Pendant arm functionalized benzamidinate titanium imido compounds: Experimental and computational studies of their reactions with CO2

Article abstract of DOI:10.1021/om049026f

The synthesis of cyclopentadienyl titanium tert-butylimido complexes containing pendant arm functionalized amidinate ligands were discussed. The approximately isosteric 'dummy arm' benzamidinate ligand^49,50 Me₃SiNC(Ph)NCH₂

Full text of DOI:10.1021/om049026f

Organometallics 2005, 24, 2347-2367
2347
Pendant Arm Functionalized Benzamidinate Titanium
Imido Compounds: Experimental and Computational
Studies of Their Reactions with CO2
Catherine L. Boyd,† Eric Clot,*,‡ Aldo E. Guiducci,† and Philip Mountford*,†
Chemistry Research Laboratory, University of Oxford, Mansfield Road, Oxford OX1 3TA, U.K.,
and LSDSMS (UMR 5636), cc 14, Universite´ Montpellier 2, 34095 Montpellier Cedex 5, France
Received December 9, 2004
The syntheses of cyclopentadienyl titanium tert-butylimido complexes containing pendant
arm functionalized amidinate ligands are reported, together with their reactions with CO2.
Mechanistic and DFT computational studies are also described. Reaction of [Ti(η-C5R4Me)-
(NtBu)Cl(py)] (R ) Me, H) with Li[Me3SiNC(Ph)N(CH2)nNMe2] (n ) 2, 3) afforded the
corresponding (dimethylamino)ethylidene or -propylidene functionalized benzamidinate
complexes [Ti(η-C5R4Me)(NtBu){Me3SiNC(Ph)N(CH2)nNMe2}] (R ) Me, n ) 2 (1), 3 (2); R )
H, n ) 2 (3), 3 (4)). The n-propyl functionalized amidinate complex [Ti(η-C5H4Me)(NtBu)-
{Me3SiNC(Ph)NCH2CH2Me}] (5) was also prepared for comparative purposes. The NMe2
donors in 3 and 4 coordinate relatively weakly, and in solution the activation Gibbs free
energy for NMe2 dissociation for 3 is ca. 10 kJ mol-1 higher than that for 4. All five compounds
react with CO2 to form the corresponding N,O-bound carbamate complexes [Ti(η-C5R4Me)-
{NtBuC(O)O}{Me3SiNC(Ph)NR′}] (R ) H, Me; R′ ) CH2CH2NMe2, CH2CH2CH2NMe2,
t
CH2CH2Me), which, in turn, extrude BuNCO (above -25 °C for R ) H and at room
temperature for R ) Me) to form the µ-oxo-bridged dimeric complexes [Ti2(η-C5R4Me)2(µ-
O)2{Me3SiNC(Ph)NR′}2], two of which have been structurally characterized. Only the C5Me5
carbamate complexes [Ti(η-C5Me5){NtBuC(O)O}{Me3SiNC(Ph)N(CH2)nNMe2}] can be isolated
and (for n ) 3) structurally characterized. The reaction of 4 and 5 with CO2 is spontaneous
at -78 °C, whereas the corresponding reaction of 3 does not occur at significant rates below
t
-35 °C. The rate of extrusion of BuNCO from [Ti(η-C5H4Me){NtBuC(O)O}{Me3SiNC(Ph)-
NR′] at -25 °C is twice as fast for R′ ) CH2CH2NMe2 as for CH2CH2CH2NMe2 and
CH2CH2Me. DFT calculations have modeled the cycloaddition/extrusion reactions of 3-5 in
conjunction with mechanistic and variable-temperature NMR experiments. The results show
that the favored route is for the cycloaddition reaction to take place on 16-valence-electron,
non pendant arm chelated isomers of 3 and 4 and that the relative ease of accessing these
intermediates in fact controls the TidNtBu/CO2 coupling reaction.
Introduction
emerged from a large body of such derivatives supported
by a range of supporting ligands. Although these
compounds have proven potential in materials chemis-
try (as a source of TiN)13-16 and olefin polymerization
catalysis (TidNR acting as a “spectator ligand”),17,18 it
is the ability of the TidNR bond to act as a reactive
site toward unsaturated substrates that has attrac-
ted the most attention. The list of potential sub-
strates is now rather extensive and includes the follow-
There continues to be much interest in the chemistry
of terminal titanium imido compounds of the type
(L)nTidNR, where (L)n represents a supporting ligand
(set) and R is almost invariably an alkyl or aryl
group.1-12 An extensive reaction chemistry has now
* To whom correspondence should be addressed. E-mail: clot@
univ-montp2.fr (E.C.); philip.mountford@chem.ox.ac.uk (P.M.).
† University of Oxford.
‡ Universite´ Montpellier 2.
(1) Wigley, D. E. Prog. Inorg. Chem. 1994, 42, 239.
(2) Mountford, P. Chem. Commun. 1997, 2127.
(3) Gade, L. H.; Mountford, P. Coord. Chem. Rev. 2001, 216-217,
65.
(12) Straub, B. F.; Bergman, R. G. Angew. Chem., Int. Ed. 2001,
40, 4632.
(13) Winter, C. H.; Sheridan, P. H.; Lewkebandara, T. S.; Heeg, M.
J.; Proscia, J. W. J. Am. Chem. Soc. 1992, 114, 1095.
(14) Lewkebandara, T. S.; Sheridan, P. H.; Heeg, M. J.; Rheingold,
A. L.; Winter, C. H. Inorg. Chem. 1994, 33, 5879.
(15) McKarns, P. J.; Yap, G. P. A.; Rheingold, A. L.; Winter, C. H.
Inorg. Chem. 1996, 35, 5968.
(4) Mountford, P. In Perspectives in Organometallic Chemistry; Royal
Society of Chemistry: London, 2003.
(5) Pohlki, F.; Doye, S. Chem. Soc. Rev. 2003, 32, 104.
(6) Li, Y.; Shi, Y.; Odom, A. L. J. Am. Chem. Soc. 2004, 126, 1794.
(7) Basuli, F.; Bailey, B. C.; Huffman, J. C.; Mindiola, D. J. Chem.
Commun. 2003, 1554.
(16) Carmalt, C. J.; Newport, A.; Parkin, I. P.; Sealey, A. J.;
Dubberley, S. R. J. Mater. Chem. 2003, 13, 84.
(8) Hanna, T. E.; Keresztes, I.; Lobkovsky, E.; Bernskoetter, W. H.;
Chirik, P. J. Organometallics 2004, 23, 3448.
(9) Johnson, J. S.; Bergman, R. G. J. Am. Chem. Soc. 2001, 123,
2923.
(10) Bolton, P. D.; Mountford, P. Adv. Synth. Catal. 2005, 347, 355.
(11) Duncan, A. P.; Bergman, R. G. Chem. Rec. 2002, 2, 431.
(17) Male, N. A. H.; Skinner, M. E. G.; Bylikin, S. Y.; Wilson, P. J.;
Mountford, P.; Schro¨der, M. Inorg. Chem. 2000, 39, 5483.
(18) Adams, N.; Arts, H. J.; Bolton, P. D.; Cowell, D.; Dubberley, S.
R.; Friedrichs, N.; Grant, C. M.; Kranenburg, M.; Sealey, A. J.; Wang,
B.; Wilson, P. J.; Cowley, A. R.; Mountford, P.; Schro¨der, M. Chem.
Commun. 2004, 434.
10.1021/om049026f CCC: $30.25 © 2005 American Chemical Society
Publication on Web 04/02/2005

2348 Organometallics, Vol. 24, No. 10, 2005
Boyd et al.
Chart 1
ing: alkenes,19 alkynes,6,20-22 allenes,21 nitriles,23 iso-
nitriles,24,25 and phosphaalkynes,23,26-28 as well as iso-
cyanates, CO2, and other heterocumulenes.29-33 The
importance of titanium imido compounds as intermedi-
ates in hydroamination is also of note.5,6,34-39 The
reactions of TidNR groups with CO2 are of the most
relevance to our present contribution.
different supporting ligand sets) with CO2: [Ti(NR)-
(Me4taa)] (Me4taa ) dianion of tetramethyldibenzo-
tetraaza[14]annulene),29 [Ti(NR)(Me2calix)] (Me2calix )
the dianion of 1,3-dimethyl ether p-tert-butylcalix[4]-
arene),30 [Ti(NR)(O2NN′)] (H2O2NN′ ) (2-C5H4N)CH2N-
(2-HO-3,5-C6H2 Bu2)2),32 and [Ti(η-C5Me5)(NR){MeC-
t
(NiPr)2}].31 In all of these cases cycloaddition of CO2 to
the TidNR bond gives rise to a carbamate product of
the type [(L)nTi{OC(O)N(R)}], although these are not
always directly observed.32 Depending on the identity
of (L)n and R, the carbamates can in turn extrude the
corresponding isocyanate RNCO via a cycloreversion
process, forming either terminal (for (L)n ) Me4taa) or
dimeric (for (L)n ) Me2calix, O2NN′, (η-C5Me5){MeC-
(NiPr)2}) titanium oxo side products. A detailed under-
standing of such bond-forming/bond-breaking reactions
is important from the point of view of the controlled
functionalization of CO2 in particular40-46 and for the
design of useful group 4 imido complexes in general.
We have previously described the reactions of four
types of titanium imido compounds (each featuring very
(19) Bennett, J. L.; Wolczanski, P. T. J. Am. Chem. Soc. 1994, 116,
2179.
(20) Haak, E.; Bytsschkov, I.; Doye, S. Angew. Chem., Int. Ed. 1999,
38, 3389.
(21) Bashall, A.; Collier, P. E.; Gade, L. H.; McPartlin, M.; Mount-
ford, P.; Tro¨sch, D. J. M. Chem. Commun. 1998, 2555.
(22) Ward, B. D.; Maisse-Franc¸ois, A.; Mountford, P.; Gade, L. H.
Chem. Commun. 2004, 704.
(23) Pugh, S. M.; Tro¨sch, D. J. M.; Wilson, D. J.; Bashall, A.; Cloke,
F. G. N.; Gade, L. H.; Hitchcock, P. B.; McPartlin, M.; Nixon, J. F.;
Mountford, P. Organometallics 2000, 19, 3205.
(24) Blake, A. J.; Collier, P. E.; Gade, L. H.; McPartlin, M.;
Mountford, P.; Schubart, M.; Scowen, I. J. Chem. Commun. 1997, 1555.
(25) Bashall, A.; Collier, P. E.; Gade, L. H.; McPartlin, M.; Mount-
ford, P.; Pugh, S. M.; Radojevic, S.; Schubart, M.; Scowen, I. J.; Tro¨sch,
D. J. M. Organometallics 2000, 19, 4784.
The reaction of the cyclopentadienyl-amidinate sys-
tem [Ti(η-C5Me5)(NR){MeC(NiPr)2}]31 (A; Chart 1) with
CO2 is particularly interesting, as it shows distinct
(26) Cloke, F. G. N.; Hitchcock, P. B.; Nixon, J. F.; Wilson, D. J.;
Mountford, P. Chem. Commun. 1999, 661.
t
selectivity as a function of the imido NR (R ) Bu, 2,6-
(27) Cloke, F. G. N.; Hitchcock, P. B.; Nixon, J. F.; Wilson, D. J.;
Tabellion, F.; Fischbeck, U.; Preuss, F.; Regitz, M.; Nyulaszi, L. Chem.
Commun. 1999, 2363.
C6H3Me2) group. In the case when R ) tBu, CO2
cycloaddition (to form the N,O-bound carbamate com-
plex B) is followed by cycloreversion (extrusion of
tBuNCO) to yield the dimeric oxo compound C, whereas
the corresponding reaction for R ) 2,6-C6H3Me2 yields
exclusively the double CO2 activation product D, where
a second CO2 molecule inserts into the first-formed
carbamate group Ti-N bond. Motivated by the potential
of cyclopentadienyl-amidinate titanium imido com-
plexes to offer further variation or control of CO2
reactivity, we set out to develop pendant arm function-
(28) Asmus, S. M. F.; Regitz, M. Tetrahedron Lett. 2001, 42, 7543.
(29) Swallow, D.; McInnes, J. M.; Mountford, P. J. Chem. Soc.,
Dalton Trans. 1998, 2253.
(30) Dubberley, S. R.; Friedrich, A.; Willman, D. A.; Mountford, P.;
Radius, U. Chem. Eur. J. 2003, 9, 3634.
(31) Guiducci, A. E.; Cowley, A. R.; Skinner, M. E. G.; Mountford,
P. Dalton 2001, 1392.
(32) Boyd, C. L.; Toupance, T.; Tyrrell, B. R.; Ward, B. D.; Wilson,
C. R.; Cowley, A. R.; Mountford, P. Organometallics 2005, 24, 309.
(33) Ong, T.-G.; Yap, G. P.; Richeson, D. S. Chem. Commun. 2003,
2612.
(34) Nobis, M.; Drieâen-Ho¨lscher, B. Angew. Chem., Int. Ed. 2001,
40, 3983.
(35) Tillack, A.; Jiao, H.; Castro, I. G.; Hartung, C. G.; Beller, M.
Chem. Eur. J. 2004, 10, 2409.
(40) Darensbourg, D. J.; Kudaroski, R. A. Adv. Inorg. Chem. 1983,
22, 129.
(41) Palmer, D. A.; van Eldik, R. Chem. Rev. 1983, 83, 651.
(42) Arakawa, H. Chem. Rev. 2001, 101, 953.
(43) Torrent, M.; Sola, M.; Frenking, G. Chem. Rev. 2000, 100, 439.
(44) Leitner, W. Coord. Chem. Rev. 1996, 153, 257.
(45) Yin, X.; Moss, J. R. Coord. Chem. Rev. 1999, 181, 27.
(46) Gibson, D. Chem. Rev. 1996, 96, 2063.
(36) Ackermann, L.; Bergman, R. G.; Loy, R. N. J. Am. Chem. Soc.
2003, 125, 11956.
(37) Bytschkov, I.; Doye, S. Eur. J. Org. Chem. 2003, 935.
(38) McGrane, P. L.; Jensen, M.; Livinghouse, T. J. Am. Chem. Soc.
1992, 114, 5459.
(39) McGrane, P. L.; Livinghouse, T. J. Am. Chem. Soc. 1993, 115,
11485.

Benzamidinate Titanium Imido Compounds
Organometallics, Vol. 24, No. 10, 2005 2349
Scheme 1. Synthesis of New Pendant Arm Cyclopentadienyl-Amidinate Titanium Imido Compounds
alized amidinate systems in order to probe the effect(s)
of an additional hemilabile donor on the TidNR/CO2
bond-making and bond-breaking reactions. In this con-
tribution we employ two dimethylamino-functionalized
benzamidinate ligands, namely Me3SiNC(Ph)NCH2-
CH2NMe2 (two methylene groups between the benz-
amidinate nitrogen and the NMe2 group)47 and Me3SiNC-
(Ph)NCH2CH2CH2NMe2 (three carbon linker).48 The
two- and three-carbon linker systems allow us to
establish the importance of the length of the pendant
donor chain (and the size of the chelate ring that it
might form). For comparison, we have also employed
the approximately isosteric “dummy-arm” benzamidi-
nate ligand49,50 Me3SiNC(Ph)NCH2CH2Me (i.e., with no
pendant donor group) in order to distinguish between
steric and donor-group effects. We have reported re-
cently on non-cyclopentadienyl titanium imido com-
pounds containing these ligands.49,51
{Me3SiNC(Ph)NCH2CH2CH2NMe2}] (2) in 83% and 89%
yields, respectively (Scheme 1). The
1H and 13C{1H}
NMR spectra for 1 and 2 support the proposed struc-
tures. Without coordination of the pendant NMe2 group,
1 and 2 are formally 16-valence-electron complexes.
Coordination of NMe2 would form an 18-valence-
electron compound and also render the methyl groups
1
of the NMe2 moiety chemically inequivalent. H NMR
spectra for both compounds at room temperature feature
a singlet (6 H) for the NMe2 protons, indicating that
either the pendant arm is not coordinated to the metal
center or that methyl group exchange is fast on the
NMR time scale at this temperature. Cooling NMR
samples of 1 and 2 to -90 °C in CD2Cl2 (500.1 MHz 1H
resonance frequency) revealed some broadening of the
NMe2 resonance but did not result in splitting into two
separate resonances. Therefore, the pendant NMe2
groups in 1 and 2 probably coordinate very weakly to
the metal but even at low temperatures easily dissoci-
ate. This was confirmed by the observation of only one
resonance for the NMe2 methyls in the 13C{1H} NMR
spectra at -95 °C.
Results and Discussion
Synthesis of Cyclopentadienyl Pendant Arm
Functionalized Amidinate tert-Butylimido Com-
pounds. The starting compounds [Ti(η-C5Me5)(NtBu)-
Cl(py)] and [Ti(η-C5H4Me)(NtBu)Cl(py)] were prepared
as previously described.52 Reaction of [Ti(η-C5Me5)-
(NtBu)Cl(py)] with Li[Me3SiNC(Ph)NCH2CH2NMe2] or
Li[Me3SiNC(Ph)NCH2CH2CH2NMe2] in C6H6 at 80 °C
yielded the new compounds [Ti(η-C5Me5)(NtBu){Me3-
SiNC(Ph)NCH2CH2NMe2}] (1) and [Ti(η-C5Me5)(NtBu)-
Corresponding reactions of [Ti(η-C5H4Me)(NtBu)-
Cl(py)] with the lithiated pendant amine functionalized
amidinates or with the “dummy arm” amidinate [Li2{Me3-
SiNC(Ph)NCH2CH2Me}2(Et2O)2] in C6H6 afforded the
compounds [Ti(η-C5H4Me)(NtBu){Me3SiNC(Ph)NCH2-
CH2NMe2}] (3), [Ti(η-C5H4Me)(NtBu){Me3SiNC(Ph)-
NCH2CH2CH2NMe2}] (4), and [Ti(η-C5H4Me)(NtBu)-
{Me3SiNC(Ph)NCH2CH2Me}] (5) in good crude yields
(Scheme 1). These three compounds are oils at room
temperature and could only be purified by high-vacuum
distillation (1 × 10-5 mbar, 120-130 °C). The room-
(47) Brandsma, M. J. R.; Brussee, E. A. C.; Meetsma, A.; Hessen,
B.; Teuben, J. H. Eur. J. Inorg. Chem. 1998, 1867.
(48) Doyle, D.; Gun’ko, Y. K.; Hitchcock, P. B.; Lappert, M. F. Dalton
2000, 4093.
1
(49) Boyd, C. L.; Guiducci, A. E.; Dubberley, S. R.; Tyrrell, B. R.;
Mountford, P. Dalton 2002, 4175.
temperature H NMR spectra of compounds 3-5 are
analogous to those of 1 and 2. However, cooling samples
of 3 and 4 in CD2Cl2 revealed decoalescence points for
the NMe2 singlets at 237 ( 2 and 208 ( 2 K, respec-
tively, below which two distinct NMe resonances were
observed. This indicates that the NMe2 groups in 3 and
(50) Boyd, C. L.; Tyrrell, B. R.; Mountford, P. Acta Crystallogr. 2002,
E58, m597.
(51) Croft, A. C. R.; Boyd, C. L.; Cowley, A. R.; Mountford, P. J.
Organomet. Chem. 2003, 683, 120.
(52) Dunn, S. C.; Mountford, P.; Robson, D. A. J. Chem. Soc., Dalton
Trans. 1997, 293.

