(BF4), 0.180 g (86%). Selected spec-
troscopic data. IR (Nujol-hexachlorobutadiene): ν(CO) 1935
(vs), ν(O2C) 1501 (m), 1495 (s), 1450 (s), 1435 (vs) cm−1
.
13C{1H} NMR (CD2Cl2): δ (ppm) 206.7 (dt, 2JPcC = 16 Hz,
2JPwC = 12 Hz, CO), 181.0 (q, JPC = 2 Hz, O2C), 38.6 (t,
3
JPwC = 12 Hz, CH of Cy), 34.4 (t, JPwC = 9.5 Hz, CH of Cy).
31P{1H} NMR (CD2Cl2): δ (ppm) 40.5 (t, 2JPcPw = 30.1 Hz,
Pc), 17.3 (d, 2JPcPw = 30.1 Hz, Pw). Anal. Found: C,
57.24; H, 7.16%. Calc. for C44H66BF4O3P3Ru: C, 57.21;
H, 7.20%.
2
2
(ppm) 205.2 (dt, JPcC = 17 Hz, JPwC = 12 Hz, CO), 166.6
3
(q, JPC ∼ 1 Hz, O2CO), 35.1 (t, JPwC = 11 Hz, CH of Cy),
34.0 (t, JPwC = 9 Hz, CH of Cy). 13P{1H} NMR (CD2Cl2): δ
(ppm) 22.5 (t, 2JPcPw = 30.5 Hz, Pc, 11.5 (d, 2JPcPw = 30.5 Hz,
Pw). MS (FAB): m/z 777 (M+, 14), 733 (M+ − CO2,
100).
2.3.5. Preparation of [mer-anti-Ru(␬2-O2COMe)(CO)
(Cyttp)]BF4 (5a(BF4))
A stirred solution of mer-anti-Ru(␬2-O2CO)(CO)(Cyttp)
(2a) (0.140 g, 0.180 mmol) in 70 ml of CH2Cl2 was treated
with [Me3O]BF4 (0.027 g, 0.182 mmol). After 1 h of stir-
ring, the solvent was evaporated, and the ivory residue was
washed consecutively with H2O (20 ml) and hexane (30 ml)
and dried overnight. Yield of 5a(BF4), 0.130 g (82%). Se-
lected spectroscopic data. IR (Nujol-hexachlorobutadiene):
ν(CO) 1940 (vs), ν(O2CO) 1505 (m), 1450 (s), 1435 (sh),
2.3.2. Preparation of mer-anti-Ru(␬2-O2CO)
(CO)(Cyttp) (2a)
A solid mixture of cis-mer-Ru(BF4)2(CO)(Cyttp) (1)
(0.428 g, 0.481 mmol) and Na2CO3 (0.065 g, 0.61 mmol) was
dissolved in acetone (40 ml)/H2O (25 ml), and the contents
were stirred for 30 min. The volume of solution was reduced
to ca. 20 ml to result in the formation of a white precipi-
tate. The solid was collected on a filter frit, washed twice
with 10-ml portions of H2O and then with 10-ml portions
of hexane, and dried under vacuum overnight. Yield of 2a,
0.353 g (95%). IR (Nujol-hexachlorobutadiene): ν(CO) 1918
1
ν(COC) 1090 (s) cm−1. H NMR (CD2Cl2): δ (ppm) 3.84
(s, 3 H, OMe). 13C{1H} NMR (CD2Cl2): δ (ppm) 206.5
(dt, 2JPcC = 12 Hz, JPwC = 11 Hz, CO), 161.3 (s, O2CO),
2
55.0 (s, OMe), 38.4 (t, JPwC = 12 Hz, CH of Cy), 34.5 (t,
JPwC = 9 Hz, CH of Cy). 31P{1H} NMR (CD2Cl2): δ (ppm)
40.1 (t, 2JPcPw = 29.8 Hz, Pc), 15.9 (d, 2JPcPw = 29.8 Hz, Pw).
MS (FAB): m/z 792 (M+, 73), 718 (M+ + H − O2COMe,
100).
When this reaction was carried out with more than 2 equiv.
of [Me3O]BF4, the only phosphorus-containing species de-
tected in solution by 31P{1H} NMR spectroscopy was cis-
mer-Ru(BF4)2(CO)(Cyttp) (1).
(vs), ν(O2CO) 1668 (m), 1603 (s), 1235 (m) cm−1
.
13C{1H}
2
NMR (CD2Cl2): δ (ppm) 209.0 (q, JPcC = 2JPwC = 12 Hz,
CO), 166.2 (s, O2CO), 38.5 (t, JPwC = 11 Hz, CH of Cy),
33.2 (t, JPwC = 9 Hz, CH of Cy). 31P{1H} NMR (CD2Cl2): δ
(ppm) 31.6 (t, JPcPw = 31.9 Hz, Pc), 15.4 (d, JPcPw = 31.9 Hz,
Pw). MS (FAB): m/z 777 (M+, 5), 733 (M+ − CO2, 100). Anal.
Found: C, 58.71; H, 8.07%. Calc. for C38H61O4P3Ru: C,
58.82; H, 7.92%.
2.3.6. Preparation of [mer-syn-Ru(␬2-O2COMe)(CO)
(Cyttp)]BF4 (5s(BF4))
2.3.3. Preparation of [mer-anti-Ru(␬2-O2CMe)(CO)
(Cyttp)]BF4 (3a(BF4))
A procedure analogous to that in Section 2.3.5 was em-
ployed using [Me3O]BF4 (0.048 g, 0.324 mmol) and mer-
syn-Ru(␬2-O2CO)(CO)(Cyttp) (2s) (0.100 g, 0.129 mmol).
Yield of 5s(BF4), 0.101 g (89%). Selected spectroscopic data.
IR (Nujol-hexachlorobutadiene): ν(CO) 1960 (vs), ν(O2CO)
The product was obtained from cis-mer-Ru(BF4)2
(CO)(Cyttp) (1) (0.200 g, 0.225 mmol) and sodium acetate
(0.050 g, 0.61 mmol) by a procedure analogous to that for
2a (Section 2.3.2). Yield of 3a(BF4), 0.161 g (83%). Se-
lected spectroscopic data. IR (Nujol-hexachlorobutadiene):
ν(CO) 1940 (vs), ν(O2C) 1522 (m), 1455 (s), 1446
1
1540 (m), 1490 (s), 1385 (s), ν(COC) 1088 (s) cm−1. H
NMR (CD2Cl2): δ (ppm) 3.83 (s, 3 H, OMe). 13C{1H} NMR
(s) cm−1
.
