Titanium hydrazides supported by diamide-amine and related ligands: A combined experimental and DFT study

Organometallics 2008, 27, 6479–6494

6479

Titanium Hydrazides Supported by Diamide-Amine and Related

Ligands: A Combined Experimental and DFT Study

Jonathan D. Selby,^†Catherine D. Manley,^†Andrew D. Schwarz,^†Eric Clot,*^,‡and

Philip Mountford*^,†

Chemistry Research Laboratory, UniVersity of Oxford, Mansﬁeld Road, Oxford OX1 3TA, U.K., and

Institut Charles Gerhardt Montpellier, UMR 5253 CNRS-UM2-UM1-ENSCM, cc 1501, Place Euge`ne

Bataillon, 34095 Montpellier Cedex 5, France

ReceiVed August 25, 2008

This paper reports a general method for the synthesis of new terminal titanium diphenyl hydrazido(2-)

complexes containing dianionic N₃- and N₄-donor ligands, along with new hydrazido synthons. Reaction

t

of Ti(N^tBu)Cl₂(py)₃or Ti(N^tBu)Cl₂(py′)₃(py′ ) 4-NC₅H₄Bu) with Ph₂NNH₂gave excellent yields of

the corresponding monomeric hydrazides Ti(NNPh₂)Cl₂(L)₃(L ) py (7) or py′), which have been

structurally characterized. Application of a dynamic vacuum to 7 formed [Ti(NNPh₂)Cl₂(py)₂]₂(4). Both

4 and 7 are entry points to new titanium hydradizo complexes on reaction with metalated reagents. In

this way, four new ﬁve-coordinate diamide-amine complexes Ti(NNPh₂)(“N₂N”)(py) were made (“N₂N”

)

MeN(CH₂CH₂NSiMe₃)₂,

Me₃SiN(CH₂CH₂NSiMe₃)₂,

MeN(CH₂CH₂CH₂NSiMe₃)₂,

(2-

NC₅H₄)C(Me)(CH₂NSiMe₃)₂) and structurally characterized. Five- and six-coordinate terminal titanium

hydrazides containing dianionic N₄- or O₂N₂-donor ligands were also synthesized from 4 by an analogous

method. The identity of the “N₂N” ligand affects the TidN_Rand N_R-N_ꢀdistances of the TidN-NPh₂

functional group. A detailed DFT analysis of the bonding in these and a range of model complexes is

presented using molecular orbital and natural bond orbital methods. The competition between N(amide)

and N(hydrazide) Ti(3d_π)-N(2p_π) interactions has an indirect and signiﬁcant effect on the N_R-N_ꢀbond.

of the biological and synthetic ﬁxation and activation of

dinitrogen.^7-15Chatt, Pickett, and co-workers showed experi-

mentally that the conversion of N₂to NH₃starting from highly

reducing zerovalent group 6 tetraphosphine complexes proceeds

via (L)M-NNH₂species.^16-19Using a tris(amide)-amine ligand

platform, Schrock et al. demonstrated the catalytic conversion

of N₂to NH₃, again via a Mo-NNH₂intermediate, employing

sequential reduction and protonation steps.^10,20Tuczek has

reported detailed DFT studies of both the Chatt²¹and Schrock²²

catalytic cycles, conﬁrming the importance of M-NNH₂species

and the reduction-assisted N_R-N_ꢀbond cleavage. The stoichio-

metric reactions of group 6 M-NNH₂species with organic

substrates have also been comprehensively studied. Reactions

typically take place at the terminal ꢀ-NH₂group, via condensa-

tion reactions or insertion into N-H.^13-15Subsequent reductive

N_R-N_ꢀbond cleavage can yield organic fragments such as

Introduction

Group 4 imido complexes (L)MdNR (R is usually alkyl or

aryl, L is a supporting ligand (set)) have been a focus of ongoing

activity for nearly 20 years.^1-6As well as being of interest from

the point of view of fundamental bonding and reactivity

(typically focusing on the MdNR bond), practical applications

have been found in the catalytic hydroamination of C-C

multiple bonds, oleﬁn polymerization, and metal organic chemi-

cal vapor deposition. In contrast, the corresponding chemistry

of terminal hydrazide complexes (L)MdNNR₂(R is alkyl or

aryl) has barely been explored for these early metals. As

summarized below, group 4 hydrazide chemistry has nonetheless

already established the potential for exciting and signiﬁcant

departures from the norms expected from the imido analogues.

However, there are currently no guiding principles for selecting

the most appropriate supporting ligand sets to maximize the

potential of this chemistry. This paper reports the synthesis and

electronic structures of a new class of titanium hydrazide

complex for advancing the chemistry of the TidNNR₂func-

tional group.

(7) Fryzuk, M. D. Chem. Rec. 2003, 3, 2.

(8) Kozak, C. M.; Mountford, P. Angew. Chem., Int. Ed. 2004, 43, 1186.

(9) MacKay, B. A.; Fryzuk, M. D. Chem. ReV. 2004, 104, 385.

(10) Schrock, R. R. Acc. Chem. Res. 2005, 38, 955.

(11) Chirik, P. J. Dalton Trans. 2007, 16.

From a more general perspective, the chemistry of metal-

bound NNR₂species is of considerable interest in the context

(12) Ohki, Y.; Fryzuk, M. D. Angew. Chem., Int. Ed. 2007, 46, 3180.

(13) Fryzuk, M. D.; Johnson, S. A. Coord. Chem. ReV. 2000, 200-

202, 379.

(14) Hidai, M. Chem. ReV. 1995, 95, 1115.

* Corresponding authors. E-mail: philip.mountford@chem.ox.ac.uk;

clot@univ-montp2.fr.

(15) Hidai, M. Coord. Chem. ReV. 1999, 185-186, 99.

(16) Chatt, J.; Pearman, A. J.; Richards, R. L. Nature 1975, 253, 39.

(17) Pickett, C. J.; Leigh, G. J. J. Chem. Soc., Chem. Commun. 1981,

1033.

^†University of Oxford.

^‡Institut Charles Gerhardt Montpellier.

(1) Wigley, D. E. Prog. Inorg. Chem. 1994, 42, 239.

(2) Duncan, A. P.; Bergman, R. G. Chem. Rec. 2002, 2, 431.

(3) Odom, A. L. Dalton Trans. 2005, 225.

(4) Hazari, N.; Mountford, P. Acc. Chem. Res. 2005, 38, 839.

(5) Bolton, P. D.; Mountford, P. AdV. Synth. Catal. 2005, 347, 355.

(6) Fout, A. R.; Kilgore, U. J.; Mindiola, D. J. Chem.-Eur. J. 2007,

13, 9428.

(18) Hussain, W.; Leigh, G. J.; Pickett, C. J. J. Chem. Soc., Chem.

Commun. 1982, 747.

(19) Pickett, C. J.; Talamarin, J. Nature 1985, 317, 652.

(20) Yandulov, D. V.; Schrock, R. R. Science 2003, 76.

(21) Stephan, G. C.; Sivasankar, C.; Studt, F.; Tuczek, F. Chem.-Eur.

J. 2008, 14, 644.

(22) Studt, F.; Tuczek, F. Angew. Chem., Int. Ed. 2005, 44, 5639.

10.1021/om8007597 CCC: $40.75

2008 American Chemical Society

Publication on Web 11/14/2008

6480 Organometallics, Vol. 27, No. 24, 2008

Selby et al.

pyrroles and pyridines. Notably, direct reaction of the M-N_R

bond with unsaturated substrates is not seen. Furthermore,

structural,^23,24spectroscopic,²⁵and DFT computational^21,22,26,27

data all suggest that the NNR₂ligand in these group 6 complexes

is best viewed as a neutral isodiazene ligand (:NdNR₂) rather

than a hydrazide(2-) species [N-NR₂]^2-

.

The ﬁrst reported terminal titanium hydrazide was

Cp₂Ti{NN(SiMe₃)₂}, prepared from Cp₂TiCl₂and the thermally

sensitive N₂(SiMe₃)₂(dec > -35 °C).²⁸There have been no

reactivity or structural reports for this system, and the incon-

venient nature of N₂(SiMe₃)₂may have limited its development.

Bergman reported and structurally characterized the ﬁrst zir-

conium hydrazide, namely, Cp₂Zr(NNPh₂)(DMAP),²⁹and Odom

was the ﬁrst to report a structurally authenticated titanium

hydrazide.³⁰Since this report in 2004, only a handful of other

terminalgroup4hydrazideshavebeenstructurallycharacterized.^31-35

As part of this early work, we reported structural and DFT

studies of homologous TidNR (imido) and TidNNPh₂systems,

the ﬁrst study of this type for group 4. These showed that for

the early transition metals, NNPh₂is best viewed as a hy-

drazide(2-) ligand analogous to imide(2-), rather than an

isodiazene moiety.³¹One could expect early and later

metal-NNR₂systems (i.e., hydrazide and isodiazene) to differ

in much the same way as Schrock alkylidenes and Fischer

carbenes.³⁶

In addition to the above stoichiometric reactions of isolated

complexes, terminal titanium hydrazides have been implicated

(butnotdirectlyobserved)inthecatalytichydrohydrazination^30,41-47

and iminohydrazination^32,46of alkynes and in the synthesis of

indoles^3,43,45and tryptamines.⁴⁴In these reactions, typically

leading to a hydrazone or its equivalent, a [2+2] cycloaddition

reaction between TidNNR₂and an alkyne (or allene) is

considered to be the key N-C bond forming step. Little, if

anything, is directly known about these processes or the

intermediates involved.

We recently reported³⁴that reaction of the diamide-amine-

supported complex Ti(N₂N^{C ,Me})(NNPh₂)(py) (2, Vide infra) with

2

PhCCMe did not give the expected [2+2] cycloaddition product,

and once again

a

N_R-N_ꢀbond cleavage product,

Ti(N₂N^{C ,Me}){NC(Ph)C(Me)NPh₂}(py) (3), was formed (eq 2).

In addition, reaction of 3 with Ph₂NNH₂regenerated 2 to form

the cis-1,2-diamino alkene PhC(NH₂)C(Me)NPh₂. This is a

unique and unprecedented transformation of alkynes, which was

also found to be catalytic in either 2 or 3.

2

Because of its apparent relationship to the corresponding

TidNR systems, one of the main interests in the group 4

MdNNR₂functional group concerns its reactivity with unsatur-

ated organic substrates.³⁷Bergman’s transient Cp₂Zr(NNPh₂)

system underwent N_R-N_ꢀbond cleavage reactions with alkynes

and CO. Surprisingly, ZrdNNR₂/RCCR cycloaddition products

(well known²for the corresponding imido systems Cp₂Zr(NR))

were not observed. Subsequent work from our own group used

the macrocycle-supported complexes Ti(NNPh₂)(Me₄taa)

(H₂Me₄taa ) tetramethyldibenzotetraaza[14]annulene³⁸). On

reaction with TolNCO or CO₂, these gave the ﬁrst well-deﬁned

MdN_R[2+2] cycloaddition reactions of any group 4 MdNNR₂

functional group.³⁹Woo et al. reported hydrazide ligand transfer

reactions for the corresponding Ti(NNR₂)(TTP) (H₂TTP ) tetra-

p-tolylporphyrin) system with p-chlorobenzaldehyde, which

gave the corresponding hydrazones, also with the N_R-N_ꢀbond

remaining intact.⁴⁰In contrast, Gade and co-workers once again

encountered hydrazide N_R-N_ꢀbond cleavage reactions on

The rate and selectivity of the new transformation in eq 2

(and Bergman and Gade’s earlier, noncatalytic analogues) should

undoubtedly depend on the nature of the strongly π-donating

diamide-donor (or related) ligand set. Indeed such chelating

anionic nitrogen ligands have been very fruitful for developing

the chemistry of early transition metal-ligand multiple bond

chemistry during the last 10-15 years.^38,48-54We note in

particular Schrock’s tris(amide)-amine (“tren”) systems, which

were crucial in that group’s discovery of the synthetic catalytic

conversion of N₂to NH₃.^10,49Polyamide-based ligands have

also been important in the growing ﬁeld of group 4 N₂activation

chemistry.¹²Therefore, as a platform for future developments

in this ﬁeld, we report in detail here the synthesis and molecular

and electronic structures of a range of new titanium hydrazide

complexes with N₃-, N₄-diamide-donor (and related) ligands,

along with new entry synthons for titanium hydrazide chemistry.

Part of this work has been communicated.³⁴

reaction of the diamide-amine-supported Zr(N₂N^py*

)

(NNPh₂)(py) (1, eq 1) with ^tBuNC, Ph₃PSe, or propylene sulﬁde

t

(N₂N^py*) (2-NC₅H₄)C(Me)(CH₂NSiMe₂Bu)₂).³⁵

(23) Allen, F. H.; Kennard, O. Chem. Des. Automation News 1993, 8,

1. and 31.

