Syntheses and characterization of titanium(IV) and titanium(III) complexes with (2-dimethylphosphino)ethane-1-thiolate and (3-dimethylphosphino)propane-1-thiolate as ligands

Article abstract of DOI:10.1021/ic034353l

Reactions of Cp₂TiCl₂ (Cp = η ⁵-cyclopentadienide) with 2 or 1 equiv of hybrid P-S ligands (L), (CH₃)₂P(CH₂)_nS^- n = 2, dmpet; n = 3, dmppt), produced new dicyclopentadienyltitanium(IV) complexes with L, Cp₂Ti(L-κS)₂ (1, L = dmpet; 2, L = dmppt) and [Cp₂Ti(L-κ²S,P)]BPh₄ (3, L = dmpet; 4, L = dmppt). The Ti(III) complexes, Cp₂Ti(L-κ²S,P) (5, L = dmpet; 6, L = dmppt), were prepared by the reaction of Cp ₂Ti(η³-C₃H₅) with 1 equiv of L. The structures of complexes 1-6 were confirmed by X-ray diffraction analyses. It was found that complexes 3 and 5 were isostructural around Ti(IV) and Ti(III) centers: the Ti(IV)-S bond length in 3 (2.3498(9) A) is shorter by 0.14 A than Ti(III)-S in 5 (2.4877(7) A), while Ti(IV)-P (2.534(1) A) was merely 0.05 A shorter than Ti(III)-P (2.5844(7) A). The redox potential between 3 and 5 in acetonitrile was -1.14 V vs the ferricinium/ferrocene couple. A heterobimetallic complex that has the frame of complex 1, [Cp₂Ti(dmpet)₂Cu]PF₆ (7), was also isolated and structurally characterized: the Ti-Cu distance (2.95(1) A) was shorter than that in [Cp₂Ti(SC₂H₄PPh ₂)₂Cu]BF₄, previously reported by White and Stephan. Structural characterization was also carried out for Cp*Ti(dmpet-κS)₂(dmpet-κ²S,P) (8) and CpTiCl₂(dmppt-κ²S,P) (9), which were obtained by the reactions of Cp*(or Cp)TiCl₃ (Cp* = η ⁵-C₅Me₅^-) with n equiv (n = 1-3) of L. The mutual site-exchange reaction between phosphorus atoms on a coordinated dmpet in the κ²S,P mode and on two other coordinated dmpet's in the κS mode within complex 8 was analyzed by the variable-temperature ³¹P{¹H} dynamic NMR method. The kinetic parameters for this process, k_ex²⁹⁸ = 1.9 × 10⁵ s ^-1, ΔH? = 48 kJ mol^-1, and ΔS? = 17 J mol^-1 K^-1, as well as the rather long Ti(IV)-P distance (2.652(1) A), indicate the fluxional nature of the coordination geometry in complex 8.

Full text of DOI:10.1021/ic034353l

Inorg. Chem. 2003, 42, 5320−5329
Syntheses and Characterization of Titanium(IV) and Titanium(III)
Complexes with (2-Dimethylphosphino)ethane-1-thiolate and
(3-Dimethylphosphino)propane-1-thiolate as Ligands
Kazutaka Matsuzaki,† Hiroyuki Kawaguchi,‡ Peter Voth,† Kyoko Noda,§ Sumitaka Itoh,§
,§
Hideo D. Takagi,† Kazuo Kashiwabara,* and Kazuyuki Tatsumi†
Research Center for Materials Science and Department of Chemistry, Nagoya UniVersity,
Nagoya 464-8602, Japan, and Institute for Molecular Science, Okazaki, Aichi, 444-8585 Japan
Received April 1, 2003
Reactions of Cp2TiCl2 (Cp ) η5-cyclopentadienide) with 2 or 1 equiv of hybrid P−S ligands (L), (CH ) P(CH ) S-
3 2
2 n
(n ) 2, dmpet; n ) 3, dmppt), produced new dicyclopentadienyltitanium(IV) complexes with L, Cp Ti(L-κS)2
2
(1, L ) dmpet; 2, L ) dmppt) and [Cp2Ti(L-κ2S,P)]BPh4 (3, L ) dmpet; 4, L ) dmppt). The Ti(III) complexes,
Cp2Ti(L-κ2S,P) (5, L ) dmpet; 6, L ) dmppt), were prepared by the reaction of Cp2Ti(η3-C H ) with 1 equiv of L.
3
5
The structures of complexes 1−6 were confirmed by X-ray diffraction analyses. It was found that complexes 3 and
5 were isostructural around Ti(IV) and Ti(III) centers: the Ti(IV)−S bond length in 3 (2.3498(9) Å) is shorter by 0.14
Å than Ti(III)−Sin 5 (2.4877(7) Å), while Ti(IV)−P (2.534(1) Å) was merely 0.05 Å shorter than Ti(III)−P (2.5844(7)
Å). The redox potential between 3 and 5 in acetonitrile was −1.14 V vs the ferricinium/ferrocene couple. A
heterobimetallic complex that has the frame of complex 1, [Cp Ti(dmpet)2Cu]PF (7), was also isolated and structurally
2
6
characterized: the Ti−Cu distance (2.95(1) Å) was shorter than that in [Cp2Ti(SC H PPh ) Cu]BF , previously reported
2
4
2 2
4
by White and Stephan. Structural characterization was also carried out for Cp*Ti(dmpet-κS)2(dmpet-κ2S,P) (8) and
CpTiCl2(dmppt-κ2S,P) (9), which were obtained by the reactions of Cp*(or Cp)TiCl3 (Cp* ) η5-C Me5-) with
5
n equiv (n ) 1−3) of L. The mutual site-exchange reaction between phosphorus atoms on a coordinated dmpet
in the κ2S,P mode and on two other coordinated dmpet’s in the κS mode within complex 8 was analyzed by the
variable-temperature 31P{1H} dynamic NMR method. The kinetic parameters for this process, kex298 ) 1.9 × 105
q
q
s-1, ∆H ) 48 kJ mol-1, and ∆S ) 17 J mol-1 K-1, as well as the rather long Ti(IV)−P distance (2.652(1) Å),
indicate the fluxional nature of the coordination geometry in complex 8.
Introduction
attention. Recent studies concerning syntheses and isomer-
ization kinetics clearly indicated that the σ-donicity and
π-acidity of phosphorus donor atoms are significantly altered
by varying the substituents on the phosphorus atom:5 methyl-
substituted phosphines are somewhat strong σ-donors, while
phenyl-substituted phosphines exhibit π-acidity, depending
on the electronic configuration of the metal ion. We
previously reported various metal complexes with hybrid
P-S ligands, (CH3)2PCH2CH2SR (R ) H, Hdmpet; R )
CH3, mtdmpe), Ni(II), Pd(II), and Pt(II) complexes with
dmpet6 and mtdmpe,7 and Co(III)8 and Ru(II)9 complexes
Although many metal complexes with diphenyl-substituted
phosphine-thiol/thioether ligands, Ph2P(CH2)nSR (n ) 2, 3),
have been reported to date,1-4 P-S ligands with a dimethyl-
substituted phosphorus donor atom seem to have drawn little
* Author to whom correspondence should be addressed. Tel/Fax: +81-
52-789-5473. E-mail: kashiwa@chem.nagoya-u.ac.jp.
† Research Center for Materials Science, Nagoya University.
‡ Institute for Molecular Science.
§ Department of Chemistry, Nagoya University.
(1) Dilworth, J. R.; Wheatley, N. Coord. Chem. ReV. 2000, 199, 89-
158.
(2) Hauptman, E.; Fagan, P.; Marshall, W. Organometallics 1999, 18,
2061-2073, and references therein.
(5) Iwatsuki, S.; Suzuki, T.; Hasegawa, A.; Funahashi, S.; Kashiwabara,
K.; Takagi, H. D. J. Chem. Soc., Dalton Trans. 2002, 3593-3602.
(6) Kita, M.; Yamamoto, T.; Kashiwabara, K.; Fujita, J. Bull. Chem. Soc.
Jpn. 1992, 65, 2272-2274.
(7) Suzuki, T.; Morikawa, A.; Kashiwabara, K. Bull. Chem. Soc. Jpn.
1996, 69, 2539-2548.
(3) Morales, D.; Poli, R.; Richard, P.; Andrieu, J.; Collange, E. J. Chem.
Soc., Dalton Trans. 1999, 867-874.
(4) Brugat, N.; Polo, A.; Alvarez-Larena, A.; Piniella, J. F.; Real, J. Inorg.
Chem. 1999, 38, 4829-4837, and references therein.
