Dithiocarbamate complexes of Ti(IV) alkoxides: Synthesis, characterization, and electrochemistry

Article abstract of DOI:10.1021/ic802394a

Isopropoxy- and ferf-butoxy-fr;'s(dithiocarbamato)titanium(IV) complexes of five dithiocarbamate ligands were prepared and characterized by LDI-MS, 1H NMR, 13C NMR, and elemental analysis, as well as by crystallographic determination of two examples. Both

Full text of DOI:10.1021/ic802394a

Inorg. Chem. 2009, 48, 4171-4178
Dithiocarbamate Complexes of Ti(IV) Alkoxides: Synthesis,
Characterization, and Electrochemistry
Alberto Donzelli and Pierre G. Potvin*
Department of Chemistry, York UniVersity, 4700 Keele Street, Toronto, Ontario, Canada M3J 1P3
Received December 16, 2008
Isopropoxy- and tert-butoxy-tris(dithiocarbamato)titanium(IV) complexes of five dithiocarbamate ligands were prepared
and characterized by LDI-MS, 1H NMR, 13C NMR, and elemental analysis, as well as by crystallographic determination
of two examples. Both showed strongly π-coordinated alkoxy groups and two separate dithiocarbamate coordination
environments that, in solution, were in rapid exchange. Cyclic voltammetry in CH2Cl2 revealed irreversible but
reproducible oxidation peaks between +1.2 and +1.6 V vs Ag/AgCl, about 1 V positive of those from the free
ligands, as well as reduction peaks in the -1.9 to -2.2 V range vs Ag/AgCl assigned to TiIV/III couples, and
second reductions in some cases. The corresponding diisopropoxy-bis(dithiocarbamato) analogues were not isolable
and slowly transformed to the more stable tris species. Indeed, these were shown to be in slow equilibrium.
Introduction
oxidation probably proceeds through high-energy ketyl anion
radicals in one-electron steps, whereas the CuII/I system of
Marko´ et al.9 is said to function via a dinuclear intermediate
by a two-electron R-deprotonation process akin to that by
CrVI or MnVIII laboratory reagents. Some mononuclear Co-
based catalysts appear to proceed by a two-electron pathway
through a formal CoIV)O species, by employing a sacrificial
reductant that, in essence, converts the four-electron oxidant
O2 into a two-electron one.
Electrochemical oxidation is an attractive alternative for
industrial scales because it is atom-economical, it obviates
transition metal-containing waste, and it can be energetically
efficient if conducted at close to the thermodynamic potential.
In fuel cells, O2 drives alcohol oxidation by being reduced
at the cathode while the alcohol is oxidized at the anode,
with migration of electrons through an external circuit and
of H+ in the electrolyte solution. Whereas the cathodic
reaction can be fast, the anodic reaction is slow, and much
research has been devoted to developing efficient electrode
materials for alcohol-fueled cells.10 Recent work with
The oxidation of alcohols is a fundamental reaction that
has elicited much attention over the decades.1 Although O2
is sufficiently strongly oxidizing, its direct reaction with
alcohols is impractically slow, in part because of the kinetic
and thermodynamic restrictions of one-electron pathways and
because of spin considerations (triplet-state O2 vs singlet-
state substrates).2 Homogeneous catalysts of aerial oxidation,3
with or without sacrificial additives,4 alter the electronic
structure of O2 to facilitate the reaction.5 When involving
one-electron couples, such as with the Wieghardt mono-
nuclear CuII/I catalyst,6 with other CuII/I systems that employ
N-oxyls as additives,7 and with CoIII/II complexes,8 alcohol
* To whom correspondence should be addressed. E-mail: pgpotvin@
yorku.ca.
(1) Homogeneous Catalysis, 2nd ed.; Parshall, G., Ittel, S., Eds.; John
Wiley & Sons, Inc.: New York, 1992.
(2) ActiVe Oxygen in Chemistry, Vol. 2; Foote, C., Valentine, J.,
Greenberg, A., Liebman, J., Eds.; Kluwer Academic Publishers:
Dordrecht, 1995; Chapter 1.
(3) Sheldon, R. A.; Arends, I. W. C. E. In AdVances in Catalytic ActiVation
of Dioxygen by Metal Complexes; Sima´ndi, L. I., Ed.; Kluwer
Academic Publishers: Dordrecht, 2003; Chapter 3.
(4) Sima´ndi, L. In AdVances in Catalytic ActiVation of Dioxygen by Metal
Complexes; Kluwer Academic Publishers: Dordrecht, 2003; Chapter
6.
(8) Cicco, S.; Latronico, M.; Mastrorilli, P.; Suranna, G.; Nobile, C. J.
Mol. Catal. A: Chem. 2001, 165, 135.
(5) Gates, B. Catalytic Chemistry; John Wiley & Sons, Inc.: New York,
1992.
(6) Paine, T.; Weyhermu¨ller, T.; Wieghardt, K.; Chaudri, P. J. Chem. Soc.,
Dalton Trans. 2004, 2092.
(7) (a) Betzemeier, B.; Cavazzini, M.; Qiuci, S.; Knochel, P. Tetrahedron
Lett. 2000, 41, 4343. (b) Gamez, P.; Arends, I.; Reedijk, J.; Sheldon,
R. J. Chem. Soc., Chem. Commun. 2003, 2414.
(9) Marko´, I.; Giles, P.; Tsukazaki, M.; Chelle´-Regnaut, I.; Gautier, A.;
Brown, S.; Urch, C. J. Org. Chem. 1999, 64, 2433.
(10) (a) McGrath, K. M.; Prakash, G. K. S.; Olah, G. A. J. Ind. Eng. Chem.
