Titanium complexes with chiral amino alcohol ligands: Synthesis and structure of complexes related to hydroamination catalysts

Article abstract of DOI:10.1016/j.ica.2004.09.051

Titanium complexes with chiral amino alcohol ligands are useful precatalysts for the intramolecular hydroamination of aminoallenes. They can be synthesized via protonolysis of titanium dimethylamide starting materials with the free ligand. In most cases, the resulting materials are not isolable due to their oily nature. However, several complexes were prepared in pure form and isolated as solid materials. [Ti(Cl)(NMe₂)(-OCH₂CH(Ph) N(CHMe₂)-]₂ was prepared at room temperature from TiCl(NMe₂)₃ and the corresponding N-substituted d-amino alcohol; the dimeric nature of the complex was established by X-ray crystallography. [Ti(NMe₂)₂(-OCH₂CH(Ph)N(2-Ad)- )]₂ (2-Ad = 2-Adamantyl) was prepared from Ti(NMe₂) ₄ and the corresponding N-substituted l-amino alcohol after prolonged heating. An intermediate complex that could not be purified or isolated is believed to be Ti(NMe₂)₃(-OCH₂CH(Ph)NH(2-Ad)). Two complexes with the composition TiCl₂(-OCH₂CH(R*) N(CHMe₂)-)(HNMe₂) (where R* = CH₂Ph or CHMe₂) were prepared at room temperature by protonolysis of TiCl ₂(NMe₂)₂ with the corresponding N-substituted l-amino alcohols. These two complexes exhibit dynamic behavior on the NMR timescale that is believed to be a dimer-monomer equilibrium, but they decompose at elevated temperatures.

Full text of DOI:10.1016/j.ica.2004.09.051

Inorganica Chimica Acta 358 (2005) 687–694
www.elsevier.com/locate/ica
Titanium complexes with chiral amino alcohol ligands: synthesis
and structure of complexes related to hydroamination catalysts
a
Juliette R. Petersen , Jessica M. Hoover , W. Scott Kassel ,
a
b
c
Arnold L. Rheingold , Adam R. Johnson
a,
*
a
Department of Chemistry, Harvey Mudd College, 301 E. 12th Street, Claremont, CA 91711, USA
b
Department of Chemistry, Villanova University, Villanova, PA 19085, USA
c
Department of Chemistry and Biochemistry, University of California, San Diego, La Jolla, CA 92093, USA
Received 25 August 2004; accepted 28 September 2004
Available online 5 November 2004
Abstract
Titanium complexes with chiral amino alcohol ligands are useful precatalysts for the intramolecular hydroamination of amino-
allenes. They can be synthesized via protonolysis of titanium dimethylamide starting materials with the free ligand. In most cases,
the resulting materials are not isolable due to their oily nature. However, several complexes were prepared in pure form and isolated
as solid materials. [Ti(Cl)(NMe2)(–OCH2CH(Ph)N(CHMe2)–]2 was prepared at room temperature from TiCl(NMe2)3 and the cor-
responding N-substituted
[Ti(NMe2)2(–OCH2CH(Ph)N(2-Ad)–)]2 (2-Ad = 2-Adamantyl) was prepared from Ti(NMe2)4 and the corresponding N-substituted
-amino alcohol after prolonged heating. An intermediate complex that could not be purified or isolated is believed to be
D-amino alcohol; the dimeric nature of the complex was established by X-ray crystallography.
L
Ti(NMe2)3(–OCH2CH(Ph)NH(2-Ad)). Two complexes with the composition TiCl2(–OCH2CH(R*)N(CHMe2)–)(HNMe2) (where
R* = CH2Ph or CHMe2) were prepared at room temperature by protonolysis of TiCl2(NMe2)2 with the corresponding N-substi-
tuted L-amino alcohols. These two complexes exhibit dynamic behavior on the NMR timescale that is believed to be a dimer–mono-
mer equilibrium, but they decompose at elevated temperatures.
ꢀ 2004 Elsevier B.V. All rights reserved.
Keywords: Titanium; Crystal structures; Amino alcohol complexes
1. Introduction
Amide and mixed amide–alkoxide ligands show in-
creased coordinative unsaturation relative to Cp com-
plexes since both ligand types occupy only a single
coordination site at the metal per donor atom. These lig-
ands also provide highly tunable electronic and steric
environments to a reactive metal center.
The hydroamination reaction (direct addition of an
N–H bond across a C–C multiple bond) is one reaction
that is currently under wide investigation, as it is a
highly atom economical method of synthesizing substi-
tuted amines. The development and study of new lig-
ands has led to enhanced or altered catalyst
performance. The inter- and intramolecular hydroami-
nation of alkenes and alkynes has been the subject of
The organometallic chemistry of the early transition
metals is dominated by complexes containing the cyclo-
pentadienyl group (Cp) and its derivatives [1]. There is
substantial interest in alternative ligands, such as amides
[2–4], alkoxides [5], aryloxides [6–8], sulfonamides [9–
12], and amidinates [13], since the dramatically different
steric and electronic environments they provide can re-
sult in novel reactivity of the resulting complexes [14].