5320 Inorganic Chemistry, Vol. 42, No. 17, 2003
10.1021/ic034353l CCC: $25.00 © 2003 American Chemical Society
Published on Web 07/29/2003

Titanium(IV) and Titanium(III) Complexes
with mtdmpe, in which our intention was to examine the
alteration in the coordination geometry around these electron-
rich metal ions by introducing a -PMe2 donor group instead
of the popular -PPh2 group since the steric requirements
(Tolman’s cone angle10) and the electronic effects (the
basicity represented by pKa of the conjugated acid and
Tolman’s ø parameter11,12) exhibited by these two phosphorus
donor sites are very different from each other. It was found
that the methyl substituents on the phosphorus donor site
unequivocally induce a strong static trans influence that
originates from the σ-donicity inherent in -PMe2.
ability of phosphines to electron-deficient Ti(IV). We also
succeeded in isolating novel isostructural pairs of [Cp2Ti-
(dmpet or dmppt)]+/0. A heterobimetallic complex, [Cp2Ti-
(dmpet)2Cu]PF6, similar to [Cp2Ti(SC2H4PPh2)2Cu]BF4 re-
ported by White and Stephan,15,17 was also isolated and
structurally characterized. To explore the nature and reactivity
of the bonding between phosphorus donor atoms and
electron-deficient metal centers, we also examined the
intramolecular site-exchange phenomenon in [Cp*Ti(dmpet)3]
in solution by means of variable-temperature 31P{1H} NMR.
Experimental Section
Studies concerning interactions between electron-deficient
metal ions with a -PMe2 donor site are still scarce, although
examination of the coordination geometry of [Mo(dmpet)2-
(StBu)2]13 and (PPh4)[W(S)3(dmpet)]14 revealed that these
compounds are excellent candidates for further syntheses of
various heterometallic clusters. To further examine the nature
of the interaction of hybrid P-S ligands with electron-
deficient metal ions, a series of Ti(IV) (d0) and Ti(III) (d1)
complexes with P-S hybrid ligands (dmpet or dmppt where
dmppt ) (CH3)2PCH2CH2CH2S-) were synthesized and the
crystal structures were determined in this study, as such
complexes have been limited to those reported by White and
Stephan, Cp2Ti(SCH2CH2PPh2)215 and Cp2Ti(SCH2CH2CH2-
PPh2)2,16,17 in which the P-S ligand binds to Ti(IV) in a κS
mode as a monodentate ligand. It was found that dmpet and
dmppt ligands coordinate in a κ2S,P mode at Ti(IV): only a
few complexes with a Ti(IV)-phosphorus bond have been
reported to date18-30 because of the rather low coordinating
General Information. All reactions were carried out using
standard Schlenk techniques under an argon atmosphere. Solvents
were dried and distilled before use according to known methods.
The following starting materials were prepared according to the
literature methods: tetramethyldiphosphine,31 Cp2TiCl2,32 Cp2Ti-
(η3-C3H5),33 [Cu(CH3CN)4]PF6,34 CpTiCl3,35 and Cp*TiCl3.36 The
1H, 31P{1H}, and 13C{1H} NMR spectra were recorded on a Varian
UNITY plus-500 and Bruker AMX-400 WB spectrometers. IR and
UV-visible spectra were measured by a Perkin-Elmer 2000FT-IR
spectrophotometer and a JASCO-V560 spectrophotometer, respec-
tively. EI-MS spectra were obtained on a Perkin-Elmer Sciex API-
300. Elemental analyses were carried out on a LECO-CHNS
microanalyzer, where the crystalline samples were sealed in thin
silver tubes. A JES-RE IX X-band spectrometer was used for the
measurements of the EPR spectra.
Electrochemical measurements were carried out by using a BAS
100B/W electrochemical analyzer at room temperature. Silver/silver
nitrate in acetonitrile was used as the reference electrode, while a
glassy carbon disk and a Pt wire were used as the working and
counter electrodes, respectively. The redox potentials of sample
solutions were calibrated by using the ferricinium/ferrocene redox
signal. Tetra-n-butylammonium tetrafluoroborate (0.3 mol kg-1) was
used as the supporting electrolyte.
(8) Kashiwabara, K.; Taguchi, N.; Takagi, H. D.; Nakajima, K.; Suzuki,
T. Polyhedron 1998, 17, 1817-1829.
(9) Taguchi, N.; Kashiwabara, K.; Nakajima, K.; Kawaguchi, H.; Tatsumi,
K. J. Organomet. Chem. 1999, 587, 290-298.
(10) Tolman, C. A. Chem. ReV. 1977, 77, 313-348.
(11) Streuli, C. A. Anal. Chem. 1960, 32, 985-987.
(12) Liu, H.; Eriks, K.; Prock, A.; Giering, W. Organometallics 1990, 9,
1758-1766, and references therein.
(13) Arikawa, Y.; Kawaguchi, H.; Kashiwabara, K.; Tatsumi, K. Inorg.
Chem. 1999, 38, 4549-4553.
(14) Arikawa, Y.; Kawaguchi H.; Kashiwabara, K.; Tatsumi, K. Inorg.
Chem. 2002, 41, 513-520.
(15) White, G. S.; Stephan, D. W. Inorg. Chem. 1985, 24, 1499-1503.
(16) White, G. S.; Stephan, D. W. Organometallics 1987, 6, 2169-2175.
(17) White, G. S.; Stephan, D. W. Organometallics 1988, 7, 903-910.
(18) Bottrill, M.; Gavens, P. D.; Kelland, J. W.; McMeeking, J. In
ComprehensiVe Organometallic Chemistry; Wilkinson, G., Stone, F.
G. A., Abel, E. W., Eds.; Pergamon Press: Oxford, 1982; Vol. 5,
Chapter 22.
(19) Bochman, M. In ComprehensiVe Organometallic Chemistry II; Abel,
E. W., Stone, F. G. A., Wilkinson, G., Eds.; Pergamon Press: Oxford,
1995; Vol. 4, Chapters 4 and 5.
Synthesis of (2-Dimethylphosphino)ethane-1-thiol (Hdmpet).
Hdmpet was prepared by the literature method.6 1H NMR (C6D6):
δ 2.36-2.27 (m, 2H, SCH2), 1.40-1.34 (m, 3H, SH and CH2P),
0.72 (d, 6H, P(CH3)2, JPH ) 2.5 Hz). 13C{1H} NMR (C6D6): δ
37.9 (d, SCH2, JPC ) 14 Hz), 22.0 (d, CH2P, JPC ) 20 Hz), 14.2
(d, P(CH3)2, JPC ) 15 Hz). 31P{1H} NMR (C6D6): δ -52.5. The
white powder of Lidmpet was obtained almost quantitatively by
n
mixing Hdmpet and BuLi in hexane followed by the removal of
solvent.
Synthesis of (3-Dimethylphosphino)propane-1-thiol (Hd-
mppt). Tetramethyldiphosphine (13.19 g, 108 mmol) was added
to liquid ammonia (200 cm3 at -78 °C) in a 500 cm3 three-necked
flask containing small pieces of metallic sodium (4.97 g, 216 mmol).
After stirring the mixture for 3 h, 3-chloropropane-1-thiol (24.0 g,
217 mmol) was added. The solution was stirred for a further 1 h,
followed by addition of 6 g of ammonium chloride (112 mmol).
After 30 min, the solvent was evaporated at room temperature. The
(20) Cotton, F. A.; Murillo, C. A.; Petrukhina. M. A. J. Organomet. Chem.
2000, 593, 1-6.
(21) Cotton, F. A.; Murillo, C. A.; Petrukhina. M. A. J. Organomet. Chem.
1999, 573, 78-86.
(22) Cotton, F. A.; Petrukhina, M. A. Inorg. Chem. Commun. 1998, 1, 195-
196.
(23) Willoughby, C. A.; Duff, R. R., Jr.; Davies, W. M.; Buchwald, S. L.
Organometallics 1996, 15, 472-475.
(30) Hues, D. L.; Leigh, G. J.; Walker, D. G. J. Organomet. Chem. 1988,
355, 113-119.
(24) Chadwell, S. J.; Coles, S. J.; Edwards, P. G.; Hursthouse, M. B. J.
Chem. Soc., Dalton Trans. 1996, 1105-1112.
(31) Butter, S. A.; Chatt, J. Inorg. Synth. 1974, 15, 185-191.
(32) Wilkinson, G.; Birmingham, J. M. J. Am. Chem. Soc. 1954, 76, 4281-
4284.