1H NMR (CD2Cl2): δ (ppm) 2.03 (s, 3
2
2
H, Me). 13C{1H} NMR (CD2Cl2): δ (ppm) 206.6 (dt,
2JPcC = 15.6 Hz, 2JPwC = 11.5 Hz, CO), 187.1 (s, O2C), 38.7
(t, JPwC = 11.9 Hz, CH of Cy), 35.0 (t, JPwC = 9.5 Hz, CH of
Cy), 25.2 (s, Me). 31P{1H} NMR (CD2Cl2): δ (ppm) 40.0 (t,
(CD2Cl2): δ (ppm) 203.1 (dt, JPcC = 21 Hz, JPwC = 11 Hz,
CO), 161.3 (s, O2CO), 55.0 (s, OMe), 36.1 (t, JPwC = 11 Hz,
CH of Cy), 34.5 (t, JPwC = 10 Hz, CH of Cy). 31P{1H} NMR
2
(CD2Cl2): δ (ppm) 30.9 (t, JPcPw = 28.5 Hz, Pc), 11.2 (d,
2JPcPw = 28.5 Hz, Pw). MS (FAB): m/z 791 (M+ − H, 100),
717 (M+ − O2COMe, 97). Anal. Found: C, 56.07; H, 7.11%.
Calc. for C39H64BF4O4P3Ru·C6H6: C, 56.55; H, 7.38%.
2JPcPw = 30.4 Hz, Pc), 17.2 (d, JPcPw = 30.4 Hz, Pw). Anal.
2
Found: C, 54.46; H, 7.27%. Calc. for C39H64BF4O3P3Ru: C,
54.36; H, 7.49%.

136
P.W. Blosser et al. / Journal of Molecular Catalysis A: Chemical 224 (2004) 133–144
2.3.7. Preparation of [mer-anti-Ru(␬2-O2COEt)(CO)
(Cyttp)]PF6 (6a(PF6))
0.129 mmol) in 40 ml of THF, and the mixture was heated
at reflux temperature for 10 h. The volatiles were then re-
moved, hexane (20 ml) was added to the residue, and the
yellow solid was filtered off and dried under vacuum. Yield
of 7s, 0.080 g (64%). Selected spectroscopic data. IR (Nu-
A procedure analogous to that in Section 2.3.5 was em-
ployed using [Et3O]PF6 (0.027 g, 0.109 mmol) and mer-
anti-Ru(␬2-O2CO)(CO)(Cyttp) (2a) (0.076 g, 0.098 mmol).
Yield of 6a(PF6), 0.079 g (86%). Selected spectroscopic data:
IR (Nujol-hexachlorobutadiene): ν(CO) 1950 (vs), ν(O2CO)
jol): ν(CO) 1940 (vs) cm−1
.
13C{1H} NMR (CDCl3): δ
(ppm) 199.1 (dt, 2JPcC = 13 Hz, 2JPwC = 12 Hz, CO), 40.9 (t,
1535 (m), 1482 (s), 1385 (m), ν(COC) 1075 (m) cm−1
.
J
PwC = 11 Hz, CH of Cy), 38.6 (t, JPwC = 11 Hz, CH of Cy).
31P{1H} NMR (CDCl3): δ (ppm) 5.4 (t, JPcPw = 27.2 Hz,
2
1H NMR (CDCl3): δ (ppm) 4.27 (q, JHH = 7.1 Hz, 2 H,
3
2
3
OCH2), 1.35 (t, JHH = 7.1 Hz, 3 H, Me). 31P{1H} NMR
Pc), −5.6 (t, JPcPw = 27.2 Hz, Pw). Anal. Found: C,
2
46.02; H, 6.44%. Calc. for C37H61I2OP3Ru: C, 45.83;
H, 6.34%.
(CDCl3): δ (ppm) 39.8 (t, JPcPw = 30.3 Hz, Pc), 15.9 (d,
2JPcPw = 30.3 Hz, Pw). MS (FAB): m/z 805 (M+ − H, 100),
717 (M+ − O2COEt, 72).
When this reaction was carried out in THF/toluene at
room temperature for 16 h, the isolated product was [mer-
syn-Ru(␬2-O2COMe)(CO)(Cyttp)] I (5s(I)), as determined
by 1H and 31P{1H} NMR spectroscopy.
2.3.8. Preparation of [mer-syn-Ru(␬2-O2COEt)(CO)
(Cyttp)]PF6 (6s(PF6))
The title complex was prepared from mer-syn-Ru(␬2-
O2CO)(CO)(Cyttp) (2s) (0.110 g, 0.142 mmol) and [Et3O]
PF6 (0.040 g, 0.161 mmol)byaprocedureanalogoustothatin
Section 2.3.5. Yield of 6s(PF6), 0.122 g (92%). Selected spec-
troscopic data. IR (Nujol-hexachlorobutadiene): ν(CO) 1950
(vs), ν(O2CO) 1548 (s), 1490 (s), 1382 (m), ν(COC) 1080 (m)
cm−1. 1H NMR (CDCl3): δ (ppm) 4.25 (q, 3JHH = 7.1 Hz, 2
H, OCH2), 1.37 (t, 3JHH = 7.1 Hz, 3 H, Me). 31P{1H} NMR
2.5. Reactions of carbonato, carboxylato, and
alkylcarbonato complexes of Ru(CO)(Cyttp) with HCl
2.5.1. Preparation and characterization of
cis-mer-syn-RuCl2(CO)(Cyttp) (8s)
Concentrated hydrochloric acid (0.25 ml, ca. 3 mmol)
was added to a stirred solution of mer-syn-Ru(␬2-O2CO)
(CO)(Cyttp) (2s) (0.270 g, 0.348 mmol) in 30 ml of CH2Cl2.
Effervescence occurred immediately, and after 5 min the vol-
ume of solution was reduced to ca. 1 ml. Pentane (25 ml)
was added with stirring, and the white solid was fil-
tered off, washed with pentane (20 ml), and dried under
vacuum overnight. Yield of 8s, 0.160 g (58%). Selected
2
(CDCl3): δ (ppm) 30.7 (t, JPcPw = 28.4 Hz, Pc), 11.2 (d,
2JPcPw = 28.4 Hz, Pw). MS (FAB): m/z 805 (M+ − H, 100),
717 (M+ − O2COEt, 11).
2.4. Reactions of carbonato complexes of
Ru(CO)(Cyttp) with MeI
spectroscopic data. IR (Nujol): ν(CO) 1930 (vs) cm−1
.
13C{1H} NMR (CD2Cl2): δ (ppm) 200.2 (dt, 2JPcC = 17 Hz,
2JPwC = 12 Hz, CO), 33.1 (t, JPwC = 10 Hz, CH of Cy), 32.5
(t, JPwC = 9 Hz, CH of Cy). 31P{1H} NMR (CD2Cl2): δ
(ppm) 11.3 (t, 2JPcPw = 30.0 Hz, Pc), 4.6 (d, 2JPcPw = 30.0 Hz,
Pw). Anal. Found: C, 56.27; H, 7.64; Cl, 8.89%. Calc. for
C37H61Cl2OP3Ru: C, 56.48; H, 7.81; Cl, 9.01%.