*
Corresponding author. Tel.: +1 909 607 8450; fax: +1 909 607
7577.
E-mail address: adam_johnson@hmc.edu (A.R. Johnson).
0020-1693/$ - see front matter ꢀ 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.ica.2004.09.051

688
J.R. Petersen et al. / Inorganica Chimica Acta 358 (2005) 687–694
several comprehensive reviews [15–21], including a re-
view of enantioselective catalytic hydroamination of
alkenes [22]. The intramolecular hydroamination of
aminoallenes has been the focus of less attention. Silver,
mercury and palladium catalysts were described more
than 20 years ago [23–26], organolanthanide complexes
have been studied at length by Marks and coworkers
[27–29], and Bergman recently investigated Cp, amide
and sulfonamide derived catalysts for the intramolecular
hydroamination of alkynes and allenes [30,31].
We recently reported the catalytic intramolecular
hydroamination of aminoallenes by in situ prepared
titanium complexes of a family of readily prepared N-,
O-donor ligands [32]. These ligands are prepared in a
few steps from the chiral amino acids valine (Val),
phenylalanine (Phe) and phenylglycine (Phg) [32,33].
The precatalysts, believed to have the structure
[TiL2(NMe2)2]2, can react with a terminal amine to form
an imido complex 1 (Scheme 1). The proposed mecha-
nistic step crucial for determining the product distribu-
tion is a [2 + 2] cycloaddition reaction of the transient
titanium imido complex 1 to form the corresponding
azametallacyclobutane 2. The mechanism of this and
the related intermolecular reaction catalyzed by tita-
nium has been studied in detail by Bergman [30,31,34–
38].
lated under an atmosphere of nitrogen using standard
Schlenk techniques, or in a glove box (MBraun UNI-
lab). The ligands D-H2PhgPrO, L-H2PhgAdO, L-H2Val-
PrO and L-H2PhePrO were prepared by the established
procedures [32,33]. Microanalyses were performed by
Desert Analytics (Tucson, AZ). All NMR spectra were
recorded at ambient temperature on a Bruker Avance
¨
400 spectrometer unless otherwise noted. Carbon assign-
ments were made using DEPT experiments. Melting
points were taken on a Meltemp melting apparatus
and are uncorrected.
2.2. Preparations
2.2.1. TiCl(
In the glove box,
D
-PhgPrO)(NMe2) (3)
-H2PhgPrO (0.502 g, 2.80 mmol),
D
was dissolved in ether (10 mL). This solution was added
to a solution of TiCl(NMe2)3 (0.606 g, 2.81 mmol) in
ether (5 mL) and the reaction mixture was allowed to
stir overnight, resulting in the formation of an orange
precipitate. The reaction mixture was concentrated and
the precipitate was collected via vacuum filtration
(0.260 g, 0.85 mmol, 30% yield). The crude product
was recrystallized from hexane and the resulting precip-
itate was collected by vacuum filtration giving 3 as a
bright orange powder. M.p. 191–192 ꢁC. 1H NMR
(400 MHz, C6D6): d = 7.37 (m, 3H, ArH), 7.25 (m,
2H, ArH), 5.299 (sept., J = 6.8 Hz, 1H, CHMe2); 5.25
(m, 2H), 4.522 (d, J = 4.8 Hz, 1H), 3.221 (s, 6H,
N(CH3)2), 1.222 (d, J = 6.8, 3H, CH3), 0.948 (d,
J = 6.4 Hz, 3H, CH3). 13C NMR (100 MHz, C6D6):
d = 146.0 (4ꢁ Ar), 129.0 (CH), 127.6 (CH), 127.2 (CH),
87.4 (CH2), 71.8 (CH), 54.9 (CH), 48.1 (CH3), 22.8
(CH3), 22.3 (CH3). Anal. Calc. for C13H21ClN2OTi: C,
51.25; H, 6.95; N, 9.20. Found: C, 51.00; H, 6.70; N,
8.94%.
We sought to understand the nature of our in situ cat-
alysts by initiating a synthetic study. The study was
hampered by the poor crystallinity properties of the Ti
complexes with most of the ligands. After much effort,
several complexes were prepared and isolated, allowing
a deeper understanding of the coordination chemistry
of titanium with these ligands.
2. Experimental
2.1. General details
2.2.2. Ti(
L
-PhgAdO)(NMe2)2 (4)
-H2PhgAdO (1.256 g, 4.63 mmol)
In the glove box,
L
All reagents were obtained from commercial suppli-
ers and were purified by standard methods [39], or used
as received. Solvents were degassed and distilled from
sodium/benzophenone and stored under nitrogen. All
air and/or moisture sensitive compounds were manipu-
was dissolved in toluene (20 mL) and added slowly to a
solution of Ti(NMe2)4 (1.053 g, 4.70 mmol) in toluene
(20 mL) in a Teflon valved glass reaction vessel. The ves-
sel was removed from glove box and heated in an oil
bath at 110 ꢁC overnight. Most of the solvent was re-
moved in vacuo and complex 4 was isolated as a fine yel-
low precipitate via vacuum filtration (0.195 g, 0.48
mmol, 10% yield). The material was recrystallized from
(Me3Si)2O to yield small yellow microcrystals. M.p.
