Oxidation of Thioether Ligands in Pseudotetrahedral Cyclopentadienylruthenium Complexes: Toward a New Stereoselective Synthesis of Chiral Sulfoxides

Article abstract of DOI:10.1021/ic961280f

Ionic ruthenium thioether complexes [Cp(LL′)Ru(SRR′)]PF₆ (LL′ = Ph₂PCH₂PPh₂ (1), Ph₂PC₂H₄PPh₂ (2),(Ph₃P, CO) (3), Me₂PC₂H₄PPh₂ (4), (S,S)-Ph₂PCHMeCHMePPh₂ (5), SRR′ = MeSPh (a), MeS-i-Pr (b), MeSBz (c), i-PrSBz (d), EtSBz (e), MeSCy (f), SC₄H₈ (g)) were synthesized from the corresponding chloro complexes [Cp-(LL′)RuCl] and thioethers. 5a crystallized in the orthorhombic system, space group P2₁2₁2₁ (No. 19), with a = 11.269(3) A°, b = 15.104(2) A?, c = 23.177(4) A?, and Z = 4. 5b crystallized in the monoclinic system, space group P2₁ (No. 4), with a = 10.539(5) A?, b = 16.216(9) A?, c = 11.011(8) A?, β = 106.04(2)°, and Z = 2. A similar ligand exchange reaction yielded the analogous sulfoxide complexes [Cp(LL′)Ru(S(O)RR′)]PF₆ (6-10). 10a crystallized in the orthorhombic system, space group P2₁2₁2₁ (No. 19), with a = 14.1664(13) A?, b = 15.792-(2) A?, c = 17.641(2) A?, and Z = 4. 10b-0.93CH₂Cl₂ crystallized in the orthorhombic system, space group P2₁2₁2₁ (No. 19), with a = 12.069(2) A?, b = 17.379(2) A?, c = 19.760(5) A?, and Z = 4. The thioether complexes can also be directly converted to sulfoxide complexes with the strong oxygen transfer reagent dimethyldioxirane (DMD). No crossover products are formed when mixtures of two thioether complexes (e.g., 1a/2c or 1c/2a) are treated with DMD, demonstrating that no Ru-S bond cleavage is involved. Moderate diastereoselectivities are observed for the oxygen transfer to chiral, racemic thioether complexes 3 (8-28%) and 4 (34-60%). Oxidation of the (S,S)-CHIRAPHOS complexes 5, however, is highly stereoselective (de = 46-98%). Treatment of the sulfoxide complexes 10 with sodium iodide removes the chiral, nonracemic sulfoxides from the metal with retention of the configuration at sulfur.

Full text of DOI:10.1021/ic961280f

2372
Inorg. Chem. 1997, 36, 2372-2378
Oxidation of Thioether Ligands in Pseudotetrahedral Cyclopentadienylruthenium
Complexes: Toward a New Stereoselective Synthesis of Chiral Sulfoxides1
Wolfdieter A. Schenk,*,2a Ju1rgen Frisch,2a Michael Du1rr,2a Nicolai Burzlaff,2a
Dietmar Stalke,2a Roland Fleischer,2a Waldemar Adam,2b Frank Prechtl,2b and
Alexander K. Smerz2b
Institut fu¨r Anorganische Chemie and Institut fu¨r Organische Chemie, University of Wu¨rzburg,
Am Hubland, D-97074 Wu¨rzburg, Germany
ReceiVed October 24, 1996X
Ionic ruthenium thioether complexes [Cp(LL′)Ru(SRR′)]PF6 (LL′ ) Ph2PCH2PPh2 (1), Ph2PC2H4PPh2 (2), (Ph3P,
CO) (3), Me2PC2H4PPh2 (4), (S,S)-Ph2PCHMeCHMePPh2 (5), SRR′ ) MeSPh (a), MeS-i-Pr (b), MeSBz (c),
i-PrSBz (d), EtSBz (e), MeSCy (f), SC4H8 (g)) were synthesized from the corresponding chloro complexes [Cp-
(LL′)RuCl] and thioethers. 5a crystallized in the orthorhombic system, space group P212121 (No. 19), with a )
11.269(3) Å, b ) 15.104(2) Å, c ) 23.177(4) Å, and Z ) 4. 5b crystallized in the monoclinic system, space
group P21 (No. 4), with a ) 10.539(5) Å, b ) 16.216(9) Å, c ) 11.011(8) Å, â ) 106.04(2)°, and Z ) 2. A
similar ligand exchange reaction yielded the analogous sulfoxide complexes [Cp(LL′)Ru(S(O)RR′)]PF6 (6-10).
10a crystallized in the orthorhombic system, space group P212121 (No. 19), with a ) 14.1664(13) Å, b ) 15.792-
(2) Å, c ) 17.641(2) Å, and Z ) 4. 10b‚0.93CH2Cl2 crystallized in the orthorhombic system, space group
P212121 (No. 19), with a ) 12.069(2) Å, b ) 17.379(2) Å, c ) 19.760(5) Å, and Z ) 4. The thioether complexes
can also be directly converted to sulfoxide complexes with the strong oxygen transfer reagent dimethyldioxirane
(DMD). No crossover products are formed when mixtures of two thioether complexes (e.g., 1a/2c or 1c/2a) are
treated with DMD, demonstrating that no Ru-S bond cleavage is involved. Moderate diastereoselectivities are
observed for the oxygen transfer to chiral, racemic thioether complexes 3 (8-28%) and 4 (34-60%). Oxidation
of the (S,S)-CHIRAPHOS complexes 5, however, is highly stereoselective (de ) 46-98%). Treatment of the
sulfoxide complexes 10 with sodium iodide removes the chiral, nonracemic sulfoxides from the metal with retention
of the configuration at sulfur.
Introduction
derived oxaziridines7 or achiral oxidants in the presence of chiral
catalysts including enzymes6,9 or the well-known Sharpless
reagent in its original8b or modified8a form. All of these
methods, however, suffer from certain drawbacks. The sulfinic
ester/amide route gives high selectivities only for aryl or tert-
butyl sulfoxides,5 enzyme reactions are often unpredictable with
respect to optical purity and yield of products,6 and enantiose-
lective oxidations including the Kagan method8a are reliable only
for the synthesis of aryl alkyl sulfoxides. Thus, although an
impressive range of methods is available for the enantioselective
generation of sulfoxides,10 there is still a need to develop novel
strategies which might favorably supplement the existing
methodologies.