(25) Pulst, S.; Arndt, P.; Baumann, W.; Tillack, A.; Kempe, R.; Rosenthal,
U. J. Chem. Soc., Chem. Commun. 1995, 1753-1754.
(26) Noth, H.; Schmidt, M. Organometallics 1995, 14, 4601-4610.
(27) Doxsee, K. M.; Garner, L. C.; Juliette, J. J. J.; Mouser, J. K. M.;
Weakley, J. R.; Hope, H. Tetrahedron 1995, 51, 4321-4332.
(28) Nadasdi, T.; Stephan, D. W. Inorg. Chem. 1993, 32, 5933-5938.
(29) Park, J. W.; Henling, L. M.; Schaefer, W. P.; Grubbs, R. H.
Orgametallics 1990, 9, 1650-1656.
(33) Martin, H. A.; Jellinek, F. J. Organomet. Chem. 1967, 8, 115-128.
(34) Kubas, G. J. Inorg. Synth. 1979, 19, 90-92.
(35) King, R. B. Organometallic Syntheses; Academic: New York, 1965;
Vol. 1, p 78.
(36) Yamamoto, H.; Yasuda, H.; Tatsumi, K.; Lee, K.; Nakamura, A.; Chen,
J.; Kai, Y.; Kasai, N. Organometallics 1989, 8, 105-119.
Inorganic Chemistry, Vol. 42, No. 17, 2003 5321

Matsuzaki et al.
product was then extracted by diethyl ether from the residue.
Evaporation of ether under reduced pressure after filtration gave a
colorless oily product, which was distilled (bp 30-35 °C/2 mmHg)
to obtain pure Hdmppt as a clear liquid. The yield was 10.98 g
2.59 (m, 1H, CH2P), 2.43 (m, 1H, CH2P), 1.51 (d, 3H, P(CH3),
JPH ) 9.8 Hz), 1.21 (d, 3H, P(CH3), JPH ) 8.8 Hz), 1.08 (m, 2H,
CH2CH2CH2). 31P{1H} NMR (CD3CN): δ 24.2. IR (Nujol mull/
KBr): 3100 w, 3040 w, 1576 s, 1424 m, 1272 s, 1260 m, 942 s,
1
925 s, 842 s, 825 s, 735 s, 710 s, 628 s, 610 s, 482 m, 464 m cm-1
.
(37%). H NMR (C6D6): δ 2.23 (dt, 2H, SCH2), 1.49 (m, 2H,
UV-visible (λ/nm (ꢀ/M-1 cm-1), CH3CN): 492 (1000), 330 (8000).
EI-MS (m/z, CH3CN): 313 ([Cp2Ti(dmppt)]+). Anal. Calcd for
C39H42BPSTi: C, 74.06; H, 6.70; S, 5.07. Found: C, 73.39; H,
6.73; S, 4.95.
CH2CH2CH2), 1.13 (m, 1H, SH), 1.13 (m, 2H, CH2P), 0.80 (d, 6H,
P(CH3)2, JPH ) 2.7 Hz). 13C{1H} NMR (C6D6): δ 31.5 (d, SCH2,
JPC ) 11.5 Hz), 31.1 (d, CH2P, JPC ) 14.4 Hz), 26.5 (d, CH2CH2-
CH2, JPC ) 12.7 Hz), 14.5 (d, P(CH3)2, JPC ) 14.4 Hz). 31P{1H}
NMR (C6D6): δ -52.8.
Synthesis of Cp2Ti(dmpet) (5). Hdmpet (144 mg, 1.18 mmol)
was added to the THF solution (30 cm3) of Cp2Ti(η3-C3H5) (247
mg, 1.06 mmol). The color of the solution changed immediately
from reddish purple to dark green. After stirring the mixture for
0.5 h, the solution was concentrated to ca. 5 cm3 and cooled to
-30 °C, yielding dark red crystals of 5 (245 mg, 77%). IR (Nujol
Synthesis of Cp2Ti(dmpet)2 (1). A THF solution (45 cm3) of
Cp2TiCl2 (306 mg, 1.23 mmol) was added to a suspension of
Lidmpet (2.90 mmol) in THF (15 cm3) at 0 °C. The mixture was
stirred for 1 h at room temperature, and the solvent was removed
under reduced pressure. The extract with diethyl ether (40 cm3)
from the deep purple residue was concentrated to ca. 5 cm3 and
cooled to -30 °C, yielding purple crystals of 1 (770 mg, 87%).
Prismatic red-purple crystals were obtained by recrystallization from
mull/KBr): 3100 w, 1435 w, 1283 s, 959 s, 870 m, 795 s cm-1
.
Anal. Calcd for C14H20PSTi: C, 56.20; H, 6.74; S, 10.72. Found:
C, 55.70; H, 6.75; S, 10.46.
1
hexane at -10 °C. H NMR (C6D6): δ 5.77 (s, 10H, C5H5), 3.39
Synthesis of Cp2Ti(dmppt) (6). Preparation of 6 was carried
out by a method similar to that used for the synthesis of 5. Yield:
56%. IR (Nujol mull/KBr): 3060 w, 1260 w, 946 m, 870 m, 795
s cm-1. Anal. Calcd for C15H22PSTi: C, 57.51; H, 7.08; S, 10.24.
Found: C, 56.92; H, 7.39; S, 10.22.
(m, 4H, SCH2), 1.79 (m, 4H, CH2P), 0.98 (d, 12H, P(CH3)2, JPH
)
2.7 Hz). 31P{1H} NMR (C6D6): δ -50.9. IR (Nujol mull/KBr):
3150 w, 1430 m, 1283 m, 933 s, 880 m, 835 m, 810 s cm-1. UV-
visible (λmax/nm (ꢀ/M-1 cm-1), hexane): 540 (2600), 364 (2700).
Anal. Calcd for C18H30P2S2Ti: C, 51.43; H, 7.19; S, 15.26.
Found: C, 51.21; H, 7.02; S, 15.32.
Synthesis of [Cp2Ti(dmpet)2Cu]PF6 (7). To a CH3CN solution
(40 cm3) of [Cu(CH3CN)4]PF6 (0.17 g, 0.456 mmol) was slowly
added a CH3CN solution (20 cm3) of 1. The dark red mixture was
stirred for a further 4 h and then concentrated to ca. 5 cm3. Diethyl
ether was slowly added to the concentrate, yielding a dark red
crystalline powder of 7, which was washed with diethyl ether (20
cm3 × 2). The yield was 84%. 1H NMR (CD3CN): δ 5.98 (s, 10H,
C5H5), 3.31 (br, 4H, SCH2), 2.10 (br, 4H, CH2P), 1.35 (d, 12H,
P(CH3), JPH ) 2.7 Hz). 31P{1H} NMR (CD3CN): δ -21.6 (PMe2),
-143.2 (quint., PF6, JPF ) 707 Hz). IR (Nujol mull/KBr): 3120
w, 1425 m, 1410 w, 1395 w, 1285 m, 1260 w, 1245 w, 953 s, 930
s, 802 s cm-1. UV-visible (λ/nm (ꢀ/M-1 cm-1), CH3CN): 405
(8900). Anal. Calcd for C18H30CuF6P3S2Ti: C, 34.38; H, 4.81; S,
10.20. Found: C, 34.82; H, 4.92; S, 10.40.
Synthesis of Cp2Ti(dmppt)2 (2). The preparation was carried
out by the method similar to that used for the synthesis of 1.
Extraction was carried out with hexane, instead of diethyl ether,
1
which was used for the synthesis of 1. The yield was 91%. H
NMR (C6D6): δ 5.87 (s, 10H, C5H5), 3.38 (m, 4H, SCH2), 1.56
(m, 4H, CH2CH2CH2), 1.93 (m, 4H, CH2P), 0.98 (d, 12H, P(CH3)2,
JPH ) 2.7 Hz). 31P{1H} NMR (C6D6): δ -53.0. IR (Nujol mull/
KBr): 3060 w, 1284 m, 1019 s, 941 m, 815 s, 722 m cm-1. UV-
visible (λmax/nm, hexane): 539. Anal. Calcd for C20H34P2S2Ti: C,
53.57; H, 7.74; S, 14.30. Found: C, 53.03; H, 7.94; S, 13.77.
Synthesis of [Cp2Ti(dmpet)]B(C6H5)4 (3). Lidmpet (1.01 mmol)
in THF (30 cm3) was added to Cp2TiCl2 (250 mg, 1.01 mmol) in
THF (40 cm3) at 0 °C. The dark red solution was stirred for 2 h at
room temperature, and the solvent was evaporated under reduced
pressure. The dark red residue was extracted with toluene (40 cm3),
followed by evaporation of the solvent. The residue was redissolved
in 30 cm3 of THF, and a THF solution of NaBPh4 (274 mg, 0.80
mmol) was added. After stirring the mixture overnight, a purple
crystalline powder was precipitated, which was dissolved in
acetonitrile, and insoluble solids were removed by centrifugation.
