Slow epimerization of stereochemically rigid diastereomers of the equatorially substituted cluster [Os3H2(μ3-S)(CO)8{(S)- PhCHMeNH2}]

Article abstract of DOI:10.1039/b203611h

Turnstile rotation is suppressed in the equatorially substituted cluster [Os₃(μ-H)₂(μ₃-S)(CO)₈{(S)- PhCHMeNH₂}] which was separated by HPLC into two diastereomers which do not interconvert at room tem

Full text of DOI:10.1039/b203611h

Slow epimerization of stereochemically rigid diastereomers of the
equatorially substituted cluster [Os3H2(m3-S)(CO)8{(S)-PhCHMeNH2}]
Antony J. Deeming,*a Caroline S. Forth,a Graeme Hogarth,a David Markham,a Jade O. Princea and
Jonathan W. Steedb
a Department of Chemistry, University College London, 20 Gordon Street, London, UK WC1H 0AJ.
E-mail: a.j.deeming@ucl.ac.uk and g.hogarth@ucl.ac.uk
b Department of Chemistry, King’s College London, Strand, London, UK WC2R 2LS
Received (in Cambridge, UK) 12th April 2002, Accepted 12th August 2002
First published as an Advance Article on the web 5th September 2002
Turnstile rotation is suppressed in the equatorially substi-
tuted cluster [Os3(m-H)2(m3-S)(CO)8{(S)-PhCHMeNH2}]
which was separated by HPLC into two diastereomers which
do not interconvert at room temperature and epimerize only
slowly at 90 °C.
doublets at d 1.57 and 1.55 with a diastereomeric ratio of
0.54+0.46 at room temperature. Attempts to fractionate the
yellow product by TLC on SiO2 led to only slight isomer
enrichment but HPLC separation of B1 and B2 was very
effective and gave the two diastereomers in 100 and 95%
isomeric purity, respectively.
Two features of the fluxionality of metal carbonyl clusters are
turnstile rotation of M(CO)3 and M(CO)2L sub-units and the
ability of hydride ligands to hop between metal–metal edges.
Both can be rapid enough to give NMR coalescences. The
dynamics of [Os3(m-H)2(m3-S)(CO)9] 11 involve rapid hydride
motion between osmium–osmium edges to give time-averaged
C3v symmetry.2 In this case turnstile rotation of the Os(CO)3
groups occurs much more slowly.
We have now studied the substitution of 1 by amines and
phosphines and have observed that the mono-substituted
products [Os3(m-H)2(m3-S)(CO)8(L)] exist as the isomers A, B1
and B2 depending upon which ligand L is used. For tertiary
The structure of the isomer B1 was determined by single-
crystal XRD (Fig. 1).‡ The equatorial co-ordination of (S)-
PhCHMeNH2 leading to the two 1H NMR signals at d 219.39
and 224.51 was confirmed. The hydride giving the signal at d
219.39 is that cis to the amine. The amine has a smaller trans
influence than CO on the Os–Os distances which are 2.8647(13)
and 2.8846(14) Å trans to amine and cis to CO and 2.9155(12)
and 2.9050(12) Å cis to amine and trans to CO, compared with
2.908(1) and 2.922(1) Å for the hydride-bridged edges in
[Os3(m-H)2(m3-S)(CO)9] 1, although in essence the structure is
closely similar to that of cluster 1.1 Note that the cluster is
equatorially substituted unlike [Os3(m-H)2(m3-Se)(CO)8(PPh3)]
which contains an axial phosphine.4
The final confirmation that the isomers are indeed B1 and B2
came from their CD spectra which are shown in Fig. 2. The
ligand (S)-PhCHMeNH2 does not absorb significantly in the
range 300 to 400 nm so that the cluster itself is the chromophore
in this region and there is a strong inverse relationship between
the CD spectra totally consistent with cluster-centred chirality.