Homochiral sulfoxides have a prominent role as intermediates
in enantioselective syntheses.3 Their methods of preparation
fall into roughly two categories: (i) nucleophilic substitution
at enantiomerically pure esters or amides of sulfinic acids4,5 and
(ii) enantioselective oxidation of thioethers.4,6-9 The latter
method may involve either chiral oxidants such as camphor-
* To whom correspondence should be addressed.
X Abstract published in AdVance ACS Abstracts, April 15, 1997.
(1) Enantioselective Organic Syntheses Using Chiral Transition Metal
Complexes. 2. Part 1: Schenk, W. A.; Frisch, J.; Adam, W.; Prechtl,
F. Angew. Chem. 1994, 106, 1699; Angew. Chem., Int. Ed. Engl. 1994,
33, 1609.
(2) (a) Institut fu¨r Anorganische Chemie. (b) Institut fu¨r Organische
Chemie.
(3) (a) Posner, G. H. Acc. Chem. Res. 1987, 20, 72. (b) Solladie, G. Pure
Appl. Chem. 1988, 60, 1699. (c) Durst, T. In ComprehensiVe Organic
Chemistry; Barton, D., Ollis, D., Eds.; Pergamon: Oxford, U.K., 1979;
Vol. 3, p 121.
Knowing that oxidations of R-chiral thioethers to sulfoxides
often proceed with high diastereoselectivities,10 we expected that
oxygen transfer to a thioether which is coordinated to a chiral
transition metal complex would also be highly stereoselective
(eq 1). Addition of a transition metal to a thioether will of
(4) Kagan, H. B.; Rebiere, F.; Samuel, O. Phosphorus Sulfur Silicon 1991,
58, 89.
(5) (a) Andersen, K. K.; Bujnicki, B.; Drabowicz, J.; Mikolajczyk, M.;
O’Brien, J. B. J. Org. Chem. 1984, 49, 4070. (b) Rebiere, F.; Samuel,
O.; Richard, L.; Kagan, H. B. J. Org. Chem. 1991, 56, 5991. (c)
Evans, D. A.; Faul, M. M.; Colombo, L.; Biaha, J. J.; Clardy, J.;
Cherry, D. J. Am. Chem. Soc. 1992, 114, 5977 and references cited
therein.
course dramatically reduce its reactivity; however, there is some
(6) Holland, H. L. Chem. ReV. 1988, 88, 473.
(7) (a) Davis, F. A.; McCauley, J. P.; Chattopadhyay, S.; Harakal, M. E.;
Towson, J. C.; Watson, W. H.; Tavanaiepour, I. J. Am. Chem. Soc.
1987, 109, 3370. (b) Melanidis, V.; Verfu¨rth, U.; Herrmann, R. Z.
Naturforsch., B 1990, 45, 1689. (c) Davis, F. A.; Reddy, R. T.; Han,
W.; Carroll, P. J. J. Am. Chem. Soc. 1992, 114, 1428. (d) Pointner,
A.; Herrmann, R. Z. Naturforsch., B 1995, 50, 1396.
(8) (a) Pitchen, P.; Dunach, E.; Deshmukh, M. N.; Kagan, H. B. J. Am.
Chem. Soc. 1984, 106, 8188. (b) DiFuria, F.; Modena, G.; Seraglia,
R. Synthesis 1984, 325.
(9) (a) Colonna, S.; Gaggero, N.; Manfredi, A.; Casella, L.; Gullotti, M.
J. Chem. Soc., Chem. Commun. 1988, 1451. (b) Colonna, S.; Gaggero,
N. Tetrahedron Lett. 1989, 30, 6233. (c) Colonna, S.; Gaggero, N.
Phosphorus Sulfur Silicon 1994, 95-96, 103.
(10) (a) Madesclaire, M. Tetrahedron 1986, 42, 5459. (b) Drabowicz, J.;
Kielbasinski, P.; Mikolajczyk, M. In The Chemistry of Sulfones and
Sulfoxides; Patai, S., Rappoport, B., Stirling, C. J. M., Eds.; Wiley:
New York, 1988; p 233.
S0020-1669(96)01280-3 CCC: $14.00 © 1997 American Chemical Society

Oxidation of Thioether Ligands in CpRu Complexes
Inorganic Chemistry, Vol. 36, No. 11, 1997 2373
were obtained by published methods or adaptations thereof.17 Thio-
ethers were obtained from Aldrich or prepared by alkylation of the
corresponding thiols. Oxidation of the thioethers with 3-chloroper-
benzoic acid gave the required sulfoxides. Dimethyldioxirane (DMD)
was employed as a freshly prepared 0.08-0.12 M solution in acetone.18
In the following, only representative examples are given. A full
description of experimental details is available as Supporting Informa-
tion.
Ruthenium Thioether Complexes 1-5. General Procedure.
[CpRu(LL′)Cl] (0.25 mmol), NH4PF6 (0.30 mmol), and the appropriate
thioether (1.00 mmol) were suspended in methanol (15 mL) and the
suspension was heated to 60 °C for 3 h (16 h for (LL′) ) (CO, PPh3)).
All volatiles were then removed under vacuum, and the residue was
extracted with several portions of dichloromethane. After filtration,
the products were precipitated by partial evaporation and addition of
diethyl ether.
precedent in the literature that electrophilic attack at coordinated
thioethers is still possible.11 Some isolated reports of oxidations
of thioether complexes do in fact exist;12 the conditions,
however, were such that these reactions almost certainly
proceeded via (i) dissociation of the thioether, (ii) oxidation,
and (iii) readdition of the sulfoxide. A mechanistically clean
oxidation of a coordinated thioether requires kinetically stable
complexes as well as a powerful yet selective oxygen transfer
reagent. We expected that dimethyldioxirane (DMD) would
(a) [CpRu(chir)(MeSPh)]PF6, 5a: yield 95%; mp 148-155 °C.