(Seoul, Repub. Korea) 2004, 10, 1063–1080. (b) Liu, H.; Song, C.;
Zhang, L.; Zhang, J.; Wang, H.; Wilkinson, D. P. J. Power Sources
2006, 155, 95–110. (c) Anthony, C. R.; Serra, D.; McElwee-White,
L. ACS Symp. Ser. 2007, 954, 296–310.
10.1021/ic802394a CCC: $40.75  2009 American Chemical Society
Inorganic Chemistry, Vol. 48, No. 9, 2009 4171
Published on Web 04/06/2009

Donzelli and Potvin
tetra-tert-butoxide, Ti(OtBu)4, was purchased in Aldrich Sure-Seal
bottles and used directly. Solvents were from Caledon Laboratories
(Georgetown, ON, Canada). Prior to use, tetrahydrofuran (THF)
was distilled over K and CH2Cl2 was distilled over P2O5. All NMR
spectra were acquired in CDCl3 at 23 °C at 300 or 400 MHz on
Bruker ARZ instruments. LDI-MS was carried out on MALDI
Voyager-DE spectrometer (PerSeptive Biosystems) equipped with
a TOF detector in the positive ion mode. High-resolution EI-MS
was performed on a Waters GCT Premier instrument by Dr. Alex
Young, University of Toronto. Elemental analyses were performed
with weighing under N2 by Guelph Chemical Laboratories (Guelph,
ON, Canada). Crystal structure data collection, structural analysis,
and refinement were carried out by Dr. Alan Lough at the University
of Toronto.
graphite electrodes covalently modified with RuIV/III/II spe-
cies11 has demonstrated the potential utility of surface-grafted
molecular electrocatalysts in alcohol oxidation.
As exemplified by the Katsuki-Sharpless catalyst,12 TiIV
alkoxide complexes are capable of rapid alcohol-alkoxide
exchanges for reactant loading and product unloading.
Because TiIV is not itself oxidizable, a TiIV-based alcohol
electro-oxidation catalyst would need to include oxidizable
ligands that would reversibly store two oxidation equivalents
to drive the R-deprotonation of a Ti-bound alkoxide, as well
as sites for alcohol attachment and surface binding. The use
of redox-active ligands is not new: the ligands in the
Wieghardt catalysts,13 for instance, cycle between two
oxidation states, and porphyrin cation radical complexes of
Fe and Ru are suspected intermediates in oxidation cata-
lyzes.14 TiIV is additionally inexpensive and environmentally
benign. While solution-phase complexes aggregate through
µ-O linkages upon hydrolysis, surface-chemisorbed TiIV
complexes should be relatively insensitive to hydrolysis and
unable to aggregate.
Sample Procedure for Tris(dithiocarbamato) Complexes:
Preparation of Isopropoxy-tris(pyrrolidine-N-carbodithioato)ti-
tanium(IV), Ti(1)3(OiPr). Ti(OiPr)4 (0.33 mL, 1.1 mmol) was
added to 40 mL of THF, followed by anhydrous CS2 (0.2 mL, 3.3
mmol) and, subsequently, pyrrolidine (0.28 mL, 3.3 mmol). The
resultant yellow mixture was stirred at room temperature for
approximately 16 h. The solvent and reaction byproduct were
removed under reduced pressure, and the product was recovered
1
We report here our initial exploration in this area, using
the well-known dithiocarbamate ligands. These are known
to be reversibly oxidized to dimeric thiuram disulfides,15
which can also act as ligands.16 We therefore aimed for
L2Ti(OR)2 species in which the dithiocarbamate/thiuram
disulfide couple can store two oxidation equivalents, and in
which one of the alkoxy groups is available for exchange
with a fuel alcohol while the other is used for anchoring to
an electrode by exchange with a surface OH group. Dithio-
carbamates are easily prepared in situ from amines of varying
basicities, and they bind symmetrically as bidentates,17 yet
their coordination chemistry at TiIV alkoxides is essentially
undeveloped (only three prior examples of mixed dithiocar-
bamate-alkoxide TiIV complexes have been reported)18,19 and
their electrochemistry is completely unexplored.
as a solid yellow residue in quantitative yield (0.587 g, 97%). H
NMR: δ 1.24 (d, 6H), 2.00 (bm, 12H), 3.71 (bm, 12H), 4.6 (h,
1H) ppm; 13C NMR: δ 24.72, 25.08, 49.86, 82.29, 201.19 ppm.
LDI-MS: m/z 485.9 (100%, M-OiPr), 398.9 (41%, M-1). Anal.
Calcd for C18H31N3OS6Ti: 39.62%C, 5.73%H, 7.70%N; Found
39.52%C, 6.05%H, 7.68%N. Epa +1.58 V. Epc -1.89, -2.43 V.
tert-Butoxy-tris(pyrrolidine-N-carbodithioato)titanium(IV),
Ti(1)3(OtBu). 1H NMR: δ 1.32 (s, 9H), 1.99 (bm, 12H), 3.70 (bm,
12H) ppm; 13C NMR: δ 25.12, 30.37, 49.86, 88.43, 201.34 ppm.
LDI-MS: m/z 485.8 (100%, M-OtBu), 412.9 (7%, M-1); Anal. Calcd
for C19H33N3OS6Ti: 40.77%C, 5.94%H, 7.51%N; Found 40.55%C,
6.13%H, 7.37%N. Epa +1.54 V. Epc -2.14, -2.59 V.
Isopropoxy-tris(N,N-diethylamine-N-carbodithioato)tita-
1
nium(IV), Ti(2)3(OiPr). H NMR: δ 1.15 (d, 6H), 1.19 (t, 18H),
3.76 (m, 12H), 4.55 (h, 1H) ppm; 13C NMR: δ 12.51, 24.57, 44.60,
81.59, 204.19 ppm. LDI-MS: m/z 491.9 (100%, M-OiPr), 402.9
(94%, M-2); Anal. Calcd for C18H37N3OS6Ti: 39.18%C, 6.76%H,
7.62%N; Found 38.98%C, 6.95%H, 7.43%N. Epa +1.49 V. Epc
-2.06, -2.66 V.
Experimental Section
General Procedures. All reactions were carried out under Ar.
All reagents were Sigma-Aldrich products. Titanium tetraisopro-
poxide, Ti(OiPr)4, was distilled under Ar prior to use. Titanium
tert-Butoxy-tris(N,N-diethylamine-N-carbodithioato)tita-
nium(IV), Ti(2)3(OtBu). 1H NMR: δ 1.18 (bm, 18H), 1.22 (s, 9H),
3.75 (bm, 12H) ppm; 13C NMR: δ 12.57, 30.22, 44.63, 87.62,
204.20 ppm. LDI-MS: m/z 491.9 (100%, M-OtBu), 417.0 (9%,
M-2); Anal. Calcd for C19H39N3OS6Ti: 40.33%C, 6.95%H, 7.43%N;
Found 39.98%C, 7.23%H, 7.45%N.