2.4.1. Preparation of cis-mer-anti-RuI2(CO)(Cyttp) (7a)
Neat MeI (0.25 ml, 0.57 g, 4.0 mmol) was added to a
stirred solution of mer-anti-Ru(␬2-O2CO)(CO)(Cyttp) (2a)
(0.105 g, 0.135 mmol) in 50 ml of benzene (and C6D6) at
room temperature, and progress of reaction was monitored
by 31P{1H} NMR spectroscopy. After a few minutes, new
signals were observed at δ (ppm) 10.8 (t) and −0.9 (d) with
2.5.2. Reactions of other Ru(CO)(Cyttp)
complexes with HCl
1
2JPP = 31.5 Hz. In the H NMR spectrum (C6D6 solution),
In a typical reaction, mer-anti-Ru(␬2-O2CO)(CO)(Cyttp)
(2a) (0.020 g, 0.026 mmol) was dissolved in 0.5 ml of
CH2Cl2 in a 5-mm NMR tube, and concentrated hydrochlo-
ric acid (0.05 ml, ca. 0.6 mmol) was added. After the tube
was shaken for about 2 min, a 31P{1H} NMR spectrum
was recorded. The only phosphorus-containing species de-
tected was cis-mer-anti-RuCl2(CO)(Cyttp) (8a) [10]. The
same product was obtained by starting with each of [mer-
anti-Ru(␬2-O2CMe)(CO)(Cyttp)]BF4 (3a(BF4)), [mer-anti-
Ru(␬2-O2CPh)(CO)(Cyttp)]BF4 (4a(BF4)), [mer-anti-Ru
(␬2-O2COMe)(CO)(Cyttp)]BF4 (5a(BF4)), and [mer-anti-
Ru(␬2-O2COEt)(CO)(Cyttp)]PF6 (6a(PF6)). Similarly, re-
actions of concentrated hydrochloric acid with each
of [mer-syn-Ru(␬2-O2COMe)(CO)(Cyttp)]BF4 (5s(BF4))
and [mer-syn-Ru(␬2-O2COEt)(CO)(Cyttp)]PF6 (6s(PF6))
afforded cis-mer-syn-RuCl2(CO)(Cyttp) (8s).
an additional resonance appeared at δ (ppm) 3.78 (s) and has
been assigned to the same intermediate. With time, 31P{1H}
resonances of 7a [10] grew in, and those of the reaction in-
termediate decreased in intensity. After 48 h, solvent was re-
moved under vacuum, and the residue was washed with hex-
ane (10 ml). The bright yellow solid was filtered off and dried
under vacuum. Yield of 7a, 0.092 g (70%). Selected spec-
troscopic data. 13C{1H} NMR (CDCl3): δ (ppm) 201.2 (q,
2JPcC = 2JPwC = 9 Hz, CO), 42.1 (t, JPwC = 11 Hz, CH of Cy),
38.1 (t, JPwC = 11 Hz, CH of Cy). 31P{1H} NMR (CDCl3): δ
(ppm) 7.8 (t, 2JPcPw = 30.5 Hz, Pc), −8.4 (t, 2JPcPw = 30.5 Hz,
Pw).
2.4.2. Preparation of cis-mer-syn-RuI2(CO)(Cyttp) (7s)
Neat MeI (0.25 ml, 0.57 g, 4.0 mmol) was added to a so-
lution of mer-syn-Ru(␬2-O2CO)(CO)(Cyttp) (2s) (0.100 g,

P.W. Blosser et al. / Journal of Molecular Catalysis A: Chemical 224 (2004) 133–144
137
2.6. Crystallographic analyses
to 0.76(3) for C(2A), with C(2A) constrained at 0.24. There
is a solvent molecule of acetone, which is refined with an
occupancy factor of 0.75. There also appear to be three H2O
molecules in the asymmetric unit, and one of them is lo-
cated on a twofold axis. The occupancy factors for all three
H2O molecules were refined and set at 0.79(3) for O(6) and
0.27(3) for O(8). The occupancy factor for the third H2O
oxygen, O(7), refined to a value of 1.02, and was fixed at 1.0
for the final cycles. Because of the large thermal parameters
for the acetone and H2O, these molecules were only refined
isotropically, and no hydrogen atoms were added.
All data were collected on a Rigaku AFC5S diffractometer
at room temperature using Mo K␣ radiation. Cell constants
were determined by a least-squares fit of the diffractometer
setting angles for 25 reflections in the 2θ range 20–30◦. All
calculations were done with the TEXSAN package [26]. Full
matrix least-squares refinement was based on F. Hydrogen
atoms are included in the models as fixed contributions in cal-
˚
culated positions with the assumption that C H = 0.98 A and
B(H) = Beq(C)·1.2. Scattering factors for neutral atoms and
anomalous dispersion terms for non-hydrogen atoms were
used [27]. A summary of the crystal data and the details of
the intensity data collection and refinement are provided in
Table 1.
2.6.3. Analysis of [mer-syn-Ru(␬2-O2COMe)(CO)
(Cyttp)]BF4·C6H6 (5s(BF4)·C6H6)
The data collection crystal, a clear, colorless rectangular
rod, was cut from a larger crystal grown from a CH2Cl2/C6H6
solution. Six standard reflections were measured after every
150 reflections and indicated that a very small amount of
crystal decomposition occurred during data collection. On
the average, the intensities of the standards decreased by 1%,
and the data were corrected for this small amount of decay. A
correction for absorption was made by the empirical ψ scan
method [28].
2.6.1. Analysis of mer-anti-Ru(␬2-O2CO)
(CO)(Cyttp)·CH2C12·2H2O (2a·CH2Cl2·2H2O)
The data collection crystal was a clear, colorless rod grown
from a CH2Cl2/hexane solution. During data collection, six
standard reflections were measured after every 150 reflec-
tions and indicated that a small amount of decomposition
was occurring. The standards decreased in intensity by an
average value of 6%. A linear decay correction was applied
to the data. A series of ψ scans indicated that an absorption
correction was unnecessary.
The asymmetric unit contains complex 5s, a BF4− ion, and
a benzene molecule of solvation. Several regions of this struc-
ture appear to be disordered. One of the trimethylene bridges,
C(4) C(5) C(6), appears to have two positions for atom
C(5). These are introduced into the model as atoms C(5A)
and C(5B), with occupancy factors of 0.62 and 0.38, respec-
tively, based on their peak heights from a difference electron
The asymmetric unit contains complex 2a, a disordered
moleculeofCH2Cl2, andtwoH2Omolecules. Thedisordered
CH2Cl2 was modeled by two orientations of the molecule
with a common carbon atom. The site occupation factor for
the primary orientation refined to a final value of 0.60(1).