L2Ti
N
L2Ti
N
1
231–232 ꢁC. H NMR (400 MHz, C6D6): d = 7.520 (m,
+ H2NR
R'
2a
2H, ArH), 7.358 (d, J = 7.2, 2H, ArH), 7.296 (m, 1H,
ArH), 5.005 (dd, J = 5.2, 9.2 Hz, 1H, CH), 4.777 (d,
J = 4.8 Hz, 1H, CH), 4.183 (d, J = 9.2 Hz, 1H, CH),
3.627 (br s, 1H, CH), 3.325 (s, 6H, N(CH3)2), 2.890 (s,
6H, N(CH3)2), 2.764 (br d, J = 12.4 Hz, 1H, CH), 2.3–
1.3 (m, 13H, Ad). 13C NMR (100 MHz, C6D6):
d = 147.2 (4ꢁ Ar), 128.5 (CH), 127.6 (CH), 126.8 (CH),
L2Ti(NMe2)2
·
- 2 HNMe2
L2Ti
N
R'
R'
1
2b
Scheme 1.

J.R. Petersen et al. / Inorganica Chimica Acta 358 (2005) 687–694
689
78.6 (CH2), 71.7 (CH), 68.6 (CH), 47.4 (CH3), 44.6
(CH3), 39.5 (CH2), 38.8 (CH2), 35.5 (CH), 33.0 (CH),
32.6 (CH2), 31.7 (CH), 28.7 (CH), 28.4 (CH). Anal. Calc.
for C22H35N3OTi: C, 65.18; H, 8.70; N, 10.37. Found:
C, 64.78; H, 8.55; N, 9.98%.
J = 6.8 Hz, 3H, CHMe2); 0.954 (d, J = 6.8 Hz, 3H,
CHMe2); 0.847 (d, J = 6.8 Hz, 3H, CHMe2); 0.758 (d,
J = 6.8 Hz, 3H, CHMe2). 13C NMR (100 MHz, CD2Cl2,
À25 ꢁC): d = 77.73 (CH2), 77.37 (CH2), 75.99 (CH),
75.34 (CH), 61.03 (CH), 58.18 (CH), 46.24 (HN(CH3)2),
43.71 (HN( CH3)2), 42.65 (HN(CH3)2), 42.57
(HN(CH3)2), 35.70 (CH), 33.35 (CH), 30.66 (CH3),
24.19 (CH3), 22.74 (CH3), 21.15 (CH3), 20.39 (CH3),
19.88 (CH3), 16.89 (CH3), 16.72 (CH3). Anal. Calc. for
C10H23Cl2N2OTi: C, 39.24; H, 7.57; N, 9.15. Found:
C, 38.95, H, 7.77, N, 9.51%.
2.2.3. TiCl2(
In the glove box,
L
-PhePrO)(HNMe2) (5a)
-H2PhePrO (0.225 g, 1.14 mmol)
L
was dissolved in ether (5 mL) and added to a stirring
solution of Cl2Ti(NMe2)2 (0.237 g, 1.15 mmol) in ether
(10 mL). A light brown precipitate formed immediately.
The solution was allowed to stir overnight at room tem-
perature and complex 5a was collected via vacuum fil-
tration (0.373 g, 1.05 mmol, 92%). M.p. 134–135 ꢁC.
NMR assignments at low temperature are based on
both 1-D and 2-D (COSY) spectra. 1H NMR (400
MHz, CD2Cl2, À5 ꢁC): d = 7.322 (m, 4H, ArH),
7.1234 (m, 2H, ArH), 7.173 (d, J = 6.8 Hz, 2H, ArH),
7.034 (d, J = 7.2 Hz, 2H, ArH), 6.402 (septet, J = 6.8
Hz, 1H, CH), 6.063 (septet, J = 6.8 Hz, 1H, CH),
4.950 (m, 2H, CH2 and NH), 4.704 (dd, J = 5.4, 11
Hz, 1H, CH), 4.546 (m, 2H, CH2 and NH), 4.318 (d,
J = 11.2, 1H, CH), 4.206 (dd, J = 6, 11 Hz, 1H, CH),
3.980 (m, 1H, CH), 2.8 (m, 3H), 2.899 (d, J = 6 Hz,
3H, HN(CH3)2), 2.682 (d, J = 6 Hz, 3H, HN(CH3)2),
2.632 (d, J = 5.6 Hz, 3H, HN(CH 3)2), 2.465 (t,
J = 12.4 Hz, 1H, CH 2), 2.268 (d, J = 5.6 Hz, 3H,
2.3. X-ray crystal structure determination for TiCl(D-
PhgPrO)(NMe2) (3)
Suitable crystals were obtained by prolonged storage
of a saturated ether solution at À35 ꢁC. An amber block
with approximate dimensions of 0.20 · 0.25 · 0.25 mm,
was used for the X-ray crystallographic analysis. Dif-
fraction intensity data were collected at 213(2) K using
a Bruker Smart Apex CCD diffractometer equipped
¨
with a graphite monochromator and a Mo Ka fine-focus
sealed tube. Crystal data and refinement parameters are
summarized in Table 2. The systematic absences in the
diffraction data are consistent with the monoclinic space
group options P21 and P21/m. E statistics suggested the
non-centrosymmetric space group, which yielded a
chemically reasonable and computationally stable re-
sult. The structure was solved using direct methods,
completed by subsequent difference Fourier syntheses,
and refined by full-matrix least-squares procedures.