(11) (a) Geneste, F.; Moinet, C New J. Chem. 2004, 28, 722–726. (b)
Geneste, F.; Moinet, C.; Ababou-Girard, S.; Solal, F Inorg. Chem.
2005, 44, 4366–4371.
(12) Katsuki, T.; Martin, V. S. Org. React. 1996, 48, 1–299.
(13) (a) Chaudri, P.; Hess, M.; Flo¨rke, U.; Wieghardt, K. Angew. Chem.,
Int. Ed. 1998, 37, 2217. (b) Paine, T.; Weyhermu¨ller, T.; Wieghardt,
K.; Chaudri, P. J. Chem. Soc., Dalton Trans. 2004, 2092. (c) Chaudri,
P.; Hess, M.; Mu¨ller, J.; Hildenbrand, K.; Eckhardt, B.; Weyhermu¨ller,
T.; Wieghardt, K. J. Am. Chem. Soc. 1999, 121, 9599.
(14) (a) Stephenson, N. A.; Bell, A. T. J. Mol. Catal. A: Chem. 2007, 275,
54–62. (b) Ezhova, M. B.; James, B. R. In AdVances in Catalytic
ActiVation of Dioxygen by Metal Complexes; Sima´ndi, L. I., Ed.;
Kluwer Academic Publishers: Dordrecht, 2003; Chapter 1.
(15) Liu, M.; Visco, S. J.; De Jonghe, L. C. J. Electrochem. Soc. 1989,
136, 2570–2575.
(16) (a) Saravanan, M.; Prakasam, B. A.; Ramalingam, K.; Bocelli, G.;
Cantoni, A. Z. Anorg. Allg. Chem. 2005, 631, 1688–92. (b) Bond,
A. M.; Hollenkamp, A. F. Inorg. Chem. 1990, 29, 284–289. (c)
Brinkhoff, H. C.; Cras, J. A.; Steggerda, J. J.; Willemse, J. Rec. TraV.
Chim. Pays-Bas 1969, 88, 633–640.
Isopropoxy-tris(morpholine-N-carbodithioato)titanium(IV),
Ti(3)3(OiPr). 1H NMR: δ 1.21 (d, 6H), 3.71 (bm, 12H), 3.97 (bm,
12H), 4.6 (h, 1H) ppm; 13C NMR: δ 24.63, 47.39, 66.90, 82.67,
205.10 ppm. LDI-MS: m/z 533.8 (100%, M-OiPr), 430.9 (48%,
M-3); Anal. Calcd for C18H31N3O4S6Ti: 36.41%C, 5.26%H, 7.08%N;
Found 36.64%C, 5.36%H, 6.90%N. Epa +1.47 V. Epc -2.17, -2.65
V.
tert-Butoxy-tris(morpholine-N-carbodithioato)titanium(IV),
Ti(3)3(OtBu). 1H NMR: δ 1.25 (s, 9H), 3.66 (bm, 12H), 3.96 (bm,
12H) ppm; 13C NMR: δ 30.25, 47.45, 66.50, 88.70, 205.15 ppm.
LDI-MS: m/z 533.7 (100%, M-OtBu), 444.9 (19%, M-3); Anal.
Calcd for C19H33N3O4S6Ti: 37.55%C, 5.47%H, 6.91%N; Found
37.42%C, 5.69%H, 6.64%N.
(17) Bond, A. M. Coord. Chem. ReV. 1984, 54, 23–98.
(18) Choukroun, R.; Gervais, D. Synth. React. Inorg., Met.-Org. Chem.
1978, 8, 137–147.
Isopropoxy-tris(N′-methylpiperidine-N-carbodithioato)tita-
nium(IV), Ti(4)3(OiPr). 1H NMR: δ 1.16 (d, 6H), 2.26 (s, 9H), 2.40
(bm, 12H), 3.94 (bm, 12H), 4.55 (h, 1H) ppm; 13C NMR: δ 24.64,
(19) Shen, H.; Chan, H.-S.; Xie, Z. Organometallics 2007, 26, 2694–2704.
4172 Inorganic Chemistry, Vol. 48, No. 9, 2009

Dithiocarbamatotitanium(IV) Alkoxides
Chart 1
45.78, 46.90, 54.33, 82.35, 204.42 ppm. LDI-MS: m/z 572.8 (100%,
M-OiPr), 456.9 (50%, M-4); HR-EI-MS: m/z 573.0550 (M-OiPr,
C18H33N6S6Ti requires 573.0570), 457.0692 (M-4, C15H29N4OS4Ti
requires 457.0704). Epa +1.32 V. Epc -2.02 V.
tert-Butoxy-tris(N′-methylpiperidine-N-carbodithioato)tita-
1
nium(IV), Ti(4)3(OtBu). H NMR: δ 1.23 (s, 9H), 2.25 (s, 9H),
2.40 (bm, 12H) 3.93 (bm, 12H) ppm; 13C NMR: δ 30.25, 45.80,
46.92, 54.43, 88.42, 204.44 ppm. LDI-MS: m/z 572.8 (100%,
M-OtBu), 470.9 (18%, M-4); Anal. Calcd for C22H42N6OS6Ti:
40.85%C, 6.54%H, 12.99%N; Found 40.80%C, 6.62%H, 12.63%N.
Isopropoxy-tris(N,N′,N′-trimethylethane-1,2-diamine-N-car-
1
bodithioato)titanium(IV), Ti(5)3(OiPr). H NMR: δ 1.15 (d, 6H),
2.21 (d, 18H), 2.53 (t, 6H), 3.30 (s, 9H), 3.80 (bm, 6H), 4.55 (h,
1H) ppm; 13C NMR: δ 24.62, 38.63, 45.62, 50.28, 55.99, 82.01,
205.71 ppm. LDI-MS: m/z 578.9 (100%, M-OiPr), 461.0 (11%,
M-5); Anal. Calcd for C21H46N6OS6Ti: 39.48%C, 7.26%H, 13.15%N;
Found 39.20%C, 6.99%H, 12.87%N. Epa +1.19 V. Epc -2.08 V.
tert-Butoxy-tris(N,N′,N′-trimethylethane-1,2-diamine-N-car-
bodithioato)titanium(IV), Ti(5)3(OtBu). 1H NMR: δ 1.23 (s, 9H),
2.21 (s, 18H), 3.28 (s, 9H) 3.85 (bm, 12H) ppm; 13C NMR: δ 30.26,
38.65, 45.62, 50.30, 56.06, 88.07, 205.79 ppm. LDI-MS: m/z 578.9
(100%, M-OtBu), 475.0 (45%, M-5); Anal. Calcd for
C22H48N6OS6Ti: 40.47%C, 7.41%H, 12.87%N; Found 40.14%C,
7.12%H, 13.16%N.
was disordered over two orientations in occupancy ratio 57.4 (
i
0.9%:42.6 ( 0.9%. Similarly, the Pr group of Ti(1)3(OiPr) was
also disordered over two orientations in occupancy ratio 82.3 (
0.9%:17.7 ( 0.9%, sharing the CH3 carbon positions. In both
structures, the disordered C atom positions were refined anisotro-
pically.