Anal. Calcd for C40H41F6P3RuS: C, 55.75; H, 4.80. Found: C, 56.32;
H, 5.10. 1H NMR (acetone-d6): δ 1.90 (s, SMe), 4.82 (s, Cp). 13C
NMR (acetone-d6): δ 31.9 (d, J(P,C) ) 4 Hz, SMe), 86.0 (dd, J(P,C)
) J(P′,C) ) 2 Hz, Cp). 31P NMR (acetone-d6): δ 63.8 (d, J(P,P) )
41 Hz), 82.1 (d, J(P,P) ) 41 Hz).
(b) [CpRu(chir)(MeS-i-Pr)]PF6, 5b: yield 98%; mp 187-189 °C.
Anal. Calcd for C37H43F6P3RuS: C, 53.69; H, 5.24. Found: C, 53.91;
H, 5.19. 1H NMR (acetone-d6): δ 0.82 (d, J(H,H) ) 6.8 Hz, Me),
1.07 (d, J(H,H) ) 6.7 Hz, Me), 1.50 (s, SMe), 1.97 (m, SCH), 4.73 (s,
Cp). 13C NMR (acetone-d6): δ 21.1 (s, Me), 21.6 (s, Me), 21.8 (s,
SMe), 45.6 (d, J(P,C) ) 5 Hz, SCH), 85.1 (dd, J(P,C) ) J(P′,C) ) 2
Hz, Cp). 31P NMR (acetone-d6): δ 64.4 (d, J(P,P) ) 42 Hz), 81.4 (d,
J(P,P) ) 42 Hz).
(c) [CpRu(chir)(MeSBz)]PF6, 5c: yield 96%; mp 212-217 °C dec.
Anal. Calcd for C41H43F6P3RuS: C, 56.23; H, 4.95. Found: C, 56.23;
H, 4.93. 1H NMR (acetone-d6): δ 1.41 (br, SMe), SCH2 signal at room
temperature too broad to be observed, 4.85 (s, Cp). 31P NMR (acetone-
d6): δ 66.8 (d, J(P,P) ) 40 Hz), 82.2 (d, J(P,P) ) 40 Hz).
Ruthenium Sulfoxide Complexes 6-10. General Procedure.
[CpRu(LL′)Cl] (0.25 mmol), NH4PF6 (0.30 mmol), and the appropriate
sulfoxide (1.50 mmol) were suspended in methanol (15 mL), and the
suspension was heated to 60 °C for 6 h (48 h for (LL′) ) (CO, PPh3)).
Reactions were then worked up as described for thioether complexes
1-5.
Oxidation of Thioether Complexes. General Procedure. To a
solution of the thioether complex (0.12 mmol) in acetone (10 mL) was
slowly added at 0 °C a 4-fold excess of a cooled (-30 °C) solution of
dimethyldioxirane in acetone. After 45 min (2 h in case of 3a-c), all
volatiles were removed under vacuum. Diastereoisomer rations were
determined from the NMR spectra of the crude reaction mixture.
Further purification was effected by crystallization from dichlo-
romethane/ether. Yields were nearly quantitative except those for 6d
(20%), 8b (45%), 8c (30%), 9d (10%), 10e (5%), 10f (70%), and 10h
(7%).
(a) [CpRu(chir)(MeS(O)Ph)]PF6, 10a,a′: yield 89%. Anal. Calcd
for C40H41F6OP3RuS: C, 54.73; H, 4.71. Found: C, 54.72; H, 4.88.
Major (93%) isomer 10a: 1H NMR (acetone-d6) δ 2.64 (s, SMe), 5.09
(s, Cp); 13C NMR (acetone-d6) δ 57.9 (s, SMe), 88.0 (dd, J(P,C) )
J(P′,C) ) 2 Hz, Cp); 31P NMR (acetone-d6) δ 60.3 (d, J(P,P) ) 36
Hz), 81.8 (d, J(P,P) ) 36 Hz). Minor (7%) isomer 10a′: 1H NMR
(acetone-d6) δ 2.82 (s, SMe), 4.90 (s, Cp); 31P NMR (acetone-d6) δ
62.3 (d, J(P,P) ) 36 Hz), 79.1 (d, J(P,P) ) 36 Hz). Careful
crystallization from dichloromethane/hexane gave a sample of pure (RS)-
10a, mp 124 °C dec.
fulfill these requirements.13 DMD had, inter alia, been used
previously for oxygen transfer to remote thioether functions such
as in A or B14,15 and to the sulfur of transition metal thiolate
complexes of type C.14 A preliminary account of the work
described here has been published.1
Experimental Section
Analytical Measurements. C, H, and S analyses were carried out
by the Analytical Laboratory of the Institute of Inorganic Chemistry,
University of Wu¨rzburg. Melting points were determined in sealed
capillaries in a copper block apparatus. Infrared spectra were run
on a Bruker IFS 25 instrument. 1H, 13C{1H}, and 31P{1H} NMR
spectra were recorded using a Bruker AMX 400 instrument. Chemical
shifts are reported relative to TMS (1H, 13C) or 85% H3PO4 (31P).
Enantiomeric excesses of the sulfoxides were determined by HPLC
(Knauer HPLC 64) using a Ciralcel OD column (DAICEL Chemical
Industries Ltd.), hexane/2-propanol (9:1) as eluent, and combined UV
(Hewlett-Packard 1040 A) and ChiraLyzer (IBM Messtechnik) detec-
tion.
Materials. RuCl3‚3H2O was purchased from Degussa AG, Hanau,
Germany; 2(S),3(S)-bis(diphenylphosphino)butane [(S,S)-CHIRAPHOS,
henceforth abbreviated as “chir”] was obtained from Strem Chemicals
and used without further purification. The phosphine ligands dppm,
dppe, and (2-(dimethylphosphino)ethyl) diphenylphosphine (dpme) were
prepared as described in the literature.16 The ruthenium complexes
[CpRu(LL′)Cl] (LL′ ) (PPh3)2; dppm; dppe; CO, PPh3; dpme; chir)
(11) (a) Adams, R. D.; Chodosh, D. F. J. Am. Chem. Soc. 1978, 100, 812.
(b) Adams, R. D.; Blankenship, C.; Segmu¨ller, B. E.; Shiralian, M. J.
Am. Chem. Soc. 1983, 105, 4319. (c) Yoshida, T.; Adochi, T.; Sato,
K.; Baba, K.; Kanokogi, T. J. Chem. Soc., Chem. Commun. 1993,
1511.