The disorder model was refined only isotropically. No hy-
drogen atoms were added to the CH2Cl2 molecule or to the
two H2O molecules. The top five peaks in the final differ-
−
density map. The BF4 ion is disordered, with two orien-
tations for this group related by rotation about the B F(1)
bond. The B and F(1) atoms each have occupancy factors of
1.0; the occupancy factor for F(2A), F(3A), and F(4A) was
refined to 0.71(1), and this then fixes the occupancy factor
for F(2B), F(3B), and F(4B) (the second orientation) at 0.29.
ence electron density map range from 0.76 to 1.54 e A−3 and
˚
are all in the immediate vicinity of the disordered CH2Cl2
molecule.
−
The BF4 ion was refined only isotropically. The benzene
2.6.2. Analysis of mer-syn-Ru(␬2-O2CO)(CO)
molecule of solvation appears to be disordered, since it ac-
quires very anisotropic atomic displacement parameters. No
attempt was made to model this disorder.
(Cyttp)·(3/4)MeC(O)Me·2H2O (2s·(3/4)MeC(O)
Me·2H2O)
The data collection crystal was initially a clear, colorless
chunk, which had been cut from a much larger crystal grown
from an acetone/hexane layered solvent system. Six standard
reflections were measured after every 150 reflections during
data collection and exhibited a wide range of intensity vari-
ation. Two of the standards decreased in intensity by about
30%, while the decrease for the other four standards ranged
from 8.5 to 13.6%. A linear decay correction was applied to
the data based on the latter four standards. Upon visual in-
spection at the end of data collection, the crystal appeared
cloudy and striated.
3. Results
3.1. Synthesis and reaction chemistry
The starting point of this investigation was synthesis of
syn and anti isomers of mer-Ru(␬2-O2CO)(CO)(Cyttp) (2s
and 2a, respectively). These preparative studies were then
extended to the related carboxylato and alkylcarbonato com-
plexes [mer-Ru(␬2-O2CR)(CO)(Cyttp)]+ (R = Me (3), Ph
(4)) and [mer-Ru(␬2-O2COR)(CO)(Cyttp)]+ (R = Me (5), Et
(6)), respectively. Carbonato, carboxylato, and alkylcarbon-
ato ligand substitution reactions with chloride and iodide
were also probed with respect to stereochemical changes.
The structure suffers from much disorder. One of the
trimethylene groups is disordered, with atom C(2) occupy-
ing two positions, labeled as C(2A) and C(2B). Atom C(2B)
was refined only isotropically. The occupancy factor refined

138
P.W. Blosser et al. / Journal of Molecular Catalysis A: Chemical 224 (2004) 133–144
Table 1
Summary of crystal data, data collection, and structure refinement parameters for 2a·CH2Cl2·2H2O, 2s·(3/4)MeC(O)Me·2H2O, and 5s(BF4)·C6H6
Compound
2a·CH2Cl2·2H2O
2s·(3/4)MeC(O)Me·2H2O
5s(BF4)·C6H6
C39H64BF4O4P3Ru·C6H6
955.84
Monoclinic
P21/n
Empirical formula
Formula weight
Crystal system
Space group
C
38H61O4P3Ru·CH2Cl2·2H2O
C38
H
61O4P3Ru·3/4C3H6O·2H2O
896.85
Triclinic
P
856.46
Monoclinic
C2/c
˚
a (A)
14.231(3)
17.364(3)
9.230(2)
93.10(2)
93.44(2)
104.83(2)
2195.1(8)
2
21.517(6)
18.714(3)
22.757(3)
14.464(2)
16.104(2)
20.822(2)
˚
b (A)
˚
c (A)
α (◦)
β (◦)
γ (◦)
101.85(1)
104.29(1)
3
˚
V (A )
8968(5)
8
1.27
4699.8(7)
4
1.35
Z
Dcalc (g cm−3
)
1.36
Crystal size (mm3)
Linear abs. coeff. (cm−1
2θ limits (◦)
0.23 × 0.27 × 0.38
0.23 × 0.31 × 0.35
0.12 × 0.35 × 0.42
)
6.20
4.88
4.80
4 ≤ 2θ ≤ 55
4 ≤ 2θ ≤ 50
4 ≤ 2θ ≤ 55
Data collected
Scan type
No. of unique data
+h, k,
ω − 2θ
10066
6301
l
+h, +k,
ω
l
+h, +k,
ω
l
8136
3335
448
0.060
0.065
1.60
11170
5674
519
0.048
0.051
1.40
No. of unique data with Fo2 > 3σ(Fo2)
Final no. of variables
R(F)a
454
0.055
0.067
1.89
1.54, −0.91
Rw(F)b
Error in observation of unit weight (e)
Maximum and minimum peaks in final map (e A
−3
˚
)
0.77, −0.50
0.65, −0.81
ꢀ
ꢀ
a
R(F) = ||Fo| − |Fc|| / |Fo|.
ꢀ
ꢀ
1/2
2
Rw(F) = [ w(|Fo| − |Fc|)2/ w|Fo| ] with w = 1/σ2(Fo).
b
imental procedure (cf. Section 2.3.1), we isolated pure 2s as
a white solid in very good yield. Its syn structure was estab-
lished by X-ray diffraction techniques (vide infra). Attempts
at isomerization of 2s by heating in toluene or methanol at
reflux for 24 h showed no evidence of such a transformation.
The selective formation of 2s upon oxidation of solid mer-
Ru(CO)2(Cyttp) by O2 may be a result of the latter complex’s
crystal packing. It is possible that a molecule of O2 can ap-
proach only one carbonyl ligand owing to Cyttp substituents
protecting the other CO from attack leading to oxidation.
The anti isomer 2a was synthesized by reaction of cis-
mer-Ru(BF4)2(CO)(Cyttp) with Na2CO3 in acetone solution
at room temperature (Scheme 1). Like 2s, it also shows stabil-
ity to syn–anti isomerization under comparable conditions.
However, to obtain isomerically pure 2a from 1 and Na2CO3,
it is necessary to dissolve the reactants simultaneously (as a
solid mixture) or to add solid 1 to Na2CO3 in solution. When a
solution of 1 in acetone was stirred for 15 min and then treated
with Na2CO3, both 2a (ca. 85%) and 2s (ca. 15%) were ob-
served by 31P{1H} spectroscopy. The experiment shows that
at the time of addition of Na2CO3, a mixture of 1a and 1s
was present in solution. Solution behavior of 1, as well as of
the related cis-mer-Ru(OSO2CF3)2(CO)(Cyttp), was inves-
tigated by variable-temperature 19F{1H} and 31P{1H} NMR
spectroscopy [10]. Owing to the complexity of the spectra,
which indicate a number of ongoing dynamic processes that
most likely involve different coordination of the anionic lig-
ands and their substitution by the solvent, the presence of
syn–anti isomerization could not be inferred with complete
Scheme 1.