Electrochemistry. All the electrochemical analyses were per-
formed in a one-compartment, three-electrode cell with a Pt disk
working electrode, a Ag/AgCl pseudo-reference electrode, and a
graphite counter-electrode, under an Ar atmosphere, connected to
an Obbligato Objectives Faraday MP potentiostat. The sample
solutions were 0.03 M in freshly distilled CH2Cl2, containing 0.1
Sample Procedure for Bis(dithiocarbamato) Complexes: Prepa-
ration of Diisopropoxy-bis(pyrrolidine-N-carbodithioato)tita-
nium(IV), Ti(1)2(OiPr)2. Ti(OiPr)4 (0.98 mL, 3.3 mmol) was added
to 30 mL of anhydrous CH2Cl2. Pyrrolidine (0.28 mL, 3.3 mmol)
was added, and the mixture was left to cool in an ice-water bath
for 1 h. Then, CS2 (0.2 mL, 3.3 mmol) dissolved in 10 mL of
anhydrous CH2Cl2 was added via a syringe pump over the course
of 2 h at 0 °C. The yellow mixture was left to stand overnight at
0 °C. The solvent and reaction byproduct were removed under
reduced pressure, and the product was recovered as a yellow oil.
1H NMR: δ 1.25 (d, 6H), 1.96 (m, 4H), 3.65 (m, 4H), 4.94 (h, 1H)
ppm. 13C NMR: δ 25.71, 25.08, 50.81, 80.77, 199.41 ppm.
Diisopropoxy-bis(N,N-diethylamine-N-carbodithioato)titanium-
n
M Bu4NPF6 as supporting electrolyte. All scans were performed
at a 200 mV s-1 sweep rate. They were either limited to oxidation
or reduction regimes or covered both in succession and were then
initiated in both anodic and cathodic directions. The potentials were
internally calibrated with ferrocene and are reported with respect
to Ag/AgCl (satd. KCl). To assess the redox potentials of the free
ligands, 1:1 mixtures of amine and CS2 were prepared in THF under
Ar. After vacuum removal of the volatiles, NMR in CDCl3 showed
the solid products to be the ammonium dithiocarbamate salts
R2NH2+-SSCNR2, and these were redissolved in CH2Cl2 for
analysis as with the complexes.
Results and Discussion
1
(IV), Ti(4)2(OiPr)2. H NMR: δ 1.25 (d, 6H), 2.26 (s, 3H), 2.40
(m, 4H), 3.96 (m, 4H), 4.93 (h, 1H) ppm. 13C NMR: δ 26.49, 45.69,
47.80, 54.33, 80.88, 203.03 ppm.
Synthesis. The direct reactions of 3:3:1 ratios of a
secondary amine, CS2, and Ti(OR)4 (R ) iPr or tBu) in THF,
CH2Cl2, or CDCl3 gave instantaneous color changes and led
to the tris(dithiocarbamato) complexes L3TiOR as the only
detected products, and these could be isolated in quantitative
yields (eq 1). These were also the only reaction products at
2:2:1 amine:CS2/Ti stoichiometry, aside from liberated HOR
and unreacted Ti(OR)4. Chart 1 lists the ligands L explored.
Of these, 3-6 presented the possibility of tridentate coor-
dination through the oxygen or free nitrogen atoms, but this
was not observed.
Diisopropoxy-bis(N,N′,N′-trimethylethane-1,2-diamine-N-car-
bodithioato)titanium(IV), Ti(5)2(OiPr)2. 1H NMR: δ 1.23 (d, 6H),
2.20 (d, 9H), 2.51 (t, 2H), 3.31 (s, 3H), 3.84 (t, 2H), 4.93 (h, 1H)
ppm. 13C NMR: δ 25.68, 39.59, 45.62, 51.29, 55.99, 80.67, 204.13
ppm.
Diisopropoxy-bis(piperidine-N-carbodithioato)titanium(IV),
1
Ti(6)2(OiPr)2. H NMR: δ 1.25 (d, 6H), 2.86 (m, 4H), 3.93 (m,
4H), 4.94 (h, 1H) ppm. 13C NMR: δ 25.71, 45.76, 49.34, 80.88,
202.74 ppm.
Crystallography. Yellow crystals of Ti(1)3(OiPr) were grown
over approximately 6 months after layering petroleum ether over a
CH2Cl2 solution. Similarly, yellow crystals of Ti(3)3(OtBu) were
grown from CDCl3. Diffraction intensities were collected on a
Bruker-Nonius Kappa CCD instrument using a fine-focus sealed
tube Mo KR source and graphite monochromator. Unique reflections
were corrected for absorption (Denzo-SMN) and used in all
calculations. Heavy-atom positions were determined by direct
methods (SHELXS-97) and refined anisotropically. The hydrogen
atoms were assigned idealized positions according to a riding model
and refined isotropically. Structure refinement used full-matrix least-
squares on F2 (SHELXL-97). The Ti(3)3(OtBu) crystal contained
a molecule of CDCl3, the non-hydrogen atom positions of which
CS2
Ti(OiPr)4
8
amine8 TiL3(OiPr)
(1)
THF/RT
The L3TiOR products were initially identified by 1H NMR.
Beyond showing the uptake of amine and the release of HOR
when conducted in CDCl3, the reactions showed the appear-
1
ance of new L and OR H NMR signals with integration
ratios consistent with a 3:1 L/Ti reaction stoichiometry, as
well as signals from unreacted Ti(OR)4 at amine/Ti ratios
below 3, or from unreacted amine, at ratios above 3. When
R was iPr, the new OiPr methine signals consistently appeared
t
were also refined anisotropically. The Bu group of Ti(3)3(OtBu)
Inorganic Chemistry, Vol. 48, No. 9, 2009 4173

Donzelli and Potvin
near 4.6 ppm, some 0.2 ppm downfield of that of Ti(OiPr)4.