(24) Fletcher, D. A.; McMeeking, R. F.; Parkin, D. J. Chem. Inf. Comput.

Sci. 1996, 36, 746 (The United Kingdom Chemical Database Service).

(25) Lehnert, N.; Tuczek, F. Inorg. Chem. 1999, 38, 1659.

(26) Mersmann, K.; Horn, K. H.; Bores, N.; Lehnert, N.; Studt, F.;

Paulat, F.; Peters, G.; Ivanovic-Burmazovic, I.; van Eldik, R.; Tuczek, F.

Inorg. Chem. 2005, 44, 3031.

Results and Discussion

(27) Lehnert, N.; Tuczek, F. Inorg. Chem. 1999, 38, 1671.

(28) Wiberg, N.; Haring, H.-W.; Huttner, G.; Friedrich, P. Chem. Ber.

1978, 111, 2708.

(29) Walsh, P. J.; Carney, M. J.; Bergman, R. G. J. Am. Chem. Soc.

1991, 113, 6343.

(30) Li, Y.; Shi, Y.; Odom, A. L. J. Am. Chem. Soc. 2004, 126, 1794.

(31) Parsons, T. B.; Hazari, N.; Cowley, A. R.; Green, J. C.; Mountford,

P. Inorg. Chem. 2005, 44, 8442.

(32) Banerjee, S.; Odom, A. L. Organometallics 2006, 25, 3099.

(33) Patel, S.; Li, Y.; Odom, A. L. Inorg. Chem. 2007, 46, 6373.

(34) Selby, J. D.; Manley, C. D.; Felix, M.; Schwarz, A. D.; Clot, E.;

Mountford, P. Chem. Commun. 2007, 4937.

Ligand Choices. Chart 1 illustrates the ligands used in this

study. These chelating, dianionic ligands should provide well-

deﬁned, relatively nonlabile environments for studying com-

pounds of the type (L)TidNNR₂, and the hard N- and O-donors

are compatible with the desired Ti(+4) oxidation state. They

have been widely used in early transition metal chemistry and

metal-ligand multiple bond chemistry^{4,23,24,38,39,50,51,55-65}and

(36) Crabtree, R. H. The Organometallic Chemistry of the Transition

Metals, 4th ed.; WileyBlackwell: London, 2005.

(37) Mindiola, D. J. Angew. Chem., Int. Ed. 2008, 47, 1557.

(35) Herrmann, H.; Fillol, J. L.; Wadepohl, H.; Gade, L. H. Angew.

Chem., Int. Ed. 2007, 46, 8426.

Titanium Hydrazides

Organometallics, Vol. 27, No. 24, 2008 6481

Chart 1. The Ligands Used in This Contribution and Their Abbreviations (Formal Anionic Charges Omitted for Clarity)

allow variation of the metal’s coordination number and steric

accessibility. They differ in bite angle, hemilability (or other-

wise), and π-donor ability.

Development of a Synthetic Strategy to New Families

of Hydrazide Complexes. While hydrazide complexes

(L)TidNNR₂are rather limited in number, several different

synthetic strategies have been used for their preparation.

Wiberg²⁸and Woo⁴⁰employed Me₃SiCl or LiCl elimination

protocols, respectively, starting from the corresponding dichlo-

ride precusors (L)TiCl₂(L ) Cp₂or TTP). We showed that

Ti(Me₄taa)(N^tBu) reacted with Ph₂NNH₂to form

t

Ti(Me₄taa)(NNPh₂) and BuNH₂via a protonolysis type reac-

tion³⁹reminiscent of the well-established tert-butyl imide/aniline

exchange protocols used to form aryl imides.^4,6,66-68Proto-

nolysis reactions of bis(dimethylamido) compounds

(L)Ti(NMe₂)₂with several different hydrazines RR′NNH₂(R

) R ) alkyl or aryl; R ) aryl, R′ ) H) have also afforded

both terminal and bridging hydrazido derivatives.^30,33

(38) Mountford, P. Chem. Soc. ReV. 1998, 27, 105.

(39) Blake, A. J.; McInnes, J. M.; Mountford, P.; Nikonov, G. I.;

Swallow, D.; Watkin, D. J. J. Chem. Soc., Dalton Trans. 1999, 379.

(40) Thorman, J. L.; Woo, L. K. Inorg. Chem. 2000, 39, 1301.

(41) Johnson, J. S.; Bergman, R. G. J. Am. Chem. Soc. 2001, 123, 2923.

(42) Cao, C.; Shi, Y.; Odom, A. L. Org. Lett. 2002, 4, 2853.

(43) Ackermann, L.; Born, R. Tetrahedron Lett. 2004, 45, 9541.

(44) Khedkar, V.; Tillack, A.; Michalik, M.; Beller, M. Tetrahedron

Lett. 2004, 45, 3123.

(45) Tillack, A.; Jiao, H.; Garcia Castro, I.; Hartung, C. G.; Beller, M.

Chem.-Eur. J. 2004, 10, 2410.

(46) Banerjee, S.; Shi, Y.; Cao, C.; Odom, A. L. J. Organomet. Chem.

2005, 690, 5066.

(47) Banerjee, S.; Barnea, E.; Odom, A. L. Organometallics 2008, 27,

1005.

(48) Brand, H.; Arnold, J. Coord. Chem. ReV. 1995, 140, 137.

(49) Schrock, R. R. Acc. Chem. Res. 1997, 30, 9.

(50) Gade, L. H. Chem. Commun. 2000, 173.

(51) Gade, L. H.; Mountford, P. Coord. Chem. ReV. 2001, 216-217,

65.

(52) Bourget-Merle, L.; Lappert, M. F.; Severn, J. R. Chem. ReV. 2002,

102, 3031.

(53) Bigmore, H. R.; Lawrence, S. C.; Mountford, P.; Tredget, C. S.

Dalton Trans. 2005, 635.

(54) Kempe, R. Angew. Chem., Int. Ed. Engl. 2000, 39, 468.

(55) Lowes, T. A.; Ward, B. D.; Whannel, R. A.; Dubberley, S. R.;

Mountford, P. Chem. Commun. 2005, 113.

(56) Mountford, P.; Swallow, D. J. Chem. Soc., Chem. Commun. 1995,

2357.

(57) Blake, A. J.; Mountford, P.; Nikonov, G. I.; Swallow, D. Chem.

Commun. 1996, 1835.

(58) Swallow, D.; Dunn, S. C.; Nikonov, G. I.; Mountford, P. Abstr.

Papers Am. Chem. Soc. 1996, 211, 78.

(59) Swallow, D.; McInnes, J. M.; Mountford, P. J. Chem. Soc., Dalton

Trans. 1998, 2253.

Although protonolysis procedures appeared attractive for

preparing the target diamide-amine hydrazide complexes

(avoiding potentially reducing anionic reagents), they were

unsatisfactory due to concomitant protonolysis side-reactions

of the supporting ligands. For example, reaction of

Ti(N₂N^{C ,SiMe})(N Bu)(py)⁶⁹with Ph₂NNH₂in C₆D₆gave less

t

2

3

than

ca.

35%

conversion

to

the

target

Ti(N₂N^{C ,SiMe})(NNPh₂)(py) (Vide infra). The remainder of the

2

3

Ti(N₂N^{C ,SiMe})(N Bu)(py) was consumed to form the protio

t

2

3

ligand H₂N₂N^{C ,SiMe}and unknown species. Analogous results

were found for certain other tert-butyl imido titanium com-

plexes of the target ligands. Gade and co-workers have encountered

similar difﬁculties starting from Ti(N₂N^py)(NMe₂)₂.⁷⁰The amido

nitrogens of N₃- and N₄-donor supporting ligands of the type

2

3

N₂N^{C ,R}, N₂N^py, and N₂NN′ are therefore prone to protonolysis

n

reactions at a rate competitive with those of the target TidN^tBu

or Ti-NMe₂leaving groups. This is in contrast to the

Ti(Me₄taa)(N^tBu) system, which underwent imide protonolysis with

Ph₂NNH₂without loss of H₂Me₄taa.^39,59The difference may be

partially attributable to the charge delocalization within the Me₄taa

backbone, making the formally anionic ligands less Brønsted basic.

Odom’s successful conversion of Ti(dmpa)(NMe₂)₂to

Ti(dmpa)(NNMe₂)(^tBubipy) might be explained in a similar way

t

(dmpa ) N,N-di(pyrrolyl-R-methyl)-N-methylamine; Bubipy )

(60) McInnes, J. M.; Swallow, D.; Blake, A. J.; Mountford, P. Inorg.

Chem. 1998, 37, 5970.

di-tert-butylbipyridine).³⁰

(61) Kisko, J. L.; Hascall, T.; Parkin, G. J. Am. Chem. Soc. 1997, 119,

7609.

(62) Amgoune, A.; Thomas, C. M.; Roisnel, T.; Carpentier, J.-F. Chem.-

Eur. J. 2006, 12, 169.

(63) Bonnet, F.; Cowley, A. R.; Mountford, P. Inorg. Chem. 2005, 44,

9046.

(66) Blake, A. J.; Collier, P. E.; Dunn, S. C.; Li, W.-S.; Mountford, P.;

Shishkin, O. V. J. Chem. Soc., Dalton Trans. 1997, 1549.

(67) Coles, M. P.; Dalby, C. I.; Gibson, V. C.; Clegg, W.; Elsegood,

M. R. J. Polyhedron 1995, 14, 2455.

(64) Suzuki, Y.; Terao, H.; Fujita, T. Bull. Chem. Soc. Jpn. 2003, 76,

1493.

(65) Mitani, M.; Saito, J.; Ishii, S.; Nakayama, Y.; Makio, H.; Mat-

sukawa, N.; Matsui, S.; Mohri, J.; Furuyama, R.; Terao, H.; Bando, H. H. T.;

Fujita, T. Chem. Rec. 2004, 4, 137.

(68) Michelman, R. I.; Bergman, R. G.; Andersen, R. A. Organometallics

1993, 12, 2741.

(69) Pugh, S. M.; Clark, H. S. C.; Love, J. B.; Blake, A. J.; Cloke,

F. G. N.; Mountford, P. Inorg. Chem. 2000, 39, 2001.

(70) Gade, L. H. Personal communication.

6482 Organometallics, Vol. 27, No. 24, 2008

Scheme 1. Synthesis of Titanium Hydrazides Supported by Tridentate Diamide-Amine Ligands^a

Selby et al.

Reagents and conditions: (i) Li₂N₂N^{C ,Me}, -78 °C, toluene, 65% or Li₂N₂N^{C ,SiMe}, -78 °C, toluene, 26%; (ii) Li₂N₂N^{C ,Me}, -78 °C, toluene,

41%; (iii) Li₂N₂N^py, -78 °C, toluene, 73%.

2

3

a

Scheme 2. Synthesis of Titanium Hydrazides Supported by Tetradentate N₄- or O₂N₂-Donor Ligands^a

a

Reagents and conditions: (i) Li₂N₂NN′, -78 °C, toluene, 31%; (ii) Li₂Me₄taa, C₆D₆, 100%; (iii) Na₂O₂NN′, -78 °C, toluene, 49%.

A wide number of titanium imido compounds^4,6are available

from synthons of the type Ti(NR)Cl₂(py)_xor

Ti(NR)Cl₂(NHMe₂)₂(R ) alkyl or aryl; x ) 2 or 3)^66,71,72by

substitution strategies. We therefore considered an analogous

approach for making new hydrazide complexes. At ﬁrst sight,

a suitable hydrazide synthon is Ti(NNPh₂)Cl₂(NHMe₂)₂(pre-

pared from Ti(NMe₂)₂Cl₂and Ph₂NNH₂).³¹Unfortunately, the

somewhat acidic NH groups of the NHMe₂ligands lead to

undesired side-reactions when the complexes are treated with

anionic reagents such as lithiated amides.⁴However, the NHMe₂

ligands of Ti(NNPh₂)Cl₂(NHMe₂)₂can be substituted by pyri-

dine to form dimeric [Ti(NNPh₂)Cl₂(py)₂]₂(4), believed to have

terminal TidNNPh₂groups and bridging Cl ligands.³¹Prelimi-

nary experiments showed that this material reacted with

metalated salts of the target ligands to form the desired

complexes in acceptable to good yields as described below

(Schemes 1 and 2).

(71) Adams, N.; Arts, H. J.; Bolton, P. D.; Cowell, D.; Dubberley, S. R.;

Friederichs, N.; Grant, C.; Kranenburg, M.; Sealey, A. J.; Wang, B.; Wilson,

P. J.; Cowley, A. R.; Mountford, P.; Schro¨der, M. Chem. Commun. 2004,

434.

