Synthesis and catalytic application of [Rh(PPh3)2([9]aneS3)]PF6?

Article abstract of DOI:10.1021/om970578q

The reaction of [RhCl(PPh₃)₃] with [9]aneS₃ (1,4,7-trithiacyclononane) and NH₄PF₆ provides [Rh(PPh₃)₂([9]aneS₃)PF₆, which undergoes ligand subsitution and o

Full text of DOI:10.1021/om970578q

Organometallics 1997, 16, 4517-4518
4517
Syn th esis a n d Ca ta lytic Ap p lica tion of
‡
[Rh (P P h 3)2([9]a n eS3)]P F 6
Anthony F. Hill* and J ames D. E. T. Wilton-Ely
Department of Chemistry, Imperial College of Science Technology and Medicine,
South Kensington, London SW7 2AY, U.K.
Received J uly 8, 1997X
Summary: The reaction of [RhCl(PPh3)3] with [9]aneS3
(1,4,7-trithiacyclononane) and NH4PF6 provides
[Rh(PPh3)2([9]aneS3)]PF6, which undergoes ligand sub-
sitution and oxidative-addition reactions and effectively
catalyzes the demercuration of bis(alkynyl)mercurials.
substitution and oxidative-addition chemistry, and its
deployment in a catalytic process, viz the demercuration
of bis(alkynyl)mercurials to form the corresponding 1,3-
diynes.7
Wilkinson’s catalyst [RhCl(PPh3)3] (1) is unique in its
versatility. This may be traced to facile interconversion
between complexes of different coordination number and
the accessibility of both mono- and trivalent rhodium.
The reaction of (1) with 1,4,7-trithiacyclononane
([9]aneS3) and NH4PF6 provides high yields of the salt
[Rh(PPh3)2([9]aneS3)]PF6 (2)8 (Scheme 1). A related
series of complexes [RhL2([9]aneS3)]+ (L2 ) (C2H4)2,
(PPh3)(CO); cod ) 1,4-cyclooctadiene) has been described
by Schro¨der,9 and crystallographic studies indicate that
the [9]aneS3 macrocycle adopts a tridentate coordination
mode, at least in the solid state. Spectroscopic data for
2 also support such a coordination. Given the preva-
The application of polythiamacrocycles (‘thiacrowns’)
as co-ligands for metal-based catalysis has been mooted
as grounds for the study of their coordination chemis-
try.1 This rationale appreciates (i) that thiacrowns
overcome the comparatively weak coordinating ability
of acyclic thioethers and (ii) that thioethers may in part
mimic phosphines, the catalytic utility of which is
enormous. Despite the substantial growth in the coor-
dination chemistry of thiacrowns,2 the application of
such complexes to catalysis has received scant attention,
with the elegant exception of studies by Kellog who
showed that Grignard cross-coupling reactions could be
effected by the addition of thiacrowns to nickel chloride
and that the process could be made enantioselective by
employing chiral thiacrowns.3 Although the nickel
complexes involved were not identified, nickel chloride
in the presence of [14]aneS4 achieved yields comparable
to those of triphenylphosphine while dibutyl sulfide
and the acyclic thioether 2,5,9,12-tetrathiatridecane
provided considerably reduced yields. More recently,
Adams has shown numerous examples of the catalytic
synthesis of thiacrowns from metal-mediated ring-
opening oligomerizations of thietanes, which clearly
involve thiacrown complex intermediates.4
(7) Bedford, R. B.; Hill, A. F.; Thompsett, A. R.; White, A. J . P.;
Williams, D. J . J . Chem. Soc., Chem. Commun. 1996, 1059.