13C NMR confirmed the presence of CdS groups and single
OiPr (or OtBu) C-O signals shifted some 5-6.5 ppm (or
8-9 ppm) downfield of the corresponding C-O signals with
Ti(OiPr)4 (or Ti(OtBu)4).20 In all cases, the single set of ligand
NMR signals indicated symmetry or rapid exchange between
inequivalent coordination sites. Fast positional exchange has
previously been witnessed in tetrakis21 and chlorotris(dithio-
carbamate)22 complexes. The only other tris(dithiocarbamate)
Ti(IV) alkoxide is a Ti(2)3(OR) analogue and its NMR
spectra (in pyridine-d5) showed similar 13CS2 and ethyl
signals.
Moreover, the L2Ti(OR)2 species were unstable and slowly
transformed to L3Ti(OR). The mass spectrum of the putative
Ti(1)2(OiPr)2 was not significantly different from that of
Ti(1)3(OiPr), perhaps because of this.
We attempted to obtain L2Ti(OR)2 products in pure state
by other means: Not surprisingly, treatment of sodium
dithiocarbamate salts with Ti(OR)4 failed to cause the
expulsion of sodium alkoxides. Bis(dithiocarbamate)diphe-
noxy complexes have reportedly been prepared from
(PhO)2TiCl2.23 Treatment of Ti(THF)2Cl4 or Ti(OR)2Cl2
(formed in situ from equal amounts of Ti(OR)4 and
Ti(THF)2Cl4) with sodium pyrazole-1-carbodithioate, pre-
pared and purified separately, led to the isolation of new
species. The spectroscopic data obtained for these products
were consistent with L2TiCl2 and L2Ti(OR)2, respectively,
but they proved to be too unstable to permit further
characterization, possibly because of disproportionation.
In both series of complexes, the alkoxy NMR signal
positions were quite independent of the basicity of the amine
component. The OC1H signals in the isopropoxy cases were
consistently at 4.55-4.60 ppm with the tris complexes,
compared to 4.93-4.94 ppm with the bis complexes. The
tert-butoxy O13CC signals also appeared in a narrow range
(87.6-88.7 ppm), as did the isopropoxy O13CH signals
(81.6-82.7 ppm with the tris complexes and 80.7-80.9 ppm
with the bis species). Besides having a diagnostic value, this
indicates a poor transmission of the amine electron density
through the metal. Conversely, there was little transmission
in the opposite direction: The 13CS2 signals were at virtually
the same positions in the spectra of the OtBu complexes as
in those of their OiPr analogues, and this was also the case
with the signals from the amine components. Although
measured in a different solvent, this was also true of the
signals from a Ti(2)3(OR) analogue. We earlier noted that
the OC1H chemical shifts from TiOiPr groups were related
to the Ti-O bond lengths, which are shortened by π-dona-
tion, with shorter bonds associated with more downfield
positions.24 (Unfortunately, there were insufficient data in
the literature to draw a correlation with the corresponding
13C NMR chemical shifts.) The chemical shift and bond
length measured for Ti(1)3(OiPr) fit within that correlation.
In the absence of a crystal structure for a bis species, the
chemical shift differences between the tris and bis species’
signals are less easily explained.
The structures were confirmed by 13C NMR, elemental
analysis and, in two cases, crystallography (vide infra). LDI-
MS showed consistent fragmentations of OR or dithiocar-
bamate ligands (L) from all L3TiOR. Only Ti(4)3(OiPr) did
not provide elemental analyses that were consistent with the
formula, indicative instead of possible carbonation and/or
partial hydrolysis. However, high-resolution measurements
of the EI fragments M-4 and M-OiPr were consistent with
the compound’s identity.
Evidence of a second species, later determined to be
bis(dithiocarbamato) complexes, was found at 1:1:1 amine/
CS2/Ti stochiometry, but this evolved to tris product
overnight. When CS2 was added last and slowly at 0 °C,
this second product was the only detectable product (eq 2),
aside from excess Ti(OR)4 (reactions in THF led to ca. 1:1
mixtures with L3Ti(OR) coproducts).
amine
CS2
Ti(OiPr)4.
8.
8 TiL2(OiPr)2
(2)
0 oC
CH2Cl2/RT
The OiPr methine signals now appeared at 4.9 ppm (0.5
ppm downfield of that with Ti(OiPr)4), and integration of
the amine and OR signals revealed a 2:2:1 reaction stoichi-
ometry. Again, single sets of signals were found in both 1H-
and 13C NMR spectra, that is, without separate signals from
bridging alkoxide groups, so they appear to be mononuclear
species of formula L2Ti(OiPr)2 with equivalence through
symmetry or rapid positional exchange. A dimer (or other
aggregate) would need to undergo fast ligand scrambling to
be consistent with the single set of NMR signals. The
observation of slow or absent exchange with the liberated
HOiPr, confirmed by the absence of any effect on the NMR
signal positions or linewidths upon removal of the HOiPr,
indicates that there is no fast exchange mechanism available.
Moreover, we noted no concentration dependence with either
the signal positions nor the signal linewidths, which would
have been a sign of reversible aggregation. On the weight
of the evidence, therefore, we favor the formulation of these
products as mononuclear species. Unfortunately, it has
proven impossible to prepare or isolate these 2:2:1 products
free of either excess Ti(OR)4 or of co-produced L3Ti(OR),
and the crude products were oily and failed to grow crystals.
Ti(OiPr)4 + HL h TiL(OiPr)3 + HOiPr
TiL(OiPr)3 + HL h TiL2(OiPr)2 + HOiPr
TiL2(OiPr)2 + HL h TiL3(OiPr) + HOiPr
(3)
(4)
(5)
Substitutions at Ti(OR)4 by an associative mechanism are
hindered by interligand repulsions involving the OR alkyl
groups, which are evidenced by the fact that Ti(OR)4 is
i
t
monomeric when R ) Pr or Bu but oligomeric when R )
(20) Holloway, C. E. J. Chem. Soc., Dalton Trans. 1976, 1050–1054.
(21) Muetterties, E. L. Inorg. Chem. 1973, 12, 1963–1966.