The clear solution was concentrated and cooled to -30 °C, yielding
Attempts to isolate and characterize heterobimetallic complexes
with Fe(II) or Ni(II) were not successful. Reactions of 1 with NiCl2-
(PPh3)2 and FeCl2 produced only Ni(dmpet)26 (ca. 94% yield) and
Fe3(dmpet)6Cl (ca. 59% yield), respectively. The reaction of 2 with
Ni(cod)2 (cod ) 1,5-cyclooctadiene) yielded merely a trace amount
of [Cp2Ti(SC3H6PMe2)2Ni], which is in contrast to the high yield
of [Cp2Ti(SC3H6PPh2)2Ni]17 (70%). The dark green solid, obtained
after removing the solvent from the reaction mixture, contained
several nonidentifiable byproducts according to the 31P{1H} and
1H NMR spectra. On the other hand, [Ni(dmppt)2] was isolated
(ca. 50% yield) from the reddish brown mixture of 2 and Ni(cod)2
in THF. The reactions of 2 with FeCl2 and [Fe(CH3CN)6](ClO4)2
yielded only Fe3(dmppt)6Cl and [Fe2(dmppt)3(CH3CN)3](ClO4)2,
respectively.
1
purple crystals of 3‚CH3CN. (180 mg, 34%). H NMR (CD3CN):
δ 7.28 (m, 8H, BPh4), 7.01 (m, 8H, BPh4), 6.83 (m, 4H, BPh4),
6.57 (d, 10H, C5H5, JPH ) 2.7 Hz), 4.01 (m, 2H, SCH2), 3.19 (m,
2H, CH2P), 1.51 (d, 6H, P(CH3)2, JPH ) 9.8 Hz). 31P{1H} NMR
(CD3CN): δ 22.6. IR (Nujol mull/KBr): 3110 m, 3050 m, 1580
m, 1428 m, 1260 s, 950 s, 928 m, 900 m, 846 m, 828 s, 752 s, 738
s, 714 s, 600 s, 485 w, 465 w cm-1. UV-visible (λ/nm (ꢀ/M-1
cm-1), CH3CN): 545 (2400), 390 (sh), 332 (8000). EI-MS (m/z,
CH3CN): 299 ([Cp2Ti(dmpet)]+). Anal. Calcd for C40H43BNPS-
Ti: C, 72.85; H, 6.57; S, 4.86. Found: C, 71.95; H, 6.46; S, 4.51.
Synthesis of [Cp2Ti(dmppt)]B(C6H5)4 (4). The preparation was
carried out by a method similar to that for 3. After removal of
THF, a red crystalline powder was obtained. The crude compound
was recrystallized from CH3CN/diethyl ether. The yield was 82%.
1H NMR (CD3CN): δ 7.28 (m, 8H, BPh4), 7.01 (m, 8H, BPh4),
6.83 (m, 4H, BPh4), 6.45 (d, 5H, C5H5, JPH ) 2.4 Hz), 6.25 (d,
5H, C5H5, JPH ) 2.7 Hz), 4.22 (m, 1H, SCH2), 3.80 (m, 1H, SCH2),
Synthesis of Cp*Ti(dmpet)3 (8). Cp*TiCl3 (360 mg, 1.24 mmol)
in THF (20 cm3) was added to a suspension of Lidmpet (3.8 mmol)
in THF (15 cm3) at 0 °C. The dark red solution was stirred at room
temperature for 1 h, followed by the removal of the solvent. The
residue was extracted with hexane (50 cm3), and the extract was
filtered. The filtrate was concentrated to ca. 5 cm3 and cooled to
-30 °C, yielding red crystals. (430 mg, 63%). 1H NMR (C6D6): δ
3.62-3.35 (m, 6H, SCH2), 2.14 (s, 15H, C5(CH3)5), 1.94-1.73 (m,
6H, CH2P), 1.12 (d, 18H, P(CH3)2, JPH ) 1.7 Hz). 31P{1H} NMR
(C6D6, 80 °C): δ -29 (br). IR (Nujol mull/KBr): 1430 w, 1285
m, 932 s, 788 (br) cm-1. UV-visible (λ/nm, hexane): 435, 374.
5322 Inorganic Chemistry, Vol. 42, No. 17, 2003

Titanium(IV) and Titanium(III) Complexes
Table 1. Crystallographic Data for 1-9
1
2
3
4
5
6
7
8
9
formula
fw
temp (K)
C18H30P2S2Ti C20H34P2S2Ti C40H43BNPSTi C39H42BPSTi C14H20PSTi C15H22PSTi C18H30CuF6P3S2Ti C22H45P3S3Ti C10H17Cl2PSTi
420.40
296.2
448.46
296.2
659.53
296.2
632.50
296.2
299.25
296.2
313.27
296.2
628.91
296.2
546.60
296.2
319.08
273.2
space group triclinic
triclinic
P1h (2)
triclinic
P1h (2)
orthorhombic monoclinic monoclinic orthorhombic
triclinic
P1h (2)
monoclinic
P121/n1 (14)
7.216(1)
13.970(3)
13.8178(7)
90
103.574(1)
90
1354.0(3)
4
P1h (2)
Pna21 (33)
29.621(5)
9.772(5)
11.557(4)
90
P121/n1 (14) P121/a1 (14) Pna21 (33)
a (Å)
b (Å)
c (Å)
R (deg)
â (deg)
γ (deg)
V (Å3)
Z
11.579(4)
13.098(4)
8.159(2)
91.44(2)
108.58(2)
107.08(3)
1111.5(7)
2
11.432(2)
13.193(3)
8.755(2)
103.68(2)
94.68(1)
70.36(1)
1208.3(4)
2
12.855(4)
13.828(6)
10.223(3)
91.09(3)
100.13(2)
78.52(3)
1752(1)
2
8.059(2)
15.155(3)
12.164(2)
90
15.692(2)
13.804(4)
16.223(5)
90
25.796(2)
9.973(2)
9.943(2)
89
13.359(3)
14.035(3)
8.586(3)
106.74(2)
106.56(2)
89.84(2)
1472.1(7)
2
90
90
102.55(2)
90
117.92(2)
90
89
90
3345(3)
4
1450.1(5)
4
3105(1)
8
2557(1)
4
Dcalcd
1.256
1.232
1.250
1.256
1.371
1.340
1.633
1.233
1.565
(g cm-3
)
µ(Mo KR) 0.714
0.661
0.378
0.393
0.821
0.770
1.543
0.675
1.265
(cm-1
)
R
0.051
0.044
0.038
0.086
0.032
0.049
0.044
0.055
0.039
Rw
0.0507
0.0507
0.0430
0.0790
0.0396
0.0498
0.0497
0.0746
0.0514
Scheme 1
Anal. Calcd for C22H45TiS3P3: C, 48.34; H, 8.30; S, 17.60.
Found: C, 48.21; H, 8.09; S, 17.27.
Synthesis of CpTiCl2(dmppt) (9). Lidmppt (1.0 mmol) in THF
(15 cm3) was added to a THF solution (20 cm3) of CpTiCl3 (215
mg, 1.0 mmol) at 0 °C, and the mixture was stirred for 2 h. After
removal of the solvent, the green oily residue was extracted with
1,2-dimethoxyethane (DME). The DME solution was concentrated
and cooled to -30 °C, yielding green crystals of 9 (177 mg, 57%).
1H NMR (CDCl3): δ 6.56 (d, 5H, C5H5, JPH ) 2.4 Hz), 3.70 (br
m, 2H, SCH2), 2.29 (br m, 2H, CH2P), 1.86 (br m, 2H, CH2CH2-
CH2), 1.49 (d, 6H, P(CH3)2, JPH ) 9.3 Hz). 31P{1H} NMR
(CDCl3): δ 13.22. EI-MS (m/z): 319 (CpTiCl2(dmppt)+), 283
(CpTiCl2(dmppt)+Cl-), 218 (TiCl(dmppt)2+). Anal. Calcd for
C10H17Cl2PSTi: C, 37.64; H, 5.37; S, 10.05. Found: C, 37.21; H,
5.07; S, 9.64.