(72) Adams, N.; Bigmore, H. R.; Blundell, T. L.; Boyd, C. L.; Dubberley,

S. R.; Sealey, A. J.; Cowley, A. R.; Skinner, M. E. G.; Mountford, P. Inorg.

Chem. 2005, 44, 2882.

(73) Lorber, C.; Choukroun, R.; Donnadieu, B. Inorg. Chem. 2002, 41,

4217.

Although [Ti(NNPh₂)Cl₂(py)₂]₂(4) can be prepared from

Ti(NNPh₂)Cl₂(NHMe₂)₂, the latter itself requires at least a two-

Titanium Hydrazides

Organometallics, Vol. 27, No. 24, 2008 6483

Figure 1. Displacement ellipsoid plots (25% probability) with H atoms omitted for clarity. (a) Ti(NNPh₂)Cl₂(py)₃(7), atoms carrying the

sufﬁx “A” are related to their counterparts by the symmetry operator -x+1, y, -z+³/₂. (b) One of the two crystallographically independent

molecules of Ti(NNPh₂)Cl₂(py′)₃(8), second orientation of one of the py′ tert-butyl groups omitted.

Table 1. Selected Bond Distances (Å) and Angles (deg) for

step synthesis, and its conversion to 4 requires heating in a large

excess of neat pyridine at 70 °C for 16 h, giving variable yields

Ti(NNPh₂)Cl₂(py)₃(7)^a

Ti(1)-N(1)

1.727(2)

2.2464(17) Ti(1)-N(4)

1.359(3)

97.25(4)

96.738(15)

Ti(1)-Cl(1)

2.4265(5)

2.402(2)

1.411(2)

and purity. Similar difﬁculties with this type of apparently

straightforward reaction have been noted for the imido systems

M(NR)Cl₂(NHMe₂)₂(M ) Ti⁷²or V⁷³). We therefore developed

an alternative route (eq 3) to 4 and well-deﬁned monomeric

homologues starting from the tert-butyl imido compounds

Ti(1)-N(3)

N(1)-N(2)

N(1)-Ti(1)-N(3)

N(1)-Ti(1)-Cl(1)

N(1)-Ti(1)-N(4) 180

N(3)-Ti(1)-Cl(1) 90.44(5)

N(2)-C(1)

Cl(1)-Ti(1)-Cl(1A) 166.52(3)

Ti(1)-N(1)-N(2)

N(1)-N(2)-C(1)

C(1)-N(2)-C(1A)

180

118.66(11)

Ti(N^tBu)Cl₂(L)₃(L ) py (5) or 4-NC₅H₄Bu (py′, 6)). These

t

^aAtoms carrying the sufﬁx “A” are related to their counterparts by

are both readily available in multigram quantities from inex-

pensive and commerically available starting materials (TiCl₄,

^tBuNH₂, and py or py′).^66,74

the symmetry operator -x+1, y, -z+³/₂.

Reaction of 5 with Ph₂NNH₂(1 equiv) in benzene at room

temperature overnight afforded a precipitate of bright yellow,

monomeric Ti(NNPh₂)Cl₂(py)₃(7) in 89% yield (ca. 5 g scale)

and in analytically pure form. The corresponding reaction of 6

gave the homologous compound Ti(NNPh₂)Cl₂(py′)₃(8) in 74%

yield, again as a pure yellow solid. The lower isolated yield for

8 reﬂects its higher solubility in hydrocarbon solvents. Further

exposure of 7 to a dynamic vacuum leads to loss of one

coordinated pyridine and eventual quantitative formation of the

bis(pyridine) complex 4. Analogous behavior is found for the

hydrazide-bridged dimer Ti₂(µ-η²,η¹-NNMe₂)₂Cl₄(py)₂, which

we found to be ineffective as a synthon.³¹

The NMR spectra of 7 and 8 show resonances for a NNPh₂

ligand and two py (or py′) ligand environments in a 2:1 ratio.

The lower intensity py or py′ resonances are shifted upﬁeld from

their counterparts and are assigned to the ligands trans to the

TidNNPh₂functional groups. The solution NMR and other

analytical data are fully consistent with the structures proposed

in eq 3, but do not uniquely deﬁne them as monomeric terminal

hydrazides. They have therefore been characterized by X-ray

crystallography in both cases. The molecular structures are

shown in Figure 1, and key distances and angles are given in

Tables 1 and 2. Molecules of 7 lie on crystallographic 2-fold

rotation axes passing through the Ti(1)-N(1)-N(2) linkage,

while compound 8 contains two crystallographically independent

molecules in the asymmetric unit. Some positional disorder of

certain py′ tert-butyl groups of 8 was satisfactorily modeled.

t

imido systems Ti(NR)Cl₂(py)₃(R ) Bu or aryl)⁶⁶and was

attributed to the strong labilizing effect of the imido groups on

the pyridine trans to the TidNR bond.⁷⁵Compound 8 does not

easily lose the trans py′ ligand in this way because of the lower

volatility of the substituted pyridine and its better donor ability.

Both 4 and 7 were subsequently used as synthons for the

syntheses described below, and there is no difference in their

performance. Compound 8 was not used further in the present

syntheses because preliminary studies found the parent pyridine

adduct products were already very soluble and prone to be oily

in some instances (due to the presence of SiMe₃groups in the

diamide complexes). Unfortunately, reaction of Me₂NNH₂with

5 did not give terminal hydrazido compounds but only the

The structures of 7 and 8 conﬁrm those anticipated in eq 3.

These are only the second and third examples of a structurally

authenticated TidNNPh₂functional group (two such examples

are now known for zirconium).^29,35The other titanium example

is Ti(NNPh₂){HC(Me₂pz)₃}Cl₂(9), which also contains an

octahedral Ti, but with fac-coordination of the N₃-donor set and

mutually cis Cl ligands. Since a number of imido analogues

(74) Dunn, S. C.; Batsanov, A. S.; Mountford, P. J. Chem. Soc., Chem.

Commun. 1994, 2007.

t

Ti(NR)Cl₂(py)₃(R ) Bu or aryl) have been structurally

(75) Kaltsoyannis, N.; Mountford, P. J. Chem. Soc., Dalton Trans. 1999,

781.

characterized,^23,247 and 8 provide an excellent opportunity for

6484 Organometallics, Vol. 27, No. 24, 2008

Selby et al.

Table 2. Selected Bond Distances (Å) and Angles (deg) for Ti(NNPh₂)Cl₂(py′)₃(8)^a

Ti(1)-N(1)

1.7240(18)

2.4079(7)

2.4224(7)

2.3942(19)

2.2417(18)

[1.7216(17)]

[2.4132(7)]

[2.3967(6)]

[2.483(2)]

[2.2314(19)]

[178.20(8)]

[93.71(8)]

[93.41(8)]

[98.30(6)]

[98.23(6)]

[163.43(3)]

Ti(1)-N(5)

N(1)-N(2)

N(2)-C(1)

N(2)-C(7)

2.2274(18)

1.351(2)

1.414(3)

1.430(3)

[2.211(2)]

[1.353(2)]

[1.423(3)]

[1.417(3)]

Ti(1)-Cl(1)

Ti(1)-Cl(2)

Ti(1)-N(3)

Ti(1)-N(4)

N(1)-Ti(1)-N(3)

N(1)-Ti(1)-N(4)

N(1)-Ti(1)-N(5)

N(1)-Ti(1)-Cl(1)

N(1)-Ti(1)-Cl(2)

Cl(1)-Ti(1)-Cl(2)

178.56(8)

N(4)-Ti(1)-N(5)

Ti(1)-N(1)-N(2)

N(1)-N(2)-C(1)

N(1)-N(2)-C(7)

C(1)-N(2)-C(7)

169.51(7)

[172.69(7)]

[178.38(16)]

[117.77(18)]

[118.54(18)]

[123.45(18)]

94.51(8)

95.97(8)

96.68(6)

99.23(6)

164.07(3)

179.54(15)

118.49(17)

118.69(17)

122.82(17)

^aValues in brackets are for the other crystallographically independent molecule.

structural comparisons with their better established imido

analogues.

compounds in general is 1.317 Å,^23,24and so in this sense the

N_R-N_ꢀdistances for 7 and 8 are comparatively long.

Synthesis of Titanium Hydrazides Containing

Dianionic N₃-, N₄-, and O₂N₂-Donor Ligands. As mentioned,

both [Ti(NNPh₂)Cl₂(py)₂]₂(4) and Ti(NNPh₂)Cl₂(py)₃(7) serve

as starting materials for the general synthesis of new families

of titanium diphenyl hydrazides containing dianionic N₃- and

N₄-donor ligands. The syntheses are summarized in Schemes 1

and 2.

Preliminary studies focused on the N₂N^{C ,SiMe}ligand system,

2

3

which we have previously used in the tert-butyl imido complex

C₂,SiMe₃

Ti(N^tBu)(N₂N^{C ,SiMe})(py). Reaction of Li₂N₂N

with 4

2

3

in benzene at room temperature gave the crude product as a

green-brown solid. The ¹H NMR spectrum showed the presence

The metric data for the structures of 7 and 8 are generally

comparable within error. They possess pseudo-octahedral

geometries at Ti(1), mutually trans Cl donors, and three py (or

py′) ligands arranged in a mer fashion. This is the same as for

the imido analogues Ti(NR)Cl₂(py)₃.^23,24There is a signiﬁcant

bond-lengthening effect of the hydrazido ligand on the py or

py′ ligand opposite it, and the magnitude of this trans inﬂuence

(av ca. 0.20 Å) is comparable to that found in the imido

analogues. The Ti-Cl and Ti-py (py′) distances for 7 and 8

are within the ranges found for these imido and other Ti(+4)

compounds.^23,24They are thus consistent with NNPh₂acting

as a formally dianionic “hydrazide(2-)” ligand (as opposed to

being considered as a neutral NdNPh₂“isodiazene”⁷⁶) in these

titanium complexes, as also found in the DFT calculations for

of both the desired complex Ti(NNPh₂)(N₂N^{C ,SiMe})(py) (10)

2

3

and free protio-ligand H₂N₂N^{C ,SiMe}in a ca. 1:2 ratio. The

2

3

presence of H₂N₂N^{C ,SiMe}is attributed to redox-coupled H atom

abstraction side-reactions, as have been found previously in

some analogous reactions of this ligand.⁶⁹These could be

partially alleviated by starting the reaction at -78 °C and adding

2

3

precooled toluene to a cold solid mixture of Li₂N₂N^{C ,SiMe}and

4, followed by a short reaction time and workup. Analytically

pure samples of 10 were obtained in 26% overall yield, the

relatively high solubility of 10 in hydrocarbon solvents also

contributing to the disappointing yield.

2

3

The reactions between 4 and Li₂N₂N^{C ,Me}or Li₂N₂N^{C ,Me}were

2

3

carried out in the same way and afforded the homologous

9.³¹

The parameters associated with TidNNPh₂linkages are

complexes

Ti(NNPh₂)(N₂N^C²3^,Me)(py)

(2)

and

Ti(NNPh₂)(N₂N^{C ,Me})(py) (11) in 65% and 41% isolated yield.

We found that Li₂N₂N^pyreacted smoothly with 4 to form

Ti(NNPh₂)(N₂N^py)(py) (12) in good yield (73%) without notice-

able redox complications. This reaction could best be performed

by adding a cold solution of the lithiated amide to 4. The NMR

and other analytical data for the new compounds 10-12 are

fully consistent with the structures proposed in Scheme 1, which

have all been conﬁrmed by X-ray crystallography (Vide infra).

3

similar to those in 9 with comparable TidN_R(1.722(2)-1.727(2)

Å) and N_R-N_ꢀ(1.351(2)-1.359(3) Å) distances (cf. 1.718(2)

and 1.369(3) Å, respectively in 9). Likewise, the hydrazide

linkages are linear (TidN-NPh₂) 178-180 °) and the N_ꢀ

nitrogen is planar (sum of the angles subtended at N(2) ∼360

° within error) for both compounds. The Ph rings are rotated

by ca. 30-40° out of the plane containing the TidN-N(C_{ipso 2}

)

atoms, and the N_ꢀ-C_ipsodistances are approximately equivalent

in all cases (1.41-1.42 Å). This suggests the possibility of

partial delocalization of electron density from the sp²-hybridized

N_ꢀonto the phenyl rings.