The correct absolute structure was unambiguously
determined, Flack parameter = 0.002(12). Data were
corrected for absorption effects using SADABS [40]. The
ratio of minimum to maximum apparent transmission
was 0.753. All non-hydrogen atoms were refined with
anisotropic displacement coefficients, and hydrogen
atoms were treated as idealized contributions. All soft-
ware and sources of scattering factors are contained in
the SHELXTL (6.10) program package (G. Sheldrick,
Bruker XRD, Madison, WI).
HN(CH ) ), 1.794 (d, J = 6.8 Hz, 3H, CH3), 1.568 (d,
3 2
J = 6.8 Hz, 3H, CH3), 1.386 (d, J = 6.4 Hz, 3H, CH3),
1.281 (d, J = 6.4 Hz, 3H, CH3). 13C NMR (100 MHz,
C6D6, À5 ꢁC): d = 138.4 (4ꢁ), 137.1 (4ꢁ), 129.3 (CH),
129.2 (CH), 128.7 (CH), 127.5 (CH), 127.1 (CH), 81.8
(CH2), 78.8 (CH2), 71.0 (CH), 70.9 (CH), 63.1 (CH),
58.5 (CH), 47.0 (HN(CH3)2), 44.6 (HN(CH3)2), 42.4
(HN(CH3)2), 42.2 (HN(CH3)2), 41.0 (CH2), 39.4
(CH2), 25.9 (CH3), 22.8 (CH3), 21.9 (CH3), 17.1
(CH3). Anal. Calc. for C14H24Cl2N2OTi: C, 47.35; H,
6.81; N, 7.89. Found: C, 47.37; H, 6.70; N, 8.20%.
2.2.4. TiCl2(
In the glove box,
L
-ValPrO)(HNMe2) (5b)
-H2ValPrO (0.182 g, 1.25 mmol)
L
was dissolved in ether (5 mL) and added to a stirring
solution of Cl2Ti(NMe2)2 (0.273 g, 1.32 mmol) in ether
(12 mL). A dark brown precipitate formed immediately.
The solution was allowed to stir overnight at room tem-
perature and complex 5b was collected via vacuum fil-
tration (0.142 g, 0.40 mmol, 32%). M.p. 100–102 ꢁC.
1H NMR (400 MHz, CD2Cl2, À25 ꢁC): d = 6.3–6.4 (br
m, 2H, CHMe2); 5.2 (m, 2H); 4.7 (br septet, 1H,
CHMe2); 4.5 (m, 2H); 4.4 (br septet, 1H, CHMe2); 4.3
(br s, 1H, NH); 4.0 (br s, 1H, NH) 2.789 (d, J = 6.4
Hz, 3H, NMe); 2.612 (d, J = 6 Hz, 3H, NMe); 2.420
(d, J = 5.6 Hz, 3H, NMe); 2.396 (d, J = 5.6 Hz, 3H,
NMe); 1.664 (d, J = 6.8 Hz, 3H, CHMe2); 1.421 (d,
J = 6.8 Hz, 3H, CHMe2); 1.254 (d, J = 6.8 Hz, 3H,
CHMe2); 1.132 (d, J = 7.2 Hz, 3H, CHMe2); 1.009 (d,
3. Results and discussion
Titanium complexes with the amino alcohol ligands
shown in Fig. 1 (among others) were prepared in situ
for the intramolecular hydroamination of 1,3-disubsti-
tuted- and 1,1,3-trisubstituted aminoallenes [32]. The
precatalyst was postulated to be the dimeric [Ti(N-
Me2)2(L2)]2 based on the similar NMR resonances to
the related [TiCl(NMe2)(L2)]2 complexes previously re-
ported [33]. A synthetic study was undertaken in order
to more fully understand the coordination chemistry
of titanium with these ligands.

690
J.R. Petersen et al. / Inorganica Chimica Acta 358 (2005) 687–694
NH
OH
NH
OH
NH
OH
NH
OH
D-H2PhgPrO
L-H2PhgAdO
L-H2PhePrO
L-H2ValPrO
Fig. 1. Amino alcohol ligands used in this synthetic study.
The reaction of
ceeds cleanly at room temperature over the course of
several hours to form the dimeric species
D
-H2PhgPrO with TiCl(NMe2)3 pro-
axis perpendicular to the Ti1–O1–Ti2–O2 heterocycle.
The remaining bond distances and angles are
unremarkable.