(12) (a) Biscarini, P.; Fusina, L.; Nivellini, G. D. J. Chem. Soc. A 1971,
1128. (b) Goggin, P. L.; Goodfellow, R. J.; Reed, F. J. S. J. Chem.
Soc., Dalton Trans. 1974, 576. (c) Fergusson, J. E.; Page, C. T.;
Robinson, W. T. Inorg. Chem. 1976, 15, 2270.
(13) (a) Adam, W.; Curci, R.; Edwards, J. O. Acc. Chem. Res. 1989, 22,
205. (b) Murray, R. W. Chem. ReV. 1989, 89, 1187. (c) Adam, W.;
Hadjiarapoglou, L. P.; Curci, R.; Mello, H. In Organic Peroxides;
Ando, W., Ed.; Wiley: New York, 1992; p 195. (d) Adam, W.; Smerz,
A. K. Bull. Chim. Soc. Belg. 1996, 105, 581.
(14) Schenk, W. A.; Frisch, J.; Adam, W.; Prechtl, F. Inorg. Chem. 1992,
31, 3329.
(15) Perez-Encabo, A.; Perrio, S.; Slawin, A. M. Z.; Thomas, S. E.;
Wierzchleyski, A. T.; Williams, D. J. J. Chem. Soc., Chem. Commun.
1993, 1059.
(16) (a) Hewertson, W.; Watson, H. R. J. Chem. Soc. 1962, 1491. (b)
Butter, S. A.; Chatt, J. Inorg. Chem. 1974, 15, 185. (c) King, R. B.;
Cloyd, J. C. J. Am. Chem. Soc. 1975, 97, 53.
(b) [CpRu(chir)(MeS(O)-i-Pr)]PF6, 10b,b′: yield 86%. Anal.
Calcd for C37H43F6OP3RuS: C, 52.67; H, 5.14. Found: C, 53.01; H,
5.28. Major (88%) isomer 10b: 1H NMR (acetone-d6) δ 0.89 (d,
(17) (a) Bruce, M. I.; Hameister, C.; Swincer, A. G.; Wallis, R. C. Inorg.
Synth. 1982, 21, 78. (b) Ashby, G. S.; Bruce, M. I.; Tomkins, I. B.;
Wallis, R. C. Aust. J. Chem. 1979, 32, 1003. (c) Davies, S. G.;
Simpson, S. J. J. Chem. Soc., Dalton Trans. 1984, 993. (d) Consiglio,
G.; Morandini, F.; Bangerter, F. Inorg. Chem. 1982, 21, 455.
(18) (a) Adam, W.; Bialas, J.; Hadjiarapoglou, L. P. Chem. Ber. 1991, 124,
2377. (b) Murray, R. W.; Jeyaraman, R. J. Org. Chem. 1985, 50,
2817.

2374 Inorganic Chemistry, Vol. 36, No. 11, 1997
Schenk et al.
Table 1. Crystallographic Data for 5a, 5b, 10a, and 10b
5a
5b
10a
10b
empirical formula
fw
temp, K
space group
a, Å
b, Å
c, Å
â, deg
V, Å3
Z
C40H41F6P3RuS
861.81
293
P212121 (No. 19)
11.269(3)
15.104(2)
23.177(4)
90
3944.8(13)
4
C37H43F6P3RuS
827.80
193
P21 (No. 4)
10.539(5)
16.216(9)
11.011(8)
106.04(2)
1809(2)
2
C40H41F6OP3RuS
877.81
293
P212121 (No. 19)
14.1664(13)
15.792(2)
17.641(2)
90
3946.5(7)
4
C37H43F6OP3RuS‚0.93CH2Cl2
843.80 + 79.0
293
P212121 (No. 19)
12.069(2)
17.379(2)
19.760(5)
90
4144.7(14)
4
λ, Å
0.710 73
1.451
2.6
0.071
0.097
0.710 73
1.520
6.8
0.038
0.098
0.710 73
1.477
2.65
0.050
0.128
0.710 73
1.479
3.72
0.048
0.136
F
calc, g cm-3
µ, cm-1
R1 [I > 2σ(I)]
wR2 [I > 2σ(I)]a
2
2
a wR2 ) {[∑w(Fc2 - Fo )2]/[∑w(Fo )2]}1/2
.
Table 2. Bond Distances (Å) and Angles (deg) within the Cations
of 5a, 5b, 10a, and 10b
J(H,H) ) 7.0 Hz, Me), 1.08 (d, J(H,H) ) 6.9 Hz, Me), 2.16 (s, SMe),
2.48 (m, SCH), 4.93 (s, Cp); 13C NMR (acetone-d6) δ 15.4 (s, Me),
16.5 (s, Me), 45.3 (d, J(P,C) ) 2 Hz, SMe), 62.7 (d, J(P,C) ) 1 Hz,
SCH), 86.7 (dd, J(P,C) ) J(P′,C) ) 2 Hz, Cp); 31P NMR (acetone-d6)
δ 57.8 (d, J(P,P) ) 37 Hz), 80.6 (d, J(P,P) ) 37 Hz). Minor (12%)
isomer 10b′: 1H NMR (acetone-d6) δ 5.01 (s, Cp), other signals not
detected; 31P NMR (acetone-d6) δ 60.4 (d, J(P,P) ) 37 Hz), 78.6 (d,
J(P,P) ) 37 Hz). Careful crystallization from dichloromethane/hexane
gave a sample of pure (RS)-10b.
(c) [CpRu(chir)(MeS(O)Bz)]PF6, 10c,c′: yield 87%. Anal. Calcd
for C41H43F6OP3RuS: C, 55.22; H, 4.86. Found: C, 54.95; H, 4.88.