We were also interested to prepare complexes with dianionic

N₄-donor ligands (Scheme 2). Thus reaction of Li₂N₂NN′ with

4 under analogous conditions for 2, 10, and 11 afforded the

ﬁve-coordinate Ti(NNPh₂)(N₂NN′) (13). We were not able to

grow diffraction-quality crystals of 13, but its NMR data are

consistent with the C_ssymmetrical, trigonal-bipyramidal struc-

ture proposed in Scheme 2. Titanium and zirconium imido

analogues of 13 have been reported previously and have

comparable structures.⁷⁷The chemical shift of the ortho-H of

the pyridyl donor in 13 appears at δ 8.98 ppm, consistent with

it being coordinated in solution (in free H₂N₂NN′ this hydrogen

The TidN_Rdistances in 7 and 8 are all slightly longer than

t

for the imido complexes Ti(NR)Cl₂(py)₃: R ) Bu, TidN )

1.705(3) Å; R ) Ph, 1.714(2) Å; R ) p-Tol, 1.705(4) Å).⁶⁶

They are, however, well within the general range (ca. 1.68-1.76

Å based on titanium imides) expected for a formal Tit N_Rtriple

bond (σ²π⁴). This triple-bond assignment is also consistent with

previous DFT calculations on 9 and also the observed linear

TidN-NPh₂linkage (sp-hybridized N_R).³¹In contrast, the

average N_R-N_ꢀdistance for transition metal diphenylhydrazido

(76) Kahlal, S.; Saillard, J.-Y.; Hamon, J.-R.; Manzur, C.; Carrillo, D.

J. Chem. Soc., Dalton Trans. 1998, 1229.

(77) Skinner, M. E. G.; Toupance, T.; Cowhig, D. A.; Tyrrell, B. R.;

Mountford, P. Organometallics 2005, 24, 5586.

Titanium Hydrazides

Organometallics, Vol. 27, No. 24, 2008 6485

Figure 2. Displacement ellipsoid plots with H atoms omitted for clarity. (a) Ti(NNPh₂)(N₂N^{C ,Me})(py) (2, 20% probability, disorder in

2

coordinated pyridine omitted). (b) Ti(NNPh₂)(N₂N^{C ,SiMe})(py) (10, 30% probability).

2

3

appears at δ 8.60 ppm).⁷⁸Signiﬁcantly, 13 is the neutral group

4 analogue of Schrock’s triamide-amine (tren)-supported

Mo-NNR₂(R ) alkyl or H) systems.^10,20,79

As mentioned, Ti(NNPh₂)(Me₄taa) (14) was previously

prepared by reaction of Ti(N^tBu)(Me₄taa) with Ph₂NNH₂.³⁹As

shown in Scheme 2, the reaction between Li₂Me₄taa and 4 in

C₆D₆gave quantitative conversion to 14 immediately at room

temperature, indicating this is a viable route to macrocycle-

supported hydrazides also. The NMR data for 14 were com-

parable to those of an authentic sample prepared by the original

imide/hydrazine exchange route.

Figures 2 (2, 10) and 3 (11, 12). Key bond distances and angles

are compared in Table 3.

All four complexes possess trigonal-bipyramidal titanium

centers with the diamide-amine ligand coordinating in a fac

manner. The hydrazide N(1) (N_R) and amide nitrogens (N(3),

N(4)) occupy the equatorial positions of the trigonal bipyramids;

the neutral donors, the axial sites. The Ti-N and intraligand

distances and angles for the diamide-amine and pyridine donors

are all within the expected ranges,^23,24although there are some

interesting variations as discussed below. The titanium tert-butyl

imido analogues of 10 and 12 have been reported previously,^69,81

as have related group 4 and 5 imido complexes Zr(N-2,6-

We have also made a comparison between the N₄-donor

systems Ti(NNPh₂)(L) (L ) N₂NN′ (13) or Me₄taa (14)) and

those of O₂NN′ (see Chart 1) as a representative bis(phenolate)-

diamine ligand. Addition of a cold toluene solution of Na₂O₂NN′

to [Ti(NNPh₂)Cl₂(py)₂]₂(4) gave the compound

Ti(NNPh₂)(O₂NN′)(py) (15) in 49% yield as summarized in

Scheme 2. The NMR data for 15 are consistent with the C_s

symmetric structure proposed and conﬁrm the presence of the

coordinated pyridine. Corresponding monomeric titanium imido

complexes of O₂NN′-type ligands have been prepared previ-

ously.⁸⁰Unfortunately we have not been able to obtain diffrac-

tion-quality crystals of 15, and for the remainder of this paper

we shall focus on the diamide-amine systems. Attempts to

prepare a terminal dimethyl hydrazide analogue of 15 are

described in the Supporting Information along with the X-ray

structure of dimeric Ti₂(µ-η²η¹-NNMe₂)₂(O₂NN′)₂(16).

i

C₆H₃Pr₂)(N₂N^py)(py) and Ta(N^tBu)(N₂N^py)Cl.^81,82These have

comparable structures with the amide and imide nitrogens in

the equatorial plane. In the hydrazide structures, as in the

aforementioned imido complexes, the amide nitrogens are

trigonal planar (sum of the angles subtended at N(3) and N(4)

) 360° within error) and sp²hybridized, which is typical for

early transition metal complexes.^50,54

Before considering the data associated with the TidNNPh₂

moieties it is necessary to examine more closely the variations

in the Ti(N₂N^R)(py) fragments in the four complexes. The ﬁve-

membered chelate rings formed by the diamide-amine ligands

in 2 and 10 lead to larger N(1)-Ti(1)-N(5) angles (av ca. 103°)

compared to those for 11 and 12 (six-membered chelate rings,

av ca. 94°), thus pulling the apical donor away from the

TidNNPh₂linkage. The N(5)-Ti(1)-N(6) angles are cor-

respondingly smaller for 2 and 10 (av ca. 161°) than for 11 and

12 (av ca. 173°). This in turn affects the Ti(1)-N(6) bond

distances for the pyridine ligand (shorter in 2 and 10 (av ca.

2.22 Å) than in 11 and 12 (av ca. 2.27 Å). The effect of replacing

the apical N-Me of the diamide ligand in 2 with the bulky

N-SiMe₃in 10 results in a 0.131(4) Å lengthening of the

Ti(1)-N(5) bond. Interestingly, the Ti(1)-N(6) distances for

the pyridine ligands do not vary signiﬁcantly between 2 and 10

(difference 0.011(5) Å). Finally it is important to note that the

variations in the diamide-amine ligand framework lead to

signiﬁcant changes in the Ti(1)-N(3) and Ti(1)-N(4) distances

to the amide nitrogens. Thus from 2 to 12 the average Ti-N_amide

Structures of Titanium Hydrazides Containing

N₃-Donor Ligands. The compounds 2 and 10-12 offer an

excellent opportunity to determine systematically the effects of

the different diamide-amine ligands on the ground-state structure

of this new class of titanium hydrazide complexes, and thereby

gain insight into the unique reactivity of these systems.

Diffraction-quality crystals of Ti(NNPh₂)(N₂N^{C ,Me})(py) (2),

2

Ti(NNPh₂)(N₂N^{C ,SiMe})(py) (10), Ti(NNPh₂)(N₂N^{C ,Me})(py)

(11), and Ti(NNPh₂)(N₂N^py)(py) (12) were grown from the

appropriate solvents. The molecular structures are shown in

2

3

(78) Skinner, M. E. G.; Li, Y.; Mountford, P. Inorg. Chem. 2002, 41,

1110.

(79) O’Donoghue, M. B.; Davis, W. M.; Schrock, R. R. Inorg. Chem.

1998, 37, 5149.

(81) Blake, A. J.; Collier, P. E.; Gade, L. H.; Mountford, P.; Lloyd, J.;

Pugh, S. M.; Schubart, M.; Skinner, M. E. G.; Tro¨sch, J. M. Inorg. Chem.

2001, 40, 870.

(80) Boyd, C. L.; Toupance, T.; Tyrrell, B. R.; Ward, B. D.; Wilson,

C. R.; Cowley, A. R.; Mountford, P. Organometallics 2005, 24, 309.

(82) Ward, B. D.; Orde, G.; Clot, E.; Cowley, A. R.; Gade, L. H.;

Mountford, P. Organometallics 2004, 23, 4444.

6486 Organometallics, Vol. 27, No. 24, 2008

Selby et al.

Figure 3. Displacement ellipsoid plots with H atoms omitted for clarity. (a) Ti(NNPh₂)(N₂N^{C ,Me})(py) (11, 20% probability, minor disorder

omitted). (b) Ti(NNPh₂)(N₂N^py)(py) (12, 20% probability).

3

Table 3. Selected Bond Distances (Å) and Bond and Torsion Angles

distances (1.733(5)-1.759(2) Å) are systematically longer than

(deg) for Ti(NNPh₂)(N₂N^{C ,Me})(py) (2), Ti(NNPh₂)(N₂N^{C ,SiMe})(py)

2

3

for 7, 8, and 9 (range 1.718(2)-1.727(2) Å). The same situation

(10), Ti(NNPh₂)(N₂N^{C ,Me})(py) (11), and Ti(NNPh₂)(N₂N^py)(py) (12)^a

3

is found when comparing the TidN^tBu distances of

Ti(N^tBu)(N₂N^{C ,SiMe})(py) and Ti(N^tBu)(N₂N^py)(py)⁸¹with

69

2

3

parameter

2

10

11

12

that of Ti(N^tBu)Cl₂(py)₃and in the ﬁrst instance can be

66

Ti(1)-N(1)

Ti(1)-N(3)

Ti(1)-N(4)

Ti(1)-N(5)

Ti(1)-N(6)

N(1)-N(2)

N(2)-C(1)

N(2)-C(7)

1.733(5)

2.029(5)

2.026(5)

2.224(4)

2.215(5)

1.359(7)

1.438(7)

1.399(7)

1.750(2)

2.014(2)

2.003(2)

2.355(2)

2.226(2)

1.362(2)

1.426(2)

1.403(2)

118.83(7) 114.89(6)

120.57(7) 116.03(6)

120.29(7) 128.98(6)

104.01(6)

93.17(7)

162.82(6) 171.49(5)

173.22(14) 168.75(13)

118.3(2)

120.2(2)

121.4(2)

38.6

1.7506(14)

1.999(2)

2.0020(14)

2.289(2)

2.281(2)

1.364(2)

1.433(2)

1.398(2)

1.759(2)

1.974(2)

2.184(2)

2.259(2)

1.367(2)

1.393(3)

1.434(3)

124.52(8)

128.74(8)

106.64(7)

95.49(7)

90.25(7)

174.26(6)

163.8(2)

121.2(2)

115.9(2)

122.7(2)

21.2

attributed to the good donor abilities of diamide-amine ligands.

The TidN_Rdistances in 10 and 12 are longer by 0.032(4) and

0.035(3) Å than the TidN^tBu distances in the corresponding

imido complexes, as expected from our earlier comparison of

7 with Ti(N^tBu)Cl₂(py)₃.

The N_ꢀ-C_ipsodistances (N(2)-C(1) and N(2)-C(7)) merit

comment. In 10 the difference between them is relatively small

(0.023(3) Å), and the phenyl groups are twisted out of the

{N(2),C(1),C(7)} least-squares plane by ca. 27° and 47° for C(7)

and C(1), respectively. In the other structures the differences

between the N_ꢀ-C_ipsodistances are signiﬁcantly larger (0.035(3)-

0.041(4) Å), and at the same time the differences in the extent

of twisting of the phenyl rings out of the {N(2),N(1),N(7)}

planes become much larger. In 2, 11, and 12 one phenyl ring is

much more coplanar (twist angle 4-9°) than the other (twist

angle 47-69°). The more coplanar phenyl rings are associated

with the shorter N_ꢀ-C_ipsodistances (range 1.393(3)-1.399(7)

Å, cf. 1.433(2)-1.438(7) Å). These trends are evidence for

efﬁcient π conjugation between N_ꢀand the phenyl substituents.⁸³

Previous ab initio calculations have found that the free model

hydrazide dianion [NNH₂]^2-has a pyramidalized ꢀ-nitrogen,

as does the monoanion and the triplet form of neutral NNH₂.⁷⁶

In contrast, the formally dianionic³¹NNPh₂groups in group 4

systems possess planar ꢀ-nitrogens. Since terminal TidNNMe₂

analogues feature pyramidal ꢀ-Ns,^30,34it appears that the

planarity of the diphenyl hydrazide ꢀ-nitrogens is due to

N_ꢀ-C_ipsoconjugation.