We believe that this is the first time isomeric clusters related by
a trigonal twist of a M(CO)2L group have been separated and
shown to interconvert only very slowly.
phosphines all three isomers are formed in equilibrium, but only
isomers B1 and B2 have been detected for amines. With simple
achiral monodentate ligands, isomers B1 and B2 are enantio-
mers.
Considering the projections shown for B1 and B2 below, the
turnstile rotation is basically a Bailar twist and known to occur
very slowly if at all for octahedral complexes. Note that, if we
N-donor ligands bind exclusively in equatorial sites, whether
the ligand is bulky or not.† Thus [Os3(m-H)2(m3-S)(CO)8(L)]
1
shows just two hydride H NMR signals (CDCl3 solutions;
signals sometimes resolved as narrow doublets) at d 219.70 and
224.73 (MeCN),3 d 218.85 and 225.57 (NMe3),3 d 219.19
i
and 225.52 (NEt3), d 219.43 and 224.38 (NH2 Pr), d 29.31
and 224.88 (NHEt2), d 219.29 and 224.39 (NH2CH2Ph).
There is no evidence for the axial isomer A in any case, even
with the most bulky amines such as NEt3, and the rate of
interconversion of the enantiomers B1 with B2 is too slow to
result in NMR coalescence. The NH2CHMe2, NHEt2, NEt3 and
NH2CH2Ph ligands in these clusters all show diastereotopic
groups as appropriate for co-ordination of the amine to a
i
stereochemically rigid chiral cluster. For example, the NH2 Pr
ligand shows signals at d 3.11 (sept., CH), 1.241 (d, J = 4.0 Hz,
Me), 1.225 (d, J = 4.0 Hz, Me), 3.01 (br, NH), 2.53 (br, NH).
1
The isopropylamine cluster is rigid in H NMR spectra up to
100 °C with no line broadening at this temperature. Turnstile
rotation of the Os(CO)2L group is therefore slow. This suggests
the possibility of separation of enantiomers of [Os3(m-H)2(m3-
S)(CO)8(L)] or diastereomers if a chiral ligand L is used.
As predicted the complex [Os3(m-H)2(m3-S)(CO)8{(S)-
PhCHMeNH2}] demonstrates clearly the presence of two
Fig. 1 Molecular structure of one of two independent molecules in the unit
cell of the diasteromer B1 of [Os3(m-H)2(m3-S)(CO)8{(S)-PhCHMeNH2}].
Selected bond lengths (Å): Os(1)–Os(2) 2.9155(12), Os(2)–Os(3)
2.7723(12), Os(1)–Os(3) 2.8647(13), Os(1)–S(1) 2.404(6), Os(2)–S(1)
2.391(6), Os(3)–S(1) 2.390(6), Os(1)–N(1) 2.17(3).
1
diastereomers in the H NMR spectrum (CDCl3); there are
hydride signals at d 219.39, 219.47, 224.48 and 224.51,
broad NH signals at d 3.49, 3.28, 2.99 and 2.85, and Me
2230
CHEM. COMMUN., 2002, 2230–2231
This journal is © The Royal Society of Chemistry 2002

propylamine, there is some conversion to the B1/B2 mixture (i.e.
some epimerization) but most of the reaction proceeds by
substitution to the isopropylamine cluster [Os3(m-H)2(m3-
S)(CO)8(Me2CHNH2}]. Thus, after 30 min reaction in refluxing
heptane, a sample of B1 and B2 (initial mol ratio 0.10+0.90)
gave a mixture of B1, B2 and the isopropylamine substituted
product in mol ratio 0.04+0.24+0.72. Epimerization and
substitution occur at similar rates under the same conditions.
We are currently carrying out a more detailed kinetic study of
both the epimerization and substitution processes. Preliminary
results indicate that epimerization is about 4 times faster than
substitution.7 Intramolecular turnstile rotation and Os–N bond
cleavage are therefore very finely balanced.
consider only the ligand positions and not the metal–metal
vectors, then the substituted osmium atom is closely octahe-
dral.