[TiCl(NMe2)(
D
-PhgPrO)]2 (3), Scheme 2. The proton
The [TiCl(NMe2)(L)]2 complexes are not very good
catalysts for intramolecular hydroamination [42],
although the structures of the active precatalysts are
probably closely related. We therefore carefully moni-
tored reactions of the free ligands with Ti(NMe2)4 while
mimicking the reaction conditions of the catalytic reac-
tion. Solutions of the ligand (any of a number of the ami-
no alcohol ligands H2L reported previously [32]) and the
titanium starting material in C6D6 were mixed in J. Young
NMR tubes. A color change occurred immediately upon
mixing, and the formation of a single major (>90%) com-
plex was observed spectroscopically. Attempts to purify
the complex with any of the amino alcohol ligands have
been hampered by the fact that all complexes are oils.
Our assignment of the complex as Ti(NMe2)3(HL) is
based on the observation of three dimethylamide peaks
in the region from 3.0 to 3.6 ppm that integrate to a total
of 18 protons in the proton NMR spectrum.
Prolonged heating of the complex in the NMR tube
at 100–100 ꢁC for 12–18 h gives rise to a new complex
[Ti(NMe2)2(L)]2 with two sharp singlets each integrating
to six protons at about 3.3 and 2.9 ppm due to the dia-
stereotopic dimethylamide groups. The appearance of
HNMe2 at 2.3 ppm also confirms the proposed reaction.
The dimeric nature of the complex is not known with
certainty, but is postulated due to the similarity of the
NMR resonances. Attempts to isolate the proposed
complex were hampered by their non-crystallinity, ex-
cept with the H2PhgAdO ligand.
NMR resonances for this complex are very similar to
those reported for related amino alcohol complexes with
titanium; four protons corresponding to the protons a
to the heteroatoms appear in the region between 4.5
and 5.5 ppm. The solid state structure of complex 3
was determined by single crystal X-ray diffraction, and
the gross structural features of the complex are very sim-
ilar to those previously reported [33]. An ORTEP [41]
and atom numbering scheme of complex 3 is shown in
Fig. 2. Table 1 lists selected bond distances and angles
for the metal coordination environment, while crystallo-
graphic data collection parameters and refinement sta-
tistics are listed in Table 2.
There are two molecules of 3 in the asymmetric unit;
only one is shown for clarity. The two molecules are not
equivalent, and the major difference between the two
molecules is a rotation by about 20ꢁ of one of the phenyl
rings (C16–C21) in one of the ligands. The ligand binds
to titanium to form a five-membered metallocycle, and
the amino alcohol oxygen bridges the two titanium
centers. Each titanium has a highly distorted trigonal bipy-
ramidal coordination environment. The two titanium-
oxygen bond distances differ, with the Ti–Oaxial
(Ti1–O2 and Ti2–O1) bonds shorter by approximately
˚
0.03 A than the Ti–Oeq (Ti1–O1 and Ti2–O2) bonds.
Each titanium center in the complex contains a terminal
chloride and dimethylamide ligand. Each independent
molecule is nearly twofold symmetric, with the rotation
NMe2
Cl
Ti
O
Ti
N
+ TiCl(NMe2)3
-2 HNMe2
0.5
O
N
NH
OH
Me2N
Cl
3
Scheme 2.

J.R. Petersen et al. / Inorganica Chimica Acta 358 (2005) 687–694
691
Fig. 2. ORTEP and partial atom numbering scheme for one of the two independent molecules of complex 3. Ellipsoids are shown with 30%
probability. Hydrogen atoms are omitted for clarity.