(8) Selected data for representative new complexes (25 °C, NMR
(CDCl3 or CDCl3:CH2Cl2 ) 1:3)). 2: Yield 93%; 1H NMR δ 1.88-1.97,
2.31-2.39 (m × 2, 12 H, SCH2), 7.18-7.45 (m, 30 H, C6H5); 31P{1H}
NMR 42.4 ppm (1J (RhP) ) 169.6 Hz); FAB-MS m/ z 807 (100, [M]+),
627 (32, [M - [9]aneS3]+), 545 (74, [M - PPh3]+). Anal. Calcd for
C42H42F6P3RhS3: C, 53.0; H, 4.4. Found: C, 52.8; H, 4.2. NB: Schro¨der
has reported that the reaction of [Rh2(µ-Cl)2(C2H4)4] with 1 equiv of
PPh3 (!) and 2 equiv of [9]aneS3 provides an inseparable mixture of
compounds, the mass spectrum of which includes a peak at m/ z )
807.9c 4: Yield 81%; IR (Nujol) 1947 cm-1 (ν(CO)); 1H NMR δ 2.31-
2.56, 2.63-2.87 (m × 2, 12 H, SCH2), 7.41-7.79 (m × 2, 15 H, C6H5);
31P{1H} NMR 42.8 ppm (J (RhP) ) 128.9 Hz), see also ref 9. 5: Yield
63%; 1H NMR Insufficiently soluble; 31P{1H} NMR 38.7 ppm (1J (RhP)
) 195.8 Hz); FAB-MS m/ z 836 (56, [M]+), 657 (12, [M - [9]aneS3]+).
Anal. Calcd for C40H40F6FeP3RhS3: C, 46.1; H, 4.0. Found: C, 45.5;
H, 3.9. 6: IR (Nujol) 1279 cm-1 (ν(CS)). 1H NMR δ 2.38-2.52, 2.76-
2.90 (m × 2, 12 H, SCH2), 7.42-7.55, 7.73-8.81 (m × 2, 15 H, C6H5);
13C{1H} NMR 293.0 (dd, RhCS, 2J (PC) ) 17.8, 1J (RhC) ) 71.4 Hz),
135.5-127.6 (C6H5), 36.4, 34.3, 32.2 (SCH2); 31P{1H} NMR 43.3 ppm
(1J (RhP) ) 149.2 Hz); FAB-MS m/ z 1276 (1.5, [M2ClO4]+), 742 (0.5,
[M + nba]+), 589 (100, [M]+), 545 (4, [M - CS]+). Anal. Calcd for
C25H27ClO4PRhS4‚1.25CH2Cl2: C, 39.7; H, 3.7. Found: 39.7; H, 3.6.
CH2Cl2 solvate confirmed by 1H NMR integration. 7: Yield 67%; 1H
NMR δ 1.22, 1.75, 2.44, 2.88, 3.72 (m × 5, 12 H, SCH2), 7.40-7.81 (m
× 2, 15 H, C6H5). 31P{1H} NMR 43.2 ppm (1J (RhP) ) 149.2 Hz); FAB-
MS m/ z 798 (7, [M]+), 671 (13, [M - I]+), 545 (3, [M - I2]+). Anal.
Calcd for C24H27ClI2O4PRhS3‚1.5CH2Cl2: C, 29.9; H, 3.0. Found: C,
29.8; H, 4.0. CH2Cl2 solvate confirmed by 1H NMR integration. 8a :
Yield 93%; IR (Nujol) 2152, 2113 (sh) cm-1 (ν(CN)). 1H NMR δ 1.03 (s,
18 H, CH3), 7.36-7.47 (m, 30 H, C6H5); 31P{1H} NMR 30.8 ppm; FAB-
MS m/ z 793 (100, [M]+), 710 (15, [M - CNR]+), 627 (12, [M - 2CNR]+),
531 (M - PPh3]+). Anal. Calcd for C54H48F6N2P3: C, 62.7; H, 4.7; N,
2.7. Found: C, 62.9; H, 4.8; N, 2.7. 8b: Yield 90%; IR (Nujol) 2129
cm-1(ν(CN)). 1H NMR δ 1.61 (s, 12 H, CH3), 6.84 (d, 4 H, H3,5(C6H3)),
7.00 (t, 2 H, H4(C6H3)), 7.32, 7.64 (m × 2, 30 H, C6H5); 31P{1H} NMR
31.2 ppm (1J (RhP) ) 124.1 Hz); FAB-MS m/ z 889 (100, [M]+), 758
We have been concerned recently with the prepara-
tion of organometallic complexes of thiacrowns, with a
focus on investigating the compatability of such mac-
rocycles with typical ‘C1’ ligands involved in conven-
tional catalytic cycles. These have included σ-vinyl,
aryl, carbonyl, and thiocarbonyl ligands,5 and we have
shown that coordination of such macrocycles can induce
ligand-coupling reactions.6 We report herein the syn-
thesis of a thiacrown complex of rhodium(I), its ligand
* Author to whom correspondence should be addressed. Email:
a.hill@ic.ac.uk.