2350 Organometallics, Vol. 24, No. 10, 2005
Boyd et al.
q
1
Table 1. Values of ∆G (from H Coalescence
4 coordinate to titanium in the ground state, giving two
chemically distinct Me group environments (eq 1), and
Tc
q
Measurements) and Activation Parameters ∆H
q
1
and ∆S (from Variable-Temperature H Linewidth
Analysis) for NMe2 Methyl Group Exchange for
[Ti(η-C5H4Me)(NtBu){Me3SiNC(Ph)NCH2CH2NMe2}]
(3) and [Ti(η-C5H4Me)(NtBu)-
{Me3SiNC(Ph)NCH2CH2CH2NMe2}] (4)
∆Gq
∆Hq
∆Sq
Tc
compd
Tc (K)
(kJ mol-1
)
(kJ mol-1
)
(J mol-1 K-1
)
3
4
237 ( 2
208 ( 2
46.4 ( 0.4
41.6 ( 0.4
81.2 ( 7.1
95 ( 27
150 ( 30
267 ( 133
enough to permit inversion at the nitrogen, and so it is
reasonable to assume that the Me exchange involves the
16-electron isomers 3′ and 4′. Such process are expected
to have positive ∆Sq values.54 ∆Gq values obtained at
Tc
different Tc’s cannot be directly compared because of the
different Tc∆Sq contributions, but since Tc (208 ( 2 K)
for the three-carbon linker compound 4 (∆Gq ) 41.6
208
( 0.4 kJ mol-1) is ca. 30 K lower than that for the two-
carbon linker compound 3 (∆Gq
) 46.4 ( 0.4 kJ
237
mol-1), the ∆Gq value for 4 would be even lower than
Tc
the value given in Table 1 at the Tc for 3. Overall, it
seems very likely that ∆Gq for NMe2 dissociation will
be significantly lower at any particular temperature for
4 than for 3.
is confirmed by the presence of two NMe resonances in
the low-temperature 13C{1H} NMR spectra of these
compounds. For 4 an NOE was observed at -90 °C
t
between the NMe resonances and the Bu group of the
To gain an additional insight into the activation
parameters, we attempted to obtain ∆Hq and ∆Sq (and
hence ∆Gq) values from 1H NMR (500.1 MHz) NMe2 line
width measurements at temperatures below the coa-
lescence temperature for both compounds. Activation
parameters for NMe group exchange in 3 and 4 were
imido ligand, confirming coordination of the NMe2
group. Coordination of the pendant arm to titanium may
be more favorable for the C5H4Me systems than for the
C5Me5 systems 1 and 2, since C5Me5 is both more
sterically demanding and electron releasing.
1
Although it is likely that the NMe2 methyls (arbi-
trarily labeled “Mea” and “Meb”) of 3 and 4 undergo
mutual chemical exchange via the 16-valence-electron
isomers 3′ and 4′, no evidence for significant equilibrium
amounts of either of these compounds was indicated by
any of the low-temperature spectra. Therefore, assum-
obtained by H NMR (500.1 MHz) line width analysis
at 3 K intervals at temperatures between 215 and 233
K (for 3) and between 186 and 201 K (for 4) in CD2Cl2.
In each case, the line widths at half-height (ν1/2(obs))
were corrected by subtracting the natural line widths
obtained from the low-temperature-limit 1H NMR spec-
trum at 183 K. These corrected values (ν1/2(corr)) were
used to calculate the observed rate constants for the
processes Mea f Meb {kobs(afb)} and Meb f Mea
{kobs(bfa)} according to kobs ) πν1/2(corr).53,55 The rate
constants were used to construct Eyring plots (see the
Supporting Information), assuming that the rate con-
stant k1 (eq 1) is equal to 2kobs, since the conversion of
the methyl group environment “a” to “b” occurs at the
same rate as the conversion from environment “b” to
“a”. Values for ∆Hq and ∆Sq were extracted and are
listed in Table 1.
It is clear from Table 1 that there are rather large
errors on the enthalpic and entropic activation data for
both systems. This arises unavoidably from the difficulty
in obtaining good-quality spectra at the very low tem-
peratures required. The errors are particularly large for
the three-carbon linker system 4. Although one there-
fore cannot make detailed comparisons of the ∆Hq and
∆Sq values between the two compounds 3 and 4, it is
indisputable that the NMe2 methyl group exchange
processes are associated with significantly positive ∆Sq
values. Likewise, the ∆Hq data are favorably comparable
1
ing that H NMR could detect a minimum of ca. 1% of
3′ (or 4′) in the presence of a 99% relative amount of 3
(or 4) (i.e. Keq ) [3′]/[3] or [4′]/[4] e 0.01), then k-1 (eq
1) must be at least 2 orders of magnitude greater than
k1 for both the 3/3′and 4/4′ dynamic equilibria (since Keq
) k1/k-1). Furthermore, a Keq value of e0.01 would
imply that ∆Grxn for the process 3 f 3′ and 4 f 4′ must
be ca. g10 kJ mol-1 at temperatures below -40 °C.
Activation Parameters for NMe2 Methyl Group
Exchange in 3 and 4. The values of ∆Gq (Gibbs free
Tc
energy of activation at the coalescence temperature Tc)
from coalescence measurements of the NMe2 resonances
1
of 3 and 4 in the H NMR spectra (at 237 ( 2 and 208
( 2 K, respectively) were extracted using standard
procedures53 and are listed in Table 1. Strictly speaking,
the ∆Gq values refer to the highest barrier to overcome
Tc
on the Gibbs free energy surface to effectively exchange
Mea and Meb. This process could be described as three
consecutive steps: NMe2 dissociation, inversion at
NMe2, and rotation about the CH2-NMe2 σ bond.
However, it is very likely that all processes are coupled
and the actual TS for exchange would be best described
as a concomitant dissociation of NMe2 and reorganiza-
tion of the pendant arm to achieve Me exchange.
However, the NMe2 group must dissociate sufficiently
(54) For a general account of thermodynamic aspects of chelate ring
opening processes, see: Hancock, R. D.; Martell, A. E. Chem. Rev. 1989,
89, 1875.
(55) Green, M. L. H.; Wong, L.-L.; Sella, A. Organometallics 1992,
11, 2660.
(53) Sandstro¨m, J. Dynamic NMR Spectroscopy; Academic Press:
London, 1982.

Benzamidinate Titanium Imido Compounds
Organometallics, Vol. 24, No. 10, 2005 2351
Scheme 2. Reactions of [Ti(η-C5Me5)(NtBu){Me3SiNC(Ph)N(CH2)nNMe2}] (n ) 2 (1), 3 (2)) with CO2
to those for related dissociative equilibria of amidinate-
supported titanium imido complexes.56
NCH2CH2NMe2}] (6) and [Ti(η-C5Me5){NtBuC(O)O}-
{Me3SiNC(Ph)NCH2CH2CH2NMe2}] (7) in 76% and 82%
isolated yields, respectively (Scheme 2). The carbamate
As a check on the internal consistency between the
∆Gq values for 4 obtained from coalescence and line
width analyses, it is reassuring that the ∆Gq values for
4 at 237 K (Tc) extracted from coalescence (46.4 ( 0.4
complexes 6 and 7 give rise to the expected 1H and 13
C
spectra, and their IR spectra contain bands at 1686 and
1666 cm-1, respectively, which are attributed to ν(CdO)
of the carbamate ligand.31,57 The single-crystal X-ray
structure of 7 is discussed below.
kJ mol-1) and the Eyring plots (45.6 ( 14.1 kJ mol-1
)
are in good agreement, despite the large errors on the
activation parameters extracted from the line width
data.
The cycloaddition products 6 and 7 are unstable in
solution and decompose over ca. 1 h with elimination
To make a comparison between ∆Gq for both systems
at the same temperature, the more precise activation
parameters extracted for the two-carbon linker system
of BuNCO, as identified by H NMR spectroscopy, to
form the oxo compounds 8 and 9. The absence of
terminal TidO stretches in the range 850-930 cm-1 in
the IR spectra of these products suggests that they are
dimeric µ-oxo species, namely [Ti2(η-C5Me5)2(µ-O)2{Me3-
SiNC(Ph)NCH2CH2NMe2}2] (8) and [Ti2(η-C5Me5)2(µ-
O)2{Me3SiNC(Ph)NCH2CH2CH2NMe2}2] (9). This is sup-
ported by the dimeric structures found by X-ray
crystallography for the methylcyclopentadienyl µ-oxo
complexes 10a and 11a (vide infra). The dimeric non
pendant arm acetamidinate analogue [Ti2(η-C5Me5)2(µ-
O)2{MeC(NiPr)2}2] (C; Chart 1) has also been crystal-
lographically characterized.31 The oxo compounds 8 and
9 are highly insoluble, and neither NMR data nor mass
spectral data could be obtained.
t
1
3 have been used to calculate ∆Gq
(208 K being Tc
208
for the three-carbon linker system 4), which can then
be compared to ∆Gq208 for 4 calculated directly from the
coalescence measurement. The values thus obtained are
∆Gq ) 50.0 ( 13.2 kJ mol-1 for 3 and 41.6 ( 0.4 kJ
208
mol-1 for 4. These data are consistent with there being
a smaller ∆Gq
value for dissociation of the three-
208
carbon linker arm in 4 than for the two-carbon linker
arm in 3. While the data in Table 1 do not permit us to
determine the origin of this effect with any certainty, it
seems likely that it might be the result of a greater gain
in entropy upon dissociation of a three-carbon arm from
the metal.54
Reactions of C5Me5 Compounds 1 and 2 with
CO2. Reaction of the compounds [Ti(η-C5Me5)(NtBu)-
{Me3SiNC(Ph)NCH2CH2NMe2}] (1) and [Ti(η-C5Me5)-
(NtBu){Me3SiNC(Ph)NCH2CH2CH2NMe2}] (2) with CO2
(1 atm) at room temperature for 5 min, followed by
immediate isolation, yielded the cherry red N,O-bound
carbamates [Ti(η-C5Me5){NtBuC(O)O}{Me3SiNC(Ph)-
The proposed mechanism for this process involves a
retrocyclization (Scheme 2) to form initially a terminal
oxo complex (possibly with NMe2 coordination). No
evidence was observed for this species in the NMR-scale
reactions, suggesting that it immediately undergoes
back-reaction with tBuNCO (re-forming 6 or 7) or
eventually self-traps irreversibly by dimerization to give
8 or 9. This is in contrast to the reactions of CO2 with
(56) Stewart, P. J.; Blake, A. J.; Mountford, P. Inorg. Chem. 1997,
36, 3616.
(57) Blake, A. J.; McInnes, J. M.; Mountford, P.; Nikonov, G. I.;
Swallow, D.; Watkin, D. J. J. Chem. Soc., Dalton Trans. 1998, 379.

2352 Organometallics, Vol. 24, No. 10, 2005
Boyd et al.
1
observation of only one isomer in the H NMR spec-
trum. The distances and angles associated with the
Ti{OC(O)N} moiety are comparable to those in the
dibenzotetraaza[14]annulene-supported
carbamate
[Ti{N(4-C6H4Me)C(O)O}(Me4taa)].29
Reactions of C5H4Me Compounds 3-5 with CO2
at Room Temperature. The behavior of the methyl-
cyclopentadienyl systems 3-5 was found to differ
significantly from that of the C5Me5 systems 1 and 2.
Upon exposure to CO2 at room temperature a color
change to cherry red was observed straight away but
the color darkened to brown almost immediately, indi-
cating decomposition to the corresponding oxo com-
pound. When the reaction was followed at room tem-
perature by 1H NMR spectroscopy, spectra after 5 min
contained resonances attributable to both the imido
starting materials and the dimeric oxo products [Ti2(η-
C5H4Me)2(µ-O)2{Me3SiNC(Ph)NR}2] (R ) CH2CH2NMe2
(10), CH2CH2CH2NMe2 (11), CH2CH2Me (12)), together
Figure 1. Displacement ellipsoid plot of [Ti(η-C5Me5)-
{NtBuC(O)O}{Me3SiNC(Ph)NCH2CH2CH2NMe2}] (7) (25%
probability). H atoms are omitted for clarity.
t
with BuNCO, but not the expected carbamate inter-
mediates 13-15. This is in contrast to the reactions
(Scheme 2) of the C5Me5 systems 1 and 2; these display
only resonances due to the intermediate carbamate
under the same reaction time and conditions. The
immediate decomposition of the proposed carbamate
intermediates 13-15 means that characterization is not
possible at room temperature. The faster rate of their
decomposition in comparison to that of the C5Me5
compounds 1 and 2 is attributed to the decreased steric
stabilization toward dimerization of a transient mono-
meric oxo intermediate provided by C5H4Me compared
to C5Me5. Scheme 3 summarizes the overall sequence
for the reaction of 3-5 with CO2.
The dimeric oxo products 10-12 have been isolated
and characterized by NMR and IR spectroscopy and
mass spectrometry. The NMR spectra were found to
contain twice as many resonances as expected for single
products and probably imply the formation of both trans
and cis isomers (10a-12a and 10b-12b; Scheme 3). It
was found that heating samples to 80 °C in C6D6 for
several hours led to a shift in the relative ratios of the
two products in favor of the trans isomer. Single crystals
of the trans isomers 10a and 11a were grown from
solutions in pentane at room temperature. The molec-
ular structures are shown in Figure 2, and selected bond
lengths and angles are listed in Tables 3 and 4.
The structures of 10a and 11a confirm that they are
µ-oxo dimers and that the NMe2 groups do not coordi-
nate to the metal. Both compounds have a pseudo four-
legged piano-stool geometry about the titanium atoms.
Compounds with Ti2(µ-O)2 cores are well established,
and several have been crystallographically character-
ized.59,60 The bond lengths and angles are of comparable
magnitudes for the two compounds. Both compounds
have slightly asymmetric Ti2(µ-O)2 cores in which the
Ti-O bond trans to the SiMe3-bearing nitrogen is
slightly longer than the Ti-O bond trans to the pendant
arm bearing nitrogen. This may be attributed to the
asymmetric amidinate ligand substitution, as the sym-
metric oxo dimer [Ti2(η-C5H4Me)2(µ-O)2{PhC(NSiMe3)2}2]
Table 2. Selected Bond Lengths (Å) and Angles
(deg) for [Ti(η-C5Me5){NtBuC(O)O}-
{Me3SiNC(Ph)NCH2CH2CH2NMe2}] (7)a
Ti(1)-Cpcent
Ti(1)-N(1)
Ti(1)-N(2)
2.058
2.087(5)
2.131(4)
Ti(1)-N(4)
Ti(1)-O(1)
1.991(5)
1.971(4)
Cpcent-Ti(1)-N(1)
Cpcent-Ti(1)-N(2)
119.3
108.7
Cpcent-Ti(1)-N(4)
Cpcent-Ti(1)-O(1)
129.3
109.0
a Cpcent refers to the computed C5Me5 ring centroid.
the dibenzotetraaza[14]annulene-supported titanium
imides [Ti(NR)(Me4taa)].29 In these systems, extrusion
of RNCO (R ) tBu, aryl) from the intermediate carbam-
ates yields stable terminal titanium oxo compounds.
Decomposition of 6 and 7 via tBuNCO extrusion to form
oxo products 8 and 9 is much faster for these pendant
arm functionalized systems (ca. 1 h) than for the
previously reported non pendant arm acetamidinate
system [Ti(η-C5Me5){NtBuC(O)O}{MeCN(iPr)2}], which
takes ca. 12 h to completely decompose to the corre-
sponding dimeric oxo product. This possibly suggests
that the pendant donor group may play a role in the
extrusion step (an anchimeric effect58). Alternatively,
the difference in steric demands of the N substituents
of the two different amidinate ligands may be respon-
sible.
The molecular structure of compound 7 is shown in
Figure 1, and selected bond distances and angles are
presented in Table 2. The pendant arm is clearly not
coordinated, and the compound has a pseudo four-legged
piano-stool structure that is slightly distorted, as re-
vealed by the Cpcent-Ti-N and -O angles (Table 2). The
Cpcent-Ti(1)-N(1) angle is 119.3°, as compared to a
Cpcent-Ti(1)-N(2) angle of 108.7°. Similarly, the Cpcent
-
Ti(1)-N(4) angle is 129.3°, while the Cpcent-Ti(1)-O(1)
angle is 109.0°. This is probably due to unfavorable
steric interactions between the C5Me5 ring and the
bulky tBu and SiMe3 groups. Unfavorable steric interac-
tions between these groups also explain the formation
t
of the observed isomer, in which the Bu and SiMe3
(59) Allen, F. H.; Kennard, O. Chem. Des. Automation News 1993,
8, 1&31.
(60) Fletcher, D. A.; McMeeking, R. F.; Parkin, D. J. J. Chem. Inf.
Comput. Sci. 1996, 36, 746.
groups are trans to each other, as confirmed by the
(58) Braunstein, P.; Naud, F. Angew. Chem., Int. Ed. 2001, 40, 680.