50% isomer 10c: 1H NMR (acetone-d6) δ 2.25 (s, SMe), 3.28 (d, J(H,H)
) 13.5 Hz, SCH2), 3.92 (d, J(H,H) ) 13.5 Hz, SCH2), 5.15 (s, Cp);
13C NMR (acetone-d6) δ 49.4 (d, J(P,C) ) 2 Hz, SMe), 69.8 (d,
J(P,C) ) 2 Hz, SCH2), 87.1 (dd, J(P,C) ) J(P′,C) ) 2 Hz, Cp);
31P NMR (acetone-d6) δ 61.8 (d, J(P,P) ) 36 Hz), 80.9 (d, J(P,P) )
36 Hz). 50% isomer 10c′: 1H NMR (acetone-d6) δ 1.96 (s, SMe),
3.78, 3.82 (AB system, J(H,H) ) 13.6 Hz, SCH2), 5.01 (s, Cp); 13C
NMR (acetone-d6) δ 48.4 (d, J(P,C) ) 2 Hz, SMe), 70.5 (dd, J(P,C) )
J(P′,C) ) 1 Hz, SCH2), 87.2 (dd, J(P,C) ) J(P′,C) ) 2 Hz, Cp); 31P
NMR (acetone-d6) δ 61.2 (d, J(P,P) ) 37 Hz), 80.8 (d, J(P,P) ) 37
Hz).
Liberation of the Sulfoxides from 10a-c. The complex (0.10
mmol), sodium iodide (0.50 mmol), and acetone (5 mL) were heated
under reflux (15 h). The mixture was then evaporated to dryness,
and the residue was extracted with dichloromethane (2 mL) and
chromatographed over a short (10 cm) silica column. First, the complex
[CpRu(chir)I] (11) was eluted with dichloromethane as a yellow band,
and then, using acetone as eluent, the sulfoxides were removed from
the column. Upon evaporation of the solvent, 12a-c remained as
colorless oils in quantitative yield (by 1H NMR). The ee’s as
determined by HPLC were identical with the de’s of the corresponding
complexes.
5a
5b
10a
10b
Ru-S
2.349(3)
2.305(3)
2.282(3)
1.877
1.795(10) 1.855(7)
1.787(12) 1.817(8)
2.380(4)
2.307(2)
2.301(2)
1.894
2.2791(14) 2.299(2)
2.3334(14) 2.329(2)
Ru-P(1)
Ru-P(2)
Ru-Cpa
S-C(Me)
S-C(R)
S-O
2.293(2)
1.891
2.284(2)
1.903
1.798(11)
1.858(11)
1.476(7)
1.799(6)
1.803(6)
1.479(4)
P(1)-Ru-P(2)
P(1)-Ru-S
P(2)-Ru-S
Ru-S-C(Me)
Ru-S-C(R)
Ru-S-O
82.51(11) 83.50(5)
95.25(11) 94.32(10) 97.07(5)
85.90(11) 82.89(11) 88.97(5)
81.85(5)
82.45(8)
92.74(8)
89.52(8)
110.7(4)
116.0(4)
117.9(3)
103.6(5)
103.5(5)
103.6(5)
110.2(4)
115.3(5)
106.1(3)
121.2(4)
110.1(3)
116.1(2)
119.7(2)
99.3(3)
104.7(3)
104.3(3)
C(R)-S-C(Me) 99.2(5)
C(R)-S-O
101.3(4)
C(Me)-S-O
a Cp Denotes the midpoint of the cyclopentadienyl ring.
unit. The disordered PF6- ion in 10a was refined to a split occupancy
of 0.53 and 0.47 with distance restraints.
(b) X-ray Measurements of 5b. Crystals were grown from
dichloromethane/hexane. The data were collected on a Stoe-Huber-
Siemens diffractometer fitted with a Siemens CCD-detector at a
temperature of 153 K20 using Mo KR radiation. A semiempirical
absorption correction was applied. The structure was solved by direct
methods with SHELXS-90.19b The disordered S(Me)-i-Pr moiety was
refined to a split occupancy of 0.66 and 0.34, while the disordered
-
PF6 anion was refined to a split occupancy of 0.70 and 0.30,
respectively, using distance and rigid-bond restraints with the anisotropic
displacement parameters being similar. All structures were refined by
full-matrix least-squares procedures on F2, with a weighting scheme
Anal. Calcd for 11 C33H33IP2Ru: C, 55.08; H, 4.62. Found: C,
55.37; H, 4.52. 1H NMR (CDCl3): δ 1.05 (dd, J(H,H) ) 7.1 Hz, J(P,H)
) 11.7 Hz, Me), 1.15 (dd, J(H,H) ) 6.9 Hz, J(P,H) ) 10.9 Hz, Me),
2.12 (m, CH), 3.07 (m, CH), 4.45 (s, Cp). 13C NMR (CDCl3): δ 15.8
2
2
w-1 ) σ2Fo + (g1P)2 + g2P, where P ) (Fo + 2Fc2)/3, using
SHELXL93.19c All non-hydrogen atoms were refined anisotropically,
and a riding model was employed in the refinement of the hydrogen
atom positions. Relevant crystallographic data can be found in Table
1, and selected bond lengths and angles are give in Table 2. Further
details on the structure investigation are reported as Supporting
Information or may be obtained from the Fachinformationszentrum
Karlsruhe, D-76344 Eggenstein-Leopoldshafen, Germany, on quoting
the depository numbers CSD 405919 (5a), CSD 406080 (5b), CSD
405920 (10a), and CSD 405921 (10b) and the full journal citation.
2
3
2
(dd, J(P,C) ) 14 Hz, J(P,C) ) 4 Hz, Me), 17.5 (dd, J(P,C) ) 16
Hz, 3J(P,C) ) 2 Hz, Me), 37.7 (dd, 1J(P,C) ) 27 Hz, 2J(P,C) ) 16 Hz,
CH), 40.3 (dd, J(P,C) ) 31 Hz, J(P,C) ) 17 Hz, CH), 81.2 (s, Cp).
31P NMR (CDCl3): δ 73.6 (d, J(P,P) ) 33 Hz), 81.9 (d, J(P,P) ) 33
Hz).
1
2
Crystallographic Studies. (a) X-ray Measurements of 5a, 10a,
and 10b. Crystals were grown from dichloromethane/hexane. The
data sets were collected on an Enraf-Nonius CAD4 using Mo KR
radiation. Semiempirical absorption corrections were applied.19a The
structures were solved by Patterson methods with SHELXS-86.19b 10b
crystallizes with one molecule of dichloromethane in the asymmetric
Results
Synthesis of Ruthenium Thioether and Sulfoxide Com-
plexes. Substitution of Cl- in a polar medium21 is the method
(19) (a) North, A. C. T.; Phillips, D. C.; Mathews, F. S. Acta Crystallogr.