We thank the Royal Society and the Swedish Academy of
Sciences for supporting collaborative work with Dr E. Nor-
lander (Lund University) on catalysis by chiral transition metal
clusters. We also thank the University of London Intercollegiate
Research Service for CD spectra provided by the chiroptical
service at KingAs College London (Dr A. F. Drake and Dr G.
Siligardi).
The situation with tertiary phosphines is very different to that
with amines. Thus we have observed mixtures of A and B1/B2
for each of the tertiary phosphines we have used: PPh3, PCy3,
PEt3, PMe2Ph with an increased preference for isomer A with
the more bulky phosphines.§ The crystal structure of [Os3(m-
H)2(m3-Se)(CO)8(PPh3)] shows that the structure A is adopted
in the solid state.4 However, we have now shown that [Os3(m-
H)2(m3-S)(CO)8(PPh3)] exists in solution as a mixture of the
equatorial isomers (B1/B2) and the axial isomer (A) as do all the
other tertiary phosphine complexes of this type we have studied.
The mol ratio A+B1/B2 varies from 7.16+1 (PiPr3), 3.92+1
(PPh3), 1.25+1 (PEt3) to 0.67+1 (PMe2Ph) at room temperature
in the order of Tolman cone angles [170° PCy3, 145° PPh3, 132°
PEt3, 122° PMe2Ph]. The cluster [Os3(m-H)2(m3-S)(CO)8(P-
Me2Ph)] probably crystallizes as the isomer B1/B2 because this
is the dominant isomer initially on dissolving in CDCl3 [mol
ratio A+B1/B2 = 0.12+1] but the equilibrium mixture [mol ratio
A+B1/B2 = 0.67+1] is formed over 27 hours at 20 °C. Our
studies on the mechanism of interconversion of isomers and
hydride exchange show that hydride migration is faster than
phosphine or CO movement but that the interconversion of
isomers can be detected by EXSY methods and by NMR
dynamic line broadening at temperatures around 100 °C in the
1H and 13C NMR spectra. Kinetic details will be published
elsewhere.5
Notes and references
† Typical synthesis of [Os3(m-H)2(m3-S)(CO)8(L)] for amines L: Cluster 1
(0.06 mmol) in dichloromethane (30 cm3) was treated with Me3NO·2H2O
(0.24 mmol) and the amine L (0.5 cm3) under nitrogen. After 10 min the
reaction mixture was separated on silica (TLC, eluent = light petroleum
spirit–dichloromethane eluent) to give traces of 1 and the yellow product
(19 to 40%). n(CO)/cm21 (cyclohexane, L = iPrNH2): 2085m, 2049vs,
2030vs, 2004vs,1989s, 1968m, 1964m. Satisfactory elemental analyses
were obtained.
‡ Crystallographic data for [Os3(m-H)2(m3-S)(CO)8{(S)-PhCHMeNH2}]:
C16H13NO8Os3S, M
= 949.93, monoclinic, space group P21, a =
9.4018(4), b = 20.5782(11), c = 11.5621(5) Å, b = 95.605(3)°, V =
2226.25(18) Å, Z = 4, Dc = 2.834 g cm23, l(Mo-Ka) = 0.71073 Å, m =
17.213 mm21, F(000)
= 1696. 11460 independent reflections were
measured in the q range 1.77 to 24.99°. 526 parameters were refined to give
R [I > 2s(I)] = 0.0683 and wR2 (all data) = 0.1781. The structure was
solved by direct methods and refined (SHELXL-97) with all non-H atoms
anisotropic. H-atoms were included using a riding model except the
hydrides which were located using HYDEX.8 Flack parameter = 0.01(2).
The two independent molecules in the unit cell do not differ significantly.
CCDC 183148. See http://www.rsc.org/suppdata/cc/b2/b203611h/ for crys-
tallographic files in CIF or other electronic format.