Benzamidinate Titanium Imido Compounds
Organometallics, Vol. 24, No. 10, 2005 2353
Scheme 3. Reactions of [Ti(η-C5H4Me)(NtBu){Me3SiNC(Ph)N(CH2)nNMe2}] (n ) 2 (3), 3 (4)) and
[Ti(η-C5H4Me)(NtBu){Me3SiNC(Ph)NCH2CH2Me}] (5) with CO2
(C; Chart 1) has equal bond lengths for both Ti-O and
Ti-Namidinate bonds.31
10. This extrusion occurred at a rate which appeared
to be qualitatively about twice as fast as for 13 and 14
at the same temperature. Note that the -78 °C NMR
spectra of 13 and 15 showed no evidence of NMe2 group
coordination and in this regard these compounds are
analogous to the C5Me5 homologues 6 and 7.
These observations suggested that the NMe2 donor
in the pendant arm systems could influence both the
rate of cycloaddition of CO2 to the TidNtBu bond of 3
and the extrusion of tBuNCO from the carbamate
intermediates. Low-temperature kinetic studies were
therefore carried out to probe the role of the pendant
donor group during the various stages of the reactions
of 3-5 with CO2.
Cycloaddition Reaction of 3 with CO2: Kinetic
and Mechanistic Studies. Low-temperature control
of the cycloaddition of CO2 to the TidNtBu bond of 3
makes this reaction convenient for determining the rate
law dependence of CO2 in the reaction to form the
carbamate 15. These studies were carried out under
pseudo-first-order conditions, whereby at least a 10-fold
excess of CO2 was present at the lowest pressure studied
(linear ln(I/I0) plots (e.g. Figure 3b and the Supporting
Information) showed no evidence for significant diffu-
sion control limitations in these experiments). NMR
Reactions of C5H4Me Compounds 3-5 with
CO2: Low-Temperature Studies. Reactions of 3-5
with CO2 at -78 °C revealed a difference in behavior
between the two-carbon and the three-carbon pendant
arm systems 3 and 4. When an NMR tube sample of 4
in CD2Cl2 at -78 °C was exposed to CO2 (ca. 1 atm),
1
the H NMR spectrum (-78 °C) recorded immediately
afterward contained only resonances attributable to the
carbamate [Ti(η-C5H4Me){NtBuC(O)O}{Me3SiNC(Ph)-
NCH2CH2CH2NMe2}] (13; Scheme 3). Subsequent warm-
ing of this solution to -25 °C led to extrusion of tBuNCO
and the trans and cis dimeric oxo compounds 11. This
behavior was mirrored by the “dummy-arm” compound
5, which gave carbamate 14 immediately at -78 °C and
then extrusion of isocyanate at -25 °C to form the oxo
compounds 12. However, when the two-carbon pendant
arm compound 3 was exposed to CO2 at -78 °C, the 1H
NMR spectrum showed only resonances for 3. Cyclo-
addition to form the carbamate [Ti(η-C5H4Me){NtBuC-
(O)O}{Me3SiNC(Ph)NCH2CH2NMe2}] (15) did not occur
at a significant rate until the solution was warmed to
-35 °C. Further warming to -25 °C resulted in iso-
cyanate extrusion to yield the dimeric oxo compounds

2354 Organometallics, Vol. 24, No. 10, 2005
Boyd et al.
Figure 2. Displacement ellipsoid plots of (a, left) one of the crystallographically independent molecules of trans-[Ti2(η-
C5H4Me)2(µ-O)2{Me3SiNC(Ph)NCH2CH2NMe2}2] (10a) (20% probability) and (b, right) trans-[Ti2(η-C5H4Me)2(µ-O)2{Me3-
SiNC(Ph)NCH2CH2CH2NMe2}2] (11a) (25% probability). H atoms are omitted for clarity. For 11a, atoms carrying the
suffix “A” are related to their counterparts by the symmetry operator -x, -y + 1, -z + 1.
Table 3. Selected Bond Lengths (Å) and Angles
(deg) for trans-[Ti2(η-C5H4Me)2(µ-O)2-
{Me3SiNC(Ph)NCH2CH2NMe2}2] (10a)a
Ti(1)-N(1)
Ti(1)-N(2)
Ti(1)-Cpcent(1)
Ti(1)-O(1)
Ti(1)-O(2)
2.180(6)
2.123(6)
2.081
1.813(5)
1.892(5)
Ti(2)-N(4)
Ti(2)-N(5)
Ti(2)-Cpcent(2)
Ti(2)-O(1)
Ti(2)-O(2)
2.202(6)
2.116(7)
2.08
1.897(5)
1.813(5)
Cpcent(1)-Ti(1)-N(1)
Cpcent(1)-Ti(1)-N(2)
Cpcent(1)-Ti(1)-O(1)
Cpcent(1)-Ti(1)-O(2)
108.5
110.9
121.6
112.2
Cpcent(2)-Ti(1)-N(4)
Cpcent(2)-Ti(1)-N(5)
Cpcent(2)-Ti(1)-O(1)
Cpcent(2)-Ti(1)-O(2)
107.9
111.9
111.7
122.2
a Cpcent(1) and Cpcent(2) refer to the computed C5H4Me ring
centroids for Ti(1) and Ti(2), respectively.
Table 4. Selected Bond Lengths (Å) and Angles
(deg) for trans-[Ti2(η-C5H4Me)2(µ-O)2-
{Me3SiNC(Ph)NCH2CH2NMe2}2] (11a)a
Ti(1)-N(1)
Ti(1)-N(2)
Ti(1)-Cpcent
2.203(1)
2.096(1)
2.086
Ti(1)-O(1)
Ti(1)-O(1A)
1.802(1)
1.907(1)
Cpcent-Ti(1)-N(1)
Cpcent-Ti(1)-N(2)
109.6
115.6
Cpcent-Ti(1)-O(1)
Cpcent-Ti(1)-O(1A)
121.7
111.3
a Cpcent refers to the computed C5H4Me ring centroid. Atoms
carrying the suffix “A” are related to their counterparts by the
symmetry operator -x, -y + 1, -z + 1.
samples of 3 in CD2Cl2 were exposed to known pressures
of CO2 at -78 °C. Samples were transferred to a
spectrometer with a probe maintained at the same
temperature. The sample was warmed to -35 °C, at
which temperature cycloaddition occurred at a signifi-
cant rate and NMR spectra were recorded over at least
3 half-lives. The reaction was carried out at four CO2
pressures in the range 0.25-1.0 atm. The relative
concentrations of the imido starting material 3 and the
carbamate 15 were measured from each spectrum, and
the results were fitted to a first-order exponential
relationship. The observed rate constant kobs was ex-
tracted from a plot of ln(I/I0) vs time for 3 (parts a and
b of Figure 3) The kobs data are summarized in Table 5,
Figure 3. (a, top) Plot of normalized concentrations vs
time for the cycloaddition reaction of 3 with CO2 (0.5 atm)
Values given are for the starting imide [Ti(η-C5H4Me)-
(NtBu){Me3SiNC(Ph)NCH2CH2NMe2}] (3) and product car-
bamate [Ti(η-C5H4Me){OC(O)NtBu}{Me3SiNC(Ph)NCH2-
CH2NMe2}] (15). (b, bottom) Plot of ln(I/I0) vs time for the
reaction of [Ti(η-C5H4Me)(NtBu){Me3SiNC(Ph)NCH2CH2-
NMe2}] (3) with CO2 (0.5 atm).
and further details are given in the Supporting Infor-
mation. The exponential decay of the starting material
and evolution of carbamate indicate that the rate-
determining step is a bimolecular reaction which is first

Benzamidinate Titanium Imido Compounds
Organometallics, Vol. 24, No. 10, 2005 2355
Table 5. Observed Rate Constants kobs for
Cycloaddition of CO2 to
[Ti(η-C5H4Me)(NtBu){Me3SiNC(Ph)NCH2CH2NMe2}]
(3) at Different CO2 Pressures
p(CO2)
(atm)
kobs × 104
p(CO2)
(atm)
kobs × 104
(s-1
)
(s-1
)
0.25
0.50
0.74 ( 0.02
1.6 ( 0.03
0.75
1.0
2.7 ( 0.03
3.7 ( 0.04
order in CO2. This is confirmed by the linear dependence
of kobs on CO2 pressure (Figure 4, Table 5).
Scheme 4 summarizes the two likely mechanisms for
addition of CO2 to the pendant arm compounds 3 and
4. In mechanism 1, cycloaddition of CO2 to the imide
takes place without decoordination of the pendant arm
from the metal. Mechanism 2 involves a preequilibrium
in which the pendant arm dissociates from the metal
to generate an intermediate species (3′ or 4′), which then
undergoes cycloaddition of CO2. Note that for the
“dummy arm” imide 5 a simple bimolecular process
analogous to the second step of mechanism 2 is the only
possibility.
Figure 4. Plot of kobs vs CO2 pressure for the reaction of
[Ti(η-C5H4Me)(NtBu){Me3SiNC(Ph)NCH2CH2NMe2}] (3)
with CO2 at -35 °C.
carbamates (vide supra). Additional studies were made
on the “dummy-arm” system (14 f 12) at three different
concentrations, to probe the dependence of the rate of
decomposition upon carbamate concentration (since
compounds 13-15 are monomeric and the ultimate
products 10-12 are dimeric, it is possible that the rate-
determining step could be bimolecular with respect to
titanium). The experiments for 13 and 15 were carried
out in duplicate and were highly reproducible. NMR
samples of in situ generated 13-15 were warmed to -30
°C, at which point all of the imido starting materials
had been consumed by reaction with CO2 and only
carbamate was present.61 Extrusion of tBuNCO from the
carbamates 13-15 to yield 10-12 was followed over 3
half-lives of carbamate. The data were fitted to a first-
order exponential relationship. The rate constants k
extracted from plots of ln(I/I0) vs. time (Figure 5 shows
data for 13 by way of example) are listed in Table 6.
The data (Table 6) for extrusion from 14 show that
the reaction is independent of [14] in the range studied;
therefore, a bimolecular rate-limiting step is unlikely.62
Examination of the rate constants for 13-15 reveals a
small but persistent increase in the rate constant for
extrusion with the two-carbon arm system 15 compared
with 13 and 14. Energetically the effect is small (2-3
kJ mol-1 difference in ∆Gq for extrusion), but none-
theless it amounts to a quantitatively established halv-
ing of the half-life (t1/2, Table 6) for extrusion on going
from the three-carbon arm system 13 to the two-carbon
arm system 15. This effect might arise from slight
differences in steric effects between the two ligands or
from a small anchimeric effect of the ligand arm. The
compounds 13 (three-carbon arm) and 14 (“dummy”
arm) have experimentally indistinguishable values for
k, suggesting that the three-carbon arm NMe2 does not
For mechanism 1, the rate law is given by eq 2 (where
“imide” ) 3 or 4).
rate ) k[imide][CO2]
(2)
Under pseudo-first-order conditions (large and ef-
fectively constant [CO2] as in the experiments described
above), this expression approximates to eq 3, where kobs
) k[CO2] and is linearly dependent upon [CO2].
rate ) kobs[imide]
(3)
The rate law for mechanism 2 may be derived by
applying the steady-state approximation to the inter-
mediates 3′ and 4′ such that the rate law in eq 4 applies
(k1, k-1, and k2 as defined in eq 1 and Scheme 4).
k1k2[imide][CO2]
rate )
(4)
k-1 + k2[CO2]
However, in the case where k-1 . k2[CO2], eq 4
approximates to eq 5, where k′obs ) (k1k2/k-1)[CO2] and
again has a linear dependence on [CO2].
rate ) k′obs[imide]
(5)
Therefore, the experimental data for the reaction of
3 with CO2 fit either mechanism 1 (eq 3) or mechanism
2 (eq 5), provided that k-1 is larger than k2[CO2]. This
issue is further addressed using DFT calculations later
in this paper.
t
accelerate measurably the rate of BuNCO extrusion
from the carbamate. Attempts to address this issue via
DFT studies are described in the following section.
Density Functional Theory Computational Stud-
ies. Transition-metal imido compounds have been the
subject of several computational studies, as this ligand
is involved in various catalytic transformations.63 They
Extrusion of tBuNCO: Mechanistic and Kinetic
Studies. Given the very different rates of cycloaddition
of CO2 to TidNtBu for the two-carbon and three-carbon
arm NMe2 donor systems 3 and 4, it was of interest to
quantify any anchimeric effect of the different arms on
t
the rate of BuNCO extrusion.
(61) In the case of the two-carbon linker carbamate 15, a small
amount of imido starting material 3 was still present at -25 °C and
data obtained until all of this had been consumed were discarded.
(62) It has recently been shown that the transient terminal titanium
oxo compound [Ti(η-C5Me5)(O){MeC(NiPr)2}] generated by tBuNCO
extrusion from the unstable carbamate B (Chart 1) can be trapped
using p-tolyl isocyanate, forming a double-substrate activation product
somewhat analogous to that in: Boyd, C. L. D. Phil. Thesis, University
of Oxford, 2004.
Extrusion of tBuNCO from all three carbamates 13-
15 to form the bridging oxo products (Scheme 3) was
monitored by low-temperature NMR at the same con-
centrations. The isosteric “dummy arm” system 14 was
used as a control to benchmark any effect of the pendant
(CH2)nNMe2 donors upon the rate of extrusion. All of
the NMR studies were carried out on in situ generated
(63) Cundari, T. R., Chem. Rev., 2000, 100, 807.