N(1)-Ti(1)-N(3) 113.8(2)

N(1)-Ti(1)-N(4) 113.7(2)

N(3)-Ti(1)-N(4) 130.5(2)

N(1)-Ti(1)-N(5) 102.1(2)

N(1)-Ti(1)-N(6) 99.3(2)

N(5)-Ti(1)-N(6) 158.4(2)

Ti(1)-N(1)-N(2) 177.7(4)

N(1)-N(2)-C(1) 116.1(4)

N(1)-N(2)-C(7) 120.9(5)

C(1)-N(2)-C(7) 119.5(5)

92.91(6)

95.57(6)

116.78(14)

121.25(14)

120.66(14)

37.4

av N(1)-Ti(1)-

53.0

N(3)-R¹

av N(1)-Ti(1)-

N(4)-R²

70.3

49.4

21.5

25.2

^a“Av N(1)-Ti(1)-N(3,4)-R^1,2” refers to the average of the torsion

angles N(1)-Ti(1)-N(3,4)-Si or N(1)-Ti(1)-N(3,4)-C for the

linkages under consideration.

distance decreases from ca. 2.027 Å to 1.974 Å. At the same

time the orientation of the amide nitrogen substituents relative

to the equatorial plane changes such that they become more

coplanar with the amide nitrogens and hydrazide N_R. This can

be quantiﬁed by the average N(1)-Ti(1)-N_amide-R torsion

angles (see Table 3), which decrease from 61.6° in 2 to 23.2°

in 12. The signiﬁcance of these variations is evaluated using

DFT later.

Turning now to the TidNNPh₂moiety, it is seen that in all

compounds the Ti(1)-N(1)-N(2) linkage is approximately

linear (163.8(2)-177.7(4)°). The ꢀ-nitrogen (N(2)) is effectively

trigonal planar, as judged by the sums of the angles subtended

at this atom (range 356(1)-360(1)°). These features are also

consistent with the structures of 7, 8, and 9. Similarly, the

N(1)-N(2) distances in 2 and 10-12 (1.359(7)-1.367(2) Å)

fall within the range of those found in the dichloride complexes

(range 1.351(2)-1.369(3) Å). In contrast, the Ti(1)-N(1)

It is interesting to note the variation in Ti(1)-N(1), N(1)-N(2),

and Ti(1)-N(1)-N(2) parameters on going from 2 to 12 (Table

3). There is an apparent trend in lengthening of both the

Ti(1)-N(1) and N(1)-N(2) distances and a decrease in the

Ti(1)-N(1)-N(2) angle along this series. At the same time, as

mentioned above, the diamide-amine Ti(1)-N(3) and Ti(1)-N(4)

distances are decreasing, and the amide nitrogen substituents

are becoming more coplanar with the {Ti(1),N(1),N(2),N(3)}

(83) Either strongly to just one Ph ring (e.g., 2, 11, 12) or in an

intermediate way to both (e.g., 7, 8, 10)

Titanium Hydrazides

Organometallics, Vol. 27, No. 24, 2008 6487

N_ꢀPh₂moiety in 2-12 is therefore understood in terms of

providing the best interaction between the NNR₂and titanium

since this provides the most favorable overlap between the

highest energy donor MO of NNPh₂(2b₂) and the appropriate

lowest energy acceptor orbital at titanium (1b₂in Figure 4).

The second π interaction between Ti and N_Ris formed between

the 1b₁-donor MO of NNPh₂and the 1b₁acceptor at the metal

center.

The trigonal planar amide nitrogens of the real compounds 2

and 10-12 suggest that they are able in principle to donate

two electrons each from the occupied 2p_πAOs (giving each

complex an 18 valence electron count). However, as discussed

previously,^81,82the situation is complicated by symmetry

restrictions, the details of which depend on the orientation of

the lone pairs with respect to the equatorial plane of the complex.

In the model complex 17 two extreme positions can be identiﬁed

as depicted in Figure 5b,c. In Figure 5b the lone pairs are lying

in the equatorial plane and their orientation is deﬁned by a

dihedral angle N_R-Ti-N-R′ of 90°. In Figure 5c the lone pairs

are perpendicular to the equatorial plane and the dihedral angle

N_R-Ti-N-R′ is 0°. In both cases it can be seen that one of

the amide SALCs competes with one of the NNR₂MOs of the

corresponding symmetry for the same π-acceptor orbital at the

metal center. This is either the 1b₁orbital in the case of Figure

5b or the 1b₂orbital when the amide SALCs are oriented as in

Figure 5c. Thus in either extreme (and at all points in between)

there is always a three-center-four-electron π-bonding conﬂict,

and so the real complexes are best described as 16 valence

electron “π-loaded” systems. This type of situation is well

established in imido chemistry with very interesting conse-

quences for metal-ligand multiple bonding properties and

reactivity patterns.^{1,81,82,85-90}

Figure 4. σ-Only frontier MOs (d orbitals only, arbitrary energies)

for hypothetical bis(dialkylamide)-hydrazide system 17. Labels are

for C_2Vsymmetry, and L_axrepresents a nominal axial ligand.^81,82

trigonal plane. In contrast, the orientation of the N_ꢀPh₂moiety

ispreservedinallthesecompounds,withthe{N(1),N(2),C(1),C(7)}

least-squares plane being effectively coplanar with {Ti(1),N(1),

N(2),N(3)} in all instances. To help account for these observa-

tions and to gain a better understanding of their bonding in

general, we analyzed the competition between the amide and

hydrazide nitrogen π-donor orbitals for the Ti(d_π) acceptor MOs

and considered how this would be affected by the orientation

of the various amide N and hydrazide N_ꢀsubstituents.

Molecular Orbital Analysis of Titanium Hydrazides

Containing Dianionic N₃-Donor Ligands. We have previously

reported detailed analyses of the metal(d_π)-N(p_π) bonding in

d⁰imido complexes of the type M(N₂N^py)(NR)(L) (L ) σ

donor).^81,82This serves as a starting point for considering the

π-bonding situation in 2 and 10-12. Figure 4 illustrates the

σ-only frontier MOs (d orbitals only, arbitrary energies) for a

hypothetical trigonal-bipyramidal complex M(NNR₂)(NR′₂)₂-

(L_ax)₂(17), where L_axis a σ-only donor. The 2a₁MO is strongly

σ* antibonding and would not engage in metal-ligand π

interactions. The 1a₁and 1b₁levels are slightly σ* metal-ligand

antibonding and higher in energy than the strictly σ-nonbonding

1b₂and 1a₂levels. Although these four orbitals are all available

for metal-ligand π-bonding interactions with correctly oriented

amide and hydrazide p_πdonors, the 1b₂and 1a₂levels are the

acceptors with the best energy match.

To test these qualitative ideas, the structure of

Ti(NNPh₂)(N₂N^{C ,Me})(py) (2) has been optimized using DFT.

2

The agreement between the computed (see Table S3 of the

Supporting Information) and experimental geometries is very

good. The four highest occupied MOs for 2 are shown in Figure

6. The orientation of the amide donors in 2 lies closer to the

situation described by Figure 5b, and the appearance and

energies of the HOMOs can be interpreted in terms of the

qualitative model described above. The HOMO of 2 is the

bonding combination between π_vof NNPh₂and 1b₂on Ti and

lies 1.9 eV above the corresponding π_h-derived orbital

(HOMO-3). Both the HOMO and HOMO-3 are heavily

ligand-based, and the NNPh₂is best described as a hy-

drazido(2-) ligand forming a Ti-N_Rtriple bond. The very large

destabilization of the HOMO is attributed to the N_R-N_ꢀ

antibonding contribution. The LUMO of 2 is essentially the

LUMO of pyridine with some weak bonding Ti-N interaction

with a 3d AO on Ti.

Figure 5a shows schematically the four highest occupied MOs

of a planar, dianionic [NNR₂]^2-ligand based on previous

EHMO and DFT calculations.^27,31,76,84The 1a₁MO is suitable

for M-N_Rσ bonding, and the 1b₂MO is N_R-N_ꢀπ bonding

and would not be expected to interact to any signiﬁcant extent

with a metal center.^31,76The 1b₁MO (which we abbreviate as

π_h) is essentially N_R-N_ꢀnonbonding and would form one of

the M-N_Rπ bonds. The 2b₂MO (abbreviated π_v) is the HOMO

The HOMO-3 also contains a minor contribution from the

out-of-phase amide nitrogen lone pairs of the N₂N^{C ,Me}ligand

2

(formally the b₁SALC in Figure 5b), which is competing with

π_hfor the single suitable acceptor MO on titanium (equivalent

of [NNR₂]^2-and lies signiﬁcantly higher in energy (g1 eV^31,76

)

(85) Green, J. C.; Green, M. L. H.; James, J. T.; Konidaris, P. C.;

Maunder, G. H.; Mountford, P. J. Chem. Soc., Chem. Commun. 1992, 1361.

(86) Dunn, S. C.; Mountford, P.; Robson, D. A. J. Chem. Soc., Dalton

Trans. 1997, 293.

(87) Parkin, G.; van Asselt, A.; Leahy, D. J.; Whinnery, L.; Hua, N. G.;

Quan, R. W.; Henling, L. M.; Schaefer, W. P.; Santarsiero, B. D.; Bercaw,

J. E. Inorg. Chem. 1992, 31, 82.

than the 1b₁MO because of its N_R-N_ꢀπ*-antibonding nature.

Of these two MOs (π_v(2b₂) and π_h(1b₁)), the former is the

better π donor due to its better energy match with the metal

acceptor orbitals. Donation of electron density from the π_v

strengthens the N_R-N_ꢀbond due to the formal removal of π*-

antibonding electron density. The consistent orientation of the

(88) Schoﬁeld, M. H.; Kee, T. P.; Anhaus, J. T.; Schrock, R. R.; Johnson,

K. H.; Davis, W. M. Inorg. Chem. 1991, 30, 3595.

(89) Benson, M. T.; Bryan, J. C.; Burrell, A. K.; Cundari, T. R. Inorg.

Chem. 1995, 34, 2348.

(90) Lin, Z.; Hall, M. B. Coord. Chem. ReV. 1993, 123, 149.

(84) Kahlal, S.; Saillard, J.-Y.; Hamon, J.-R.; Manzur, C.; Carrillo, D.

New J. Chem. 2001, 25, 231.

6488 Organometallics, Vol. 27, No. 24, 2008

Selby et al.

Figure 5. (a) Four highest occupied MOs of a C_2Vsymmetric dianion [NNR₂]^2-. (b) Amide lone pair SALCs when the lone pair 2p AOs

lie in the equatorial plane. (c) Amide lone pair SALCs when the lone pair 2p AOs are perpendicular to the equatorial plane. Symmetry

labels are given for the C_2Vpoint group.

Figure 7. Simpliﬁed models based on Ti(NR)(NMe₂)₂(py)₂(R )

NMe₂, Me) used to evaluate the effect of changing the orientation

or type of various substituents. The sufﬁxes “90” and “0” refer to

the magnitude of the torsion angles N_R-Ti-N_am-Me (cf. the

torsion angles N(1)-Ti(1)-N(3,4)-Si or N(1)-Ti(1)-N(3,4)-C

in Table 3).

Figure 6. Four highest occupied MOs of Ti(NNPh₂)(N₂NC₂,Me)-

(py) (2) and their DFT-computed energies.

(lone pairs lying in the equatorial plane), and on going from

one to the other, the Ti-N_amdistances progressively shorten

while both the TidN_Rand N_R-N_ꢀlengthen (Table 3). A clear

understanding of these effects will be important for understand-

ing the reactivity patterns of such compounds and will aid

selection of suitable supporting ligands. Therefore, using a series

of simpliﬁed models based on Ti(NR)(NMe₂)₂(py)₂(R ) NMe₂,

Me; Figure 7), we investigated the differing π-donor effects of

the various diamide-amide ligands in 2 and 10-12. We varied

the orientation of the ꢀ-NMe₂group and compared the bonding

between model hydrazide systems (L)TidNNMe₂and the

corresponding imides (L)TidNMe. The model compounds are

shown in Figure 7 along with the abbreviations used. Table 4

lists key geometrical parameters and relative energies for the

various systems.

The model systems (Figure 7; see the Computational Details

for full details) were constrained to C_2Vsymmetry with planar

ꢀ-nitrogens in the case of the hydrazides (in the real systems 2

and 10-12 the phenyl substituents stabilize the trigonal-planar

N_ꢀgeometry). To understand the structural variations (Table

4) between the different models, we analyzed the electronic

structures using the NBO (natural bond orbital) method.^91,92The

advantage of this approach is that it aids interpretation of the

bonding in a transparent and intuitive way in terms of the closest

to 1b₁in Figure 4). The HOMO-2 of 2 represents a Ti-N_amide

π-bonding interaction from the in-phase amide nitrogen lone

pairs of N₂N^{C ,Me}(formally the a₁SALC in Figure 5b), which

2

does not compete with either π_vor π_h. The HOMO-1 is

essentially a Ti-N nonbonding, ligand-based lone pair in nature.