Crystal Structure Determinations. Crystallographic parameters
for 1-9 are summarized in Table 1. Single crystals were obtained
as described in the Experimental Section. The crystals suitable for
X-ray analyses were mounted in glass capillaries and sealed under
argon. Diffraction data of compounds except for 9 were collected
at room temperature on a Rigaku AFC7R diffractometer with
graphite-monochromated Mo KR radiation (λ ) 0.710690 Å) using
the ω-2θ scan technique. Refined cell dimensions and their
standard deviations were obtained by the least-squares method with
25 randomly selected centered reflections. These standard reflec-
tions, monitored periodically for crystal decomposition or move-
ment, exhibited slight intensity variation over the course of the data
collections. The raw intensities were corrected for Lorentz and
polarization effects. Empirical absorption corrections based on Ψ
scans were applied.
Data collections of 9 were carried out on a Rigaku AFC7
equipped with a MSC/ADSC Quantum 1 CCD detector at room
temperature. Four preliminary data frames were measured at 0.5°
increments of ω, to assess the crystal quality and calculate
preliminary unit cell parameters. The intensity image was measured
at 0.5° intervals of ω for a duration of 82 s for 9. The frame data
were integrated using the d*TREK program package, and the data
sets were corrected empirically for absorption using the REQAB
program.
one methyl group and one P atom of dmpet are disordered with
occupancy factors of 83:17. Additional information is available as
Supporting Information.
Results and Discussion
A schematic diagram for the syntheses of complexes 1,
3, 5, and 7 is shown in Scheme 1. The synthetic routes for
the corresponding dmppt complexes 2, 4, and 6 can be
described by a similar scheme.
Ti(IV) Complexes of Cp2TiL2 and [Cp2TiL]BPh4 (L )
dmpet and dmppt). Complexes 1 and 2 were obtained by
the reactions of Cp2TiCl2 with 2 equiv of Lidmpet and
Lidmppt, respectively, in ca. 90% yields. Reactions of Cp2-
TiCl2 with 1 equiv of Lidmpet and Lidmppt produced
complexes 3 and 4, respectively, in moderate yields, 35-
80%. It was not possible to isolate Cp2TiCl(dmpet), the
probable precursor of 3.
All calculations were carried out by using the TEXSAN program
package. All structures were solved by direct methods, and then
the structures were refined by full-matrix least-squares analyses.
Anisotropic refinement was applied to all non-hydrogen atoms, and
all the hydrogen atoms were placed at calculated positions. For 1,
The crystal structures of 1, 2, 3, and 4 were determined
by X-ray diffraction methods. ORTEP drawings of the
molecular structures of 1 and 2 and the cationic parts of 3
and 4 are shown in Figure 1 with selected bond lengths (Å)
Inorganic Chemistry, Vol. 42, No. 17, 2003 5323

Matsuzaki et al.
Figure 1. ORTEP drawings of 1 (Ti-S(1) 2.390(2), Ti-S(2) 2.383(3) Å, S(1)-Ti-S(2) 93.62(8)°) and 2 (Ti-S(1) 2.387(1), Ti-S(2) 2.397(1) Å, S(1)-
Ti-S(2) 94.74(4)°), and the cationic parts of 3 (Ti-S 2.3498(9), Ti-P 2.534(1) Å, S-Ti-P 72.63(3)°) and 4 (Ti-S 2.332(9), Ti-P 2.54(2) Å, S-Ti-P
79.7(5)°).
and angles (deg). The dmpet and dmppt in 1 and 2 coordinate
to the Ti(IV) center as a monodentate ligand through S,
whereas those in 3 and 4 bind to the Ti(IV) center with both
phosphorus and sulfur atoms. The Ti-S bond lengths and
the S-Ti-S bond angles in 1 and 2 are almost identical to
those (2.378(6) and 2.384(6) Å, and 93.3(2)°) in Cp2Ti(SCH2-
CH2CH2PPh2)2, which was reported by White and Stephan.16
The 31P{1H}NMR spectra of 1 and 2 in C6D6 exhibited sharp
singlet signals at -50.9 and -53.0 ppm, respectively. These
values are close to those for the free ligands, -52.5 (Hdmpet)
and -52.8 ppm (Hdmppt) in C6D6, indicating no coordination
of phosphorus donor atoms in solution. On the other hand,
the 31P{1H}NMR spectra of 3 and 4 in CD3CN exhibited
singlet signals at 22.6 and 24.2 ppm, respectively. The large
downfield shifts (75-77 ppm) of the 31P{1H}NMR signals
of the -P(Me)2 group confirm the coordination of dmpet
and dmppt through both phosphorus and sulfur atoms: the
structures observed in the solid state are retained in solution.37
The 1H NMR spectra of 3 and 4 also confirmed the
observations by the 31P{1H}NMR method. All signals due
to the Cp moieties in 3 and 4 appeared as doublets: a doublet
signal with JPH ) 2.7 Hz for 3 and two doublet signals with
JPH ) 2.4 and 2.7 Hz for 4. Nonequivalence of the two Cp
rings in 4 may be attributed to the fixed chair conformation
of the dmppt chelate ring, which is very much like that
observed for Cp2Ti(dmppt) 6, while the flexibility of the
distorted gauche (envelope) conformation of the dmpet
chelate ring in 3 readily provides an equivalent environment
for the two Cp rings in 3. The rigidity of the dmppt chelate
ring in 4 stems from (1) the significant steric congestion
around Ti(IV) in 4 as discussed below and (2) the large
energy barrier for the inversion.
Complexes 3 and 4 are rare examples of Cp2Ti(IV) species
with a Ti-phosphorus bond.25,29 The Ti(IV)-P bond length
in 3, 2.534(1) Å, is notably shorter than those in Cp2Ti(η2-
CH2S)(PMe3)29(2.601(1) Å) and [Cp2Ti(η2-CH2SMe)(PMe3)]-
I29 (2.581(1) Å), although it is similar to the Ti-PPh2 bond
length in the phosphide-bridged Cp2Ti(µ-σ,η2-CCPh)(µ-
PPh2)Ni(PPh3)25 (2.516(2) Å). The Ti(IV)-P bond found in
3 is the shortest of all of the reported analogous complexes
in which the Ti(IV)-phosphine bonds are in the range 2.57-
2.65 Å.20-30 The Ti(IV)-S bond length in 3 (2.3498(9) Å)
is shorter by ca. 0.04 Å than that (av 2.387(2) Å) in 1,
probably because of the positive charge on the cationic part
of 3. The Ti(IV)-S and Ti(IV)-P bond lengths in 4 are
similar to those in complex 3, while the S-Ti(IV)-P angle
was wider in 4 (79.7°) than in 3 (72.6°), reflecting the larger
bite angle of the dmppt ligand. However, relatively small
S-Ti(IV)-P angles (,90°) as well as the long Ti(IV)-P
(37) Garrou, P. E. Chem. ReV. 1981, 81, 229-266.
5324 Inorganic Chemistry, Vol. 42, No. 17, 2003

Titanium(IV) and Titanium(III) Complexes
Figure 2. Cyclic voltammogram of the [Cp2Ti(dmpet)]+/0 couple in
acetonitrile. [Cp2Ti(dmpet)BPh4] ) 1.3 × 10-3 mol kg-1. Sweep rate is
0.1 V s-1. T ) room temperature. I ) 0.3 mol kg-1(TBABF4).
bond lengths in these complexes indicate that the coordina-
tion environment around Ti(IV) in these complexes is
sterically congested because of the two Cp groups and that
the Ti(IV)-P bonds in these complexes are rather weak.
The redox potential of the [Cp2Ti(dmpet)]+/0 couple was
measured in acetonitrile. As shown in Figure 2, the vol-
tammogram was quasi-reversible with Ep-p ) 96 mV and
I
cathodic/Ianodic ≈ 1.0. No other signal was observed within the
experimental potential window, from -2.0 to +1.0 V.
Therefore, the redox potential for the [Cp2Ti(dmpet)]+/0
couple was determined as -1.14 V vs the ferricinium/
ferrocene couple. Such a low redox potential confirms the
instability of the Ti(III) oxidation state as described in the
Experimental Section. However, the quasi-reversible signals
of the voltammogram even at a slower scan rate than 100
mV s-1 indicates that the Ti(III) complex is not labile.
Therefore, it seems possible to isolate the corresponding Ti-
(III)-dmpet complex, although it is predicted that such a
species is highly air-sensitive. We will report the successful
isolation and characterization of [Cp2TiIII(dmpet)] in the next
section.
Figure 3. ORTEP drawings of 5 (Ti-S 2.4877(7), Ti-P 2.5844(7) Å,
S-Ti-P 74.71(2)°) and 6 (Ti-S 2.503(2), Ti-P 2.568(2) Å, S-Ti-P
79.22(6)°).
anion from MAO and by olefins. The catalytic activity was
the largest for 1, and it was not significant for all the other
complexes. Moreover, the catalytic activity of 1 was well
below the performance of other standard catalysts.