2356 Organometallics, Vol. 24, No. 10, 2005
Boyd et al.
Scheme 4. Two Alternative Mechanisms for the Formation of Carbamates 15 and 13 from the Starting
Imido Complexes 3 or 4 on Reaction with CO2: (Mechanism 1) Cycloaddition at the Complex with the
NMe2 Group Coordinated; (Mechanism 2): Dissociation of NMe2 Precedes Cycloaddition
can act as spectator ligands in olefin polymerization
catalysts.64 However, the TidNR linkage may also
promote C-H and H-H bond cleavage. Carbon-
hydrogen bond activation by titanium imido complexes
has been studied computationally by Cundari et al.65
Chirik et al. have compared the C-H and H-H bond
activation promoted by Ti-alkylidene and Ti-imido
complexes and have related the absence of reactivity
toward C-H of the latter to the respective shape of the
frontier orbitals.8 Bergman and Straub have used
density functional theory calculations to elucidate the
mechanism of hydroamination of allenes, alkynes, and
alkenes catalyzed by cyclopentadienyl titanium imido
complexes.12 Mountford and co-workers have used den-
sity functional theory in combination with photoelectron
spectroscopy to elucidate the bonding in “pogo stick”
complexes of the type [Ti(η-C8H8)(NR)].66 In the present
work, density functional theory (DFT) calculations on
model systems were performed using hybrid DFT
(B3PW91).67,68 The main objective of the theoretical
study is to elucidate the potential role played by the
pendant arm on the reactivity with CO2 of the amidinate
titanium imido complexes.
The reaction of the model compound [Ti(η-C5H5)-
(NMe){MeC(NMe)2}] (I) with CO2 was used as a refer-
ence. The global pathway for the overall reaction as
summarized in eq 6 is given in Figure 6. The final
dimeric µ-oxo product [Ti2(η-C5H5)2(µ-O)2{MeC(NMe)2}2]
was not calculated by DFT, since this would be expen-
sive in terms of CPU time. Furthermore, the dimeriza-
tion of the terminal oxo products (as modeled by [Ti(η-
C5H5)(O){MeC(NMe)2}] (IV)) is not the main point of
interest from an experimental or mechanistic point of
view.
(64) Jensen, V. R.; Børve, K. J. Chem. Commun. 2002, 542.
(65) Cundari, T. R.; Klinckman, T. R.; Wolczanski, P. T. J. Am.
Chem. Soc. 2002, 124, 1481.
(66) Blake, A. J.; Cowley, A. R.; Dunn, S. C.; Green, J. C.; Hazari,
N.; Jones, N. M.; Moody, A. G.; Mountford, P. Chem. Eur. J., in press.
(67) Perdew, J. P.; Wang, Y. Phys. Rev. B 1992, 45, 13244.
(68) Becke, A. D. J. Chem. Phys. 1992, 98, 5648.

Benzamidinate Titanium Imido Compounds
Organometallics, Vol. 24, No. 10, 2005 2357
Figure 5. (a, top) Plot of normalized concentrations vs
time for the tBuNCO extrusion reaction of [Ti(η-C5H4Me)-
{NtBuC(O)O}{Me3SiNC(Ph)NCH2CH2CH2NMe2}] (13). Val-
ues given are for the carbamate 13 and tBuNCO. (b,
t
bottom) Plot of ln(I/I0) vs time for the BuNCO extrusion
reaction of [Ti(η-C5H4Me){NtBuC(O)O}{Me3SiNC(Ph)NCH2-
CH2CH2NMe2}] (13).
Figure 6. DFT computed pathway for the reaction of [Ti(η-
C5H5) (NMe){MeC(NMe)2}] (I) with CO2 showing Gibbs free
energies at 298 K (kJ mol-1). Electronic energies are given
in parentheses. Representations of the DFT structures are
shown (left to right) for I, TSI-II, II, TSII-III, III, and IV.
See the text for further details.
Table 6. First-Order Rate Constants k and
Associated Half-Lives (t1/2) for Extrusion of
tBuNCO from in Situ Generated Carbamate
Complexes 13-15a
initial concn × 108
carbamate
(mol dm-3
)
k × 104 (s-1
)
t1/2 × 10-3 (s)
14
1.9
2.9
3.8
1.9
1.3 ( 0.03
1.2 ( 0.02
1.2 ( 0.03
2.4 ( 0.03
2.1 ( 0.03
1.1 ( 0.08
1.1 ( 0.02
5.63 ( 1.30
5.94 ( 0.99
6.16 ( 1.54
2.93 ( 0.36
3.37 ( 0.48
6.43 ( 0.55
6.30 ( 0.11
{NMeC(O)O}{MeC(NMe)2}] (II) at -113 kJ mol-1 (∆GI-II
) -55.4 kJ mol-1). The very low activation energy
barrier and the relatively low energy of the carbamate
are in agreement with the experimental results, where
cycloaddition of CO2 to the imide to form the carbamate
occurs readily, even at -78 °C for the three-carbon and
“dummy” arm systems 14 and 15.
15
13
1.9
a Values of “initial concentration” refer to the imido precursors
3-5.
Isocyanate extrusion was found to be a two-step
process. The first step is the breaking of the C-O bond
of carbamate II to yield the isocyanate adduct [Ti(η-
C5H5)(O)(MeNCO){MeC(NMe)2}] (III). The activation
The reaction leading to the titanium terminal oxo
complex IV and MeNCO is exothermic by 25.4 kJ mol-1
(∆GI-IV ) -32.2 kJ mol-1). The first step in the reaction
is the side-on cycloaddition of CO2 to the TidNMe bond.
The direction of CO2 attack and geometry of the transi-
tion state (TSI-II) is explained by the shape of the
frontier orbitals of I (Figure 7). The HOMO is one of
the π-components of the TidN bond, resulting from the
interaction of the titanium 3dyz orbital with the nitrogen
2pz orbital (Figure 7 gives the coordinate system), and
energy barrier of 88 kJ mol-1 (∆GqII-III ) 85.1 kJ mol-1
)
is associated with the concerted breaking of the C-O
carbamate bond and formation of the TidO oxo bond
(Ti-O ) 1.945 Å, C-O ) 1.346 Å (II), Ti-O ) 1.666 Å,
C-O ) 2.222 Å (TSII-III)). The second step involves
crossing TSIII-IV, which is essentially a transition state
corresponding to breaking the Ti-N bond. In the case
of the symmetric system III, the transition state for this
step has not been located but the activation barrier is
likely to be no greater than ca. 10 kJ mol-1, as was
found to be the case for the pendant arm substituted
complexes (vide infra and Table 7). Moreover, the
transformation from III to IV is exothermic (∆GIII-IV
) -53.9 kJ mol-1); thus, the associated TS is expected
to be reactant-like and to be at low energy.
2
the LUMO is essentially a 3dz orbital. Consequently,
the frontier orbitals of the imido group are ideally suited
for efficient interaction with both the HOMO and the
LUMO of CO2, leading to a very low electronic energy
activation barrier for cycloaddition of 0.6 kJ mol-1
(∆Gq
) 51.7 kJ mol-1, taking into account entropy
I-II
effects). The reaction proceeds exothermically to form
the very stable carbamate intermediate [Ti(η-C5H5)-

2358 Organometallics, Vol. 24, No. 10, 2005
Boyd et al.
Figure 7. Representations of the HOMO (left) and LUMO (right) of the model complex I together with the chosen coordinate
system.
Table 7. Relative Electronic Energies (kJ mol-1) of Extrema along the Pathway for CO2 Cycloaddition/
a
MeNCO Extrusion on the Two Pendant Arm Functionalized Titanium Amidinate Complexes IC2 and IC3
IC2
IC3
I
SiH3
C2 arm
4.5 (54.0)
SiH3
C3 arm
TSI-II
II
TSII-III
III
TSIII-IV
IV + MeNCO
0.6 (51.7)
3.5 (51.1)
3.0 (51.0)
4.6 (51.9)
-113.1 (-55.4)
-25.1 (29.7)
-29.7 (21.7)
not located
-105.7 (-44.4)
-20.6 (31.6)
-23.7 (27.4)
-15.3 (30.5)
-24.6 (-33.1)
-110.5 (-49.1)
-25.2 (29.7)
-29.4 (25.3)
-21.4 (19.7)
-24.6 (-33.1)
-105.8 (-46.4)
-21.9 (29.2)
-25.0 (31.6)
-15.6 (27.1)
-23.7 (-33.0)
-109.1 (-53.3)
-23.7 (25.0)
-27.8 (22.3)
-20.4 (19.5)
-23.7 (-33.0)
-25.4 (-32.2)
a Gibbs free energy values at 298 K are given in parentheses. For IC2 and IC3, SiH3 and C2 arm are the sides of CO2 attack. Energies
are computed with respect to the separated imido-amidinate titanium complex and CO2. Values for the corresponding reaction of I and
CO2 are presented for ease of comparison.
(a) Influence of the Amidinate Pendant Arm
Substituent: NMe2 Not Chelated. The critical issue
in the reactions of 3-5 concerns the effect of the
pendant donor group (for 3 and 4) and chain length, and
whether the key TidNtBu/CO2 coupling reaction takes
place on a chelated (i.e., 3 or 4) or nonchelated (3′ or 4′)
complex. Further DFT model complexes have therefore
been studied (with and without pendant group coordi-
nation). We consider first the pendant arm noncoordi-
nated model complexes IC2 and IC3. These correspond
(SiMe3 in the real compounds 3′ and 4′), leading to the
carbamate IIC2-Si or IIC3-Si, or from the pendant arm
side, leading to IIC2-N and IIC3-N (Figure 8). The
pathways for the two regiochemistries of addition have
been computed. The energies of the extrema are re-
ported in Table 7, and the associated activation barriers
for CO2 addition and MeNCO extrusion are given in
Table 8.
In terms of the energies associated with the various
pathways, the identity of the noncoordinated pendant
arm substituent has no significant influence. The reac-
tion is very similar to I + CO2 (Figure 6), except that
in all but one case, the isocyanate adduct (for example
III in Figure 6) is not stable enough to create a local
minimum in the Gibbs free energy surface (indicated
by the values given in italics in Table 7) and so extrusion
could equally well be considered to be a one-step process.
The difference in behavior between the energy and
Gibbs free energy surfaces for the extrusion could be
explained by the influence of rotational energy contribu-
tions on dissociation.
In terms of the regiochemistry, the barriers for CO2
cycloaddition and for isocyanate extrusion are very close
in energy for all four cases. For both the two-carbon and
three-carbon arm systems there is a small energetic
preference for CO2 attack from the SiH3 side to form a
carbamate intermediate where the imido NMe substitu-
to the postulated intermediates in NMe2 methyl group
exchange, namely 3′ and 4′ (eq 1).
As expected, the overall pathway is analogous to that
computed for I. However, in this case the cycloaddition
of CO2 can proceed either from the SiH3-substituted side

Benzamidinate Titanium Imido Compounds
Organometallics, Vol. 24, No. 10, 2005 2359
Figure 8. Various DFT computed carbamate complexes formed from the reaction of CO2 with IC2 and IC3
.
a ∆∆Gq value of less than 2 kJ mol-1, which is below
the accuracy associated with these calculations.
Table 8. Electronic Activation Energy Barriers
(kJ mol-1) along the Pathway for CO2
Cycloaddition/MeNCO Extrusion Reactions of
Pendant Arm Functionalized Titanium Amidinate
(b) Influence of the Amidinate Pendant Arm
Substituent: NMe2 Chelated. It is important to
determine whether CO2 cycloaddition or/and isocyanate
extrusion could take place on pendant arm coordinated
complexes. Model imido amidinate complexes with the
pendant NMe2 coordinated to titanium have been
optimized, and the resulting geometries are shown in
a
Complexes IC2 and IC3
IC2
IC3
SiH3
C2 arm
4.5 (54.0)
SiH3
C3 arm
4.6 (51.9)
I to TSI-II
3.5 (51.1)
3.0 (51.0)
II to TSII-III 85.1 (76.0) 85.3 (78.8) 83.9 (75.6) 85.4 (78.3)
a Gibbs free energy values at 298 K are given in parentheses.
For IC2 and IC3, SiH3 and C2 arm are the sides of CO2 attack.
Figure 9(a). The chelated isomer (IC2-chel and IC3-chel
is more stable than the nonchelated complex (IC2 and
C3) in both cases by 25.4 and 23.3 kJ mol-1, respec-
)
I
tively. In IC3-chel, the six-membered ring is in a “chair”
conformation. The alternative “boat” conformation has
been optimized and lies 12.2 kJ mol-1 higher in energy.
When Gibbs free energy values are considered at 298
K, the chelated isomer IC2-chel is still more stable than
ent and the SiH3 group are on opposite sides. The
dissociation of MeNCO from the pendant arm side that
necessarily follows CO2 addition from the SiH3 direction
is also slightly kinetically preferred (Table 8). While
these differences are not significant within the accuracy
of DFT calculations, it might be expected that the small
energetic preference for this regiochemistry of addition
and extrusion would be increased if the actual steric
bulk were to be considered (i.e. SiMe3 in place of SiH3
I
C2, but only by 6.2 kJ mol-1. For the three-carbon
system, however, the nonchelated isomer IC3-chel is now
only marginally more stable (0.4 kJ mol-1) than IC3. The
Gibbs free energy differences between IC2 and IC2-chel
(and IC3 and IC3-chel) are lower than the electronic ∆E
values, because the chelated form is entropically dis-
favored. Since the three-carbon chain has more rota-
tional degrees of freedom, the impact of the decoordi-
nation upon G is more pronounced. This is in agreement
with the experimental observation that NMe2 group
methyl exchange occurs at a lower temperature for the
real C3 system 14.
Interestingly, the HOMOs for the chelated complexes
(Figure 9b) are not orientated in the same plane as for
the nonchelated complexes (Figure 7). For the chelated
complexes the frontier orbital preferred trajectory for
CO2 addition is therefore from below, trans to the
methylcyclopentadienyl ring. The corresponding transi-
tion states (C2chelTSI-II and C3chelTSI-II) are shown
in Figure 9c, and their electronic energies are 24.5 and
39.5 kJ mol-1, respectively, with respect to the isolated
reactants. Therefore, in terms of electronic energies, CO2
attack on the chelated isomers IC2-chel and IC3-chel is
disfavored by some 20-35 kJ mol-1 compared to attack
at the nonchelated isomers IC2 and IC3. Furthermore,
it is the chelated two-carbon arm species that should,
according to the calculations, react more readily with
CO2, which is in contrast to the experimental results.
Taken together, the experimental and computational
data imply that preferred attack is via the nonchelated
t
and Bu in place of NMe). The crystallographically
characterized carbamate complex 7 (Figure 1) formed
from CO2 and the real pentamethylcyclopentadienyl C3
arm system 2 is consistent with CO2 attack from the
SiMe3 side.
The calculations of the nonchelating model systems
C2 and IC3 imply that the CO2 addition reactions of the
I
real complexes 3 and 4 (and obviously 5) would also be
very similar if the pendant NMe2 groups played only
an indirect role. However, since the two-carbon arm
system 3 does not react with CO2 at significant rates
below -35 °C and the three-carbon and “dummy-arm”
homologues react rapidly even at -78 °C, it is clear that
there must be a direct effect of the pendant NMe2 that
is different (or significantly amplified) for the two-carbon
system. This will be addressed in the following section.
We turn now briefly to the isocyanate extrusion step.
This occurs at the same temperature for all three of the
real carbamate systems 13-15, with rate constants that
are very close in value. From the calculated values, the
activation barrier for the isocyanate extrusion is ca. 85
kJ mol-1 for both the two-carbon and the three-carbon
arm model systems (energy difference between II and
TSII-III). Although the experimental results indicate
that the rate of extrusion is twice as fast for the two-
carbon pendant arm carbamate 15, this translates into

2360 Organometallics, Vol. 24, No. 10, 2005
Boyd et al.
Figure 9. (a) DFT minimized geometries of the chelated imido amidinate complexes IC2-chel (left) and IC3-chel viewed
along the Cpcent‚‚‚Ti vectors. (b) Representations of the HOMO of each compound viewed approximately perpendicular to
these vectors. (c) Geometries of the transition states for CO2 cycloaddition to TidNMe (C2chel-TSI-II for addition to IC2-chel
and C3chel-TSI-II for addition to IC3-chel).
isomers and that it is the activation barriers to accessing
these that control the reactions.
been established (Table 1), it has been possible to make
a “hybrid” comparison of computational and experimen-
tal data for the two possibilities for CO2 attack (Figure
10). The Gibbs free energies in Figure 10 were computed
for T ) 200 K, slightly above the temperature at which
the real systems 4 (three-carbon arm) and 5 (“dummy-
arm”) react. All energies are from DFT calculations on
the species or transition states indicated, except for
those for the transition state between IC2-chel and IC2
(51.2 kJ mol-1) and between IC3-chel and IC3 (41.6 kJ
mol-1), which are taken from the experimental data for
the real complexes 3 and 4, respectively (Table 1)
Figure 10 unambiguously indicates that CO2 attack
would occur on the nonchelated isomers IC2 (corre-
sponding to the real system 3′, eq 1) and IC3 (real system
4′). Furthermore, this preference would be reinforced
as the temperature is increased (recall that 3 only reacts
Unfortunately, it was not possible computationally to
find a transition state for dissociation of NMe2 in the
model systems (i.e. for IC2-chel f IC2 or IC3-chel f IC3).
Such a transition state would correspond to a dissocia-
tion of NMe2 (coupled with subsequent pendant arm
chain reorganization) and would be associated with a
very soft motion (low value of the imaginary fre-
quency at the TS). Given the large number of degrees
of freedom associated with the different conformations
of the decoordinating CH2CH2NMe2 and CH2CH2CH2-
NMe2 chains, it is not at all straightforward to estimate
a starting geometry for the TS structure. Therefore,
despite many attempts, it was not possible to locate such
a transition-state structure. However, since the experi-
mental values for NMe2 methyl group exchange have