1968, A24, 351. (b) Sheldrick, G. M. Acta Crystallogr. 1990, A46,
467. (c) Sheldrick, G. M. Program for crystal structure refinement.
University of Go¨ttingen, 1993.
(20) Kottke, T.; Stalke, D. J. Appl. Crystallogr. 1993, 26, 615.
(21) Davies, S. G.; McNally, J. P.; Smallridge, A. J. AdV. Organomet. Chem.
1990, 30, 1.

Oxidation of Thioether Ligands in CpRu Complexes
Inorganic Chemistry, Vol. 36, No. 11, 1997 2375
of choice for the preparation of the ionic thioether complexes
1-5 (eq 2; 2c and 3c had been obtained previously by
methylation of the corresponding phenylmethanethiolate com-
plexes).22 The products are bright yellow, moderately air-stable
compounds which are quite readily soluble in polar organic
solvents.
Figure 1. Ortep plot and space-filling model of the cation [CpRu-
(chir)(MeSPh)]+ (5a+).
At room temperature, thioether complexes undergo a rapid
pyramidal inversion at sulfur.23 Therefore, complexes 1 and 2
are achiral on the NMR time scale and accordingly give the
the aryl groups at the chelate ligand occupy considerably more
space than does the Cp ring,27 and even for the less bulky ligand
combinations PPh3/CO and PPh3/NO, this is the preferred
orientation.25,26 In 5b, the organic groups at the sulfur atom
(Me and i-Pr) are of very different sizes. Consequently, the
diastereoisomer ratio at -60 °C is quite high [94:6. Major
isomer: 1H NMR δ 0.58 (d, J(H,H) ) 7.0 Hz, i-Pr), 1.08 (d,
J(H,H) ) 6.6 Hz, i-Pr), 1.33 (s, SMe), 4.76 (s, Cp); 31P NMR
δ 62.7, 81.7 (AX system, J(P,P) ) 41 Hz). Minor isomer: 1H
NMR δ 1.06 (d, J(H,H) ) 7.3 Hz, i-Pr), 1.14 (d, J(H,H) ) 6.7
Hz, i-Pr), 1.47 (s, SMe), 4.84 (s, Cp); 31P NMR δ 63.4, 81.5
(AX system, J(P,P) ) 42 Hz)]. Crystal structure determinations
of 5a and 5b (Figures 1 and 2) unequivocally corroborate the
spectroscopically determined conformations and indicate that
in all likelihood this rotational orientation is the preferred one
for all complexes of this kind. Bond distances and angles
around the ruthenium atom (Table 2) are within the expected
range.21 The steric strain between the thioether and phosphine
ligands is apparent from an increased Ru-S bond distance in
5b. The orientation of the phenyl groups of the phosphine
ligand deserves some comment. In square-planar complexes
relevant to catalysis, it is often observed that chelating chiral
phosphines adopt an “edge-face” conformation, which means
that the coordinated substrate molecule is exposed to the edge
of a phenyl group on one side and to the face of a phenyl group
on the opposite side. Indeed, it is frequently emphasized that
this “edge-face” arrangement is primarily responsible for
diastereoselection.28 While the structure of 5a conforms to this
expectation, it is easily seen that the thioether ligand in 5b is
on both sides framed by the faces of phenyl groups. Thus
2(S),3(S)-bis(diphenylphosphino)butane is, despite the fixed
configuration at the two carbon atoms of the backbone, a
remarkably flexible ligand.29 The R configuration imposed on
the thioether ligands is, therefore, mainly due to the puckering
of the five-membered chelate ring, which pushes one phenyl
1
expected simple H, 13C, and 31P NMR spectra. For the same
reason, chiral racemic complexes 3 and 4 do not form
distinguishable diastereoisomers. In some cases, line-broaden-
ing can be observed even at room temperature, but we have
not investigated this in much detail since this has been done
previously for many similar cases,23 including the closely related
ruthenium complexes [CpRu(dppe)(SRR′)]+ (R, R′ ) Et, Ph)24
as well as a number of chiral rhenium cations [CpRe(NO)(PPh3)-
(SRR′)]+.25 A noteworthy feature of the 13C NMR spectra of
3
complexes 3 is the lack of observable coupling J(P,C) of the
SMe group. X-ray structure determinations of [CpRu(CO)-
(PPh3)(MeS(O)-t-Bu)]SbF626 and the above-mentioned rhenium
complexes25 indicate that the thioether ligand assumes a
preferred rotational orientation with the dihedral angle Me-
S-Ru-P close to 90°. For similar reasons, the two possible
couplings, 3J(P,C) in chiral, racemic complexes 4 as well as in
chiral, enantiomerically pure complexes 5 are very different.
For two examples of the latter case, we have recorded low-
temperature NMR spectra. In 5c, the two groups R and R′ are
of similar sizes, and consequently two diastereoisomers are
observed at -40 °C in a 60:40 ratio. [Major isomer: 1H NMR
δ 1.40 (s, SMe), 2.37, 3.54 (AB system, J(H,H) ) 13.0 Hz,
SCH2), 4.95 (s, Cp); 31P NMR δ 66.4, 82.4 (AX system, J(P,P)
) 39 Hz). Minor isomer: 1H NMR δ 1.20 (s, SMe), 2.89,
3.50 (AB system, J(H,H) ) 13.2 Hz, SCH2), 4.85 (s, Cp); 31P-
NMR δ 65.7, 82.1 (AX system, J(P,P) ) 40 Hz)]. NOE
measurements at this temperature indicate that, for each of the
two diastereoisomers, the rotamer with both the methyl and the
benzyl groups oriented toward the Cp ring is the preferred form
(irradiation of the Cp resonances gives approximately equal
enhancements of the methyl and one of the benzyl signals, and
Vice Versa). This rotational preference is quite expected since
(22) Schenk, W. A.; Stur, T. Z. Naturforsch., B 1990, 45, 1495.
(23) Abel, E. W.; Bhargava, S. K.; Orrell, K. G. Prog. Inorg. Chem. 1984,
32, 1.
(24) Ohkita, K.; Kurosawa, H.; Hirao, T.; Ikeda, I. J. Organomet. Chem.
1994, 470, 179.