(22) Hawthorne, S. L.; Fay, R. C. J. Am. Chem. Soc. 1979, 101, 5268–
5277.
(23) Sangari, H. S.; Sodhi, G. S.; Kaushik, N. K.; Singh, R. P. Synth. React.
Inorg., Met.-Org. Chem. 1981, 11, 373–382.
(24) Wallace, W. A.; Potvin, P. G. Inorg. Chem. 2007, 46, 9463–9472.
4174 Inorganic Chemistry, Vol. 48, No. 9, 2009

Dithiocarbamatotitanium(IV) Alkoxides
Et or Me.25 As no evidence of monosubstituted product was
observed during substitutions of OR with dithiocarbamates
L, while TiL2(OiPr)2 and Ti(OiPr)4 can evidently coexist, it
appears that the second substitution (eq 4) was easier than
the first (eq 3), as that initial substitution relieved some of
the steric congestion. This is consistent with the kinetics of
analogous substitutions by other monobasic bidentates that
produce stable, hexacoordinate TiL2(OR)2 species without
detectable monosubstituted intermediates.26,27 A relief of
steric congestion is also likely the driving force for the ligand
redistribution that accompanies the insertion of CS2 into
Ti(NEt2)(OiPr)3 to produce Ti(2)2(OiPr)2. The third substitu-
tion (eq 5) raises the coordination number from six to seven,
which the small dithiocarbamate bite angle allows, but it
appears to be sufficiently slowed to permit the accumulation
of TiL2(OiPr)2 under some conditions, although TiL3(OiPr)
accumulates over time and appears to be more stable, as it
suffers no OR-OR repulsions. The fourth substitution, raising
the coordination number to eight, appears to be too difficult
in these complexes, probably because the consequent build
up of congestion at the metal is no longer justified by a relief
of interligand repulsions, though other driving forces can be
exploited to prepare tetrakis(dithiocarbamate) complexes,
from TiCl4 and excess sodium dithiocarbamate salts28,29 or
by the insertion of CS2 into tetrakis(dialkylamido) com-
plexes.30
We also used the diamines piperidine and N,N′-dimethyl
ethylenediamine in attempts to form multinuclear variations
containing bridging ligands at 1:2:1 and other amine/CS2/
Ti(OR)4 stoichiometries. Unfortunately, much insoluble
material was observed, suggestive of oligomerization, as well
as only very broad signals in the NMR spectra from what
remained in solution. Given the results with non-bridging
ligands, 1 equiv of a bis(dithiocarbamate) should have given
TiL2(OR)2 and unreacted Ti(OR)4, then either a macrocycle
or a polymer with all metals doubly substituted. The observed
outcome could mean that the geometry of the TiL2(OR)2
intermediates was not conducive to macrocyclization, perhaps
simply because the L-Ti-L angle was too large. Parallel
attempts to combine two disubstituted species with a half-
equivalent of a bis(dithiocarbamate) also failed.
Figure 1. Evolution of a mixture of Ti(4)2(OiPr)2 [A] and Ti(4)3(OiPr) [B]
treated with an excess of Ti(OiPr)4, followed through the diagnostic 1H
NMR heptets of the OCH groups.
added, or simply over time. These results might indicate
kinetic and thermodynamic control, which requires an
equilibration between the two products, with L3Ti(OR) being
more stable. To test this hypothesis, we sought to force a
shift in the putative equilibrium. A mixture of Ti(4)2(OiPr)
and Ti(4)3(OiPr) was generated in CH2Cl2, then placed in a
sealed NMR tube, dried under vacuum, and redissolved in
CDCl3. To this was added a large excess of Ti(OiPr)4, and
the time course of the NMR spectra was monitored over a
period of 6 days. There was very little sign of decomposition
or polymerization over this period. The sample completely
converted the Ti(4)3(OiPr) component into Ti(4)2(OiPr)2
(Figure 1). This is consistent with the two species being in
equilibrium (eq 6, a combination of eqs 3-5):
3Ti(4)2(OiPr)2 h 2Ti(4)3(OiPr) + Ti(OiPr)4
(6)
In contrast, a sample treated with excess CS2 completely
converted into Ti(4)3(OiPr) in the same time frame, but we
do not have a satisfactory explanation for this shift. In another
sample treated with excess amine, the proportions of
Ti(4)3(OiPr) and Ti(4)2(OiPr)2 were not much affected, but
a new heptet appeared near 4.7 ppm along with a broader
multiplet near 4.4 ppm. More work will be needed to identify
this new species. Interestingly, these phenomena were only
observed in CDCl3. Reactions conducted in THF showed,
after transfer to CDCl3, the trisubstituted complex as the
exclusive product. Solvent coordination and/or solvent polar-
ity may therefore play a role in these reactions.
Equilibration. The formation of L2Ti(OR)2 to the exclu-
sion of L3Ti(OR) by addition of the CS2 last and in the
presence of excess Ti(OR)4 could be because the third HL/
HOR substitution on L2Ti(OR)2 was rendered kinetically
uncompetitive with the earlier substitutions under these
conditions. On the other hand, we found that the proportion
of the L3Ti(OR) co-product increased when the amine was
added as the last component, or when an excess of CS2 was
Crystal Structures. Diffraction-quality crystals of Ti(1)3-
(OiPr) and Ti(3)3(OtBu) were obtained, and Table 1 reports
the crystallographic data. As with tris(dithiocarbamate)
(25) Bradley, D. C.; Holloway, C. E. J. Chem. Soc., A 1968, 131, 6–9;
Inorg. Chem. 1964, 3, 1163-1164.
(26) Holloway, C. E.; Sentek, A. E. Can. J. Chem. 1971, 49, 519–522.
(27) Potvin, P. G.; Fieldhouse, B. G. Can. J. Chem. 1995, 73, 401–413.
(28) Bhat, A. N.; Fay, R. C.; Lewis, D. F.; Lindmark, A. F.; Strauss, S. H.
Inorg. Chem. 1974, 13, 886–892.
(29) (a) Kumar, S.; Kaushik, N. K. Synth. React. Inorg., Met.-Org. Chem.
1982, 12, 159–171. (b) Colapietro, M.; Vaciago, A.; Bradley, D. C.;
Hursthouse, M. B.; Rendall, I. F. J. Chem. Soc., Dalton Trans. 1972,
1052–1057.