Benzamidinate Titanium Imido Compounds
Organometallics, Vol. 24, No. 10, 2005 2361
Figure 10. Schematic representation of free energies for CO2 addition for both the chelated and nonchelated imido-
amidinate systems at 200 K, relative to the respective chelated isomers. The upper value in each pair is for the two-carbon
arm system and the lower value (in italics) is for the three-carbon arm systems (TS ) transition state). All energies are
from DFT calculations except for “TS chelate f nonchelate” which are derived from the ∆Hq and ∆Sq data for 3 and 4
(Table 1). DFT minimized geometries for the two-carbon arm system are illustrated.
at a reasonable rate above 238 K) because attack by CO2
is entropically disfavored (thus the Gibbs free energy
of the TS for addition to IC2chel and IC3chel would be
higher than 62.9 and 77.7 kJ mol-1, respectively)
whereas ∆Gq for decoordination (leading to NMe2 meth-
yl exchange) is entropically favored and would lead to
lower values than the 51.2 and 41.6 kJ mol-1 calculated
at 200 K.
Although it is known experimentally that the real
carbamate complexes 15 and 13 do not have a chelated
NMe2 group, attempts were made to elucidate compu-
tationally a mechanism for MeNCO extrusion from a
carbamate complex with a chelating pendant arm. It
was found that the κ2-carbamates formed from CO2
cycloaddition to IC2chel or IC3chel have to be converted
to κ1-bound carbamates before MeNCO extrusion could
occur. However, it was not possible to find a transition
state for this κ2 to κ1 interconversion which did not
involve dissociation of the NMe2 group. Therefore, not
only is CO2 attack on the chelate isomer kinetically less
favored but also there is no exit channel from the κ2-
carbamate so formed, which leads to the observed
products. This also further militates against the likeli-
hood of a pendant arm anchimeric effect on extrusion
of isocyanate from the carbamate complexes.
tween the chelated and nonchelated isomers would lie
in favor of the nonchelated isomer for the experimental
C3 system. This is in contrast to the clear evidence for
chelation of the NMe2 group in the low-temperature
NMR spectra of this compound. Thus, while the calcu-
lated energies do not entirely reflect the experimental
behavior of compound 4, they do confirm the overall
trend that the stabilization upon chelation is greater for
the two-carbon arm system and that this trend is
amplified when steric factors are considered. Again, the
less favorable dissociation of the C2 arm from titanium
may account for the higher temperature required for
CO2 cycloaddition if the nonchelated isomer is the
reactive species.
Conclusions
Reactions of cyclopentadienyl amidinate tert-butyl-
imido complexes with CO2 all ultimately yield oxo-
bridged dimers and tBuNCO. The stabilities of the N,O-
bound carbamate intermediates depend mainly on the
cyclopentadienyl ring substituents, but there is also a
small dependence on the nature of the length of the
(dimethylamino)alkyl chain for the C5H4Me systems.
This might be due either to small steric differences
between the amidinate ligand systems or from a slight
anchimeric effect of the two-carbon arm on the extru-
sion. The differences in the DFT calculated activation
barriers for isocyanate extrusion from the two-carbon
and three-carbon arm carbamate complexes without the
pendant arms coordinated are too small to be signifi-
cant. The origin of the small increase in the rate of
isocyanate extrusion for the C2 arm systems is therefore
still somewhat unclear.
(c) Influence of Steric Factors. ONIOM calcula-
tions to take into account the influence of steric bulk
have been performed. These indicate that the chelated
isomer IC2-chel is more stable than the nonchelated
isomer IC2 by 3.8 kJ mol-1 for the C2 system. In the case
of the C3 system, the chelated isomer IC3-chel is slightly
less stable than the nonchelated isomer IC3 by 4.9 kJ
mol-1. These values suggest that the equilibrium be-

2362 Organometallics, Vol. 24, No. 10, 2005
Boyd et al.
referenced internally to residual protio solvent (1H) or solvent
(13C) resonances and are reported relative to tetramethylsilane
(δ 0 ppm). Chemical shifts are quoted in δ (ppm) and coupling
constants in hertz. Infrared spectra were prepared as Nujol
mulls or thin films between KBr or NaCl plates and were
recorded on Perkin-Elmer 1600 and 1700 series spectrometers.
Infrared data are quoted in wavenumbers (cm-1). Mass spectra
were recorded by the mass spectrometry service of the Uni-
versity of Oxford’s Inorganic Chemistry Laboratory. Combus-
tion analyses were recorded by the analytical services of the
University of Oxford’s Inorganic Chemistry Laboratory.
Literature Preparations. The compounds [Ti(η-C5Me5)-
(NtBu)Cl(py)],52 [Ti(η-C5H4Me)(NtBu)Cl(py)],52 Li[Me3SiNC-
(Ph)NCH2CH2NMe2],47 Li[Me3SiNC(Ph)NCH2CH2CH2NMe2],48
and Li[Me3SiNC(Ph)NCH2CH2Me]49 were prepared according
to published methods. All other compounds and reagents were
purchased and used without further purification.
Much more pronounced are the differences in reaction
temperature for onset of CO2 cycloaddition to the
TidNtBu bonds of 3 and 4. The kinetics studies on the
cycloaddition of CO2 to TidNtBu for the two-carbon
linker arm system (3) have established that the rate-
determining step is bimolecular with a first order
dependence on CO2. In the case of the “dummy-arm”
system 5, cycloaddition necessarily occurs to a non-
chelated pendant arm species. However, the first-order
CO2 dependence in the case of the pendant amine
functionalized compound 3 (and probably 4), is equally
consistent with both direct cycloaddition to the coordi-
nated isomer and pendant arm dissociation followed by
cycloaddition. The DFT studies suggest that, for the
three-carbon atom linker system 4, CO2 cycloaddition
occurs to the nonchelated isomer and is essentially
barrierless. This is supported by the similar behavior
of 4 and the “dummy-arm” system 5, which both
undergo rapid cycloaddition of CO2 even at -78 °C. For
the two-carbon arm system 3 it is possible that CO2
addition to the chelated isomer could be competitive, but
comparison of the activation barriers for methyl group
exchange (experimentally determined) and CO2 cyclo-
addition (calculated) at 200 K suggest that decoordina-
tion of the pendant arm prior to CO2 cycloaddition is
energetically favored. It was not possible to optimize a
transition state for NMe2 group dissociation by DFT;
thus, theoretical values are not available for comparison
with the experimentally determined activation param-
eters. The greater energy required to decoordinate the
pendant amine from titanium for the two-carbon arm
system explains the higher temperature needed before
CO2 cycloaddition to 3 occurs.
The mechanistic studies of isocyanate extrusion from
the “dummy-arm” system at different concentrations
suggest a first-order dependence on the carbamate
complex in the rate-determining step. This is consistent
with extrusion to form a monomeric terminal oxo
complex which then self-traps by dimerization. The flat
nature of the energy profile between TSII-III and the
oxo compound IV suggests that there may be some
reversibility in this part of the mechanism, whereby the
transient terminal oxo complex can undergo back-
cycloaddition of tBuNCO to re-form the carbamate
intermediate unless it is trapped by dimerization. These
results also suggest that it might be possible to intercept
the transient TidO species with other unsaturated
substrates.62
Synthesis of [Ti(η-C5Me5)(NtBu){Me3SiNC(Ph)NCH2-
CH2NMe2}] (1). To a mixture of [Ti(η-C5Me5)(NtBu)Cl(py)]
(1.46 g, 3.97 mmol) and Li{Me3SiNC(Ph)NCH2CH2NMe2} (1.07
g, 3.97 mmol) was added benzene (40 mL). The resulting dark
brown solution was stirred at room temperature for 4 h and
refluxed at 80 °C for 15 h; the solution became brown, and a
white precipitate formed. Volatiles were removed under re-
duced pressure, and the resulting brown oil was extracted into
benzene (40 mL). The brown solution was filtered, and volatiles
were removed under reduced pressure to yield a brown oil,
which was triturated with pentane (20 mL). Yield: 1.70 g
(83%). The product was purified by distillation (1 × 10-5 mbar,
120-130 °C) to yield 1 as a brown waxy solid. Distilled yield:
0.25 g (45%).
1H NMR data (C6D6, 500.1 MHz, 293 K): 7.3-7.0 (5 H, m,
C6H5), 3.42 (1 H, m, CN2CH2), 3.14 (1 H, m, CN2CH2), 2.52 (1
H, m, CH2NMe2), 2.43 (1 H, m, CH2NMe2), 2.12 (15 H, s,
C5Me5), 2.04 (6 H, s, NMe2), 1.22 (9 H, s, CMe3), -0.01 (9 H,
s, SiMe3). 13C{1H} NMR data (C6D6, 125.7 MHz, 293 K): 166.79
(CN2), 135.74 (i-C6H5), 128.29, 128.11, 128.20 (5 × C of C6H5*),
119.26 (C5Me5), 66.89 (CMe3), 62.30 (CH2NMe2), 46.76
(CN2CH2), 46.02 (NMe2), 33.10 (CMe3), 12.02 (C5Me5), 2.88
(SiMe3); the asterisk indicates that other resonances were
obscured by solvent. IR data (KBr plates, Nujol mull, cm-1):
2766 (w), 2726 (w), 1578 (m), 1510 (s), 1485 (s, br), 1402 (s),
1348 (m), 1314 (w), 1245 (s), 1207 (w), 1189 (w), 1158 (w), 1120
(w), 1100 (w), 1057 (w, br), 1024 (m), 936 (w), 920 (m), 906
(m), 860 (w), 838 (m), 802 (w), 778 (w), 759 (w), 723 (w), 702
(w), 633 (w), 587 (w), 542 (m), 507 (w), 470 (w), 417 (w). Anal.
Found (calcd for C28H48N4SiTi): C, 65.3 (65.1); H, 9.3 (9.3); N,
10.6 (10.7).
Synthesis of [Ti(η-C5Me5)(NtBu){Me3SiNC(Ph)NCH2-
CH2CH2NMe2}] (2). To a mixture of [Ti(η-C5Me5)(NtBu)Cl-
(py)] (1.25 g, 3.40 mmol) and Li{Me3SiNC(Ph)NCH2CH2CH2-
NMe2} (0.96 g, 3.40 mmol) was added benzene (40 mL). The
mixture was stirred at room temperature for 1 h and refluxed
at 80 °C for 15 h, giving a cloudy red-brown solution. Volatiles
were removed under reduced pressure, and the resulting red-
brown oil was extracted into benzene (30 mL) and filtered. The
residual solids were washed with benzene (2 × 20 mL).
Volatiles were removed from the combined extracts under
reduced pressure, giving a red-brown oil, which was triturated
with pentane (30 mL) to yield 2 as a red-brown waxy solid.
Yield: 1.61 g (89%). The product was purified by distillation
(1 × 10-5 mbar, 120-130 °C) to yield 1 as a red-brown waxy
solid. Distilled yield: 0.21 g (35%).
Experimental Section
General Methods and Instrumentation. All manipula-
tions were carried out using standard Schlenk line or drybox
techniques under an atmosphere of argon or of dinitrogen.
Solvents were predried over 4 Å molecular sieves and were
refluxed over appropriate drying agents under a dinitrogen
atmosphere and collected by distillation. Deuterated solvents
were dried over appropriate drying agents, distilled under
reduced pressure, and stored under dinitrogen in Teflon valve
ampules. NMR samples were prepared under dinitrogen in 5
mm Wilmad 507-PP tubes fitted with J. Young Teflon valves.
1H, 13C{1H}, and 13C NMR spectra were recorded on Varian
Unity Plus 500 and Varian Mercury spectrometers. 1H and
13C assignments were confirmed where necessary with the use
of NOE, DEPT-135, DEPT90, DEPT-45, and two-dimensional
1H-1H and 13C-1H NMR experiments. All spectra were
1H NMR data (C6D6, 300.1 MHz, 293 K): 7.3-7.0 (5 H, m,
C6H5), 3.31 (1 H, m, CN2CH2), 3.06 (1 H, m, CN2CH2), 2.18 (2
H, m, CH2NMe2), 2.15 (15 H, s, C5Me5), 2.08 (6 H, s, NMe2),
1.74 (2 H, m, CH2CH2CH2), 1.26 (9 H, s, CMe3), 0.06 (9 H, s,
SiMe3). 13C{1H} NMR data (C6D6, 75.5 MHz, 293 K): 166.52
(CN2), 135.91 (i-C6H5), 128.82, 128.30 (5 × CH of C6H5*),
119.21 (C5Me5), 66.92 (CMe3), 57.55 (CH2NMe2), 46.00
(CN2CH2), 45.51 (NMe2), 33.10 (CMe3), 30.94 (CH2CH2CH2),