Ti(III) Complexes of Cp2TiL (L ) dmpet and dmppt).
Complexes 5 and 6, Cp2Ti(dmpet or dmppt), were prepared
by the reactions of Cp2Ti(η3-C3H5) with 1 equiv of Hdmpet
or Hdmppt, in 60-80% yields (Scheme 1). Both complexes
are highly sensitive to air and moisture, in contrast to 3 and
4, as predicted by the very low redox potential for the
[Cp2Ti(dmpet)]+/0 couple in acetonitrile (see Figure 2).
Complexes 1 and 2 are somewhat unstable upon exposure
to air due to the oxidation of the dangling -PMe2 moieties,
while complexes 3 and 4 are considerably resistant to air-
oxidation. Both 3 and 4 did not react with acetone, benzal-
dehyde, and methanol. However, they decomposed gradually
in the presence of water. The reaction of 3 dissolved in CH3-
CN with 1 equiv of Lidmpet in THF at 0 °C gave 1 in a
61% yield. The reaction of 1 in THF with 3 equiv of MeI
produced [Cp2Ti(dmpetMe)2]I2 in a 78% yield, while the
According to the X-ray crystallographic analyses, there
are two independent molecules of 6 in an asymmetric unit.
The ORTEP drawings of 5 and one of two molecules in 6
are shown in Figure 3 with the selected bond lengths and
angles. The dmpet and dmppt ligands in 5 and 6 coordinate
to the Ti(III) center through both phosphorus and sulfur
atoms and form a distorted gauche and a typical chair
conformation, respectively. Complexes 5 and 3 exhibited
identical coordination geometry. The Ti(III)-S and Ti(III)-P
bond lengths in 5 are 2.4877(7) and 2.5844(7) Å, respec-
tively, and are longer by ca. 0.14 and ca. 0.05 Å than those
in 3, which may be compared with the difference of the ionic
radii for Ti3+ (0.81 Å) and Ti4+ (0.75 Å),38 although the
formation of the phosphonium salt was confirmed by the
1
31P{1H} NMR (27.3 ppm) and H NMR (δ 1.90, d (JPH
)
14 Hz), 18H, -P(CH3)3) spectra and elemental analysis. A
similar reaction was also reported for Cp2Ti(SC2H4PPh2)2
by White and Stephan.15 We also examined the catalytic
activity of 1 for the ethylene polymerization reaction with
excess MAO (methylaluminoxane): 1.58 × 105 g (mol Ti)-1
h-1 at 50 °C in toluene. The number average molecular
weight, Mn, was 2.75 × 105, and Mw/Mn was 2.27, indicating
that the thiol ligand in 1 is readily substituted by a methyl
Inorganic Chemistry, Vol. 42, No. 17, 2003 5325

Matsuzaki et al.
elongation of the Ti-S bond is somewhat more significant.
The Ti-P and Ti-S bond lengths in 6 are similar to those
in 5.
tamination with such diamagnetic species. Therefore, we
decided not to report unreliable magnetic parameters in this
article, as it is difficult to avoid partial oxidation of Ti(III)
during preparation of samples for the measurement of the
magnetic susceptibility. Complexes 5 and 6 react with
diphenyl disulfide (molar ratio ) 1:1.5) to yield Cp2Ti-
Limited numbers of results have been reported for the
crystallographic structures of the complexes with Ti(III)-
phosphorus bonds to date, and the Ti(III)-phosphorus bond
lengths in such complexes range from 2.46 to 2.70 Å.26,39-51
Interestingly, the Ti(III)-P bond lengths in 5 and 6 are
somewhat longer than the Ti(II)-phosphorus bond lengths
in (η5-MeC5H4)2Ti(Me2PCH2CH2PMe2) (2.527(4) and 2.540-
(4) Å)52 and Cp2Ti(PMe3)2 (2.524(4) and 2.527(3) Å),53
despite the larger ionic radius of Ti2+ (1.00 Å) in the d2
electronic configuration.38 It seems that the back-donation
from Ti(II) to the phosphorus atom is responsible for the
short Ti(II)-P bond: the larger covalent character of Ti-
(II)-P than that of Ti(III)-P is indicated. In a similar way,
the Ti(III)-P bond seems stronger than the bonding between
Ti(IV) and P.
55
(SPh)2 along with other byproducts, which indicates that
the relatively large Ti(III) center tends to suffer from an
attack by diphenyl disulfide to allow oxidative addition.
An alternative synthetic procedure, on the basis of the
method by Bochmann et al.56 for the synthesis of [Cp*2TiMe-
(THF)]BPh4 from Cp*2TiMe, was also examined in this
study. The dark green suspension of AgBPh4 and 1 equiv of
5 or 6 in THF was stirred for 4 h to yield a dark purple-red
solution with metallic silver. After removal of silver, the
solvent was evaporated to dryness. Complexes 3 and 4 were
successfully isolated from acetonitrile solutions of the residue
in ca. 40% yields.
Heterobimetallic Complexes, [Cp2Ti{S(CH2)2or3PMe2}2-
M]. Attempts were made to prepare heterobimetallic com-
plexes, following White and Stephan’s method for [Cp2Ti-
Complexes 5 and 6 are paramagnetic. The EPR spectrum
of 5 in THF at room temperature exhibited an intense doublet
signal due to 31P with Ti satellites on both sides: A(P) )
21.3, g ) 1.991. The spectrum of 6 under the same conditions
was intense but broad: the g value was 1.990, although
estimation of A(P) was not possible. These A(P) and g values
are similar to those for Cp2TiCl(PMe3) (A(P) ) 20.2, g )
1.986 at room temperature) and for Cp2TiMe(PMe3)
(A(P) ) 20.2, g ) 1.986 at 263 K) for which coordination
of PMe3 is confirmed in solution,54 indicating that the
structures of 5 and 6 in the solid state are retained in THF
solutions. Paramagnetic complexes 5 and 6 are very unstable
(air sensitive), as indicated by the very low redox potential.
Although the EPR parameters are not sensitive to contamina-
tion with diamagnetic Ti(IV) (air-oxidized species), the
magnetic susceptibility is significantly influenced by con-
15
(SC2H4PPh2)2Cu]BF4 and [Cp2Ti(SC3H6PPh2)2Ni]:17 dark
red crystals of [Cp2Ti(SC2H4PMe2)2Cu]PF6 (7) were obtained
from 1 (Scheme 1). While complex 7 was easily synthesized
by the reaction of 1 with [Cu(CH3CN)4]PF6 in 84% yield,
the preparation of other heterobimetallic complexes with Ni-
(II) or Fe(II) was not successful: reactions of 1 with NiCl2-
6
(PPh3)2 and FeCl2 gave Ni(dmpet)2 (ca. 94% yield) and
Fe3(dmpet)6Cl (ca. 59% yield), together with Cp2TiCl2. The
structure of Fe3(dmpet)6Cl was clarified by X-ray analysis
to be a linear trinuclear complex with two fac-[Fe(II)P3S3]
and one [Fe(III)S6] unit.57 The reaction of 2 with Ni(cod)2
(cod ) 1,5-cyclooctadiene) produced only a trace amount
of [Cp2Ti(SC3H6PMe2)2Ni], which is in contrast to the high
yield of [Cp2Ti(SC3H6PPh2)2Ni]17 (70%). [Ni(dmppt)2] was
isolated (ca. 50% yield) from the reddish brown mixture of
2 and Ni(cod)2 in THF. Therefore, it seems that the [NiII-
(dmppt)2] complex is the more stable than [Cp2Ti(SC3H6-
PMe2)2Ni], and the dissociation of dmppt from 2 took place
upon mixing 2 and Ni(cod)2. The reactions of 2 with FeCl2
and [Fe(CH3CN)6](ClO4)2 yielded Fe3(dmppt)6Cl and [Fe2-
(dmppt)3(CH3CN)3](ClO4)2, respectively. These reactions of
1 and 2 with metal ions indicate that the Ti(IV)-S bonds in
1 and 2 are readily cleaved by addition of late transition metal
ions to form the more stable [Mn(dmpet or dmppt)m]
complexes.
(38) Shannon, R. D. Acta Crystallogr. 1976, A32, 751-767.
(39) Troyanov, S. I.; Meetsma, A.; Teuben, J. H. Inorg. Chim. Acta 1998,
271, 180-186.
(40) Chen, L.; Cotton, F. A.; Dunbar, K. R.; Feng, X.; Heints, R. A.;
Uzelmeir, C. Inorg. Chem. 1996, 35, 7358-7363.