31
complexes obtained from TiCl4 and (Cp)2TiCl2,32 these
structures (Figures 2 and 3) show a distorted pentagonal
pyramidal structure, in these cases with the alkoxy group
(31) Lewis, D. F.; Fay, R. C. J. Am. Chem. Soc. 1974, 96, 3843–3846.
(32) Steffen, W. L.; Chun, H. K.; Fay, R. C. Inorg. Chem. 1978, 17, 3498–
3503.
(30) Bradley, D. C.; Gitlitz, M. H. J. Chem. Soc. 1969, 1152–1156.
Inorganic Chemistry, Vol. 48, No. 9, 2009 4175

Donzelli and Potvin
Table 1. Crystallographic Dataa
Ti(1)3(OiPr)
Table 2. Selected Bond Lengths (Å) and Angles (deg), with
Uncertainties in the Least Significant Digits in Brackets
Ti(3)3(OtBu)
Ti(1)3(OiPr)
Ti(3)3(OtBu)
formula
M
C18H31N3OS6Ti
545.72
C19H33N3O4S6Ti·CDCl3
728.12
Ti1-O1
1.769(2)
2.5765(9)
2.5583(10)
2.5221(10)
2.5771(10)
2.5766(10)
2.5308(9)
1.718(3)
1.711(3)
1.720(3)
1.700(3)
1.707(3)
1.721(3)
164.0(3)
135.5(7)
69.27(3)
75.72(3)
75.43(3)
95.02(3)
82.88(3)
85.71(3)
89.33(3)
68.36(3)
71.40(3)
67.93(3)
116.61(18)
113.77(18)
112.71(17)
163.68(8)
1.742(2)
2.5780(12)
2.5556(11)
2.5242(12)
2.5460(10)
2.5566(12)
2.5313(12)
1.716(4)
1.706(4)
1.714(4)
1.708(4)
1.697(4)
1.714(4)
159.2(4)
165.2(5)
68.98(3)
74.77(4)
75.14(4)
88.03(4)
84.82(4)
85.36(4)
91.55(4)
68.52(4)
72.05(4)
68.13(3)
116.3(2)
113.0(2)
113.34(19)
167.45(12)
Ti1-S1
space group
a (Å)
b (Å)
P21/c (No. 14)
11.0462(3)
14.5216(5)
15.7630(5)
98.811(2)
2498.68(14)
4
1.451
0.86
0.0491
0.115
Pna21 (No. 33)
11.8259(3)
23.5185(6)
11.6880(3)
90
Ti1-S2
Ti1-S3
Ti1-S4
c (Å)
Ti1-S5
ꢀ (deg)
V (Å3)
Z
Ti1-S6
3250.75(14)
4
S1-C1
Dcalc (g cm-3
)
1.486
0.926
S2-C1
µ(Mo KR) (mm-1
)
S3-C2
b
S4-C2
R(Fo)
0.0383
b
2
S5-C3
Rw(Fo )
0.082
a In both cases, T ) 150(1) K and λ ) 0.71073 Å. Esd’s are expressed
S6-C3
as uncertainties in the least significant digits in brackets. b For reflections
Ti1-O1-C16
Ti1-O1-C16A
S1-Ti1-S2
S1-Ti1-S4
S1-Ti1-S5
S2-Ti1-S3
S2-Ti1-S4
S2-Ti1-S5
S2-Ti1-S6
S3-Ti1-S4
S3-Ti1-S6
S5-Ti1-S6
S1-C1-S2
S3-C2-S4
S5-C3-S6
S2-Ti1-O1
2
where I > 2σ(I). The weights w used in the calculation of Rw(Fo ) are given
by w ) 1/[σ2(Fo ) + (0.0699P)2 + 0.7976P] where P ) (Fo + 2Fc2)/3.
2
2
also disordered; both orientations showed large Ti-O-C
angles (165° and 167.5°) which, along with an even shorter
Ti-O distance (1.742 Å), also indicated π donation. The
only other tris(dithiocarbamate) Ti(IV) alkoxide complex
showed a shorter average Ti-S bond length (2.532 Å vs
2.557 Å in Ti(1)3(OiPr) and 2.549 Å in Ti(3)3(OtBu)) but
no sign of π-donation by the primary alkoxide (Ti-O 1.833
Å).
Figure 2. ORTEP plot of the crystal structure of Ti(1)3(OiPr), with the iPr
group shown in its major orientation. H atoms have been omitted for clarity.
Otherwise, the bond lengths and angles (Tables 2 and 3)
and the degrees of distortion were very similar to those
measured in the previous dithiocarbamate complexes. Be-
cause of these similarities, it appears that the nature of the
dithiocarbamate ligand does not much affect the structure
of the complex.
Electrochemistry. To our knowledge, no electrochemical
studies of TiIV dithiocarbamate complexes have yet been
conducted. The new L3Ti(OiPr) complexes were therefore
examined by standard cyclic voltammetry (CV) in CH2Cl2.
For comparisons, we prepared the corresponding ammonium
dithiocarbamate salts and examined them under the same
conditions. Figure 4 presents typical traces, all involving
ligand 1.
Scanning the complex in the positive direction and then
in the negative direction (panel a) showed an oxidation
(process I) which gave rise to a re-reduction (process II) that
was absent when scanning only in the negative direction
(panel b), while the second anodic scan (gray trace in panel
a) revealed two new, lower-potential oxidations (processes
III and IV). Both this and the negative-only scan (panel b)
showed a reduction (process V) that probably does not
involve the ligand and which we attribute to the TiIV/III
couple. The free ligands underwent oxidations which differed
Figure 3. ORTEP plot of the crystal structure of Ti(3)3(OtBu), with the
tBu group shown in its major orientation. The solvate and H atoms have
been omitted for clarity.
located at one of the axial positions while the six sulfur atoms
account for the second axial and the five equatorial positions.
i
The Pr group in Ti(1)3(OiPr) was disordered between two
orientations; the major orientation was associated with a large
Ti-O-C angle (164°), but not the minor one (135.5°), while
both shared the oxygen at a short Ti-O distance (1.769 Å)
indicative of π-bonding. The tBu group in Ti(3)3(OtBu) was
4176 Inorganic Chemistry, Vol. 48, No. 9, 2009

Dithiocarbamatotitanium(IV) Alkoxides
Similar low-potential oxidations have been measured at Pt
for Li+ salts of dithiocarbamates in DMSO (scan-rate-
dependent values between +0.14 and +0.38 V),33 where
strong solvation of the cation leaves a more naked and more
easily oxidized anion. We therefore surmise that oxidation
of the complex (process I) formed the thiuram disulfide
which, although known to bind in bidentate fashion to
otherwise neutral TiIV,34 dissociated and was reduced back
to free dithiocarbamate anion (process II) then reoxidized at
a lower potential (process III). On the basis of these low-
potential oxidations, the binding to TiIV apparently retarded
the ligand oxidations by about 1.0-1.2 V. Such retardation
by coordination to a metal has been earlier noted in organic
solvent.35 Finally, we attribute the second new oxidation
(process IV), observed on the second anodic scan in panel
a, to an oxidation of pyrrolidine liberated upon reduction of
the thiuram disulfide and subsequent fragmentation.