Its major contribution comes from the out-of-phase amide

nitrogen lone pairs of N₂N^{C ,Me}, although there is also a

2

contribution from π_h. Therefore each amide nitrogen of N₂N^{C ,Me}

2

is effectively a two-electron donor as a consequence of the

π-loaded nature of 2. It appears that π_hcompetes more

effectively than the amide b₁SALC for the single b₁-type

acceptor orbital at titanium, consistent with the shorter Ti-N_R

compared to Ti-N_ambond distances (calculated values 1.735

and 2.015 (av) Å, respectively). However, the most remarkable

feature of the bonding in 2 is the signiﬁcant destabilization of

the HOMO. This not only lies substantially above the other

Ti-N_Rπ bond (HOMO-3) but is even ca. 1 eV above the

nonbonding HOMO-1. The high energy of this MO (which is

simultaneously Ti-N_Rbonding and N_R-N_ꢀantibonding) may

be responsible for the very novel reactivity found between 2

and alkynes (eq 2).

NBO Analysis of Model Hydrazido and Imido

Complexes. As mentioned, 2 lies closer to the situation in Figure

5b with the amide lone pairs perpendicular to the equatorial

plane. On the other hand, 12 lies closer to that in Figure 5c

(91) Reed, A. E.; Curtiss, L. A.; Weinhold, F. Chem. ReV. 1988, 88,

899.

Titanium Hydrazides

Organometallics, Vol. 27, No. 24, 2008 6489

Table 4. Key Geometrical Parameters and Relative Energies (E_rel

)

to the stabilization of isomer Hyd_0 compared with Hyd_90,

and the variation in the TidN_Rdistance is a composite of

increased σ(N_R) and π_hdonation and diminished π_vdonation.

for the Various Model Systems Ti(NR)(NMe₂)₂(py)₂(R ) NMe₂, Me)

Studied^a

parameter

Hyd_90 Hyd_0 Hyd_0_rot Imide_90 Imide_0

We have also calculated the NPA (natural population analysis)

charges for the fragments {NR}, {(NMe₂)₂} and {Ti(py)₂} for

the ﬁve model systems Ti(NR)(NMe₂)₂(py)₂(see Table S4 of

the Supporting Information). On going from Hyd_90 to Hyd_0,

the {Ti(py)₂} fragment charge decreases from +1.753e to

+1.585e while the {(NMe₂)₂} charge changes from -1.173e

to -1.000e. The {NNMe₂} charge on the other hand decreases

by only 0.005e. Similar variations were found for the imido

systems and again illustrates the effect of rotating the amide

lone pairs on NMe₂fTi electron donation.

TidN_R(Å)

N_R-N_ꢀ(Å)

Ti-N_am(Å)

Ti-py (Å)

N_R-Ti-N_am

(deg)

1.734

1.331

2.048

2.288

1.729

1.341

1.975

2.306

1.719

1.355

1.980

2.317

1.705

1.708

2.050

2.269

114.2

1.970

2.307

113.7

114.4

114.2

110.6

N_am-Ti-N_am131.1

(deg)

131.5

0.0

138.8

33.6

131.8

93.6

132.5

0.0

E_rel

72.2

(kJ mol^-1

)

^aThe energies for Hyd_90 and Hyd_0_rot are expressed relative to

that of Hyd_0. The E_relof Imide_90 is expressed relative to that of

Imide_0. N_amrefers to the NMe₂amide nitrogens. See Figure 7 for the

geometries.

Rotation of the N_ꢀMe₂Group. All of the experimental

systems 2-12 have the hydrazide NPh₂substituents lying in

the equatorial plane. As discussed above, this should allow best

overlap of the hydrazide π_vorbital with the lower energy

acceptor orbital on Ti. To quantify this analysis, we examined

the model Hyd_0_rot (Figure 7), which has the N_ꢀMe₂group

rotated by 90°. This was accompanied by a destabilization

(∆E_rel) of 33.6 kJ mol^-1(Table 4) relative to Hyd_0, consistent

with experimental observations. The NLMO analysis for

Hyd_0_rot is similar to that for Hyd_0 with regard to the

Ti-N_aminteractions and Ti-N_Rσ interactions. However, there

is a signiﬁcant effect on the Ti-N_Rπ interactions as expected.

Whereas the π_v(Ti-N_R) bond is signiﬁcantly more covalent than

π_h(Ti-N_R) in Hyd_0 (and also Hyd_90) for the reasons

discussed above, in Hyd_0_rot the Ti-N_Rπ bonding is much

more cylindrical. This is because the higher energy NNMe₂π

donor (π_v) now has to interact with a higher energy Ti d_π

acceptor (1b₁, which is also slightly Ti-N_amσ* antibonding),

while the π_hdonor accesses the lower energy Ti 1b₂orbital.

Overall there is a net gain in negative charge of the {NNMe₂}

and {(NMe₂)₂} fragments (Table S4 in the SI) and increase in

positive charge for {Ti(py)₂}, showing that this orientation of

the N_ꢀMe₂group results in a reduced net donation from both

the diamide and hydrazide ligands, despite the slight shortening

of the Ti-N_Rbond length.

N_r-N_ꢀInteraction. The weakening of the N_R-N_ꢀbond is

an important aspect of metal hydrazide chemistry from the point

of view of its eventual cleavage. Several reports from group 4

have found facile cleavage of this bond, especially in π-loaded

systems.^29,34,35Experimentally (2-12) and computationally it

is found that rotation of the amide lone pairs results in a slightly

longer N_R-N_ꢀdistance (∆ ≈ 0.01 Å). Furthermore, the model

Hyd_0_rot shows a further ca. 0.015 Å lengthening of N_R-N_ꢀ

and shortening of Ti-N_R. The NBO and NLMO analyses offer

an explanation of these trends in Ti-N_Rand N_R-N_ꢀdistances.

As mentioned, the NBO analyses of Hyd_90, Hyd_0, and

Hyd_0_rot are consistent with the Lewis structure 18, in which

there is a formal “lone pair” mainly localized on N_ꢀ. This is

conﬁrmed by the NLMO analysis for this orbital (Table 5; %N

ca. 87-89%). However, this also ﬁnds that the electron pair is

partially delocalized into the antibonding π*_v(Ti-N_R) orbital,

accounting for the nonzero %Ti and %N_Rcontributions to the

“lone pair” NLMO (see Figure S2 of the Supporting Informa-

tion). Clearly, this N_ꢀfπ*_v(Ti-N_R) interaction therefore si-

multaneously weakens the Ti-N_Rbond and strengthens the

N_R-N_ꢀbond. The extent of the interaction will vary according

to the properties (energy and %N_Rcharacter) of the π*_v(Ti-N_R)

orbital. On going from Hyd_90 to Hyd_0, the π*_v(Ti-N_R)

orbital is destabilized due to competition with the amide lone

pairs for the Ti 1b₂acceptor orbital, and in Hyd_0_rot the

NNMe₂π_vdonor has to use the higher energy 1b₁orbital.

idealized Lewis structure. For all of the models

Ti(NR)(NMe₂)₂(py)₂the NBO analysis found NLMOs (natural

localized molecular orbitals) corresponding to Ti-N_Rand

Ti-N_amσ bonds, two Ti-N_Rπ bonds (one for each of π_vand

π_h), and a Ti-N_amπ bond to each NMe₂. For the hydrazide

models a “lone pair” essentially localized (see below) on N_ꢀ

was also found. These NLMOs clearly correspond to the

expected Lewis structures 18 and 19 for the hydrazido(2-) and

imido models, respectively. The main % atomic contributions

to these NLMOs are given in Table 5.

Orientation of Amide Lone Pairs. Going from model

Hyd_90 to Hyd_0 rotates the NMe₂lone pairs from lying in

the equatorial plane to being perpendicular to it. This models

the general trend on going from 2 to 12 in the real systems.

For the model systems this results in a very slight shortening

of TidN_R, a signiﬁcant shortening of Ti-N_am, and a lengthening

of N_R-N_ꢀ. There is an associated stabilization of 72.2 kJ mol^-1

.

The shorter Ti-N_ammay be attributed to the better match

between the amide lone pairs and the 1b₂and 1a₂acceptor MOs

of Ti (cf. Figure 4) than the 1a₁and 1b₁alternatives. Consistent

with this, the NBO analysis (Table 5) shows an increased

covalency in both the σ and π Ti-N_amNLMOs.⁹³The

competition between the amide lone pairs and the hydrazide π_v

for the Ti 1b₂acceptor MO results in a diminished π_v(Ti-N_R)

interaction. Conversely, the corresponding π_h(Ti-N_R) bonding

is improved since the Ti 1b₁no longer overlaps with the amide

lone pairs. Importantly, the higher %Ti in σ(Ti-N_R) shows that

the NNMe₂ligand is a better σ donor in Hyd_0 because there

is no longer destabilization of the Ti 1a₁orbital (Ti-N_Rσ*

antibonding) through π donation from the amide a₁lone pair.

Thus there is a combination of σ and π effects that contribute

(92) Weinhold,F.; Landis, C. R. Valency and Bonding: A Natural Bond

Orbital Donor-Acceptor PerspectiVe; Cambridge University Press: Cam-

bridge, 2005.

(93) Because of the higher electronegativity of nitrogen, all of the

occupied NLMOs have a higher %N contribution than %Ti. Therefore an

increased %Ti and decreased %N contribution for a given NLMO on going

from one model to another is evidence of improved covalency in the orbital

under consideration. From the perspective of fragment MO interactions this

relates to a decreased energy separation and/or improved overlap between

the participating orbitals.

6490 Organometallics, Vol. 27, No. 24, 2008

Selby et al.

Table 5. NLMO Analysis for the Various Model Systems Ti(NR)(NMe₂)₂(py)₂(R ) NMe₂, Me) Studied^a

electron pair

Hyd_90

Hyd_0

Hyd_0_rot

Imide_90

Imide_0

σ(Ti-N_R)

11.9% Ti

86.2% N_R

36.3% Ti

62.1% N_R

16.7% Ti

80.7% N_R

5.2% Ti

15.7% Ti

83.6% N_R

32.7% Ti

65.6% N_R

18.4% Ti

78.8% N_R

4.5% Ti

15.8% Ti

83.0% N_R

28.5% Ti

69.5% N_R

22.5% Ti

75.3% N_R

3.6% Ti

16.1% Ti

83.2% N_R

27.6% Ti

69.9% N_R

21.2% Ti

76.3% N_R

18.4% Ti

81.2% N_R

25.4% Ti

72.0% N_R

22.0% Ti

75.2% N_R

π_v(Ti-N_R)

π_h(Ti-N_R)

“lone pair” (N_ꢀ)

3.3% N_R

86.7% N_ꢀ

9.2% Ti

87.9% N_am

8.9% Ti

2.5% N_R

88.3% N_ꢀ

11.9% Ti

85.6% N_am

9.2% Ti

1.8% N_R

89.0% N_ꢀ

11.3% Ti

86.1% N_am

9.8% Ti

σ(Ti-N_am

)

9.0% Ti

87.3% N_am

8.1% Ti

12.0% Ti

85.4% N_am

8.8% Ti

π(Ti-N_am

)

85.2% N_am

83.6% N_am

83.1% N_am

85.7% N_am

83.4% N_am

^aThe % contributions are for the stated atoms in the respective NLMOs. See Figure 7 for the geometries. NPA (natural population analysis) charges

for the {Ti(py)₂}, {NR}, and {(NMe₂)₂} fragments are given in Table S4 of the Supporting Information.