Complex mer-Ru(␬2-O2CO)(CO)(Cyttp) was reported by
Meek et al. to be formed by reaction of the solid mer-
Ru(CO)2(Cyttp) with molecular oxygen [25] (Scheme 1) but
was not characterized for the syn or anti configuration of its
Cyttp phenyl group. By employing a slightly modified exper-

P.W. Blosser et al. / Journal of Molecular Catalysis A: Chemical 224 (2004) 133–144
139
toward the CO [10]. In contrast, use of the weaker ligands
such as F− and H2O led to syn–anti isomerization; however,
no rigorous order of addition of reactants was maintained in
these substitutions.
The carboxylato complexes [mer-anti-Ru(␬2-O2CR)(CO)
(Cyttp)]BF4 (R = Me (3a(BF4)), Ph(4a(BF4)))wereobtained
by treatment of solutions of sodium acetate and benzoate, re-
spectively, with 1 (Scheme 2). By following this order of
introduction of reactants into solution, only a single syn–anti
isomer was detected for each product by 31P{1H} NMR spec-
troscopy. Based on the considerations presented above, and
from an analysis of the NMR data (cf. Section 3.2), both
products are formulated as anti. No anti to syn isomerization
was observed for 3a(BF4) and 4a(BF4).
The methylation reaction of 2a with 1 equiv. of
[Me3O]BF4 affords the methylcarbonato complex [mer-anti-
Ru(␬2-O2COMe)(CO)(Cyttp)]BF4 (5a(BF4)) (Scheme 3).
Surprisingly, when more than 2 equiv. of [Me3O]BF4 were
employed, the only phosphorus-containing species detected
in solution by 31P{1H} NMR spectroscopy was 1. In
contrast, methylation of 2s either with 1 equiv. or with
more than 2 equiv. of [Me3O]BF4 generates only [mer-syn-
Ru(␬2-OCO2Me)(CO)(Cyttp)]BF4 (5s(BF4)) in high yield
(Scheme 4). The different reactivities of 2a and 2s with
>2 equiv. of [Me3O]BF4 may be attributed to 2a being
Scheme 2.
confidence. However, for the data to be consistent, such an
isomerization would have to be relatively slow compared to
the other occurring reactions.
The structure of 1 in the solid has not been elucidated;
however, it is probably anti, like the structure of the re-
lated cis-mer-Ru(OSO2CF3)(CO)(Cyttp), established by X-
ray diffraction techniques [10]. This is because both of these
complexes were prepared by very similar reactions, viz.,
treatment of cis-mer-syn-RuH2(CO)(Cyttp) (Cyttp Ph group
syn to CO) with the appropriate acid, HBF4 or CF3SO3H.
Moreover, 1 and cis-mer-Ru(OSO2CF3)2(CO)(Cyttp) un-
dergosubstitutionofBF4− andCF3SO3− withother, stronger
ligands, for example, H− and I−, in reactions that also pro-
ceed with retention of the orientation of the Cyttp Ph group
Scheme 3.
Scheme 4.

140
P.W. Blosser et al. / Journal of Molecular Catalysis A: Chemical 224 (2004) 133–144
sterically less hindered than 2s in the vicinity of Ru(␬2-
O2CO) owing to the Cyttp Ph group’s orientation away from
this site. As a result, a second equivalent of [Me3O]BF4
methylates Ru-bonded oxygen in precursor 5a(BF4), thus
leading to elimination of (MeO)2CO−and replacement of
the two oxygen donors with two BF4 ions at ruthenium.
With the sterically more encumbered 2s, methylation stops
upon formation of 5s(BF4). These reactions represent, to our
knowledge, the first example of different reactivity of syn
and anti isomers of metal Cyttp complexes. Treatment of
each 2a and 2s with 1 equiv. of [Et3O]PF6 resulted, as ex-
pected, in the formation of the ethylcarbonato complexes
[mer-Ru(␬2-O2COEt)(CO)(Cyttp)]PF6 as the anti (6a(PF6))
and syn (6s(PF6)) isomers, respectively. All of the complexes
5(BF4) and 6(PF6) were isolated as pure isomers, and no syn-
to-anti or anti-to-syn conversion was observed. The structure
of 5s(BF4) was determined by X-ray crystallographic tech-
niques; the other syn and anti assignments are based on the
(safe) assumption that the stereochemistry of the Cyttp Ph
group is the same as in the precursor carbonato complex 2.
All assignments are supported by the NMR spectroscopic
generalizations presented in Section 3.2.
Complex 2a reacts slowly with an excess of MeI
in benzene at room temperature to yield cis-mer-anti-
RuI2(CO)(Cyttp) (7a), previously prepared from cis-mer-
Ru(OSO2CF3)2(CO)(Cyttp) and NaI [10] (Scheme 3). When
the reaction with MeI was carried out in C6D6 solution, a
1
long-lived intermediate was detected by H and 31P{1H}
NMR spectroscopy which showed presence of mer-Cyttp
and an OMe group. This intermediate may be a mer-
RuI(OCO2Me)(CO)(Cyttp) containing ␬1-methylcarbonato
ligand; it is probably formed by alkylation of ␬2-carbonate
in 2a by Me+ and substitution of one ligated oxygen by I−.
Further reaction with another molecule of MeI would give 7a
and (MeO)2CO (not identified).
Heating a solution of 2s and an excess of MeI in THF
at reflux temperature afforded cis-mer-syn-RuI2(CO)(Cyttp)
(7s) (Scheme 4). In contrast, by running the reaction at room
temperature for 16 h, 5s(I) was obtained. Possible conversion
of 5s(I) to 7s with MeI under more forcing conditions was
not investigated.
Reaction of 2s with an excess of concentrated aqueous
HCl in CH2Cl2 rapidly produced the dichloride cis-mer-syn-
RuCl2(CO)(Cyttp) (8s). This previously unreported isomer
Fig. 1. Molecular structure of 2a. Non-hydrogen atoms are represented by 50% probability ellipsoids. The hydrogen atoms are omitted for clarity.