(25) Mendez, N. Q.; Arif, A. M.; Gladysz, J. Organometallics 1991, 10,
2199.
(27) (a) Seeman, J. I.; Davies, S. G. J. Chem. Soc., Chem. Commun. 1984,
1019. (b) Davies, S. G.; Seeman, J. I. Tetrahedron Lett. 1984, 25,
1845. (c) Seeman, J. I. Pure Appl. Chem. 1987, 59, 1661.
(28) Halpern, J. Science 1982, 217, 401 and references cited therein.
(29) Orpen, A. G. Chem. Soc. ReV. 1993, 22, 191.
(26) Faller, J. W.; Ma, Y. Organometallics 1992, 11, 2726.

2376 Inorganic Chemistry, Vol. 36, No. 11, 1997
Schenk et al.
Figure 2. Ortep plot and space-filling model of the cation [CpRu-
(chir)(MeS-i-Pr)]+ (5b+).
Figure 3. Ortep plot and space-filling model of the cation [CpRu-
(chir)(MeS(O)Ph)]+ (10a+).
group toward the thioether ligand. Thus, the coordination site
occupied by the thioether ligand breaks down into three sectors
of decreasing size which are taken up by the substituent R, the
CH3 group, and the remaining lone pair at sulfur. Since there
exist no pronounced intermolecular contacts in the lattices of
5a and 5b, we can be sure that also in solution the R
configuration at sulfur is thermodynamically favored.
For the synthesis of the sulfoxide complexes 6-10 consider-
ably longer reaction times are required to compensate for the
reduced nucleophilicity of sulfoxides (eq 3). The sulfoxide
been made with [CpRu(CO)(PPh3)(MeS(O)-t-Bu)]+ 26 and
related rhenium sulfoxide complexes.25 Complexes 10 finally
are formed as mixtures of diastereoisomers; the observed de’s
(10a,a′, 86%; 10b,b′, 76%; 10c,c′, 0%) are the result of a
preference of the chiral, enantiomerically pure metal fragment
[Cp(chir)Ru]+ for one enantiomer of the sulfoxide. NOE-
difference spectra of 10a and 10c,c′ were recorded; in all cases,
strong signal enhancements were found between the Cp ligand
and the methyl and benzyl protons of the groups at sulfur. This
means that in the sulfoxide complexes, too, the organic
substituents at sulfur are oriented toward the Cp ligand
while oxygen, the smallest substituent, occupies the crowded
space between the phenyl groups of the bidentate phosphine
ligand, as expected. This is again corrobated by the crystal
structure determination of the major diastereoisomers 10a and
10b with the R configuration at sulfur (Figures 3 and 4). Bond
distances and angles (Table 2) are again within the expected
ranges.21,26 The Ru-S bonds are shorter than those in the
corresponding thioether complexes due to contraction of the
sulfur valence orbitals brought about by the electronegative
oxygen atom.
Oxygen Transfer from Dimethyldioxirane to Thioether
Complexes. To test the feasibility of the oxidation concept
outlined in eq 1, complexes 1 and 2 were treated with an excess
of DMD at temperatures between -40 and 0 °C (eq 4). No
complexes are light yellow, moderately air-stable compounds
which in their physical properties closely resemble the analogous
thioether complexes 1-5. The formation of 7d and 10d,d′ could
only be observed spectroscopically. Apparently, the combina-
tion of a bulky sulfoxide and a bulky metal complex makes the
Ru-S bond quite labile.
Sulfoxides are configurationally stable. Consequently, com-
plexes 6 and 7 are chiral which can be seen, inter alia, from
the nonequivalence of the two phosphorus nuclei. Complexes
8 and 9 are formed as pairs of enantiomeric diastereoisomers.
The dpme ligand imparts only negligible diastereo-discriminat-
ing ability to the complex. Quite high diastereomeric excesses,
however, can be found for the ligand combination CO/PPh3
(8a,a′, 4%; 8b,b′, 72%; 8c,c′, 72%). Similar observations have
reaction was observed between 2d and DMD, and 1d gave only

Oxidation of Thioether Ligands in CpRu Complexes
Inorganic Chemistry, Vol. 36, No. 11, 1997 2377
much better too (eq 6) but were still too low to be synthetically
useful. Since, additionally, the diastereomer separation of the
related starting material [NmcpRu(dpme)Cl] (Nmcp ) neo-
menthylcyclopentadienyl) was more tedious than expected,30 we
finally abandoned the concept of chirality-at-the-metal.
Much better results, in terms of yield and diastereoselectivity,
were obtained with oxidation of the 2(S),3(S)-bis(diphenylphos-
phino)butane complexes 5 (eq 7). In the case of the methyl
Figure 4. Ortep plot and space-filling model of the cation [CpRu-
(chir)(MeS(O)-i-Pr)]+ (10b+).
low yields of 6d, while the other sulfoxide complexes were
produced in good yields and without noticeable side reactions.
When the DMD oxidations were carried out under an atmo-
sphere of carbon monoxide, no CO incorporation was found.
In a crossover experiment, an equimolar mixture of 1a and 2c
was treated with DMD, giving only the sulfoxide complexes
6a and 7c and none of the crossover products 6c and 7a.
Analogously, when a mixture of 1c and 2a was treated with
DMD, only 6c and 7a and none of the crossover products 6a
and 7c were obtained. To compare the reactivities of coordi-
nated and free thioether, an equimolar mixture of 1a and
thioanisol was treated with a slight excess of DMD at 0 °C.
The thioether was oxidized to sulfoxide and sulfone, while most
of 1a remained unreacted. Not unexpectedly, coordinated
thioethers are much more difficult to oxidize than uncoordinated
ones.
thioether complexes, a 4-fold excess of DMD at 0 °C in acetone
was sufficient to give nearly quantitative conversions to the
sulfoxide complexes 10. Conversions drop sharply when both
substituents at sulfur are sterically more demanding (5e). Using
a larger excess of DMD in this case leads to increased
decomposition. Diastereoselectivities are, with the conspicuous
exception of the thioanisole complex 10a, excellent. A distinct
advantage of DMD as an oxidant is the fact that any excess
can be readily removed under vacuum. This greatly simplifies
the isolation of the sulfoxide complexes.