Table 2
Crystallographic data collection parameters and refinement statistics
for 3
Table 1
Selected bond lengths (A) and angles (ꢁ) for 3
˚
Ti1–O1
Ti1–O2
Ti1–N1
Ti1–Cl
Ti1–N2
2.0337(15)
1.9943(15)
1.8927(18)
2.3100(7)
Ti2–O2
Ti2–O1
Ti2–N3
Ti2–Cl2
Ti2–N4
2.0190(15)
1.9958(15)
1.9039(17)
2.3153(7)
Empirical formula
Formula weight
Color, habit
C26H42Cl2N4O2Ti2
609.34
amber, block
0.20 · 0.25 · 0.25
Monoclinic
P2(1)
Approximate crystal size (mm)
Crystal system
Space group
1.8628(18)
1.8693(19)
N1–Ti1–N2
N1–Ti1–O2
N2–Ti1–O2
N1–Ti1–O1
N2–Ti1–O1
O2–Ti1–O1
N1–Ti1–Cl1
N2–Ti1–Cl1
O2–Ti1–Cl1
O1–Ti1–Cl1
102.53(8)
N3–Ti2–N4
N3–Ti2–O1
N4–Ti2–O1
N3–Ti2–O2
N4–Ti2–O2
O1–Ti2–O2
N3–Ti2–Cl2
N4–Ti2–Cl2
O1–Ti2–Cl2
O2–Ti2–Cl2
104.80(8)
143.88(7)
103.10(8)
78.02(7)
108.33(8)
70.00(6)
103.31(6)
107.06(7)
92.99(5)
143.34(5)
140.80(7)
106.63(8)
78.10(7)
108.21(8)
70.27(6)
100.84(6)
105.08(7)
93.02(5)
145.80(5)
˚
a (A)
13.7315(9)
13.5649(9)
17.1022(11)
99.7130(10)
3139.9(4)
4
˚
b (A)
˚
c (A)
b (ꢁ)
3
˚
V (A )
Z
T (ꢁC)
213(2)
1.289
1280
Dcalc (g/cm3)
F(000)
Radiation
MoKa
0.71073
Ti1–O1–Ti2
107.59(7)
Ti1–O2–Ti2
108.23(7)
l (mmÀ1
)
Scan mode
h Range (ꢁ)
Index ranges
/ and x
1.76–28.20
À18 6 h 6 18
À18 6 k 6 18
À21 6 l 6 22
24,240
A preparative scale reaction between
L
-H2PhgAdo
and Ti(NMe2)4 was carried out in toluene at 110 ꢁC as
shown in Scheme 3. After 18 h, the complex [Ti(N-
Me2)2(L-PhgAdO)]2 (4) was isolated in 10% yield. The
Total number of reflections
Number of unique reflections
Number of reflections with I > 2r(I)
Rint
12,901
12,001
0.0235
low yield is due to the extremely high solubility of the
complex in all common solvents; the reaction is essen-
tially quantitative by NMR spectroscopy. The complex
exhibits two sharp singlets integrating to six protons at
3.325 and 2.890 ppm due to the diastereotopic dimethyl-
amide groups, and there are four well spaced multiplets
between 3.8 and 5.2 ppm due to the protons a to the
heteroatoms in the ligand. The complex has resisted
crystallization attempts to date and has only been iso-
lated as fine microcrystals or as thin fibers that do not
diffract sufficiently for structural determination.
Data/restraints/parameters
Final R indices [I > 2r(I)]
Weighted R [I > 2r(I)]
Goodness-of-fit
12,901/1/665
0.0312
0.0786
1.012
À0.158, 0.376
3
˚
Residual density (eA )
Complex 4 is stable indefinitely at 100–110 ꢁC and is
presumably the active precatalyst for the hydroamina-
tion reaction. Importantly, it has two dimethylamide

692
J.R. Petersen et al. / Inorganica Chimica Acta 358 (2005) 687–694
Me2N
NMe2
N
2-Ad
Ti
O
Ti
+ Ti(NMe2)4
0.5
O
N
-2 HNMe2
2-Ad
NH
OH
Me2N
NMe2
4
Scheme 3.
groups that can be substituted with the aminoallene to
form the transient imido complex.
the resonances sharpen into clean signals. The spec-
trum at this temperature exhibits two resonances of
equal intensity for each proton in the molecule. There
is no further change in the spectra at lower
temperatures.
We postulated that the different reactivity between
the amino alcohol ligands and Ti(NMe2)4, which re-
quires extended heating for ligand replacement, and
TiCl(NMe2)3, which is complete in a few hours, is due
to an electronic effect. The presence of chlorine atoms
on titanium increases the Lewis acidity of the metal,
and thus the rate of reaction with the Lewis basic lig-
ands. We therefore chose to examine the reactivity of
TiCl2(NMe2)2 with the amino alcohol ligands. There is
a dramatic difference in reactivity with this starting
material; a light brown product forms and precipitates
The complex is so insoluble in toluene that accurate
peak assignments could not be made, but the complex
exhibits the same fluxional behavior at approximately
the same temperatures in deuterated methylene chloride.
Assignment of the proton NMR spectrum and confir-
mation of the molecular formula was aided with a
heteronuclear COSY experiment. The complex has four
inequivalent isopropyl methyl resonances appearing as
doublets at 1.281, 1.386, 1.568 and 1.794 ppm. The
two isopropyl methyne protons appear as clean and
well-resolved septets at 6.063 and 6.402 ppm. The
dimethylamine methyl resonances appear as doublets
at 2.268, 2.632, 2.682 and 2.899 ppm. This splitting is
due to the NH proton, confirming the structural assign-
ment. The amine NH resonances appear as broad reso-
nances, overlapping other peaks at 4.5 and 4.9 ppm. The
two resonances for each proton in the molecule could be
due to the presence of two inequivalent molecules with
equal population at the lower temperature. However,
based on the relatively high temperatures for coales-
cence, effectively ruling out simple pseudorotation
[44,45], as well as the structure of the related complex
3, we postulate that the dynamic behavior is due to a di-
mer/monomer equilibrium.
immediately upon mixing etherial solutions of L-H2Phe-
PrO and TiCl2(NMe2)2, as shown in Scheme 4. The
complex that is isolated (complex 5a) is extremely insol-
uble in non-polar solvents (benzene, hexane, ether)
although it has limited solubility in toluene. The com-
plex is quite soluble in thf and methylene chloride, but
attempts to crystallize the complex from these solvents
have been unsuccessful to date. Fortunately, the com-
plex as isolated from the reaction mixture is analytically
pure, and has the formula TiCl2(NHMe2)(L-PhePrO).