‡ Dedicated to Dr. R. M. Scrowston in gratitude for his encourage-
ment and guidance.
X Abstract published in Advance ACS Abstracts, October 1, 1997.
(1) For a general review on thiacrown complexes see: Cooper, S.
R.; Rawle, S. C. Struct. Bonding 1990, 72, 1.
(2) Blake, A. J .; Schro¨der, M. Adv. Inorg. Chem. 1990, 35, 1.
(3) Kellogg, R. M. Angew. Chem., Int. Ed. Engl. 1984, 23, 782.
Vriesema, B. K.; Lemaire, M.; Butler, J .; Kellogg, R. M. J . Org. Chem.
1986, 51, 5169.
(4) Adams, R. D.; Falloon, S. B. Chem. Rev. 1995, 95, 2587.
(5) Cannadine, J . C.; Hill, A. F.; White, A. J . P.; Williams, D. J .;
Wilton-Ely, J . D. E. T. Organometallics 1996, 15, 5409. Hector, A. L.;
Hill, A. F. Inorg. Chem. 1995, 34, 3797. Alcock, N. W.; Cannadine, J .
C.; Clark, G. R.; Hill, A. F. J . Chem. Soc., Dalton Trans. 1993, 1131.
Hill, A. F.; Alcock, N. W.; Cannadine, J .; Clark, G. R. J . Organomet.
Chem. 1992, 426, C40.
(14, [M - CNR]+), 627 (23, [M - 2CNR]+), 496 (18, [M - PPh3
-
CNR]+). 9: IR (Nujol) 2165 cm-1(ν(CN)). 1H NMR 1.51 (s, 36 H, CH3);
FAB-MS m/ z 435 (100, [M]+), 352 (5, [M - CNR]+), 266 (4, [M -
2CNR]+), 184 (5, [M - 3CNR]+). See also ref 10 for alternative
synthesis.
(9) (a) Blake, A. J .; Halcrow, M. A.; Schro¨der, M. J . Chem. Soc.,
Chem. Commun. 1991, 253. (b) Blake, A. J .; Halcrow, M. A.; Schro¨der,
M. Acta Crystallogr., Sect. C, 1993, 49, 85. (c) Blake, A. J .; Gould, R.
O.; Halcrow, M. A.; Schro¨der, M. J . Chem. Soc., Dalton Trans. 1994,
2197.
(6) Cannadine, J . C.; Hector, A. L.; Hill, A. F. Organometallics 1992,
11, 2323.
(10) Mann, K. R.; Lewis, N. S.; Williams, R. M.; Gray, H. B.; Gordon,
J . G., II Inorg. Chem. 1978, 17, 829.
S0276-7333(97)00578-5 CCC: $14.00 © 1997 American Chemical Society

4518 Organometallics, Vol. 16, No. 21, 1997
Communications
Sch em e 1a
Sch em e 2a
a
R ) C6H4Me-4.
the macrocycle to provide [Rh(CNtBu)4]PF6 (9). This
lability of the thiacrown is somewhat surprising. Com-
plex 2 is also prone to oxidative addition: reaction with
iodine results in the formation of the octahedral complex
[RhI2(PPh3)([9]aneS3)]PF6 (7).