Table 6. X-ray Data Collection and Processing Parameters for Ti(NNPh₂)Cl₂(py)₃(7), Ti(NNPh₂)Cl₂(py′)₃(8), Ti(NNPh₂)(N₂N^{C ,Me})(py) (2),

2

Ti(NNPh₂)(N₂N^{C ,SiMe})(py) (10), Ti(NNPh₂)(N₂N^{C ,Me})(py) (11), and Ti(NNPh₂)(N₂N^py)(py) (12)

2

3

7

8

2

10

11

12

empirical formula

fw

temp/K

C₂₇H₂₅Cl₂N₅Ti

538.33

150

C₃₉H₄₉Cl₂Ti

706.66

150

C₂₈H₄₄N₆Si₂Ti

568.77

150

C₃₀H₅₀N₆Si₃Ti

626.93

150

C₃₀H₄₈N₆Si₂Ti

596.82

150

C₃₂H₄₄N₆Si₂Ti

616.81

150

wavelength/Å

space group

a/Å

b/Å

c/Å

0.71073

C2/c

14.0110(4)

19.8007(6)

9.4425(3)

90

93.9572(12)

90

2613.36(14)

4

0.71073

P2₁/c

22.2899(2)

17.6264(2)

19.8028(2)

90

91.9807(6)

90

7775.69(14)

8

0.71073

P₁

0.71073

P1

0.71073

P 2₁/c

11.5182(2)

14.5639(3)

20.0751(4)

90

100.6813(8)

90

3309.25(11)

4

0.71073

P1

j

10.1794(5)

11.1800(7)

13.8952(8)

89.726(3)

77.178(2)

85.912(3)

1537.9(2)

2

1.228

0.383

0.0728

0.0800

10.3113(2)

11.9118(3)

14.0949(3)

92.1330(10)

90.7564(10)

90.6550(10)

1729.75(7)

2

1.204

0.380

0.0350

0.0369

10.1894(2)

10.3302(2)

17.9754(4)

95.5849(10)

103.2464(11)

111.5826(10)

1677.77(6)

2

1.221

0.357

0.0524

0.0448

R/deg

ꢀ/deg

γ/deg

V/Å³

Z

d(calcd)/Mg · m^-3

abs coeff/mm^-1

R indices: R₁)

1.368

0.557

0.0337

0.0367

1.207

0.390

0.0397

0.0474

1.198

0.359

0.0429

0.0387

[I > 3σ(I)]:^aR_w

)

^aR₁) ∑||F_o| - |F_c||/∑|F_o|; R_w) ꢀ{∑w(|F_o| - |F_c|)²/∑w|F_o|²}.

Therefore the N_ꢀfπ*_v(Ti-N_R) interaction is diminished along

this distortion coordinate, consistent with the computed bond

length trends. Indeed, the second-order perturbation energy

associated with the N_ꢀfπ*_v(Ti-N_R) interaction decreases from

Hyd_90 to Hyd_0 and ﬁnally to Hyd_0_rot (141.9, 124.1, and

103.4 kJ mol^-1, respectively) due to both a larger energy gap

(0.26, 0.27, and 0.28 au, respectively) and a smaller Fock matrix

element F(i,j) (0.086, 0.080, and 0.074 au, respectively). This

further explains the trends in the composition of the lone pair

on N_ꢀin Table 5.

Comparison with Imido Systems. The NLMO analysis for

the imido systems Imide_90 and Imide_0 reveals analogous

trends, and in fact the latter is 93.6 kJ mol^-1more stable. The

Ti-N_amdistances shorten signiﬁcantly (∆(Ti-N_am) ) -0.08

Å), whereas (as for the hydrazido compounds) no signiﬁcant

change is registered for Ti-N_R. These small changes in Ti-N_R

for the hydrazido and imido systems show there is no signiﬁcant

net electronic driving force for Ti-N_Rbond length change,

despite varying the supporting diamide ligand set. Examination

of the π_v(Ti-N_R) and π_h(Ti-N_R) NLMOs shows a much more

cylindrical π-bonding manifold than is found in the correspond-

ing hydrazide systems. As expected, the π_h(Ti-N_R) NLMO is

more based on N_Rowing to the higher energy of the Ti 1b₁

acceptor MO.

0.032(4)-0.035(3) Å).^69,81One Ti-N_Rbond-lengthening effect

is due to the N_ꢀfπ*_v(Ti-N_R) interaction noted above. A second

contribution is the reduced σ-donor ability of NNR₂compared

to NR because of the increased electron-withdrawing inductive

effect of N_ꢀR₂compared to alkyl. This is evident in the %Ti,N

contributions for the σ(Ti-N_R) NLMOs in the two types of

model system. The NLMOs for the hydrazido models are more

localized on N_Rin each case compared to the imido analogues.

A similar conclusion with regard to σ-donor abilities was

reached previously (based on fragment MO Mulliken population

analyses) in comparing the Ti-N_Rbonding in the model

complexes Ti(NMe){HC(pz)₃}Cl₂and Ti(NMe){HC-

(pz)₃}Cl₂.³¹

Comparison of Model and Experimental Systems. The

shortening of the av Ti-N_amdistances from 2.028 to 1.974 Å

from 2 to 12 (Table 3) is reproduced well by the models Hyd_90

(Ti-N_am) 2.048 Å) and Hyd_0 (Ti-N_am) 1.975 Å). The

increase in the N_R-N_ꢀdistance is also well reproduced (0.01

Å in both experiment and theory). However, on going from 2

to 12 in the experimental systems, the Ti-N_Rdistance lengthens

from 1.733(5) to 1.759(2) Å, whereas in the model systems there

is a very small decrease from 1.734 to 1.729 Å. This apparent

discrepancy can be attributed to signiﬁcant steric interactions

between the hydrazide phenyl substituents and the equatorial

NSiMe₃groups, as the latter move up into the equatorial plane

and become more oriented toward the NNPh₂. This effect is

probably the origin of the relatively large jump of ca. 0.02 Å

The Ti-N_Rbond distances in Imide_90 and Imide_0 are

ca. 0.020-0.030 Å shorter than the corresponding values for

Hyd_90 and Hyd_0. This is consistent with the differences

in Ti-N_Rbond length between Ti(NNPh₂)(N₂N^{C ,Me})(py) (2)

2

found in the real systems Ti(“N₂N”)(NR)(py) (“N₂N” )

t

N₂N^{C ,SiMe}or N₂N^py; R ) Bu or NNPh₂; ∆(Ti-N_R) )

and Ti(NNPh₂)(N₂N^{C ,SiMe})(py) (10) as the diamide-amine

2

3

2

3

Titanium Hydrazides

Organometallics, Vol. 27, No. 24, 2008 6491

were dried over potassium (C₆D₆) or P₂O₅(CD₂Cl₂), distilled under

reduced pressure, and stored under dinitrogen in Young’s Teﬂon

valve ampules.

N-Me group is replaced by SiMe₃. Thus a balance of real system

steric factors and underlying electronic factors leads to the

observed structural trends and (most likely) controls the

reactivity patterns.

Solution NMR samples were prepared under a dinitrogen

atmosphere in a drybox, in 5 mm Wilmad NMR tubes possessing

1

Young’s Teﬂon valves. H and ¹³C NMR spectra were recorded

Conclusions

on a Varian Mercury 300 spectrometer at ambient temperature. All

spectra were internally referenced to either residual protio-solvent

(¹H) or solvent (¹³C) resonances and are reported relative to

tetramethylsilane (δ ) 0 ppm). Chemical shifts are quoted in δ

Far from being simple “imido analogues”, terminal group 4

hydrazides can exhibit novel and potentially catalytic reactivity

of the MdN-NR₂functional group.^{29,34,35,94,95}In this contribu-

tion, based on new readily available synthons, we have reported

a comprehensive study of the synthesis and bonding of a new

family of titanium diphenyl hydrazides using the important

diamide-amine class of supporting ligands, among others.

Through this series, the ligand coordination number, steric

requirements, chelate ring size, and amide π-donor ability may

be controlled. Detailed structural and computational studies of

a series of closely related model complexes have deﬁned the

parameters, both steric and electronic, which inﬂuence the

energies of these systems and the Ti-N_am, N_R-N_ꢀ, and Ti-N_R

distances. Depending on their relative orientations, competition

between the amide and hydrazide nitrogen 2p_πlone pairs

modiﬁes the π_v(Ti-N_R) NLMOs, which in turn leads to poorer

N_R-N_ꢀπ bonding and a less stabilized ꢀ-nitrogen. The

weakening of the N_R-N_ꢀbond through N_amfTi π donation to

the relevant d_πAO echoes the dramatic weakening that occurs

in the Schrock and Chatt catalytic cycles on addition of an

external electron, just prior to proton-induced cleavage.

(ppm) and coupling constants in Hz. Where necessary, ¹H and ¹³

C

assignments were assisted by the use of two-dimensional H-¹H

and ¹H-¹³C correlation experiments. IR spectra were recorded on

a Perkin-Elmer 1710 FTIR spectrometer. Samples were prepared

in a drybox as Nujol mulls between NaCl plates. IR data are quoted

as wavenumbers (cm^-1) within the range 4000-400 cm^-1. Mass

spectra were recorded by the departmental service, and elemental

analyses were carried out by the Elemental Analysis Service at the

London Metropolitan University.

1

Starting Materials. The compounds Ti(N^tBu)Cl₂(py)₃,⁶⁶

98

3

[Ti(NNMe₂)Cl₂(py)₂]₂,³¹

Li₂N₂N^{C ,Me 97}

,

Li₂N₂N^{C ,SiMe}

,

3

2

Li₂N₂N^{C ,Me 97}

,

Li₂N₂N^py,⁹⁹Li₂N₂NN′,⁷⁸Li₂Me₄taa,¹⁰⁰and

3

Na₂O₂NN′ · 0.16(THF)¹⁰¹were prepared according to the literature

methods. Ti(N^tBu)Cl₂(py′)₃was prepared by analogy with the

literature methods for Ti(N^tBu)Cl₂(py)₃and Ti(N^tBu)Cl₂(py′)₂.^66,74

Diphenylhydrazine was obtained from Sigma-Aldrich as the

hydrogen chloride salt, from which the free hydrazine was obtained

by basiﬁcation, drying, and removal of residual solvent, followed

by distillation under inert atmospheric conditions. Pyridine was dried

over freshly ground CaH₂and distilled before use. Other reagents

were obtained commercially and used as received.

Ligands that deploy steric bulk in the vicinity of the

TidNNPh₂functional group can also lead to lengthened Ti-N_R

bonds. Our synthetic and computational results also shed light

on the high reactivity so far found for the terminal zirconium

Ti(NNPh₂)Cl₂(py)₃(7). To a stirred solution of Ti(N^tBu)Cl₂(py)₃

(5.00 g, 11.7 mmol) in benzene (50 mL) was added H₂NNPh₂(2.15

g, 11.7 mmol) in benzene (10 mL) dropwise over 15 min. The

orange solution was stirred for 14 h, resulting in a dark brown

solution and a yellow precipitate. The solution was ﬁltered, and

the precipitate washed with pentane (3 × 20 mL) and ﬁltered. The

yellow solid product was dried in Vacuo. Yield: 5.60 g (89%).

Diffraction-quality crystals were grown from a saturated benzene

solution.

hydrazides

Cp₂Zr(NNPh₂)(DMAP)²⁹

and

Zr(N₂N^py*)(NNPh₂)(py), which have reactive N_R-N_ꢀbonds that

undergo unique insertion reactions.³⁵In these systems we again

expect the inherent “π-loaded” nature⁹⁶to result in destabilized

π_v(Zr-N_R) NLMOs just as for the diamide-amine-supported

titanium systems described above. Furthermore, the less stabi-

lized 4d_πAO of Zr is also expected to lead to poorer

N_ꢀfπ*_v(Zr-N_R) interactions. Indeed, the N_R-N_ꢀbond distance

of 1.398(3) Å in Zr(N₂N^py*)(NNPh₂)(py), which also features

3

¹H NMR (CD₂Cl₂, 299.9 MHz, 293 K): δ 9.01 (4 H, d, J ) 5

Hz, o-H of cis-NC₅H₄), 8.59 (2 H, br m, coupling not resolved,

o-H of trans-NC₅H₄), 7.86 (2 H, t, ³J ) 7.7 Hz, p-H of cis-NC₅H₄),

7.68 (1 H, br m, coupling not resolved p-H of trans-NC₅H₄), 7.41

(4 H, d, ³J ) 7.3 Hz, o-C₆H₅), 7.29-7.13 (10 H, overlapping peaks,

m-C₆H₅, m-H of cis- and trans-NC₅H₄), 6.99 (2 H, t, ³J ) 7.3 Hz,

p-C₆H₅). IR (NaCl plates, Nujol mull, cm^-1): ν 1653 (w), 1604

(s), 1594 (s), 1586 (s), 1559 (w), 1543 (m), 1487 (m), 1447 (s),

1337 (w), 1313 (w), 1269 (s), 1219 (m), 1168 (w), 1154 (w), 1108

(w), 1070 (m), 1043 (m), 1028 (w), 1013 (m), 842 (w), 753 (m),

693 (s), 636 (w). Anal. Found (calcd for C₂₇H₂₅Cl₂N₅Ti): C, 60.0

(60.2); H, 4.8 (4.7); N, 12.9 (13.0). A ¹³C NMR spectrum could

not be obtained due to the low solubility of 7 in suitable solvents.

t

bulky SiMe₂Bu substituents, is the longest known for a terminal

diphenyl hydrazide.

Finally we recall that terminal metal hydrazides are commonly

ﬁrst-formed products in the functionalization of molecular

dinitrogen.^7-12Incorporating some of these above-mentioned

design principles into transition metal compounds suitable for

N₂activation may lead ultimately to new dinitrogen function-

alization processes via activated terminal hydrazides.