The ability of Ti(IV) centers to tenaciously retain amine
substituents has been reported previously [43].
The proton NMR spectrum of complex 5a in tolu-
ene-d8 exhibits extremely broad resonances at 20–25
ꢁC. When the solution is heated, the broad resonances
begin to coalesce to sharper signals, but the complex
also decomposes to unknown products at temperatures
around 50 ꢁC. Upon cooling the solution to À5 ꢁC,
The complicated NMR spectrum and decomposi-
tion at elevated temperature have hampered our ability
NHMe2
Cl
R
O
Ti
- HNMe2
+ TiCl2(NMe2)2
Cl
NH
OH
N
R
5a, R = CH2Ph
5b, R = CHMe2
Scheme 4.

J.R. Petersen et al. / Inorganica Chimica Acta 358 (2005) 687–694
693
to fully simulate and interpret the dynamic behavior of
the complex. We therefore prepared a derivative with
simpler NMR signals by mixing L-H2ValPrO with
complex 4. We are also continuing to model the dynamic
behavior of complexes 5a and 5b.
TiCl2(NMe2)2 in ether (Scheme 4). A dark red–brown
precipitate formed immediately, which was collected
by filtration. Complex 5b is significantly more soluble
in ether than complex 5a, and the yield of the complex
is correspondingly lower. 5b is also very soluble in thf
and methylene chloride. The proton NMR spectrum of
complex 5b in CD2Cl2 exhibits broad resonances at
20–25 ꢁC, again possibly indicating fluxional behavior.
When the solution is heated, the broad resonances be-
gin to coalesce to sharper signals, but the complex
decomposes to unknown products at temperatures
much above room temperature. Upon cooling the solu-
tion, the resonances begin to sharpen, and the low
temperature limiting spectrum is at À25 ꢁC. The spec-
trum at this temperature again exhibits two resonances
of equal intensity for each proton in the molecule, and
there is no further change in the spectra at lower
temperatures.
Complex 5b has eight inequivalent isopropyl methyl
resonances appearing as doublets at 1.664, 1.421,
1.254, 1.132, 1.009, 0.954, 0.847, and 0.758 ppm. Two
isopropyl methyne protons appear downfield as a
broadened resonance at 6.3 ppm and the others are sep-
arated, but broad, at 4.7 and 4.3 ppm. The four dimeth-
ylamine methyl resonances appear as doublets at 2.789,
2.612, 2.420 and 2.396 ppm. The splitting of the N-me-
thyl resonances by the NH proton confirms the presence
of the coordinated amine. The amine NH resonances ap-
pear as broad resonances at 4.0 and 4.3 ppm.
Acknowledgements
The authors gratefully acknowledge the financial sup-
port of the Harvey Mudd College Research Fund, the
Donors of the American Chemical Society Petroleum
Research Fund and the National Science Foundation
(REU-CHE-0097262).
Appendix A. Supplementary data
CCDC 246787 contains the supplementary crystallo-
graphic data for this paper. These data can be obtained
free of charge via www.ccdc.cam.ac.uk/data_request/cif,
by emailing data_request@ccdc.cam.ac.uk, or by con-
tacting The Cambridge Crystallographic Data Centre,
12, Union Road, Cambridge CB2 1EZ, UK; fax: +44
1223 336033. Supplementary data associated with this
article can be found, in the online version, at
doi:10.1016/j.ica.2004.09.051.
References
[1] E.W. Abel, F.G.A. Stone, G. Wilkinson (Eds.), Comprehensive
Organometallic Chemistry II, vol. 4, Elsevier Science Ltd., New
York, 1995.
[2] C.C. Cummins, Chem. Commun. (1998) 1777.
[3] C.C. Cummins, Prog. Inorg. Chem. 47 (1998) 685.
[4] R.R. Schrock, Acc. Chem. Res. 30 (1997) 9.
[5] P.T. Wolczanski, Polyhedron 14 (1995) 3335.
[6] I.P. Rothwell, Chem. Commun. (1997) 1331.
[7] I.P. Rothwell, Acc. Chem. Res. 21 (1988) 153.
[8] M.G. Thorn, P.E. Fanwick, I.P. Rothwell, Organometallics 18
(1999) 4442.
Although complex 5b was prepared in order to sim-
plify the NMR signals for interpretation of the dynamic
behavior, a full simulation and analysis has not yet been
possible. Detailed interpretation of the NMR signals for
5b is also hampered by its thermal instability.
[9] E. Royo, J.M. Betancort, T.J. Davis, P. Carroll, P.J. Walsh,
Organometallics 19 (2000) 4840.
´
[10] L.T. Armistead, P.S. White, M.R. Gagne, Organometallics 17
4. Summary and conclusions
(1998) 4232.