P.W. Blosser et al. / Journal of Molecular Catalysis A: Chemical 224 (2004) 133–144
141
of 2 was also prepared by a similar treatment of the alkyl-
carbonate 5s(BF4) or 6s(PF6) with HCl. The known anti iso-
mer of 8, cis-mer-anti-RuCl2(CO)(Cyttp) (8a), resulted when
each of the carbonato complex 2a, carboxylato complexes
3a(BF4) and 4a(BF4), or alkylcarbonato complexes 5a(BF4)
and 6a(PF6) was treated similarly with hydrochloric acid. All
of the dihalide complexes cis-mer-RuX2(CO)(Cyttp) (X = Cl
or I) are stable with respect to syn–anti isomerization.
or similar coupling of the carbonyl carbon to the central (Pc)
and the wing (Pw) phosphorus atoms (2JPcC ∼ JPwC). These
2
assignments leave the remaining two cis-octahedral positions
for occupancy by two halides or a bidentate carbonato, car-
boxylato, or alkylcarbonato ligand. The IR spectra of 2a and
2s [25] exhibit ν(O2CO) absorptions expected for bidentate
carbonate [30], and those of 3a(BF4) and 4a(BF4) and of
5a(BF4), 5s(BF4), 6a(PF6), and 6s(PF6) show related absorp-
tions of bidentate carboxylate and alkylcarbonate, respec-
tively [31]. The IR spectra of the aforementioned complexes
also display a strong (CO) absorption, which occurs, as ex-
pected, at 1918–1915 cm−1 for the electrically neutral metal
carbonates 2a and 2s [25] and at 1960–1935 cm−1 for the
cationic metal carboxylates and alkylcarbonates. The pres-
ence of these ligands is also reflected by the appearance in the
13C{1H} NMR spectra of the appropriate signal at δ 166 ppm
for 2, 187 ppm for 3 and 4, and 161 ppm for 5 and 6.
The crystallographically determined structures of 2a, 2s,
and 5s are presented with ORTEP drawings in Figs. 1–3,
respectively. Selected bond distances and angles are given in
Table 2.
3.2. Characterization of products
All new complexes were characterized by a combina-
tion of IR and NMR (1H, 31P{1H}, and 13C{1H}) spec-
troscopy, FAB mass spectrometry, and elemental analy-
sis. The structures of 2a (as 2a·CH2Cl2·2H2O), 2s (as
2s·(3/4)MeC(O)Me·2H2O), and 5s (as 5s(BF4)·C6H6)) were
elucidated by X-ray diffraction techniques.
The meridional orientation of the Cyttp ligand in all com-
plexes is indicated by both the A2X pattern in the 31P{1H}
NMR spectra [6,7,9,10] and the 13C{1H} NMR resonances
for the CH cyclohexyl carbon atoms, which appear as vir-
tual triplets [29] at δ 42.1–33.1 ppm (JPwC = 10–12 Hz) and
38.6–32.5 ppm (JPwC = 9–11 Hz). The lone CO ligand is in
all cases cis to the three P donors, as shown by either equal
The three structures show the anti orientation of the Ph
group in 2a and the syn orientation in 2s and 5s, but reveal no
unusual features. All bond distances and angles fall within
Fig. 2. Molecular structure of 2s. Non-hydrogen atoms are represented by 50% probability ellipsoids. The hydrogen atoms are omitted for clarity.

142
P.W. Blosser et al. / Journal of Molecular Catalysis A: Chemical 224 (2004) 133–144
Fig. 3. Molecular structure of 5s. Non-hydrogen atoms are represented by 50% probability ellipsoids. The hydrogen atoms are omitted for clarity.
normal ranges for comparable complexes. The Ru O bond
phenyl group on P(2) and CO in 2a. Interestingly, the struc-
turally characterized syn and anti isomers of mer-RuCl(␩3-
PhCH C C CPh)(Cyttp) show a pronounced difference in
the angle Pc Ru Cl, which is greater for the former, and a
less pronounced difference in the angle Pc Ru (vinyl C),
which is greater for the latter [8].
˚
distances vary from 2.119(4) to 2.143(4) A for the carbonates
˚
2a and 2s, being similar to the distance 2.087(5) A found in
Ru(␬2-O2CO)(bipy)2 [32]. For 5s, the Ru O bond distances
˚
of 2.185(4) and 2.205(4) A are consistent with those found in
related complexes [33–35]. The carbon–oxygen bond lengths
of 2a and 2s (cf. Table 2) are close to those found in other
carbonato complexes [36,37] and indicate considerable lo-
calization of these bonds, viz., Ru( O)2C O. The corre-
sponding bonds of 5s – O(2) C(38) 1.261(6), O(3) C(38)
Complex 2a also differs from 2s and 5s with respect to
displacement of the Ru center from the plane of three P donor
atoms. Whereas Ru is nearly coplanar with the P atoms in 2s
˚
and 5s (a displacement of 0.0321 and 0.0028 A, respectively,
˚
1.251(7), O(4) C(38) 1.335(6) A – suggest a delocalized lig-
toward the CO), it is substantially out of that plane in 2a
˚
and representation, Ru( O)2C OMe. The C O bond length
(0.1488 A toward the CO). Furthermore, the position of Ru
˚
of 1.335(6) A compares well with that reported for organic
with reference to the P3 plane is reflected in the magnitude of
the P(1) P(2) P(3) bond angle, which is 172.85(6)◦ for 2a,
but closer to 180◦, viz., 176.6(1) and 176.11(5)◦, for 2s and
5s, respectively.
˚
carbonates (1.34 A av. [38]). Concerning Ru Cyttp bond-
ing, the shorter Ru P(2) bond as compared to Ru P(1) and
Ru P(3) in each of 2a, 2s, and 5s is typical of meridional
Cyttp complexes with weakly trans-labilizing ligands oppo-
site to Pc [3,10].
The aforementioned structural differences between the
syn and the anti complexes affect certain aspects of their
13C{1H} and 31P{1H} NMR spectra. Interestingly, resultant
differences in NMR spectroscopic properties are observed
not only for 2a, 2s, and 5s, but also for the other syn and
anticomplexesofthetype[mer-Ru(␬2-O2CX)(CO)(Cyttp)]+
(3–6) and cis-mer-RuX2(CO)(Cyttp) (7, 8) obtained in this
study. Such general behavior would seem to indicate that the
syn–anti differences in structure extend to this whole group
of complexes.
Apart from orientation of the Ph group, the most impor-
tant difference in structure between the anti isomer 2a and
the syn complexes 2s and 5s is the magnitude of the an-
gle P(2) Ru C(37). This angle is substantially larger for 2a
(96.3(2)◦) than for 2s (84.4(4)◦) or 5s (84.2(2)◦). Signifi-
cantly, the angle P(2) Ru O(2) is larger for 2s (109.0(2)◦)
and 5s (107.8(1)◦) than for 2a (101.9(1)◦); however, the
difference is not as pronounced as that for the angle
P(2) Ru C(37). These structural results may be rationalized
by a greater steric repulsion between the phenyl group on P(2)
and the carbonate oxygen O(2) in 2s and 5s than between the
The values of 2JPcC for the 13C resonance of CO are invari-
ably larger for the syn than for the corresponding anti isomers
(∆ = 4–9 Hz). This appears to be a consequence of the relative

P.W. Blosser et al. / Journal of Molecular Catalysis A: Chemical 224 (2004) 133–144
143
Table 2
ruthenium(II) Cyttp complexes in isomerization and ligand
substitution reactions. Previous studies revealed that some
complexes display high configurational stability while other
complexes show propensity to undergo syn–anti isomeriza-
tion [7,8,10,20].