The following experiments were aimed at the development
of a novel strategy for the enantioselective oxidation of
thioethers. First, the chiral racemic carbonyl complexes 3a-c
were treated with DMD at 0 °C (eq 5). Monitoring the reaction
Liberation of the Sulfoxides. To liberate the sulfoxides from
the metal, complexes 10a-c were refluxed with sodium iodide
in acetone (eq 8). The iodo complex 11 was isolated from the
by NMR revealed that the oxidation is much slower in this case
than in the case of complexes 1 and 2. Even with a 10-fold
excess of DMD, conversions were less than quantitative, and
the diastereoselectivity was disappointing. Next, the more
electron-rich complexes 4a-d were chosen as substrates. Here,
indeed, a quantitative oxidation to the sulfoxide complexes
9a-d could be readily achieved. Diastereoselectivities were
crude reaction mixture by chromatography in almost quantitative
yield. With AgPF6 as the halide abstraction reagent, 11 was
employed again to prepare the thioether complexes 5. The
enantiomeric purity of sulfoxides 12 was checked by HPLC
(30) (a) Bezler, J. Dissertation, University of Wu¨rzburg, 1995. (b) Schenk,
W. A.; Bezler, J. Unpublished results.

2378 Inorganic Chemistry, Vol. 36, No. 11, 1997
Schenk et al.
using a Chiralcel OD column in combination with UV and optical
rotation detectors. The reaction according to eq 8 was repeated
by employing diastereomerically pure samples of 10a and 10b.
In all cases, the ee’s of the sulfoxides were identical to the de’s
of the complexes. The absolute configurations follow from the
structures of 10a and 10b and the known specific rotations of
12a and 12c.3a,6a In our previous publication,1 we erroneously
assigned the R configuration to the major enantiomer of 12b
by analogy to the reported3a absolute configurations of (R)-(+)-
i-BuS(O)Me and (S)-(-)-n-PrS(O)Me.
Scheme 1
Discussion
A kinetically stable metal-sulfur bond is a necessary
prerequisite for chirality transfer from a metal complex to sulfur
as outlined as eq 1. A low-spin, d6-cationic complex of a
second-row late transition element offers the best probability
to fulfill this requirement. We22 and others24,26 found previously
that neither thioethers nor sulfoxides dissociate readily from 18-
electron complexes of the type described here. Nevertheless,
the presence of a strong oxidant in the reaction mixture could
lead to the formation of 17-electron intermediates which might
then undergo rapid ligand exchange, possibly even in a catalytic
cycle.31 Good evidence that this does not occur comes from
the observation that no carbonyl complexes are formed when
the oxidations are carried out in the presence of carbon
monoxide. The crossover experiments described in the preced-
ing section finally prove that the oxygen transfer takes place at
the complex, while the metal-sulfur bond remains intact.
Furthermore, diastereoselectivities of the formation of sulfoxide
complexes 8-10 by oxidation (eqs 5-7) are very different from
those obtained by ligand exchange. Striking examples are
10c,c′, which are formed by ligand exchange with 0% de and
by oxidation with >98% de, and 8b,b′, for which opposite
diastereoisomers are favored by the two different routes. Not
only is this additional proof that oxidation does not involve
ligand dissociation and readdition, but it also demonstrates that
the observed selectivities are a result of kinetic control. A
mechanism which involves the formation of a RudO intermedi-
ate followed by an O atom shift to sulfur can be ruled out on
the basis of the observation that the de of the oxidation depends
markedly on the nature of the oxidant.32
A detailed interpretation of the oxygen transfer reaction has
to take into account all possible rotamers/diastereomers of the
thioether complexes 5 (Scheme 1). Of these, only (R)-A and
(S)-A may be observed by NMR at low temperature. The fact
that there is no correlation between (R)/(S) equilibria and the
diastereoselectivity of the oxidation strongly suggests that
rotamers A are not the reactive species. Of the remaining
conformers, (R)-B and (S)-B should be present in higher
proportions than (R)-C and (S)-C, in which the largest sub-
stituent on sulfur occupies the narrowest space of the [CpRu-
(chir)] complex. The final outcome of the reaction will,
therefore, be determined by the ratio (R)-B/(S)-B (which is in
favor of (S)-B since there the largest substituent R occupies the
largest sector around ruthenium) and the relative reactivity of
these two conformers. The predominant formation of (S)-
sulfoxides indicates that oxygen transfer is fastest for the
conformer (R)-B, in which the oxidant can approach the sulfur
atom from the sterically least encumbered direction. (Note that
if all steps in the sequence [Ru]-SMeR f [Ru]-S(O)MeR f
RS(O)Me proceed with retention of configuration at sulfur, the
stereochemical descriptor changes as (R) f (R) f (S)). This
interpretation also explains why the oxygen transfer is largely
limited to methyl thioethers: for substituents larger than CH3
even conformers B become unaccessible.33
The limitation to methyl sulfoxides can probably be overcome
in two ways: first, oxidants which are more reactive or thermally
more stable than DMD should be able to attack rotamers B even
if present in only very small concentrations; second, other chiral
diphosphines might favor rotamers analogous to B while at the
same time give high diastereomeric excesses. Work in both
directions is in progress in our laboratory.
Acknowledgment. This work was generously supported by
the Deutsche Forschungsgemeinschaft (SFB 347). We are
indebted to Dr. U. Hoch, Institute of Pharmacy and Food
Chemistry, University of Wu¨rzburg, for carrying out the HPLC
separations of the sulfoxides.
Supporting Information Available: Analytical and spectroscopic
textual data for the thioether and sulfoxide complexes 1-10 (14 pages).
X-ray crystallographic files, in CIF format, for compounds 5a, 5b, 10a,
and 10b are available on the Internet only. Ordering and access
information is given on any current masthead page.
IC961280F
(33) A reviewer suggested that in a mechanism such as described by Scheme
1 the ee’s might be temperature dependent. We have found no
pronounced changes in the -20 to 0 °C range (at higher temperatures
DMD decomposes too rapidly), but are addressing this question using
other oxidants.
(31) (a) Astruc, D. Angew. Chem. 1988, 100, 662; Angew. Chem., Int. Ed.
Engl. 1988, 27, 643. (b) Baird, M. C. Chem. ReV. 1988, 88, 1217.
(32) Schenk, W. A.; Du¨rr, M. To be published.