[11] L.T. Armistead, P.S. White, M.R. Gagne, Organometallics 17
Titanium complexes with amino alcohol ligands have
been previously shown to be effective catalysts for the
intramolecular hydroamination of aminoallenes. In or-
der to more fully understand the catalytically active spe-
cies in the reaction, we investigated the synthesis of a
variety of titanium complexes with these ligands. Most
of the complexes have poor solubility and crystallinity
behavior. We have prepared and purified three discrete
titanium(IV) complexes with the general formulas
[Ti(L)(NMe2)2]2, [Ti(L)(Cl)(NMe2)]2 and Ti(L)Cl2 with
chiral amino alcohol ligands derived from phenylala-
nine, phenylglycine and valine. We are continuing our
efforts to understand the coordination chemistry of tita-
nium with these ligands. Studies are currently underway
in order to trap a terminal imido complex starting from
´
(1998) 216.
[12] S. Pritchett, P. Gantzel, P.J. Walsh, Organometallics 18 (1999)
823.
[13] J.R. Hagadorn, J. Arnold, Organometallics 17 (1998) 1355.
[14] A special issue of the Journal of Organometallic Chemistry is
devoted to p-donor ligands P. Mountford, J. Organometal. Chem.
591 (1999) 1.
[15] T.E. Muller, M. Beller, Chem. Rev. 98 (1998) 675.
¨
[16] M. Nobis, B. Driessen-Ho¨lscher, Angew. Chem., Int. Ed. 40
(2001) 3983.
[17] F. Alonso, I.P. Beletskaya, M. Yus, Chem. Rev. 104 (2004) 3079.
[18] F. Pohlki, S. Doye, Chem. Soc. Rev. 32 (2003) 104.
[19] I. Nakamura, Y. Yamamoto, Chem. Rev. 104 (2004) 2127.
[20] I. Bytschkov, S. Doye, Eur. J. Org. Chem. (2003) 935.
[21] Y. Yamamoto, U. Radhakrishnan, Chem. Soc. Rev. 28 (1999)
199.

694
J.R. Petersen et al. / Inorganica Chimica Acta 358 (2005) 687–694
[22] P.W. Roesky, T.E. Muller, Angew. Chem., Int. Ed. 42 (2003)
¨
2708.
[35] A.M. Baranger, P.J. Walsh, R.G. Bergman, J. Am. Chem. Soc.
115 (1993) 2753.
[23] S. Arseniyadis, J. Gore, Tetrahedron Lett. 24 (1983) 3997.
[24] D.N.A. Fox, T. Gallagher, Tetrahedron 46 (1990) 4697.
[25] R. Kinsman, D. Lathbury, P. Vernon, T. Gallagher, J. Chem.
Soc., Chem. Commun. (1987) 243.
[36] J.L. Polse, R.A. Andersen, R.G. Bergman, J. Am. Chem. Soc. 120
(1998) 13405.
[37] J.S. Johnson, R.G. Bergman, J. Am. Chem. Soc. 123 (2001)
2923.
[26] M. Meguro, Y. Yamamoto, Tetrahedron Lett. 39 (1998) 5421.
[27] V. Arredondo, F.E. McDonald, T.J. Marks, Organometallics 18
(1999) 1949.
[38] B.F. Straub, R.G. Bergman, Angew. Chem., Int. Ed. 40 (2001)
4632.
[39] D.D. Perrin, L.F. Armarego, Purification of Laboratory Chem-
icals, Pergamon Press, New York, 1988.
[40] G.M. Sheldrick, Bruker/Siemens Area Detector Absorption
[28] V. Arredondo, F.E. McDonald, T.J. Marks, J. Am. Chem. Soc.
120 (1998) 4871.
¨
[29] V. Arredondo, S. Tian, F.E. McDonald, T.J. Marks, J. Am.
Chem. Soc. 121 (1999) 3633.
Correction Program (2.01), Bruker AXS, Madison, WI, USA,
¨
1998.
[30] L. Ackermann, R.G. Bergman, Org. Lett. 4 (2002) 1475.
[31] L. Ackermann, R.G. Bergman, R.N. Loy, J. Am. Chem. Soc. 125
(2003) 11956.
[41] L.J. Farrugia, J. Appl. Cryst. 30 (1997) 565.
[42] J.M. Hoover, A.R. Johnson, unpublished results.
[43] K. Kirschbaum, O. Conrad, D.M. Giolando, Acta Crys., C C56
(2000) e541.
[32] J.M. Hoover, J.R. Petersen, J.H. Pikul, Organometallics 23 (2004)
4614–4620.
[44] L.M. Jackman, F.A. Cotton (Eds.), Dynamic Nuclear Magnetic
Resonance Spectroscopy, vol., Academic Press, New York,
1975.
[33] C.K. Ho, A.D. Schuler, C.B. Yoo, S.R. Herron, K.A. Kan-
tardjieff, A.R. Johnson, Inorg. Chim. Acta 341 (2002) 71.
[34] P.J. Walsh, A.M. Baranger, R.G. Bergman, J. Am. Chem. Soc.
114 (1992) 1708.
[45] J. Sandstro¨m, Dynamic NMR Spectroscopy, Academic Press,
London, 1982.