(41) Cotton, F. A.; Wojtczak, W. A. Gazz. Chim. Ital. 1993, 123, 499-
507.
(42) Britten, J.; Mu, Y.; Harrod, J. F.; Polowin, J.; Baird, M. C.; Samuel,
E. Organometallics 1993, 12, 2672-2676.
(43) Christou, V.; Wuller, S. P.; Arnold, J. J. Am. Chem. Soc. 1993, 115,
10545-10552.
(44) Binger, P.; Muller, P.; Philips, P.; Gabor, B.; Mynott, A. T.; Hermann,
A. H.; Langhauser, F.; Kruger, C. Chem. Ber. 1992, 125, 2209-2212.
(45) Gomez, R.; Cuenca, T.; Royo, P.; Pellinghelli, M. A.; Tripicchio, A.
Organometallics 1991, 10, 1505-1510.
(46) Fenske, D.; Grissinger, A.; Hey-Hawkins, E. M.; Magull, J. Z. Anorg.
Allg. Chem. 1991, 595, 57-66.
The 31P{1H} NMR spectrum of 7 exhibited a singlet signal
at -21.6 ppm: the notably small downfield shift (∆ ) 29.3
ppm) from the signal of free dmpet can be explained by the
higher electron density on Cu(I) compared with that on Ti-
(IV) in 3 for which a large chemical shift of the 31P{1H}
(47) Samuel, E.; Mu, Y.; Harrod, J. F.; Dromzee, Y.; Jeannin, Y. J. Am.
Chem. Soc. 1990, 112, 3435-3439.
(48) Jensen, J. A.; Wilson, S. R.; Girolami, G. S. J. Am. Chem. Soc. 1988,
110, 4977-4982.
(49) Razavi, A.; Mallin, D. T.; Day, R. O.; Rausch, M. D.; Alt, H. J.
Organomet. Chem. 1987, 333, C48-C52.
(50) Jensen, J. A.; Girolami, G. S. J. Chem. Soc., Chem. Commun. 1986,
1160-1162.
(55) Giddings, S. A. Inorg. Chem. 1967, 6, 849-850.
(56) Bochmann, M.; Jaggar, A. J.; Wilson, L. M.; Hursthouse, M. B.;
Motrvalli. Polyhedron 1989, 8, 1838-1843.
(57) Crystal data for [Fe3(dmpet)6]Cl‚CH2Cl2 prepared by different meth-
ods: C25H62S6P6Cl3Fe3, fw ) 1014.87, space group C2/m, a ) 17.856-
(9) Å, b ) 10.101(7) Å, c ) 13.857(5) Å, â ) 111.46(4)°, V ) 2326(2)
Å3, Z ) 4, Dcalc ) 2.898 g cm-3, µ(Mo KR) ) 31.75 cm-1, R )
0.047, Rw ) 0.055.
(51) Kool, L. B.; Rausch, M. D.; Alt, H. G.; Herberhold, M.; Wolf, B.;
Thewalt, U. J. Organomet. Chem. 1985, 297, 159-169.
(52) Girolami, G. S.; Wilkinson, G.; Thornton-Pett, M.; Hursthouse, M.
B. J. Chem. Soc., Dalton Trans. 1984, 2347-2350.
(53) Kool, L. B.; Rausch, M. D.; Alt, H. G.; Herberhold, M.; Thewalt, U.;
Wolf, B. Angew. Chem., Int. Ed. Engl. 1985, 24, 394-401.
(54) Samuel, E.; Henique, J. J. Organomet. Chem. 1996, 512, 183-187.
5326 Inorganic Chemistry, Vol. 42, No. 17, 2003

Titanium(IV) and Titanium(III) Complexes
Figure 4. ORTEP drawings of 7 (Ti-Cu 2.95(1), Cu-S(1) 2.271(2), Cu-
S(2) 2.292(10), Cu-P(1) 2.26(2), Cu-P(2) 2.24(2), Ti-S(1) 2.48(1), Ti-
S(2) 2.467(5) Å, S(1)-Ti-S(2) 97.3(4), S(1)-Cu-S(2) 109.1(3), Ti-S(1)-
Cu 76.6(3), Ti-S(2)-Cu 76.6(2), P(1)-Cu-P(2) 117.6(4)°).
signal was observed upon coordination (∆ ) 73.5 ppm). A
similarly small downfield shift (∆ ) 21.3 ppm) of the 31P-
{1H} signal was reported for [Cp2Ti(SC2H4PPh2)2Cu]BF4.15
The crystal structure of 7 was determined by X-ray
analysis, and the ORTEP drawing of the cationic part of 7
is shown in Figure 4 with selected bond lengths and angles.
The coordination geometry around the Cu center in 7 is a
distorted tetrahedron with the dihedral angle between S1-
Cu-S2 and P1-Cu-P2 planes being 70.0°. The Ti-S bond
lengths (2.48(1) and 2.467(5) Å) in 7 are longer than those
in Ti(IV) complexes 1 (2.390(2) and 2.383(3) Å) and 3
(2.3498(9) Å): they were rather close to the bond length of
Ti-S (2.4877(7) Å) in Ti(III) complex 5. The Ti-Cu
distance (2.95(1) Å) in 7 is shorter by ca. 0.07 Å than that
in analogous [Cp2Ti(SC2H4PPh2)2Cu]PF6 (3.021(1) Å),15
along with the shorter Cu-S and Cu-P bond lengths in 7.
Furthermore, the average Ti-S-Cu bond angle (76.6(2)°)
in 7 is slightly smaller than that (78.1(1)°) in [Cp2Ti(SC2H4-
PPh2)2Cu]PF6.
As the existence of metal-to-metal bonding has been
suggested for [Cp2Ti(SC2H4PPh2)2Cu]PF6,15 a similar dative
interaction is also possible in complex 7, by considering the
shorter Cu-Ti distance in 7 than in [Cp2Ti(SC2H4PPh2)2-
Cu]PF6. In such a case, the Cu(I)(d10) to Ti(IV)(d0) dative
bond seems to be more significant in 7 than in [Cp2Ti(SC2H4-
PPh2)2Cu]PF6, due to the larger electron density on Cu(I) in
7 than that in [Cp2Ti(SC2H4PPh2)2Cu]PF6. The difference
in the electron density on Cu(I) is caused by the difference
in the σ-donicity of the -PMe2 donor sites in 7 and that of
the -PPh2 site in [Cp2Ti(SC2H4PPh2)2Cu]PF6. Therefore, the
elongated Ti(IV)-S bonds in 7, the bond length of which is
similar to Ti(III)-S in 5, seem to reflect the increased
electron density on Ti(IV) in 7 through dative interaction
with Cu(I).
Figure 5. ORTEP drawings of 8 (Ti-S(1) 2.425(2), Ti-S(2) 2.339(1),
Ti-S(3) 2.385(1), Ti-P 2.652(1) Å, S(1)-Ti-S(2) 82.98(5), S(1)-Ti-
S(3) 124.23(5), S(2)-Ti-S(3) 93.49(5), S(1)-Ti-P(1) 72.65(4), S(2)-
Ti-P(1) 135.89(5), S(3)-Ti-P(1) 72.33(4)°) and 9 (Ti-S 2.361(1), Ti-P
2.556(1), Ti-Cl(1) 2.342(1), Ti-Cl(2) 2.370(1) Å, S-Ti-P 80.15(4), Cl-
(1)-Ti-Cl(2) 86.68(4), Cl(1)-Ti-S 82.33(4), Cl(2)-Ti-P 77.25(4)°).
produced complex 9 in 57% yield, while unknown products
were obtained by the reactions with 2 and 3 equiv of
Lidmppt.
The crystal structures of complexes 8 and 9 determined
by X-ray analyses are shown in Figure 5 with selected bond
lengths and angles. In complex 8, only one of the three dmpet
ligands coordinates to the Ti(IV) center in a κ2S,P mode,
while two other dmpet ligands bind to Ti(IV) only through
S. It seems that the dmpet ligand with κ2S,P coordination
mode in 8 rather weakly binds to the Ti(IV) center: the Ti-
(1)-S(1) bond length (2.425(2) Å) is very long compared
with Ti(1)-S(2) (2.339(1) Å) and Ti(1)-S(3) (2.385(1) Å),
while the lengths of the latter two bonds are comparable to
those in complex 1 (2.390(2) and 2.383(2) Å) and complex
3 (2.3498(9) Å). Complex 9 has a structure similar to that
of 8 with a dmppt ligand coordinated to Ti(IV) through P
and S. The Ti-P(2.556(1) Å) and Ti-S(2.361(1) Å) bond
lengths are shorter than those for the chelating dmpet in
complex 8. Such a difference in Ti-P and Ti-S lengths may
reflect the sterically less demanding nature of Cp than Cp*.