Benzamidinate Titanium Imido Compounds
Organometallics, Vol. 24, No. 10, 2005 2363
12.04 (C5Me5), 2.96 (SiMe3); the asterisk indicates that other
resonances were obscured by solvent. IR data (KBr plates,
Nujol mull, cm-1): 2724 (m), 1612 (w), 1576 (m), 1510 (m), 1484
(s, br), 1402 (m, br), 1310 (m), 1245 (m), 1205 (w), 967 (w),
920 (w), 904 (w), 875 (w), 838 (m), 802 (w), 782 (w), 759 (w),
723 (w), 701 (w), 633 (w), 587 (w), 543 (w), 507 (w). Anal. Found
(calcd for C29H50N4SiTi): C, 65.5 (65.6); H, 9.4 (9.5); N, 10.5
(10.6).
3-C5H4Me), 5.87 (1 H, m, 2- or 5-C5H4Me), 5.72 (1 H, m, 5- or
2-C5H4Me), 3.20 (1 H, m, CN2CH2), 2.81 (1 H, m, CN2CH2),
2.29 (1 H, m, CH2NMe2), 2.17 (6 H, s, NMe2), 2.04 (3 H, s,
C5H4Me), 1.95 (1 H, m, CH2NMe2), 1.60 (1 H, m, CH2CH2CH2),
1.28 (9 H, s, CMe3), 1.23 (1 H, m, CH2CH2CH2), 0.08 (9 H, s,
SiMe3). 13C{1H} NMR data (C6D6, 125.7 MHz, 293 K): 169.39
(CN2), 137.12 (i-C6H5), 128.35, 128.30, 127.47 (5 × CH of
C6H5*), 121.20 (1-C5H4Me), 113.27 (2- or 5-C5H4Me), 111.40
(5- or 2-C5H4Me), 106.53 (3- or 4-C5H4Me), 106.42 (4- or
3-C5H4Me), 67.38 (CMe3), 60.02 (CH2NMe2), 48.96 (NMe2),
46.27 (CN2CH2), 33.31 (CMe3), 28.26 (CH2CH2CH2), 14.84
(C5H4Me), 2.95 (SiMe3); the asterisk indicates that other
resonances were obscured by solvent resonance. 1H NMR data
(CD2Cl2, 500.1 MHz, 193 K): 7.38 (1 H, m, p-C6H5), 7.31 (2 H,
m, m-C6H5), 7.21-7.06 (2 H, m, o-C6H5), 6.32 (1 H, m, 3- or
4-C5H4Me), 6.13 (1 H, m, 3- or 4- C5H4Me), 5.48 (1 H, m, 2- or
5- C5H4Me), 5.44 (1 H, m, 5- or 2-C5H4Me), 2.74 (1 H, m,
CN2CH2), 2.74 (1 H, m, CH2NMe2), 2.57 (1 H, m, CN2CH2),
2.57 (3 H, s, NMe), 2.43 (3 H, s, NMe), 2.07 (1 H, m, CH2NMe2),
1.89 (3 H, s, C5H4Me), 1.83 (1 H, m, CH2CH2CH2), 1.23 (1 H,
m, CH2CH2CH2), 0.99 (9H, s, CMe3), -0.31 (9 H, s, SiMe3).
13C{1H} NMR data (CD2Cl2, 125.7 MHz, 193 K): 171.42 (CN2),
136.87 (i-C6H5), 127.62 (p-C6H5), 127.05 (m-C6H5), 126.51 (o-
C6H5), 125.52 (o-C6H5), 117.42 (1-C5H4Me), 113.16 (2- or
5-C5H4Me), 110.37 (5- or 2-C5H4Me), 102.34 (3- or 4-C5H4Me),
101.06 (4- or 3-C5H4Me), 66.84 (CMe3), 61.39 (CH2NMe2), 53.57
(NMe), 53.36 (NMe), 44.00 (CN2CH2), 32.29 (CMe3), 24.44
(CH2CH2CH2), 13.67 (C5H4Me), 1.85 (SiMe3). IR data (KBr
plates, thin film, cm-1): 3050 (m), 2856 (s), 2824 (s, br), 2766
(s), 1422 (s, br), 1348 (s), 1298 (s), 1226 (s), 1200 (s), 1112 (m),
1098 (m), 1072 (m), 1018 (m), 980 (m), 944 (m), 898 (m, br),
836 (m, br), 782 (m), 758 (w), 700 (w), 678 (w), 626 (w), 548
(w), 530 (w), 508 (w). EI-MS: m/z 279 [Me3SiNC(Ph)NCH2-
CH2CH2NMe2]+ 13%, 58 [CHMe3]+ 100%. Anal. Found (calcd
for C25H42N4SiTi): C, 62.7 (63.2); H, 8.8 (8.9); N, 11.4 (11.8).
Synthesis of [Ti(η-C5H4Me)(NtBu){Me3SiNC(Ph)NCH2-
CH2Me}] (5). To a stirred solution of [Ti(η-C5H4Me)(NtBu)-
Cl(py)] (1.53 g, 4.90 mmol) in benzene (30 mL) was added a
solution of [Li2{Me3NSiNC(Ph)NCH2CH2Me}2(Et2O)2] (1.54 g,
4.90 mmol) in benzene (30 mL). The deep red solution became
slightly lighter in color and was stirred for 15 h. Volatiles were
removed under reduced pressure, and the red oily residue was
extracted into benzene (30 mL), the extract was filtered, and
the volatiles were evaporated to yield 5 as a red oil. Yield: 1.96
g (94%).
1H NMR data (C6D6, 500.1 MHz, 293 K): 7.16 (2H, m,
o-C6H5), 7.05-7.10 (3H, m, m- and p-C6H5), 6.83 (2H, m, 3-
and 4-C5H4Me), 5.83 (2H, m, 2- and 5-C5H4Me), 3.12 (1H, m,
CN2CH2), 2.81 (1H, m, CN2CH2), 1.98 (3H, s, C5H4Me), 1.33
(2H, m, CH2CH3), 1.19 (9H, s, CMe3), 0.80 (3H, t, CH2CH3),
-0.05 (9H, s, SiMe3). 13C{1H} NMR data (C6D6, 125.5 MHz,
293 K): 165.24 (CN2), 135.94 (i-C6H5), 128.68 (m- or p-C6H5),
128.49 (p- or m-C6H5), 127.78 (o-C6H5), 113.29 (2- or 5-C5H4Me),
112.19 (2- or 5-C5H4Me), 109.14 (3- or 4-C5H4Me), 108.98 (3-
or 4-C5H4Me), 67.46 (CMe3), 51.72 (CN2CH2), 33.51 (CMe3),
26.52 (CH2CH3), 15.44 (C5H4Me), 13.13 (CH2CH3), 3.24 (SiMe3).
1H NMR data (CD2Cl2, 500.1 MHz, 195 K): 7.29-7.45 (3H,
m, br, o- and m-C6H5), 7.12 (1H, m, p-C6H5), 6.92 (2H, s, 3-
and 4-C5H4Me), 5.59 (1H, s, 5- or 2-C5H4Me), 5.55 (1H, s, 2- or
5-C5H4Me), 3.02 (1H, m, CN2CH2), 2.77 (1H, m, CN2CH2), 1.81
(3H, s, C5H4Me), 1.22 (2H, m, CH2CH3), 0.89 (CMe3), 0.77
(CH2CH3), -0.35 (SiMe3). IR data (KBr plates, thin film, cm-1):
3024 (m), 2932 (s), 2893 (s), 2854 (s), 1602 (w), 1578 (w), 1510
(s), 1480 (s), 1414 (s), 1376 (m), 1348 (s), 1292 (m), 1244 (s),
1206 (m), 1156 (w), 1120 (m), 1072 (s), 1016 (m), 922 (s), 878
(m), 836 (s), 804 (m), 782 (s), 758 (m), 686 (m), 632 (m), 594
(m), 426 (m), 412 (m). High-resolution EI-MS, found (calcd for
C23H37N3SiTi): m/z 431.2227 (431.2236). Anal. Found (calcd
for C23H37N3SiTi): C, 62.9 (64.0); H, 8.4 (8.6); N, 9.8 (9.7).
Synthesis of [Ti(η-C5Me5){OC(O)NtBu}{Me3SiNC(Ph)-
NCH2CH2NMe2}] (6). A solution of [Ti(η-C5Me5)(NtBu)-
Synthesis of [Ti(η-C5H4Me)(NtBu){Me3SiNC(Ph)NCH2-
CH2NMe2}] (3). To a stirred solution of [Ti(η-C5H4Me)(NtBu)-
Cl(py)] (1.81 g, 5.78 mmol) in benzene (30 mL) was added a
solution of Li{Me3SiNC(Ph)NCH2CH2NMe2} in benzene (30
mL), dropwise over 15 min. When it was mixed, the dark
brown solution turned orange and a white precipitate formed.
The mixture was stirred for 15 h, volatiles were removed under
reduced pressure, and the resulting red-brown oil was ex-
tracted into benzene (40 mL). The brown solution was filtered,
and volatiles were removed under reduced pressure to give a
red-brown oil, which was triturated with pentane (20 mL).
Yield: 2.42 g (91%). The product was purified by distillation
(1 × 10-5 mbar, 120-130 °C) to yield 3 as a brown oil. Distilled
yield: 0.32 g (65%).
1H NMR data (C6D6, 300.1 MHz, 293 K): 7.00-7.12 (5 H,
m, C6H5), 6.66 (1 H, m, 3- or 4-C5H4Me), 6.21 (1 H, m, 4- or
3-C5H4Me), 5.91 (1 H, m, 2- or 5-C5H4Me), 5.52 (1 H, m, 5- or
2-C5H4Me), 2.94 (1 H, m, CN2CH2), 2.89 (1 H, m, CN2CH2),
2.34 (1 H, m, CH2NMe2), 2.17 (3 H, s, C5H4Me), 2.07 (6 H, s,
NMe2), 1.80 (1 H, m, CH2NMe2), 1.22 (9 H, s, CMe3), 0.15 (9
H, s, SiMe3). 13C{1H} NMR data (C6D6, 75.5 MHz, 293 K):
172.91 (CN2), 138.46 (i-C6H5), 128.34, 128.19, 127.99, 127.80,
127.28 (5 × CH of C6H5), 120.28 (1-C5H4Me), 112.27 (2- or
5-C5H4Me), 110.49 (5- or 2-C5H4Me), 105.30 (3- or 4-C5H4Me),
105.01 (4- or 3-C5H4Me), 67.01 (CMe3), 63.65 (CH2NMe2), 49.68
(NMe2), 44.84 (CN2CH2), 33.46 (CMe3), 15.17 (C5H4Me), 3.53
1
(SiMe3). H NMR data (CD2Cl2, 500.1 MHz, 193 K): 7.38 (3
H, m, o- and p-C6H5), 7.28 (2 H, m, m-C6H5), 6.34 (1H, s, 3- or
4-C5H4Me), 5.95 (1H, s, 4- or 3-C5H4Me), 5.68 (1H, s, 2- or
5-C5H4Me), 5.52 (1H, s, 5- or 2-C5H4Me), 3.16 (1H, m, CN2CH2),
3.09 (1H, m, CN2CH2), 2.63 (3H, s, NMe), 2.35 (3H, s, NMe),
2.21 (2H, m, CH2NMe2), 2.05 (3H, s, C5H4Me), 0.99 (9H, s,
CMe3), -0.19 (9H, s, SiMe3). 13C{1H} NMR data (CD2Cl2, 125.7
MHz, 193 K): 173.03 (CN2), 137.24 (i-C6H5), 128.14 (p-C6H5),
129.91 (m-C6H5), 126.91 (o-C6H5), 110.68 (2- or 5-C5H4Me),
107.26 (5- or 2-C5H4Me), 104.15 (3- and 4-C5H4Me, overlap-
ping), 66.67 (CMe3), 63.53 (CH2NMe2), 51.48 (NMe), 48.25
(NMe), 44.45 (CN2CH2), 32.50 (CMe3), 14.74 (C5H4Me), 2.64
(SiMe3). IR data (KBr plates, thin film, cm-1): 3641 (m), 3060
(m), 2959 (s, br), 2785 (s, br), 2727 (s, br), 2648 (m), 2361(w),
2342 (w), 1952 (w), 1883 (w), 1812 (w), 1621 (w), 1602 (m),
1579 (s, br), 1402 (m), 1489 (s), 1362 (m), 1348 (m), 1331 (s),
1235 (s), 1205 (s), 1174 (w), 1159 (w), 1094 (m), 1052 (m), 1027
(s), 953 (m), 934 (w), 919 (m), 872 (s), 834 (s), 779 (s), 702 (m),
678 (m), 628 (m), 589 (w), 569 (m), 532 (w), 505 (w), 466 (w),
445 (w), 420 (w). Anal. Found (calcd for C24H40N4SiTi): C, 62.4
(62.6); H, 8.7 (8.7); N, 12.0 (12.2).
Synthesis of [Ti(η-C5H4Me)(NtBu){Me3SiNC(Ph)NCH2-
CH2CH2NMe2}] (4). To a stirred solution of [Ti(η-C5H4Me)-
(NtBu)Cl(py)] (1.04 g, 3.33 mmol) in benzene (30 mL) was
added a solution of Li{Me3SiNC(Ph)NCH2CH2CH2NMe2} (0.94
g, 3.33 mmol) in benzene (30 mL), dropwise over 15 min. The
brown mixture was stirred for 15 h. Volatiles were removed
under reduced pressure, and the resulting brown oil was
extracted into benzene (30 mL) and the extract filtered.
Volatiles were removed under reduced pressure, and the solid
brown residues were washed with benzene (2 × 20 mL).
Volatiles were removed from the combined extracts under
reduced pressure to give a dark brown oil. Yield: 1.33 g (84%).
The product was distilled under high vacuum (1 × 10-5 mbar,
120-130 °C) to yield 4 as a brown oil. Yield: 1.21 g (79%).
1H NMR data (C6D6, 500.1 MHz, 293 K): 7.2-7.0 (5 H, m,
C6H5), 6.81 (1 H, m, 3- or 4-C5H4Me), 6.57 (1 H, m, 4- or

2364 Organometallics, Vol. 24, No. 10, 2005
Boyd et al.
{Me3SiNC(Ph)NCH2CH2NMe2}] (1; 106 mg, 20.6 mmol) in
pentane (10 mL) was exposed to ca. 1 atm of CO2 via a dual-
manifold Schlenk line. The solution was stirred and volatiles
removed under reduced pressure as soon as the brown solution
became cherry red (ca. 5 min). The resulting red powder was
stored at -30 °C in a drybox. Yield: 87 mg (76%).
Synthesis of [Ti2(η-C5H4Me)2(µ-O)2{Me3SiNC(Ph)NCH2-
CH2NMe2}2] (10).
A
solution of [Ti(η-C5H4Me)(NtBu)-
{Me3SiNC(Ph)NCH2CH2NMe2}] (3; 0.40 g, 0.86 mmol) in
benzene (15 mL) was exposed to ca. 1 atm of CO2 via a dual-
manifold Schlenk line. The solution immediately changed color
from red-brown to cherry red and after 24 h turned brown.
Volatiles were removed under reduced pressure, giving a
brown oil, which was triturated with pentane (20 mL) to give
a waxy brown solid that was shown by NMR to be a mixture
of two isomers, 10a and 10b. Yield: 0.19 g (56%). A sample of
the crude product (0.16 g, 0.35 mmol) was recrystallized from
a mixture of CH2Cl2 (10 mL) and hexanes (20 mL) to yield
analytically pure product. Yield: 6.6 mg (5%).
1H NMR data (C6D6, 500.1 MHz, 293 K): 7.0-6.9 (5 H, m,
C6H5), 3.56 (2 H, m, CN2CH2), 2.38 (1 H, m, CH2NMe2), 2.10
(1 H, m, CH2NMe2), 1.99 (15 H, s, C5Me5), 1.84 (6 H, s, NMe2),
1.48 (9 H, s, CMe3), -0.05 (9 H, s, SiMe3). 13C{1H} NMR data
(C6D6, 125.7 MHz, 293 K): 175.75 (NC(O)O), 158.88 (CN2),
135.11 (i-C6H5), 129.17 (C6H5), 128.80 (C5Me5), 128.49, 128.37,
127.45, 126.92 (4 × CH of C6H5), 60.51 (CH2NMe2), 57.76
(CMe3), 48.22 (CN2CH2), 45.50 (NMe2), 31.30 (CMe3), 12.69
(C5Me5), 2.34 (SiMe3). IR data (KBr plates, Nujol mull, cm-1):
2726 (m), 2670 (w), 1686 (m), 1304 (m, br), 1260 (m, br), 1246
(m), 1156 (m, br), 1024 (m), 918 (m), 892 (m), 840 (m), 722 (s),
688 (s). Anal. Found (calcd): C, 59.9 (62.1); H, 8.0 (8.6); N, 9.8
(10.0).
IR data (KBr plates, Nujol mull, cm-1): 1516 (w), 1488 (s),
1350 (m), 1304 (w, br), 1246 (m), 1204 (w), 1198 (w), 1154 (w),
1098 (w), 1024 (m), 938 (m), 918 (m), 900 (m, br), 862 (m), 836
(m), 796 (w), 722 (m, br), 700 (w), 680 (w). Anal. Found (calcd
for C40H62N6O2Si2Ti2): C, 59.1 (59.2); H, 7.6 (7.7); N, 10.6 (10.4).
1H NMR data for 10a (C6D6, 500.1 MHz, 293 K): 7.03-7.14
(10 H, m, C6H5), 6.63 (2 H, m, 3- or 4-C5H4Me), 6.59 (2 H, m,
4- or 3-C5H4Me), 6.30 (2 H, m, 2- or 5-C5H4Me), 6.24 (2 H, m,
5- or 2-C5H4Me), 3.34 (2 H, m, CN2CH2), 3.17 (2 H, m,
CN2CH2), 2.65 (2 H, m, CH2NMe2), 2.33 (2 H, m, CH2NMe2),
2.04 (6 H, s, C5H4Me), 0.11 (18 H, s, SiMe3). 13C{1H} NMR data
for 10a (C6D6, 125.7 MHz, 293 K): 173.93 (CN2), 136.54
(i-C6H5), 128.68, 128.42, 128.29, 127.00, 126.96 (5 × CH of
C6H5), 114.25 (2- or 5-C5H4Me), 113.0-113.1 (3- and 4-C5H4Me),
112.18 (5- or 2-C5H4Me), 60.90 (CH2NMe2), 60.89 (CMe3), 47.00
(CN2CH2), 45.89 (C5H4Me), 15.75 (NMe), 2.84 (SiMe3), 1-C5H4Me
not observed. 1H NMR data (CD2Cl2, 500.1 MHz, 293 K):
7.48-7.39 (10H, m, br, C6H5), 6.30 (2H, m, 3- or 4-C5H4Me),
6.25 (4- or 3-C5H4Me), 6.04 (2H, m, 2-or 5-C5H4Me), 6.01 (2H,
m, 5- or 2-C5H4Me), 3.14 (2H, m, CN2CH2), 3.00 (2H, m,
CN2CH2), 2.43 (2H, m, CH2NMe2), 2.27 (2H, m, CH2NMe2),
2.21 (6H, s, C5H4Me), 2.02 (12H, s, NMe2), -0.11 (18H, s,
SiMe3).
Synthesis of [Ti(η-C5Me5){OC(O)NtBu}{Me3SiNC(Ph)-
NCH2CH2CH2NMe2}] (7). A solution of [Ti(η-C5H5)(NtBu)-
{Me3SiNC(Ph)NCH2CH2CH2NMe2}] (2; 100 mg, 18.9 mmol) in
pentane (10 mL) was exposed to ca. 1 atm of CO2 via a dual-
manifold Schlenk line. The solution was stirred, and volatiles
were removed under reduced pressure as soon as the brown
solution became cherry red (ca. 5 min). The resulting red
powder was stored at -30 °C in a drybox. Yield: 90 mg (82%).
1H NMR data (C6D6, 500.1 MHz, 293 K): 7.0-6.9 (5 H, m,
C6H5), 3.42 (2 H, m, CH2CH2CH2), 1.96 (15 H, s, C5Me5), 1.81
(6 H, s, NMe2), 1.75 (2 H, m, CH2NMe2), 1.61 (2 H, m, CN2CH2),
1.49 (9 H, s, CMe3), -0.05 (9 H, s, SiMe3). 13C{1H} NMR data
(C6D6, 125.7 MHz, 293 K): 175.31 (NC(O)O), 159.13 (CN2),
135.25 (i-C6H5), 129.12 (C6H5), 128.74 (C5Me5), 128.59, 128.51,
128.27, 126.85 (4 × CH of C6H5), 58.15 (CH2NMe2), 57.66
(CMe3), 48.17 (CH2CH2CH2), 45.30 (NMe2), 31.17 (CMe3), 30.24
(CN2CH2), 12.71 (C5Me5), 2.36 (SiMe3). IR data (KBr plates,
Nujol mull, cm-1): 2762 (w), 2722 (w), 2670 (w), 1666 (s, br),
1514 (m), 1356 (m), 1306 (w), 1246 (m), 1180 (w), 1158 (w),
1058 (w), 1024 (w), 1010 (w), 936 (m), 918 (w), 886 (m), 874
(m), 840 (w), 794 (w), 778 (w), 756 (w), 722 (m). Anal. Found
(calcd): C, 61.6 (62.7); H, 9.2 (8.8); N, 9.5 (9.8).
Synthesis of [Ti2(η-C5Me5)2(µ-O)2{Me3SiNC(Ph)NCH2-
CH2NMe2}2] (8). A solution of [Ti(η-C5Me5)(NtBu){Me3SiNC-
(Ph)NCH2CH2NMe2}] (1; 0.18 g, 0.35 mmol) in benzene (15
mL) was exposed to ca. 1 atm of CO2 via a dual-manifold
Schlenk line. The solution was stirred and left to stand at room
temperature for 4 days, during which time the brown solution
became red-brown and some precipitation of brown powder
was observed. Volatiles were removed under reduced pressure
to give 8 as a red-brown waxy solid. Yield: 0.13 g (74%).
Too insoluble to obtain NMR data. IR data (KBr plates,
Nujol mull, cm-1): 2207 (w), 1691 (w), 1559 (w), 1521 (w), 1246
(s), 1187 (w), 1156 (w), 1097(m), 1023 (m), 964 (m), 899 (s, br),
838 (s, br), 799 (w), 750 (w), 722 (m), 701 (w), 638 (w), 610
(w), 596 (w), 503 (w), 468 (w). Anal. Found (calcd for C48H78N6O2-
Si2Ti2): C, 55.7 (68.0); H, 9.4 (9.3); N, 8.3 (9.9).
1H NMR data for 10b (C6D6, 500.1 MHz, 293 K): 7.03-7.14
(10 H, m, C6H5), 6.63 (2 H, m, 3- or 4-C5H4Me), 6.59 (2 H, m,
4- or 3-C5H4Me), 6.30 (2 H, m, 2- or 5-C5H4Me), 6.24 (2 H, m,
5- or 2-C5H4Me), 3.34 (2 H, m, CN2CH2), 3.17 (2 H, m,
CN2CH2), 2.65 (2 H, m, CH2NMe2), 2.33 (2 H, m, CH2NMe2),
2.01 (6 H, s, C5H4Me), 0.88 (18 H, s, SiMe3). 13C{1H} NMR data
for 10b (C6D6, 125.7 MHz, 293 K): 173.82 (CN2), 136.45 (i-
C6H5), 128.64, 128.44, 127.91, 127.00, 126.84 (5 × CH of C6H5),
114.12 (2- or 5-C5H4Me), 113.0-113.1 (3- and 4-C5H4Me),
112.50 (5- or 2-C5H4Me), 60.90 (CH2NMe2), 60.89 (CMe3), 47.00
(CN2CH2), 45.84 (C5H4Me), 15.75 (NMe), 2.79 (SiMe3). 1-C5H4Me
not observed. 1H NMR data for 10b (CD2Cl2, 500.1 MHz, 293
K): 7.48-7.39 (10H, m, br, C6H5), 6.30 (2H, m, 3- or 4-C5H4Me),
6.25 (4- or 3-C5H4Me), 6.04 (2H, m, 2-or 5-C5H4Me), 6.01 (2H,
m, 5- or 2-C5H4Me), 3.14 (2H, m, CN2CH2), 3.00 (2H, m,
CN2CH2), 2.43 (2H, m, CH2NMe2), 2.27 (2H, m, CH2NMe2),
2.23 (6H, s, C5H4Me), 2.03 (12H, s, NMe2), -0.15 (18H, s,
SiMe3).
Synthesis of [Ti2(η-C5H4Me)2(µ-O)2{Me3SiNC(Ph)NCH2-
CH2CH2NMe2}2] (11). A solution of [Ti(η-C5H4Me)(NtBu)-
{Me3SiNC(Ph)NCH2CH2CH2NMe2}] (4) in benzene (150 mg,
0.32 mmol) was exposed to ca. 1 atm of CO2 via a dual-manifold
Schlenk line. The solution was stirred and immediately
changed color from red-brown to cherry red; after 5 days the
solution was brown and some precipitation of brown powder
was observed. Benzene was removed under reduced pressure
to give a red-brown solid which was shown by NMR to be a
mixture of two isomers, 11a and 11b. Attempts to recrystallize
a portion of this product from a mixture of CH2Cl2 and hexanes
failed to yield an analytically pure sample. Yield: 70 mg (50%).
IR data (KBr plates, Nujol mull, cm-1): 1515 (s), 1309 (m),
1261 (s), 1203 (m), 1154 (w), 1090 (w), 1024 (w), 964 (m), 919
(m), 902 (m), 871 (w), 839 (s), 798 (s), 752 (w), 723 (m), 701
(m), 509 (m). Anal. Found (calcd for C42H66N6O2Si2Ti2): C, 58.9
(60.1); H, 7.8 (7.9); N, 10.1 (10.0).
Synthesis of [Ti2(η-C5Me5)2(µ-O)2{Me3SiNC(Ph)NCH2-
CH2CH2NMe2}2] (9).
A
solution of [Ti(η-C5Me5)(NtBu)-
{Me3SiNC(Ph)NCH2CH2CH2NMe2}] (2; 234 mg, 0.44 mmol) in
15 mL of benzene was exposed to ca. 1 atm of CO2 via a dual-
manifold Schlenk line. The solution was stirred and left to
stand at room temperature for 4 days, after which it turned
brown and some precipitation of brown powder was observed.
Volatiles were removed under reduced pressure to give 9 as a
brown waxy solid. Yield: 0.13 g (64%).
Too insoluble for NMR data. IR data (KBr plates, Nujol
mull, cm-1): 1603 (m), 1532 (m), 1311 (m), 1261 (s), 1247 (s),
1155 (s, br), 1094 (s, br), 1010 (s, br), 981 (m), 912 (s), 839 (s),
723 (m), 701 (m), 630 (w), 609 (w), 475 (w). EI-MS: m/z 936
[M - Me]+ 29%. Anal. Found (calcd for C50H82N6O2Si2Ti2): C,
60.2 (63.1); H, 8.3 (8.7); N, 8.1 (8.8).