A preliminary theoretical study has shown that the isomers A
and B1/B2 of [Os3(m-H)2(m3-S)(CO)8(PH3)] are almost equal in
energy whereas the equatorial isomers B1/B2 of [Os3(m-H)2(m3-
S)(CO)8(NH3)] are lower in energy than the unobserved isomer
A, consistent with our observations.6
§ Typical synthesis of [Os3(m-H)2(m3-S)(CO)8(L)] for tertiary phosphines
L: a solution of [Os3(m-H)2(m3-S)(CO)8(MeCN)] (0.192 g) and PMe2Ph
(0.035 g) in toluene (75 cm3) was refluxed under nitrogen for 8.5 h.
Removal of the solvent, preliminary column chromatography on silica, and
final purification by TLC (eluent: CH2Cl2–pentane; 1+6 v/v) gave [Os3(m-
H)2(m3-S)(CO)8(PMe2Ph)] (0.102 g) as the main product. n(CO)/cm21
(cyclohexane): 2081s, 2045s, 2038(sh), 2001s, 1997(sh), 1986s, 1981(sh),
1977(sh), 1961m. Hydride 1H NMR (CDCl3): isomer A: d 220.62 (d, JPH
= 11.5 Hz, JOsH = 30.4 Hz); isomers B1/B2: d 219.99 (d, JPH = 27.5 Hz,
trans to PMe2Ph, JOsH = 29.2, 37.4 Hz), d 221.42 (d, JPH = 8.5 Hz, cis to
PMe2Ph, JOsH = 29.6 Hz).
Epimerization is very slow. There is a 26% conversion of
isomer B1 to B2 in chloroform over 142 days in the dark at room
temperature. However, if a solution of predominantly diaster-
eomer B1 in heptane is kept at 96 °C, there is significant
conversion to the B1/B2 mixture within 15 min and after 45 min
the composition is 0.52+0.48, close to the composition of the
synthesised mixture. If a similar treatment of one diastereomer
of [Os3(m-H)2(m3-S)(CO)8{(S)-PhCHMeNH2}] in refluxing
heptane solution is carried out in the presence of iso-
1 (a) A. J. Deeming and M. Underhill, J. Organomet. Chem., 1972, 42,
C60; (b) A. J. Hempleman, I. A. Oxton, D. B. Powell, P. Skinner, A. J.
Deeming and L. Markó, J. Chem. Soc., Faraday Trans. 2, 1981, 77, 1669;
(c) G. Granozzi, R. Benoni, M. Acampora, S. Aime and D. Osella, Inorg.
Chim. Acta, 1984, 84, 95; (d) B. F. G. Johnson, J. Lewis, D. A. Pippard,
P. R. Raithby, G. M. Sheldrick and K. D. Rouse, J. Chem. Soc., Dalton
Trans., 1979, 616.
2 (a) A. Forster, B. F. G. Johnson, J. Lewis and T. W. Matheson, J.
Organomet. Chem., 1976, 104, 225; (b) J. Bracker-Novak, S. Hajela, M.
Lord, M. Zhang, E. Rosenberg, R. Gobetto, L. Milone and D. Osella,
Organometallics, 1990, 9, 1379.
3 B. F. G. Johnson, J. Lewis and D. A. Pippard, J. Organomet. Chem.,
1978, 160, 263.
4 W. K. Leong, W. L. J. Leong and J. Zhang, J. Chem. Soc., Dalton Trans.,
2001, 1087.
5 A. J. Deeming and J. O. Prince, unpublished results.
6 N. Kaltsoyannis, unpublished results.
7 D. M. Rowley and C. S. Forth, unpublished results.
8 A. G. Orpen, J. Chem. Soc., Dalton Trans., 1980, 2509.
Fig. 2 CD spectra of isomers B1 and B2 of [Os3(m-H)2(m3-S)(CO)8{(S)
PhCHMeNH2}].
CHEM. COMMUN., 2002, 2230–2231
2231