◦
˚
Selected bond distances (A) and bond angles ( ) of 2a·CH2Cl2·2H2O,
2s·(3/4)MeC(O)Me·2H2O, and 5s(BF4)·C6H6
2a
2s
5s
Ru P(1)
Ru P(2)
Ru P(3)
Ru O(2)
2.400(2)
2.314(2)
2.444(2)
2.121(4)
2.119(4)
1.821(6)
1.160(7)
1.320(7)
1.296(7)
1.236(7)
2.3402(3)
2.303(3)
2.404(3)
2.126(7)
2.143(6)
1.809(14)
1.170(13)
1.323(12)
1.299(13)
1.263(12)
2.397(1)
2.279(1)
2.439(2)
2.185(4)
2.205(3)
1.824(6)
1.151(6)
1.261(6)
1.251(7)
1.335(6)
1.430(8)
We find in this investigation that octahedral ruthenium(II)
complexes of the type [mer-Ru(␬2-O2CX)(CO)(Cyttp)]
n
(n = 0, 2; n = +1, 3–6) and cis-mer-RuX2(CO)(Cyttp) (7, 8),
all containing ligands that do not readily dissociate, display
high stability toward syn-to-anti and/or anti-to-syn isomer-
ization. This behavior may be contrasted with that reported
for 1 and cis-mer-Ru(OSO2CF3)2(CO)(Cyttp) with weakly
coordinating anionic ligands, for which syn–anti isomeriza-
tion has been implicated by solution NMR spectra [10]. Fur-
ther examples of observed isomerization include syn-to-anti
conversions of cis-mer-RuH2(P(OPh)3)(Cyttp) [20] and mer-
RuCl(␩3-PhCH C C CPh)(Cyttp) [7,8]. The former has
been attributed to steric factors (bulky P(OPh)3 ligand syn
to the Ph group of Cyttp in the reactant), whereas the latter is
likely driven by a stronger Ru (␩2-C C) interaction in the
anti product than in the syn reactant, as inferred from crystal-
lographic data [8]. Thus, it appears that formation of a five-
coordinate intermediate in these reactions triggers syn–anti
isomerization.
Ligand substitution reactions of ruthenium(II) Cyttp com-
plexes support the foregoing generalization. Reactions of
freshly prepared solutions (to forestall isomerization) of the
anti isomer of 1 or of cis-mer-Ru(OSO2CF3)2(CO)(Cyttp)
with various ligands, for example, CO32−, RCO2−, Cl−, I−
[10], and H− [10], lead to retention of the orientation of the
Ph group of Cyttp in the products. The same stereochemi-
cal result was obtained when 2a and 2s reacted with HCl to
give 8a and 8s, respectively. Likewise, replacement of ␩2-H2
in cis-mer-RuH2(␩2-H2)(Cyttp) with CO, N2, or PhCH2NC
proceeds to a single isomer of each product [20]. In contrast,
corresponding reaction with P(OMe)3 affords a mixture of
the syn and anti isomers of syn-mer-RuH2(P(OMe)3)(Cyttp),
probably owing to syn-to-anti conversion for steric reasons
of the initially formed product [20], as for the analogous
P(OPh)3 complex (vide supra).
Ru O(3)
Ru C(37)
O(1) C(37)
O(2) C(38)
O(3) C(38)
O(4) C(38)
O(4) C(39)
P(1) Ru P(2)
P(1) Ru P(3)
P(1) Ru O(2)
P(1) Ru O(3)
P(1) Ru C(37)
P(2) Ru P(3)
87.69(6)
172.85(6)
82.6(1)
92.2(1)
91.9(2)
93.05(6)
101.9(1)
163.6(1)
96.3(2)
90.3(1)
85.1(1)
95.1(2)
61.9(2)
160.7(2)
100.0(2)
92.1(4)
92.6(1)
176.6(1)
88.5(2)
87.4(2)
92.3(3)
90.5(1)
109.0(2)
171.3(2)
84.9(4)
89.0(2)
89.4(2)
89.6(3)
62.3(2)
166.0(4)
103.8(4)
91.5(7)
91.4(6)
115(5)
88.77(5)
176.11(6)
90.4(1)
88.1(1)
90.0(2)
95.12(5)
107.8(1)
167.2(1)
84.2(2)
88.3(1)
88.1(1)
90.5(2)
59.8(1)
168.0(2)
108.2(2)
89.7(3)
P(2) Ru O(2)
P(2) Ru O(3)
P(2) Ru C(37)
P(3) Ru O(2)
P(3) Ru O(3)
P(3) Ru C(37)
O(2) Ru O(3)
O(2) Ru C(37)
O(3) Ru C(38)
Ru O(2) C(38)
Ru O(3) C(38)
O(2) C(38) O(3)
O(2) C(38) O(4)
O(3) C(38) O(4)
Ru C(37) O(1)
Ru P(2) C(19)
C(38) O(4) C(39)
92.9(3)
89.1(3)
113.0(5)
122.9(7)
124.1(6)
172.9(5)
119.9(2)
121.3(5)
122.3(6)
116.4(5)
178.7(5)
116.8(2)
116.6(5)
122(1)
123(1)
178(1)
118.7(4)
size of the angle Pc Ru C(of CO). Carty and coworkers have
shownthat2JPcC dependsontheP M Cbondangle, increas-
ing as the angle progressively decreases below 110◦ [39].
In the 31P{1H} NMR spectra, the resonance of Pc is down-
field from the resonance of Pw in each of the complexes 2–8
(∆ = 4–24 ppm, and generally 16–24 ppm). This relationship
results from a weak trans-labilizing ligand being opposite to
Pc [3]. In addition, the resonance of Pc for each anti isomer
is downfield with respect to the corresponding resonance for
the syn isomer (generally by 7–9 ppm), and 2JPcPw is slightly
larger for the anti isomer (∆ = 1.3–3.3 Hz). These trends are
a direct consequence of the displacement of Ru from the P3
plane. For example, a larger displacement of the Ru atom for
2a than for 2s is expected to result in a poorer overlap of
the phosphorus and metal orbitals, thus leading to a greater
deshielding of Pc in the former compared to the latter isomer.
Dissociatively stable six-coordinate ruthenium(II) Cyttp
complexes do not undergo syn–anti isomerization, since no
low energy intramolecular process exists that would effect
such a transformation. However, for five-coordinate trigo-
nal bipyramidal complexes, syn–anti isomerization can take
place by an exchange process that interconverts two of the
three equatorial as well as the two axial ligands. This mech-
anism will be considered in a subsequent publication deal-
ing with additional chemistry of ruthenium Cyttp complexes
[40].