Benzamidinate Titanium Imido Compounds
Organometallics, Vol. 24, No. 10, 2005 2365
1H NMR data for 11a (C6D6, 500.1 MHz, 293 K): 7.03-7.22
(10 H, m, C6H5), 6.61 (2 H, m, 3- or 4- C5H4Me), 6.55 (2 H, m,
4- or 3-C5H4Me), 6.28 (2 H, m, 2- or 5-C5H4Me), 6.24 (2 H, m,
5- or 2-C5H4Me), 3.25 (2 H, m, CN2CH2), 3.05 (2 H, m,
CN2CH2), 2.36 (6 H, s, C5H4Me), 2.05 (6 H, s, NMe), 2.00 (4 H,
m, CH2NMe2), 1.67 (2 H, m, CH2CH2CH2), 1.42 (2 H, m,
CH2CH2CH2), 0.15 (18 H, s, SiMe3). 13C{1H} NMR data for 11a
(C6D6, 125.7 MHz, 293 K): 173.54 (CN2), 136.62 (i-C6H5 or
i-C5H4Me), 128.48, 128.10, 127.91, 126.93, 126.91 (5 × CH of
C6H5), 124.68 (1-C5H4Me), 113.99 (2- or 5-C5H4Me), 113.02 (3-
and 4-C5H4Me), 112.40 (5- or 2-C5H4Me), 57.82 (CH2NMe2),
55.12 (CMe3), 47.08 (CN2CH2), 45.50 (NMe), 31.46 (CH2CH2-
CH2), 15.79 (C5H4Me), 2.84 (SiMe3). 1H NMR data for 11a
(CD2Cl2, 500.1 MHz, 293 K): 7.12-7.50 (10 H, m, br, C6H5),
6.28 (2 H, m, 3- or 4-C5H4Me), 6.21 (2 H, m, 4- or 3-C5H4Me),
6.04 (4 H, m, 2- and 5-C5H4Me), 2.7-3.2 (4 H, m, CN2CH2),
2.27 (6 H, s, C5H4Me), 2.04 (12 H, s, NMe2), 2.00 (4 H, m,
CH2NMe2, overlapping), 1.60 (2 H, m, CH2CH2CH2), -0.10 (18
H, s, SiMe3).
(18H, s, SiMe3). 1H NMR data for 12b (C6D6, 500.1 MHz, 293
K): 7.0-7.2 (m, 6H, C6H5), 3.27 (m, 2H, CN2CH2), 3.11 (m, 2H,
CN2CH2), 2.34 (s, 6H, C5H4Me), 1.60 (4H, m, CH2CH3), 0.78
(6H, t, CH2CH3, 3J ) 7.3), 0.15 (18H, s, SiMe3). 13C{1H} NMR
data for 12b (C6D6, 125.7 MHz, 293 K): 173.58 (CN2), 136.64
(i-C6H5), 128.62, 127.92, 127.34, 126.87, 126.82 (C6H5), 114.02
(2- or 5-C5H4Me), 113.26 (3- or 4-C5H4Me), 112.73 (4- or
3-C5H4Me), 112.44 (5- or 2-C5H4Me), 57.86 (CN2CH2), 45.53
(C5H4Me), 29.84 or 29.66 (CH2CH3), 15.79 (CH2CH3), 2.80
(SiMe3), 1-C5H4Me not observed. 1H NMR data for 12b
(CD2Cl2, 500.1 MHz, 293 K): 7.4-7.5 (10H, m, C6H5), 6.21 (4H,
m, 3- and 4-C5H4Me), 6.03 (4H, s, br, 2- and 5-C5H4Me), 2.99
(2H, m, CN2CH2), 2.87 (2H, m, CN2CH2), 2.19 (6H, s, C5H4Me),
3
1.43 (4H, m, CH2CH3), 0.71 (6H, t, CH2CH3, J ) 7.3), -0.13
(18H, s, SiMe3).
NMR Tube Scale Synthesis of [Ti(η-C5H4Me){OC(O)-
NtBu}{Me3SiNC(Ph)NCH2CH2CH2NMe2}] (13). A solution
of [Ti(η-C5H4Me)(NtBu){Me3SiNC(Ph)NCH2CH2CH2NMe2}] (4;
20 mg, 4.2 µmol) in CD2Cl2 (0.5 mL) at -78 °C in a 5 mm NMR
tube was exposed to ca. 1 atm of CO2 via a dual-manifold
Schlenk line. The brown solution immediately became cherry
1H NMR data for 11b (C6D6, 500.1 MHz, 293 K): 7.0-7.1
(10 H, m, C6H5), 6.61 (2 H, m, 3- or 4-C5H4Me), 6.55 (2 H, m,
4- or 3-C5H4Me), 6.28 (2 H, m, 2- or 5-C5H4Me), 6.24 (2 H, m,
5- or 2-C5H4Me), 3.25 (2 H, m, CN2CH2), 3.05 (2H, m, CN2CH2),
2.35 (6 H, s, C5H4Me), 2.06 (6 H, s, NMe), 2.05 (4 H, m,
CH2NMe2), 1.67 (2 H, m, CH2CH2CH2), 1.42 (2 H, m, CH2CH2-
CH2), 0.12 (18 H, s, SiMe3). 13C{1H} NMR data for 11b (C6D6,
125.7 MHz, 289 K): 173.62 (CN2), 136.62 (i-C6H5 or i-C5H4Me),
128.61, 128.10, 127.91, 126.87 (5 × CH of C6H5), 124.68 (1-
C5H4Me), 114.52 (2- or 5-C5H4Me), 113.02 (3- and 4-C5H4Me),
112.02 (5- or 2-C6H4Me), 57.82 (CH2NMe2), 55.12 (CMe3), 47.08
(CN2CH2), 45.49 (NMe), 31.44 (CH2CH2CH2), 15.79 (C5H4Me),
2.79 (SiMe3), 1 × C6H5 obscured by solvent. 1H NMR data for
11b (CD2Cl2, 500.1 MHz, 293 K): 7.12-7.50 (10 H, m, br,
C6H5), 6.45 (2 H, m, 3- or 4-C5H4Me), 6.37 (2 H, m, 4- or
3-C5H4Me), 6.04 (4 H, m, 2- and 5-C5H4Me), 2.7-3.2 (4 H, m,
CN2CH2), 2.22 (6 H, s, C5H4Me), 2.01 (12 H, s, NMe2), 2.00 (4
H, m, CH2NMe2, overlapping), 1.60 (2 H, m, CH2CH2CH2), 1.35
(2 H, m, CH2CH2CH2), -0.12 (18 H, s, SiMe3).
Synthesis of [Ti2(η-C5H4Me)2(µ-O)2{Me3SiNC(Ph)NCH2-
CH2Me}2] (12). A solution of [Ti(η-C5H4Me)(NtBu){Me3SiNC-
(Ph)NCH2CH2Me}] (5; 0.29 g, 0.67 mmol) in benzene (30 mL)
was exposed to ca. 1 atm of CO2 via a dual-manifold Schlenk
line. The solution was stirred and immediately changed color
from red-brown to cherry red and after 15 min turned brown.
Volatiles were removed under reduced pressure to give a
brown oil, which was triturated with pentane (20 mL) to give
a brown waxy solid that was shown by NMR to be a mixture
of two isomers, 12a and 12b. Yield: 0.135 g (37%).
1
red. The H NMR spectrum recorded at -78 °C after 5 min
showed quantitative formation of 13.
1H NMR data (CD2Cl2, 500.1 MHz, 195 K): 7.40 (2H, app s,
br, o- or m-C6H5), 7.32 (1H, app s, br, p-C6H5), 7.16 (2H, app
d, o- or m-C6H5, app J ) 5.50), 6.84 (1H, s, 3- or 4-C5H4Me),
6.39 (2H, s, 2- or 5- and 4- or 3-C5H4Me), 6.09 (1H, s, 5- or
2-C5H4Me), 3.23 (1H, m, CN2CH2), 3.06 (1H, m, CN2CH2), 2.30
(3H, s, C5H4Me), 2.02 (1H, m, CH2NMe2), 1.89 (1H, m,
CH2NMe2), 1.53 (6H, s, NMe2), 1.41, (2H, m, CH2CH2CH2), 1.15
(9H, s, CMe3), -0.27 (9H, s, SiMe3). 13C{1H} NMR data
(CD2Cl2, 75.5 MHz, 195 K): 171.96 (NC(O)O), 158.33 (CN2),
132.79 (i-C6H5 or 1-C5H4Me), 131.99 (1-C5H4Me or i-C6H5),
128.93 (o- or m-C6H5), 128.34 (o- or m-C6H5), 128.04 (p-C6H5),
126.25 (o- or m-C6H5), 125.64 (o- or m-C6H5), 118.07 (2- or
5-C5H4Me), 117.48 (5- or 2- or 3- or 4-C5H4Me), 116.70 (5- or
2- or 3- or 4-C5H4Me), 116.04 (4- or 3-C5H4Me), 56.89
(CH2NMe2), 56.45 (CMe3), 47.27 (CN2CH2), 44.31 (NMe2), 29.32
(CMe3), 28.46 (CH2CH2CH2), 15.39 (C5H4Me), 0.76 (SiMe3).
NMR Tube Scale Synthesis of [Ti(η-C5H4Me){OC(O)-
NtBu}{Me3SiNC(Ph)NCH2CH2Me}] (14). A solution of [Ti(η-
C5H4Me)(NtBu){Me3SiNC(Ph)NCH2CH2Me}] (5; 23 mg, 5.3
µmol) in CD2Cl2 (0.5 mL) at -78 °C in a 5 mm NMR tube was
exposed to ca. 1 atm of CO2 via a dual-manifold Schlenk line.
1
The brown solution immediately became cherry red. The H
NMR spectrum recorded at -78 °C after 5 min showed
quantitative formation of 14.
1H NMR data (CD2Cl2, 500.1 MHz, 195 K): (7.30-7.42, 3H,
m, o- or m-C6H5 and p-C6H5), 7.05-7.08 (2H, m, o- or m-C6H5),
6.85 (1H, s, 3- or 4-C5H4Me), 6.44 (1H, s, 2- or 5- or 4- or
3-C5H4Me), 6.41 (1H, s, 2- or 5- or 4- or 3-C5H4Me), 6.11 (1H,
s, 5- or 2-C5H4Me), 3.09 (1H, m, CN2CH2), 2.94 (1H, m,
CN2CH2), 2.30 (3H, s, C5H4Me), 1.49 (1H, m, CH2Me), 1.34 (1H,
IR data (KBr plates, Nujol mull, cm-1): 1514 (m), 1488 (s),
1436 (s), 1356 (m), 1246 (m), 1184 (w), 1154 (w), 1106 (m),
1096 (m), 1036 (m), 962 (m), 902 (s), 870 (m), 836 (s), 798 (m),
752 (m), 702 (s), 620 (w). EI-MS: m/z 673 [M - C5H4Me]+ 25%,
m/z 233 [Me3SiNC(Ph)NCH2CH2CH2Me]+ 32%. Anal. Found
(calcd for C38H56N6O2Si2Ti2): C, 60.6 (60.6); H, 7.4 (7.5); N, 7.3
(7.4).
3
m, CH2Me), 1.15 (9H, s, CMe3), 0.60 (3H, t, CH3, J ) 7.32),
-0.29 (9H, s, SiMe3). 13C{1H} NMR data (CD2Cl2, 75.5 MHz,
195 K): 172.15 (NC(O)O), 158.50 (CN2), 133.12 (i-C6H5 or
1-C5H4Me), 132.30 (1-C5H4Me or i-C6H5), 128.93 (o-, m-, or
p-C6H5), 128.32 (o-, m-, or p-C6H5), 128.10 (o-, m-, or p-C6H5),
126.24 (o- or m-C6H5), 125.77 (o- or m-C6H5), 118.33 (2- or
5-C5H4Me), 117.53 (5- or 2- or 3- or 4-C5H4Me), 116.69 (5- or
2- or 3- or 4-C5H4Me), 116.22 (4- or 3-C5H4Me), 56.96 (CMe3),
51.47 (CN2CH2), 29.42 (CMe3), 24.64 (CH2CH3), 15.38 (CH2CH3),
11.06 (C5H4Me), 0.73 (SiMe3).
1H NMR data for 12a (C6D6, 500.1 MHz, 293 K): 7.2-7.0
(10H, m, C6H5), 6.62 (2H, m, 2- or 5-C5H4Me), 6.53 (2H, m, 5-
or 2-C5H4Me), 6.29 (2H, m, 3- or 4-C5H4Me), 6.23 (2H, m, 4-
or 3-C5H4Me), 3.27 (2H, m, CN2CH2), 3.11 (2H, m, CN2CH2),
2.35 (6H, s, C5H4Me), 1.60 (4H, m, CH2CH3), 0.65 (4H, t,
3
CH2CH3, J ) 13.2), 0.16 (18H, s, SiMe3). 13C{1H} NMR data
for 12a (C6D6, 125.7 MHz, 293 K): 173.64 (CN2), 136.62
(i-C6H5), 128.65, 128.45, 128.30, 126.94 (C6H5), 114.17 (2- or
5-C5H4Me), 113.01 (3- or 4-C5H4Me), 112.53 (4- or 3-C5H4Me),
112.06 (5- or 2-C5H4Me), 57.97 (CN2CH2), 47.11 (C5H4Me),
29.84 or 29.66 (CH2CH3), 15.79 (CH2CH3), 2.85 (SiMe3). 1 ×
C6H5 and 1-C5H4Me not observed. 1H NMR data for 12a
(CD2Cl2, 500.1 MHz, 293 K): 7.4-7.5 (10H, m, C6H5), 6.28 (4H,
m, 3- and 4-C5H4Me), 6.03 (4H, s, br, 2- and 5-C5H4Me), 2.99
(2H, m, CN2CH2), 2.87 (2H, m, CN2CH2), 2.19 (6H, s, C5H4Me),
NMR Tube Scale Synthesis of [Ti(η-C5H4Me){OC(O)-
NtBu}{Me3SiNC(Ph)NCH2CH2NMe2}] (15). A solution of
[Ti(η-C5H4Me)(NtBu){Me3SiNC(Ph)NCH2CH2NMe2}] (3; 24 mg,
5.2 µmol) in CD2Cl2 (0.5 mL) at -78 °C in an NMR tube
equipped with a J. Young valve was exposed to ca. 1 atm of
CO2 via a dual-manifold Schlenk line. The sample was warmed
to -35 °C until the 1H NMR spectrum showed quantitative
3
1.43 (4H, m, CH2CH3), 0.71 (6H, t, CH2CH3, J ) 7.3), -0.16