The complex has been shown above to demonstrate
two useful properties for a catalyst: facile ligand
substitution and the ready accessibility of two inter-
changeable oxidation states. We have, therefore, in-
vestigated the suitability of 2 for a catalytic process
which has recently attracted our attention.7 We have
shown that the complexes [RhCl(CO)(PPh3)2] (10) and
[OsHCl(CO)(PPh3)3] catalyze the demercuration of bis-
(alkynyl)mercurials (Scheme 2).7 In the case of 10 a
mechanism has been proposed involving oxidative ad-
dition of a mercury acetylide bond, extrusion of elemen-
tal mercury from the Rh-Hg-C≡ group, and reductive
elimination of the diyne from the resulting cis-bis-
(alkynyl) intermediate. Thus, this process is particu-
larly convenient for assessing catalytic activity in that
it requires oxidative-addition and reductive-elimination
steps and its ensuite is immediately visible by virtue of
the deposition of elemental mercury. Bis(4-tolylethy-
nyl)mercury (1.2 mmol) was treated with 5 mol % of
metal catalyst as a point of reference. We find that as
previously reported, 10 effects demercuration over 2 h
in refluxing tetrahydrofuran in 75% yield. Wilkinson’s
catalyst (1) achieves a comparable yield (82%), offering
the trade-off of higher activity (proceeds at room tem-
perature) against the required use of anerobic condi-
tions. Gratifyingly, we find that 2 effects demercuration
in 94% yield over 2 h in refluxing thf. Furthermore,
when the reaction is run for 1 h with 15 mol % of 2, the
31P NMR spectrum of the product mixture reveals 2 to
be the only organometallic complex present, indicating
that 2 is a true catalyst and not merely a precatalyst.
Thus, not only has a catalytic process based on a well-
defined thiacrown complex been illustrated for the first
time, the yields are found to exceed those for more
conventional catalysts 1 and 10. We are now investi-
gating the deployment of 2 and its derivatives in more
classical metal-mediated catalytic processes and will
report subsequently.
a
R ) tBu, C6H3Me2-2,6; dppf ) 1,1′-bis(diphenylphosphi-
no)ferrocene; L ) PPh3.
lence for four-coordinate square-planar geometries for
d8 complexes, the possibility of variable and reduced
denticity should not, however, be disregarded in sub-
sequent reactions. Surprisingly, attempts to prepare 2
from the reaction of [Rh(cod)(PPh3)2]PF6 led to phos-
phine rather than diene displacement and the formation
of [Rh(cod)([9]aneS3)]PF6 (3). This salt has been pre-
pared previously via the reaction of [9]aneS3 with
[Rh2(µ-Cl)2(cod)4].9 Complex 2 is labile with respect to
ligand substitution, e.g., carbonylation at room temper-
ature provides the known salt [Rh(CO)(PPh3)([9]aneS3)]-
PF6 (4). The related and sparingly soluble salt [Rh(dppf)-
([9]aneS3)]PF6 (dppf ) 1,1′-bis(diphenylphosphino)-
ferrocene, 5) results from reaction of 2 with dppf.
The reaction of 2 with carbon disulfide provides
[Rh(η2-SCS)(PPh3)([9]aneS3)]PF6 rather than the thio-
carbonyl analogue of 4. The salt [Rh(CS)(PPh3)-
([9]aneS3)]ClO4 (6) may, however, be obtained in good
yield from the reaction of [RhCl(CS)(PPh3)2] with [9]aneS3
and NaClO4 (Scheme 1).8 Somewhat suprisingly, the
thiocarbonyl ligand is cleaved from 6 by iodine to
provide [RhI2(PPh3)([9]aneS3)]PF6 (7). Attempts to
prepare the isonitrile analogue of 4 were, however,
unsuccessful: Treating 2 with pivaloisonitrile or 2-
xylylisonitrile provided the salts [Ru(CNR)2(PPh3)2]PF6
(R ) tBu (8a ), C6H3Me2-2,6 (8b)), while treating 3 with
pivaloisonitrile leads to loss of both cyclooctadiene and
Ack n ow led gm en t. We wish to thank the Engineer-
ing and Physical Sciences Research Council (U.K.) for
the award of a studentship (to J .D.E.T.W.-E.) and
Leverhulme Trust and the Royal Society for the award
of a Senior Research Fellowship (to A.F.H.).
Su p p or tin g In for m a tion Ava ila ble: Text giving details
of the syntheses and characterization data for the complexes
prepared in this paper (3 pages). Ordering information is
given on any current masthead page.
OM970578Q