Acknowledgements
4. Discussion and conclusion
This study was supported in part by the National Science
Foundation and The Ohio State University.
The impetus for this research was to gain further in-
sight into syn–anti configurational behavior of six-coordinate

144
P.W. Blosser et al. / Journal of Molecular Catalysis A: Chemical 224 (2004) 133–144
[22] D.D. Perrin, W.L.F. Armarego, D.R. Perrin, Purification of Labora-
References
tory Chemicals, Pergamon Press, Oxford, 1991.
[23] R.R. Willis, M. Calligaris, P. Faleschini, J.C. Gallucci, A. Wojcicki,
J. Organomet. Chem. 193/194 (2000) 465.
[24] R.T. Dunsizer, V.M. Marsico, V. Plantevin, A. Wojcicki, Inorg. Chim.
Acta 342 (2003) 279.
[25] J.B. Letts, T.J. Mazanec, D.W. Meek, Organometallics 2 (1983) 695.
[26] TEXSAN, Single Crystal Structure Analysis Software, Version 5.0,
Molecular Structure Corp., The Woodlands, TX, 1989.
[27] (a) International Tables for X-ray Crystallography, vol. IV, Kynoch
Press, Birmingham, UK, 1974, pp. 71 and 148;
[1] F.A. Cotton, B. Hong, Prog. Inorg. Chem. 40 (1992) 179.
[2] D.W. Allen, in: D.W. Allen, J.C. Tebby (senior reporters), Organ-
ophosphorus Chemistry, RSC Specialist Periodical Report, vol. 32,
Cambridge, UK, 2002, p. 1 and previous volumes in this series.
[3] (a) R. Mason, D.W. Meek, Angew. Chem., Int. Ed. Engl. 17 (1978)
183;
(b) D.W. Meek, T.J. Mazanec, Acc. Chem. Res. 14 (1981) 266.
[4] R.B. King, in: E.C. Alyea, D.W. Meek (Eds.), Catalytic Aspects
of Metal Phosphine Complexes, Advances in Chemistry Series,
vol. 196, American Chemical Society, Washington, DC, 1982,
p. 313.
(b) R.F. Stewart, E.R. Davidson, W.T. Simpson, J. Chem. Phys. 42
(1965) 3175.
[5] R. Uriarte, T.J. Mazanec, K.D. Tau, D.W. Meek, Inorg. Chem. 19
(1980) 79.
[28] A.C.T. North, D.C. Phillips, F.S. Mathews, Acta Crystallogr. A 24
(1968) 351.
[6] G. Jia, D.W. Meek, J.C. Gallucci, Organometallics 9 (1990) 2549.
[7] G. Jia, D.W. Meek, Organometallics 10 (1991) 1444.
[8] G. Jia, J.C. Gallucci, A.L. Rheingold, B.S. Haggerty, D.W. Meek,
Organometallics 10 (1991) 3459.
[9] G. Jia, D.W. Meek, Inorg. Chem. 30 (1991) 1953.
[10] P.W. Blosser, J.C. Gallucci, A. Wojcicki, Inorg. Chem. 31 (1992)
2376.
[11] Y. Kim, D.E. Rende, J.C. Gallucci, A. Wojcicki, Inorg. Chim. Acta
352 (2003) 171.
[12] Y. Kim, D.E. Rende, J.C. Gallucci, A. Wojcicki, J. Organomet.
Chem. 682 (2003) 85.
[13] R.S. Cahn, C. Ingold, V. Prelog, Angew. Chem., Int. Ed. Engl. 5
(1966) 385.
[14] J.B. Letts, T.J. Mazanec, D.W. Meek, J. Am. Chem. Soc. 104 (1982)
3898.
[15] L.M. Koczon, R.C. Tisdale, K. Gebreyes, T.A. George, L. Ma, S.N.
Shaikh, J. Zubieta, Polyhedron 9 (1990) 545.
[29] L.M. Wilkes, J.H. Nelson, L.B. McCusker, K. Seff, F. Mathey, Inorg.
Chem. 22 (1983) 2476.
[30] K. Nakamoto, Infrared and Raman Spectra of Inorganic and Coor-
dination Compounds, 3rd ed., Wiley, New York, 1978, pp. 243–245.
[31] K. Nakamoto, Infrared and Raman Spectra of Inorganic and Coor-
dination Compounds, 3rd ed., Wiley, New York, 1978, pp. 232–233.
[32] T. Kimlera, T. Sakurai, M. Shima, T. Nagai, K. Mizumachi, Acta
Crystallogr. Sect. B 38 (1982) 112.
[33] (a) T.V. Ashworth, M. Nolte, E. Singleton, J. Organomet. Chem. 121
(1976) C57;
(b) T.V. Ashworth, M. Nolte, E. Singleton, J. Chem. Soc., Dalton
Trans. (1976) 2184.
[34] E.B. Boyar, A. Harding, S.D. Robinson, C.P. Brock, J. Chem. Soc.,
Dalton Trans. (1986) 1771.
[35] R.A. Sanchez-Delgado, U. Thewalt, N. Valencia, A. Andriollo, R.-
L. Marquez-Silvin, J. Puga, H. Schollhorn, H.-P. Klein, B. Fontal,
Inorg. Chem. 25 (1986) 1097.
[16] T. Zhu, W.G. Jackson, Inorg. Chim. Acta 343 (2003) 147.
[17] T. Zhu, W.G. Jackson, Inorg. Chem. 42 (2003) 88.
[18] E. Arpec, L. Dahlenberg, Z. Naturforsch. 36 (1981) 672.
[19] C. Yang, S.M. Socol, D.J. Kountz, D.W. Meek, Inorg. Chim. Acta
114 (1986) 119.
[20] G. Jia, D.W. Meek, J.C. Gallucci, Inorg. Chem. 30 (1991) 403.
[21] D.F. Shriver, M.A. Drezdzon, The Manipulation of Air-sensitive
Compounds, 2nd ed., Wiley, New York, 1986.
[36] M.C. Favas, D.L. Keport, J.M. Patrick, A.H. White, J. Chem. Soc.,
Dalton Trans. (1983) 571.
[37] M.D. Curtis, K.R. Han, Inorg. Chem. 24 (1985) 378.
[38] (a) F. Cser, Acta Chim. Acta Sci. Hung. 80 (1974) 49;
(b) A.M. Sorensen, Acta Chem. Scand. 25 (1971) 169.
[39] L.H. Randall, A.A. Cherkas, A.J. Carty, Organometallics 8 (1989)
568.
[40] P.W. Blosser, A. Wojcicki, in preparation.