2366 Organometallics, Vol. 24, No. 10, 2005
Boyd et al.
Table 9. X-ray Data Collection and Processing
Parameters for [Ti(η-C5Me5){NtBuC(O)O}-
{Me3SiNC(Ph)NCH2CH2CH2NMe2}] (7),
trans-[Ti2(η-C5H4Me)2(µ-O)2{Me3SiNC(Ph)-
NCH2CH2NMe2}2] (10a), and
formation of 15 (ca. 10 min) and then cooled back down to -78
°C, at which temperature NMR data were collected.
1H NMR data (CD2Cl2, 500.1 MHz, 195 K): 7.43 (3H, app s,
br, o- or m-C6H5 and p-C6H5), 7.16 (2H, app d, o- or m-C6H5,
app J ) 21.48), 6.89 (1H, s, 3- or 4-C5H4Me), 6.44 (2H, s, 2- or
5- and 4- or 3-C5H4Me), 6.14 (1H, s, 5- or 2-C5H4Me), 3.22 (1H,
m, CN2CH2), 3.16 (1H, m, CN2CH2), 2.30 (3H, s, C5H4Me), 2.26
(1H, m, CH2NMe2), 2.19 (1H, m, CH2NMe2), 1.90 (6H, s, NMe2),
1.14 (9H, s, CMe3), -0.30 (9H, s, SiMe3). 13C{1H} NMR data
(CD2Cl2, 75.5 MHz, 195 K): 172.59 (NC(O)O), 158.37 (CN2),
133.54 (i-C6H5 or 1-C5H4Me), 131.99 (1-C5H4Me or i-C6H5),
129.01 (C6H5), 128.29 (C6H5), 128.04 (C6H5), 126.18 (C6H5),
125.99 (C6H5), 118.74 (2- or 5-C5H4Me), 117.27 (3- or 4-C5H4Me),
116.86 (5- or 2- or 4- or 3-C5H4Me), 116.68 (5- or 2- or 4- or
3-C5H4Me), 59.20 (CMe3), 56.92 (CH2NMe2), 47.42 (CN2CH2),
45.05 (NMe2), 29.56 (CMe3), 15.38 (C5H4Me), 0.77 (SiMe3).
Kinetic Studies of the Cycloaddition Reaction of 3
with CO2. A standard solution of [Ti(η-C5H4Me)(NtBu)-
{Me3SiNC(Ph)NCH2CH2NMe2}] (3) was prepared (11.5 mg,
25.0 µmol) with 1.5 equiv of 1,4-dimethoxybenzene (5.4 mg,
37.0 µmol) in CD2Cl2 (4.8 mL). Aliquots (0.6 mL) of this
solution were transferred to 5 mm NMR tubes equipped with
J. Young Teflon valves. The solutions were freeze-pump-
thawed three times and cooled to -78 °C. Samples were
exposed to CO2 at pressures of 0.25, 0.50, 0.75, and 1.0 atm
(>10-fold excess CO2 at the lowest pressure studied) via a
calibrated gas manifold. Samples were transferred to an NMR
spectrometer maintained at a probe temperature of -78 °C
trans-[Ti2(η-C5H4Me)2(µ-O)2{Me3SiNC(Ph)NCH2CH2
CH2NMe2}2] (11a)
7
10a
11a
empirical
formula
fw
temp/K
wavelength/Å
space group
a/Å
C
30H50N4O2-
C
40H62N6O2-
Si2Ti2
C
42H66N6O2-
SiTi
Si2Ti2
574.74
150
0.710 73
P21/n
9.0352(3)
20.1007(8)
18.2507(7)
90
104.140(2)
90
810.95
150
0.710 73
P1h
838.99
150
0.710 73
P1h
9.3597(2)
10.7886(2)
12.1749(3)
72.8419(8)
76.2046(8)
72.8192(9)
1106.74(4)
1
12.7045(4)
16.1774(4)
22.8120(8)
108.200(1)
94.168(1)
90.162(2)
4440.5
b/Å
c/Å
R/deg
â/deg
γ/deg
V/Å3
3214.2(2)
4
Z
4
d(calcd)/Mg m-3 1.19
1.21
1.26
0.45
R1 ) 0.0367
abs coeff/mm-1
R indices
0.33
R1 ) 0.0597
wR2 ) 0.0496 wR2 ) 0.0575 wR2 ) 0.0387
0.45
R1 ) 0.0522
(I > 3σ(I))a
2
a R1 ) ∑||Fo| - |Fc||/∑|Fo|; wR2 ) {∑w(|Fo| - |Fc|)2/∑(w|Fo| }1/2
.
atomic coordinates, bond lengths and angles, and displacement
parameters for 7, 10a, and 11a have been deposited at the
Cambridge Crystallographic Data Center. See the guidelines
for authors in the first issue of this year.
1
and then warmed to -35 °C. H NMR spectra were recorded
at 3 min intervals until >90% of 3 had been consumed (8 h
for 0.25 atm of CO2, 2 h for 1 atm of CO2).
Kinetic Studies of the Extrusion of tBuNCO from
Carbamate Compounds 13-15. Solutions of 3-5 at 1.98 mol
dm-3 concentration (5.2 mg (3), 5.4 mg (4), 4.9 mg (5), 11.4
µmol) with ca. 1.5 equiv of 1,4-dimethoxybenzene (2.3 mg, 16
µmol) in CD2Cl2 (0.6 mL) were prepared and transferred to 5
mm NMR tubes equipped with J. Young Teflon valves. The
solution was freeze-pump-thawed three times and cooled to
-78 °C. Samples were exposed to CO2 at ca. 1.0 atm pressure
(>10-fold excess CO2) via a dual-manifold Schlenk line.
Samples were transferred to an NMR spectrometer maintained
at -78 °C and warmed to -30 °C, at which point all of the
imido starting materials 3-5 had been consumed and only the
carbamate compounds 13-15 were present. 1H NMR spectra
were recorded at 3 min intervals until >90% of 13-15 had
been consumed (ca. 4 h for 13 and 14, ca. 2 h for 15).
Experiments were carried out in duplicate for 13 and 15 at
the same concentration. The “dummy-arm” system 14 was
examined at 1.9 × 10-8, 2.9 × 10-8, and 3.8 × 10-8 mol dm-3
concentrations (4.9 mg (11.4 µmol), 7.4 mg (17.4 µmol), and
9.8 mg (2.3 µmol), respectively, in 0.6 mL of CD2Cl2). N.B. In
the case of the two-carbon linker system 15, a small amount
of imido starting material 3 was still present at -30 °C and
the data obtained until this had been consumed were dis-
carded.
Computational Details. All calculations were performed
with the Gaussian 98 set of programs72 within the framework
of hybrid DFT (B3PW91)53,54 on the model systems CpTi(NMe)-
{MeC(NR)(NR′)} (Cp ) η5-C5H5; R ) R′ ) Me; R ) SiH3, R′ )
CH2CH2NMe2; R ) SiH3, R′ ) CH2CH2CH2NMe2) and their
reactions with CO2. The titanium atom was represented by
the relativistic effective core potential (RECP) from the
Stuttgart group (12 valence electrons) and its associated
(8s7p6d)/[6s5p3d] basis set,73 augmented by an f polarization
function (R ) 0.869).74 The silicium atom was represented by
RECP from the Stuttgart group and the associated basis set,75
augmented by a d polarization function.76 For the reference
system where R ) R′ ) Me, a 6-31G(d,p) basis set was used
for all the remaining atoms of the complex (C, H, N).77 For
(69) Otwinowski, Z.; Minor, W. Processing of X-ray Diffraction Data
Collected in Oscillation Mode; Academic press: New York, 1997.
(70) Altomare, A.; Cascarano, G.; Carrithers, J. R.; Betteridge, P.
W.; Cooper, R. I. J. Appl. Crystallogr. 1994, 27, 435.
(71) Watkin, D. J.; Prout, C. K.; Carruthers, J. R.; Betteridge, P.
W.; Cooper, R. I. CRYSTALS: Chemical Crystallography Laboratory,
Oxford, U.K., 2001.
(72) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J. R.; Zakrzewski, V. G.; Montgomery, J. A.,
Jr.; Stratmann, R. E.; Burant, J. C.; Dapprich, S.; Millam, J. M.;
Daniels, A. D.; Kudin, K. N.; Strain, M. C.; Farkas, O.; Tomasi, J.;
Barone, V.; Cossi, M.; Cammi, R.; Mennucci, B.; Pomelli, C.; Adamo,
C.; Clifford, S.; Ochterski, J.; Petersson, G. A.; Ayala, P. Y.; Cui, Q.;
Morokuma, K.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.;
Foresman, J. B.; Cioslowski, J.; Ortiz, J. V.; Stefanov, B. B.; Liu, G.;
Liashenko, A.; Piskorz, P.; Komaromi, I.; Gomperts, R.; Martin, R. L.;
Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.;
Gonzalez, C.; Challacombe, M.; Gill, P. M. W.; Johnson, B. G.; Chen,
W.; Wong, M. W.; Andres, J. L.; Head-Gordon, M.; Replogle, E. S.;
Pople, J. A. Gaussian 98; Gaussian, Inc.: Pittsburgh, PA, 1998.
(73) Andrae, D.; Ha¨ussermann, U.; Dolg, M.; Stoll, H.; Preuss, H.
Theor. Chim. Acta 1990, 77, 123.
Crystal Structure Determinations of [Ti(η-C5Me5)-
{NtBuC(O)O}-{Me3SiNC(Ph)NCH2CH2CH2NMe2}]
(7),
trans-[Ti2(η-C5H4Me)2(µ-O)2{Me3SiNC(Ph)NCH2CH2-
NMe2}2] (10a), and trans-[Ti2(η-C5H4Me)2(µ-O)2{Me3SiNC-
(Ph)NCH2CH2CH2NMe2}2] (11a). Crystal data collection and
processing parameters are given in Table 9. Crystals were
mounted on a glass fiber using perfluoropolyether oil and
cooled rapidly to 150 K under a stream of cold N2 using an
Oxford Cryosystems CRYOSTREAM unit. Diffraction data
were measured using an Enraf-Nonius KappaCCD check
diffractometer. Intensity data were processed using the DEN-
ZO package.69 The structures were solved with SIR92,70 and
subsequent full-matrix least-squares refinements were carried
out using CRYSTALS.71 Coordinates and anisotropic thermal
parameters of all non-hydrogen atoms were refined. Hydrogen
atoms were placed in calculated positions. Full listings of
(74) Ehlers, A. W.; Bohme, M.; Dapprich, S.; Gobbi, A.; Ho¨llwarth,
A.; Jonas, V.; Ko¨hler, K. F.; Stegmann, R.; Veldkamp, A.; Frenking,
G. Chem. Phys. Lett. 1993, 208, 111.
(75) Bergner, A.; Dolg, M.; Kuchle, W.; Stoll, H.; Preuss, H. Mol.
Phys. 1993, 30, 1431.
(76) Hollwarth, A.; Bohme, H.; Dapprich, S.; Ehlers, A. W.; Gobbi,
A.; Jonas, V.; Kohler, K. F.; Stagmann, R.; Veldkamp, A.; Frenking,
G. Chem. Phys. Lett. 1993, 203, 237.
(77) Hariharan, P. C.; Pople, J. A. Theor. Chim. Acta 1973, 28, 213.

Benzamidinate Titanium Imido Compounds
Organometallics, Vol. 24, No. 10, 2005 2367
the mixed system (R ) SiH3), only the nitrogen atoms (even
that of the chelated arm) and the carbon atoms of the Cp ring
were represented by a 6-31G(d,p) basis set; the remaining
atoms (C and H) were represented by a 6-31G basis set to save
computational resources. For all systems, CO2 was represented
by 6-31G(d,p) basis sets. The ONIOM calculations78 were
performed on the experimental systems 3, 3′, 4, and 4′ and
the Ph group as well as the Me groups on Si and on the imido
ligand were treated at the MM level using the UFF force
field.79 The QM part was treated at the B3PW91 level, and
the atoms were represented as described above. In both QM
and ONIOM calculations, full optimizations of geometry
without any constraint were performed, followed by analytical
computation of the Hessian matrix to confirm the nature of
the located extrema as minima or transition states on the
potential energy surface. The nature of the transition states
was checked by slightly perturbing the TS geometry along the
TS vector in both directions and optimizing the resulting
geometries as local minima to ensure the nature of the
connected intermediates. Gibbs free energy values are given
at T ) 298 K unless otherwise stated, and the values at other
temperatures were calculated using the freqchk utility in the
Gaussian 98 package.
Acknowledgment. We thank the EPSRC and the
CNRS for support and Professor J. D. Protasiewicz for
helpful discussions. E.C. and P.M. thank the Royal
Society of Chemistry for International Authors Travel-
ling Grants. We acknowledge the use of the EPSRC
National Service for Computational Chemistry Software
and the UK Computational Chemistry Facilty. We
thank the British Council for an Alliance grant.
Supporting Information Available: X-ray crystallo-
graphic files in CIF format for the structure determinations
of 9, 10a, and 11a, graphs used for the determination of kinetic
parameters, and tables giving further details of the DFT
calculations. This material is available free of charge via the
Internet at http://pubs.acs.org.
(78) Svensson, M.; Humbel, S.; Froese, R. D. J.; Matsubara, T.;
Sieber, S.; Morokuma, K. J. Phys. Chem. 1996, 100, 19357.
(79) Rappe´, A. K.; Casewitt, C. J.; Colwell, K. S.; Goddard, W. A.;
Skiff, W. M. J. Am. Chem. Soc. 1992, 114, 10024.
OM049026F