CpTi(IV) and Cp*Ti(IV) Complexes with dmpet and
dmppt. The reaction of Cp*TiCl3 with 3 equiv of Lidmpet
produced complex 8 in a 63% yield, while the reactions with
1 or 2 equiv of Lidmpet resulted in the formation of unstable
and uncharacterized products in a very low yield. On the
other hand, the reaction of CpTiCl3 with 1 equiv of Lidmppt
1
The H NMR signal at 1.12 ppm (doublet) observed at
20 °C for complex 8 (JPH ) 1.7 Hz, 18H, P(CH3)2) split
into multiple signals at -80 °C in toluene-d8, indicating that
Inorganic Chemistry, Vol. 42, No. 17, 2003 5327

Matsuzaki et al.
Scheme 2
Scheme 3
three distinct -PMe2 moieties in two different coordination
environments at -80 °C become indistinguishable at elevated
temperatures: a rapid site-exchange between dmpet ligands
in κ2S,P and κS coordination modes takes place at room
temperature (Scheme 3). To examine the rate of this site-
exchange process, variable-temperature 31P{1H} NMR mea-
surements in toluene-d8 solution of 8 were carried out in the
temperature range from -80 to 80 °C. A somewhat broad
singlet signal was observed at -29 ppm at 80 °C. As the
solution was cooled, the singlet signal broadened further and
disappeared at 0 °C. However, three sharp singlet signals
were observed at -80 °C at 17.5, -51.2, and -51.5 ppm
with the intensity ratio of 1:1:1 (Figure 6). The resonance at
17.5 ppm was attributed to the coordinated phosphorus on
dmpet in a κ2S,P coordination mode, as the chemical shift
is close to the value observed for complex 3, 22.6 ppm. The
two higher field signals were assigned to the two uncoor-
dinated phosphorus atoms on the dmpet ligands in a κS
coordination mode, as the chemical shifts of these two signals
are almost identical to those for complex 1.
Figure 6. Dynamic NMR measurements: 31P{1H} NMR spectra of 8 in
toluene-d8 at -80 to 80 °C. Calculated NMR spectra at each temperature
are shown in the right column, together with the best-fit values of the rate
constants.
As it is known that a coordination number of more than 6
is common for Ti(IV), a mechanism through the associative
transition state may be possible for the mutual site-exchange
reaction observed in this study.59 On the other hand, the
relatively long Ti(IV)-P bond may indicate dissociative
activation for this reaction. Traditionally, the reaction mech-
anisms for solvent exchange and/or ligand substitution
reactions (intramolecular mutual site-exchange process is also
included in this category) have been elucidated either from
the dependence of the observed rate constant on the
concentration of the entering ligand (by means of rate law),
by comparing the obtained activation parameters with those
for other known reactions, or through discussions on the
characteristics of entering and leaving donor sites.60 However,
such a classical method for elucidating reaction mechanisms
of the solvent exchange/site-exchange processes may not be
valid any more:60 recent calculations by Cooper and Ziegler
confirmed61 that even in the solvent exchange and ligand
substitution reactions of d8 Pt(II), for which the reaction has
long been believed to proceed through the associative
mechanism, the contribution of the dissociation mode to the
activation process is significant (>90%) although the ex-
By using the DNMR v4.1.0 program,58 the line shapes of
the NMR spectra at each temperature were calculated and
fitted to the experimentally obtained 31P{1H} NMR spectra
in order to evaluate the rate constants for the mutual site-
exchange process and the corresponding activation param-
298
eters: kex ) 1.9 × 105 s-1, ∆Hq ) 48 kJ mol-1, and
∆Sq ) 17 J mol-1 K-1. No exchange sequence other than
the mechanism shown in Scheme 3 reproduced the experi-
mental spectral change: the exchange process must involve
dissociation of the Ti(IV)-P bond.
(59) Greenwood, N. N.; Earnshaw, A. Chemistry of the Elements, 2nd ed.;
Butterworth: Oxford, U.K., 1997.
(60) Jordan, R. B. Reaction Mechanisms of Inorganic and Organometallic
Systems, 2nd ed.; Oxford University Press: Oxford, U.K., 1998.
(61) Cooper, J.; Ziegler, T. Inorg. Chem. 2002, 41, 6614-6622.
(58) The GNMR4 program was modified to work on a Macintosh computer.
The original program was “Program for Iterative Analysis of Exchange
Broadened NMR Spectra” by David. S. Stephenson and Gerhard
Binsch, Munich, 1978.
5328 Inorganic Chemistry, Vol. 42, No. 17, 2003

Titanium(IV) and Titanium(III) Complexes
perimentally observed exchange and substitution rates obvi-
ously depended on the concentration of the entering ligand.
On this basis, most of the exchange/substitution processes
on metal complexes in solution may be dissociative in nature,
and the apparent dependence of the observed rate constant
on the individual characteristics and the concentration of each
entering ligand is explained by the concerted nature of the
exchange/substitution processes. Moreover, the crystal struc-
ture of complex 8 confirms that the large thiol moieties as
well as the bulky Cp* ligand provide a coordinatively
saturated environment around the Ti(IV) center: the Ti-P
bond is long (2.652(1) Å) and the bite angle of the chelating
dmpet ligand is rather small (72.33(4)°). We, therefore,
conclude that the intramolecular mutual site-exchange reac-
tion within complex 8 is regarded to proceed through the
dissociation of coordinated P atom on the chelating dmpet
ligand. Then the activation enthalpy, ∆Hq, is a measure of
the Ti(IV)-P bond strength in complex 8: the relatively
small activation enthalpy of this reaction is certainly related
to the long Ti(IV)-P bond length in 8.
phorus as donor atoms in this study. It was found that Ti-
(IV) tends to bind more strongly to the S- moiety than to
-PMe2. The Ti(IV)-P bonds in complexes 3, 4, 8, and 9
were unequivocally longer than the Ti(IV)-S bonds, reflect-
ing the inferior basicity of the phosphorus donor site to the
S- moiety. However, the Ti(IV)-P bond lengths in these
complexes were shorter than those for the other complexes
reported to date, probably because of the sterically less
demanding nature of -PMe2. We also reported novel
isostructural Ti(IV)/Ti(III) pairs with hybrid P-S ligands,
complexes 3/5 and 4/6, although the redox potential for the
3/5 couple was rather low. However, the difference in the
Ti(III)-P and Ti(IV)-P bond lengths was merely 0.04 Å,
while the Ti-S lengths in complexes 5 and 6 were some
0.14-0.17 Å longer than those in complexes 3 and 4. Such
a small elongation of the Ti-P bond may be explained by
the stabilization of the Ti(III)-P bond through π back-
bonding, while it seems that the large elongation of the Ti-S
bond directly reflects the decrease of the electronic charge
on the central metal ion. The weak nature of the Ti(IV)-P
interaction was also demonstrated by the rapid mutual site-
exchange reaction in complex 8.
1
The H NMR signal at 6.56 ppm (doublet, 5H, C5H5,
JPH ) 2.4 Hz) and the 31P{1H} signal at 13.22 ppm observed
for complex 9 in chloroform indicate that the coordination
of dmppt to Ti(IV) through P and S atoms observed in the
crystal is retained in solution. The shorter Ti(IV)-P and Ti-
(IV)-S bond lengths in 9 than those in 8 ensure resistance
to dissociation of the dmppt ligand in 9: the Ti(IV)-P bond
is shorter in 9 than in 8, since (1) the smaller steric demand
of chloride ions than the dangling P-S moiety in 8 and (2)
the less bulky nature of the Cp ring compared with Cp*
provide sufficient space around Ti(IV) for tighter Ti-P
bonding in 9.
Acknowledgment. This work was supported in part by
a Grant-in-Aid for Scientific Research on Priority Areas
(No. 14078101, 14078211 “Reaction Control of Dynamic
Complexes”) from the Ministry of Education, Culture, Sports,
Science, and Technology, Japan, and H.D.T. expresses his
thanks to the Kurata Foundation, Hitachi Corp, for financial
support.
Supporting Information Available: Lists of X-ray crystal-
lographic files, in CIF format, are available free of charge via the
Internet at http://pubs.acs.org.
Conclusion
We reported nine novel titanium complexes bearing hybrid
bidentate ligands with sulfur and dimethyl-substituted phos-
IC034353L
Inorganic Chemistry, Vol. 42, No. 17, 2003 5329