Overall, the L3Ti(OiPr) complexes underwent a variable
and irreversible reduction (peak cathodic currents at -1.87
to -2.17 V) and, in some cases, a second, irreversible
reduction near the solvent limit, as well as an irreversible
oxidation (peak anodic currents at +1.19 to +1.58 V) that
resulted in decomposition. The potentials are reported in the
Experimental Section. Similarly irreversible but higher-
potential oxidations have been measured with ZnII and CdII
bis(dithiocarbamates) in CH2Cl2 solution,36 also generating
thiuram disulfide complexes, but CoIII, RhIII and IrIII tris-
(dithiocarbamate) analogues oxidized at lower potentials in
the same solvent (E1/2 +0.68 to +1.12 V).37 It is noteworthy
that the complexes of the dithiocarbamates formed from the
two most basic amines (1, 2) were counter-intuitively
oxidized at higher potentials than those formed from the less
basic amines (3-5). This might be attributed to stronger
binding by 1 and 2, yet the average Ti-S bond lengths in
the complexes of 1 and 3 were not statistically significantly
different. In particular, those complexes formed from the
diamines (with 4 and 5) were the easiest to oxidize; this may
be a consequence of a preferential oxidation at the second
nitrogen atom.
With a more electron-rich alkoxy group, Ti(1)3(OtBu)
(Figure 4, panel d) underwent a slightly less positive
oxidation but a significantly more negative reduction than
did the OiPr analogue. The electronic effect is as expected
for the TiIV/III couple but is poorly transmitted to the
dithiocarbamate ligand, in confirmation of this same finding
by NMR. The putative bis(dithiocarbamate) species
Ti(1)2(OiPr)2 oxidized at an even more positive potential,
starting with a shoulder near +1.66 V (panel e).
Figure 4. Cyclic voltammograms at 200 mV s-1 of 0.03 M solutions in
n
CH2Cl2 containing 0.1 M Bu4NPF6: (a) full scan of Ti(1)3(OiPr) (second
scan in gray), (b) positive-only and negative-only scans of Ti(1)3(OiPr), (c)
free ligand, (d) Ti(1)3(OtBu) and (e) Ti(1)2(OiPr)2. In panels d and e, the
corresponding plot with Ti(1)3(OiPr) (panel b) is superposed in gray. Arrows
indicate the scan directions.
on the first and subsequent scans. For instance, the pyrro-
lidinium salt of 1- (panel c) showed an oxidation peak at
+0.67 V (all potentials vs Ag/AgCl) on the first scan but
also a second oxidation peak at +0.32 V on subsequent
scans. We attribute the higher-potential process (process VI)
to an oxidation of the hydrogen-bonded pyrrolidinium salt
which would be absent in panel a and which, after producing
the neutral, S-S linked thiuram disulfide, would no longer
be associated with a cation until re-reduction occurred on
the cathodic scan (process II), whence a more loosely
associated nBu4N+ salt could then undergo the lower-potential
oxidation (process III) on the subsequent anodic scans.
Conclusions
Ten alkoxy-tris(dithiocarbamato)titanium(IV) complexes
were prepared, isolated, and characterized. The crystal
(33) Lieder, M. Electrochim. Acta 2004, 49, 1813–1822.
(34) Cuadrado, I.; Moran, M. Transition Metal Chemistry; Kluwer:
Dordrecht, The Netherlands, 1986; Vol. 11, pp 375-381.
(35) Jovanoviæ, V. M.; Babiæ-Samardˇıja, K.; Sovilj, S. P. Electroanalysis
2001, 13, 1129–1135.
(36) Bond, A. M.; Hollenkamp, A. F. Inorg. Chem. 1990, 29, 284–289.
(37) Bond, A. M.; Colton, R.; Mann, D. R. Inorg. Chem. 1990, 29, 4665–
4671.
Inorganic Chemistry, Vol. 48, No. 9, 2009 4177

Donzelli and Potvin
structures of two of these were obtained, revealing strongly
π-coordinated alkoxy groups. While we have strong NMR
evidence that the target diisopropoxy-bis(dithiocarbamato)
analogues can be prepared, these were not isolable in pure
state and were instead prone to transform to the tris species.
Indeed, equilibria were apparent between the bis- and tris-
substituted complexes. Cyclic voltammetry in homogeneous
aprotic solvent revealed irreversible oxidation waves at
potentials shifted positive of those from the free dithiocar-
bamate salts, coupled to loss of the ligand.
Although the desired bis(dithiocarbamate) complexes proved
to be unviable and their oxidation potentials are perhaps too
high, these experiments have demonstrated the potential
utility of sulfur-bearing TiIV complexes in storing electro-
chemical oxidation equivalents. Work with other catalyst
designs is underway.
Acknowledgment. We thank the Natural Sciences and
Engineering Research Council (Canada) for funding and
Imperial Oil for a University Research Grant.
Undoubtedly, higher-dielectric and protic media such as
would be used in fuel cells would facilitate these oxidations
and help stabilize the coordinated disulfide products. For
some of our tris(dithiocarbamate) complexes, oxidation
would perhaps occur at sufficiently low potentials under those
conditions as to be driven by O2 reduction at a cathode.
Supporting Information Available: Crystallographic Informa-
tion Format (CIF) file, NMR signal assignments and coupling
constants. This material is available free of charge via the Internet
at http://pubs.acs.org.
IC802394A
4178 Inorganic Chemistry, Vol. 48, No. 9, 2009