Experimental Section

Alternative Synthesis of [Ti(NNPh₂)Cl₂(py)₂]₂(4).

Ti(NNPh₂)Cl₂(py)₃(7, 0.50 g, 0.93 mmol) was held under a

dynamic vacuum (4 × 10^-2mbar) for 14 h, then washed with

pentane (2 × 25 mL) to quantitatively yield [Ti(NNPh₂)Cl₂(py)₂]₂

General Methods and Instrumentation. All manipulations were

carried out under an inert atmosphere of argon or dinitrogen using

standard Schlenk-line or drybox procedures. Solvents were predried

over activated 4 Å molecular sieves and reﬂuxed over sodium

(toluene), sodium/potassium (pentane, diethyl ether), potassium

(tetrahydrofuran), or calcium hydride (dichloromethane) under a

dinitrogen atmosphere and collected by distillation. Alternatively,

solvents were degassed by sparging with dinitrogen and dried by

passing through a column of activated alumina. Deuterated solvents

(97) Ward, B. D.; Dubberley, S. R.; Maisse-Francois, A.; Gade, L. H.;

Mountford, P. Dalton Trans. 2002, 4649.

(98) Cloke, F. G. N.; Hitchcock, P. B.; Love, J. B. J. Chem. Soc., Dalton

Trans. 1995, 25.

(99) Galka, C. H.; Tro¨sch, D. J. M.; Schubart, M.; Gade, L. H.;

Radojevic, S.; Scowen, I. J.; McPartlin, M. Eur. J. Inorg. Chem. 2000,

2577.

(94) Herrmann, H.; Wadepohl, H.; Gade, L. H. Dalton Trans. 2008,

2111.

(95) Mindiola, D. J. Angew. Chem., Int. Ed. 2008, 47, 1557.

(96) Green, J. C. Chem. Soc. ReV. 1998, 27, 263.

(100) Black, D. G.; Swenson, D. C.; Jordan, R. F.; Rogers, R. D.

Organometallics 1995, 14, 3539.

(101) Toupance, T.; Dubberley, S. R.; Rees, N. H.; Tyrrell, B. R.;

Mountford, P. Organometallics 2002, 21, 1367.

6492 Organometallics, Vol. 27, No. 24, 2008

Selby et al.

(4) as a yellow-green solid. The NMR data for 4 were consistent

0.36 g (26%). Diffraction-quality crystals were grown from a

with those previously reported.³¹

saturated pentane solution at -30 °C.

3

¹H NMR (C₆D₆, 299.9 MHz, 293 K): δ 8.66 (2 H, d, J ) 6.4

Ti(NNPh₂)Cl₂(py′)₃(8). To

a

stirred solution of

Ti(N^tBu)Cl₂(py′)₃(2.00 g, 3.37 mmol) in benzene (50 mL) was

added H₂NNPh₂(0.62 g, 3.37 mmol) in benzene (10 mL) dropwise

over 15 min. The orange solution was stirred for 14 h, resulting in

a dark brown solution. Volatiles were removed under reduced

pressure, and the dark brown solid was washed with pentane (3 ×

20 mL) and ﬁltered. The yellow solid product was dried in Vacuo.

Yield: 1.75 g (74%). Diffraction-quality crystals were grown from

a saturated pentane solution.

Hz, o-NC₅H₅), 7.78 (4 H, d, ³J ) 8.3 Hz, o-C₆H₅), 7.28 (4 H, app.

3

t, app. J ) 7.1 and 8.3 Hz, m-C₆H₅), 6.89 (2 H, t, J ) 7.1 Hz,

3

p-C₆H₅), 6.72 (1 H, t, J ) 7.6 Hz, p-NC₅H₅), 6.38 (2 H, app. t,

3

app. J ) 6.4 and 7.6 Hz, m-NC₅H₅), 3.76 (2 H, m, CH₂NSiMe₃

inner protons cis to py), 3.21 (2 H, m, CH₂NSiMe₃outer protons

cis to py), 3.06 (2 H, m, CH₂NSiMe₃inner protons trans to py),

2.46 (2 H, m, CH₂NSiMe₃outer protons trans to py), 0.22 (9 H, s,

SiMe₃trans to py), 0.03 (18 H, s, SiMe₃cis to py). ¹³C{¹H} NMR

(C₆D₆, 75.4 MHz, 293 K): δ 152.5 (o-NC₅H₅), 147.7 (i-C₆H₅), 138.4

(p-NC₅H₅), 129.2 (m-C₆H₅), 123.9 (m-NC₅H₅), 122.4 (p-C₆H₅),

119.9 (o-C₆H₅), 59.5 (CH₂NSiMe₃trans to py), 48.7 (CH₂SiMe₃

cis to py), 1.8 (SiMe₃cis to py), 0.6 (SiMe₃trans to py). IR (NaCl

plates, Nujol mull, cm^-1): ν 1600 (m), 1583 (m), 1489 (m), 1444

(s), 1345 (w), 1306 (w), 1295 (w), 1276 (m), 1256 (m), 1245 (m),

1211 (w), 1167 (w), 1067 (m), 1041 (w), 1024 (m), 1012 (w), 993

(w), 939 (s), 908 (w), 872 (m), 840 (s), 797 (w), 779 (m), 754 (w),

745 (m), 696 (m), 678 (w), 635 (w), 624 (w). EI-MS: m/z 168

(98%) [NPh₂]⁺, 77 (77%) [C₆H₆]⁺. Anal. Found (calcd for

3

¹H NMR (C₆D₆, 299.9 MHz, 293 K): δ 9.34 (4 H, d, J ) 6.5

3

Hz, o-H of cis-NC₅H₄), 9.03 (2 H, d, J ) 4.7 Hz, o-H of trans-

3

NC₅H₄), 8.19 (4 H, d, J ) 7.7 Hz, o-C₆H₅), 7.29 (4 H, dd, J )

7.7 and 7.0 Hz, m-C₆H₅), 6.88 (2 H, t, ³J ) 7.0 Hz, p-C₆H₅), 6.65

(2 H, d, ³J ) 4.7 Hz, m-H of trans-NC₅H₄), 6.60 (4 H, d, ³J ) 6.5

Hz, m-H of cis-NC₅H₄), 0.90 (9 H, s, trans-CMe₃), 0.81 (18 H, s,

cis-CMe₃). ¹³C{¹H} NMR (C₆D₆, 75.4 MHz, 293 K): δ 162.2 (p-C

of cis-NC₅H₄), 160.1 (p-C of trans-NC₅H₄), 152.9 (o-C of cis-

NC₅H₄), 152.2 (o-C of trans-NC₅H₄), 145.8 (i-C₆H₅), 129.4 (m-

C₆H₅),123.3 (p-C₆H₅), 121.2 (m-C of cis-NC₅H₄), 120.6 (m-C of

trans-NC₅H₄), 120.0 (o-C₆H₅), 34.9 (cis-CMe₃), 34.7 (trans-CMe₃),

30.5 (trans-CMe₃), 30.3 (cis-CMe₃). IR (NaCl plates, Nujol mull,

cm^-1): ν 1613 (s), 1585 (s), 1543.7 (s), 1461 (s), 1418 (m), 1368

(m), 1314 (s), 1274 (s), 1230 (s), 1201 (m), 1184 (w), 1172 (w),

1070 (s), 1022 (m), 995 (w), 843 (s), 803 (s), 750 (s), 725 (s) 695

(m), 637 (w), 572 (w). Anal. Found (calcd for C₃₉H₄₉Cl₂N₅Ti): C,

C₃₀H₅₀N₆Si₃Ti): C, 57.4 (57.5); H, 7.9 (8.0); N, 13.4 (13.4).

Ti(NNPh₂)(N₂N^{C ,Me})(py) (11). To

a stirred mixture of

3

[Ti(NNPh₂)Cl₂(py)₂]₂(1.00 g, 2.18 mmol) and Li₂N₂N^{C ,Me}(0.657 g,

2.18 mmol) was added toluene (60 mL), all at -78 °C. The green

suspension was allowed to warm to room temperature and stirred for

a further hour, resulting in a dark brown solution. Volatiles were

removed under reduced pressure, and the dark oily solid was extracted

into Et₂O (2 × 30 mL) and ﬁltered. Volatiles were removed under

reduced pressure to yield a brown solid, which was washed with hexane

(3 × 5 mL) cooled to -78 °C. The brown solid product was dried in

Vacuo. Yield: 0.53 g (41%). Diffraction-quality crystals were grown

from a saturated hexane solution at -30 °C.

3

66.4 (66.3); H, 7.1 (7.0); N, 9.8 (9.9).

Ti(NNPh₂)(N₂N^{C ,Me})(py) (2). To

a stirred mixture of

2

Ti(NNPh₂)Cl₂(py)₃(2.00 g, 3.72 mmol) and Li₂N₂N^{C ,Me}(1.01 g,

3.72 mmol) was added toluene (60 mL), all at -78 °C. The green

suspension was allowed to warm to room temperature and stirred

for a further hour, resulting in a dark brown solution. Volatiles

were removed under reduced pressure, and the dark oily solid was

extracted into Et₂O (3 × 30 mL) and ﬁltered twice. Volatiles were

removed under reduced pressure to yield a brown solid, which was

washed with pentane (3 × 5 mL) cooled to -78 °C. The green

solid product was dried in Vacuo. Yield: 1.37 g (65%). Diffraction-

quality crystals were grown from a saturated pentane solution at

-30 °C.

3

¹H NMR (C₆D₆, 299.9 MHz, 293 K): δ 8.66 (2 H, d, J ) 6.4

Hz, o-NC₅H₅), 7.67 (4 H, d, ³J ) 7.6 Hz, o-C₆H₅), 7.25 (4 H, app.

3

t, app. J )7.6 and 7.8 Hz, m-C₆H₅), 6.90 (3 H, m, p-C₆H₅and

p-NC₅H₅), 6.61 (2 H, app. t, app. ³J ) 6.4 and 7.6 Hz, m-NC₅H₅),

3.93 (2 H, m, CH₂NSiMe₃inner protons), 3.28 (2 H, m, CH₂NSiMe₃

outer protons), 2.55 (3 H, s, NMe), 2.23 (2 H, m, CH₂NMe inner

protons), 1.99, (2 H, m, CH₂NMe outer protons), 1.33 (4 H, m, 2

× CH₂), 0.23 (18 H, s, SiMe₃). ¹³C{¹H} NMR (C₆D₆, 75.4 MHz,

293 K): δ 151.8 (o-NC₅H₅), 147.6 (i-C₆H₅), 136.6 (p-NC₅H₅), 129.3

(m-C₆H₅), 123.7 (m-NC₅H₅), 122.3 (p-C₆H₅), 119.3 (o-C₆H₅), 60.6

(CH₂NMe), 50.7 (NMe), 46.5 (CH₂NSiMe₃), 30.8 (CH₂), 2.5

(SiMe₃). IR (NaCl plates, Nujol mull, cm^-1): ν 1596 (m), 1585

(m), 1488 (s), 1457 (s), 1442 (m), 1297 (m), 1255 (m), 1245 (s),

1209 (w), 1158 (w), 1146 (w), 1135 (m), 1078 (m), 1061 (m), 1036

(m), 1013 (m), 991 (w), 961 (w), 947 (m), 920 (w), 888 (m), 879

(m), 854 (s), 828 (s), 797 (m), 776 (m), 756 (m), 745 (s), 695 (s),

662 (w), 625 (m), 614 (m). EI-MS: m/z 517 (11%) [M - py]⁺,

168 (100%) [NPh₂]⁺. Anal. Found (calcd for C₃₀H₄₈N₆Si₂Ti): C,

60.4 (60.4); H, 8.2 (8.1); N, 13.9 (14.1).

3

¹H NMR (C₆D₆, 299.9 MHz, 293 K): δ 8.68 (2 H, d, J ) 6.4

Hz, o-NC₅H₅), 7.87 (4 H, d, ³J ) 7.6 Hz, o-C₆H₅), 7.31 (4 H, app.

3

t, app. J ) 7.1 and 7.6 Hz, m-C₆H₅), 6.91 (2 H, t, J ) 7.1 Hz,

3

p-C₆H₅), 6.68 (1 H, t, J ) 7.6 Hz, p-NC₅H₅), 6.38 (2 H, app. t,

app. ³J ) 6.4 and 7.6 Hz, m-NC₅H₅), 3.56, 3.39 (2 × 2 H, 2 × m,

CH₂NSiMe₃), 2.74, 2.56 (2 × 2 H, 2 × m, CH₂NMe), 2.62 (3 H,

s, NMe), 0.03 (18 H, s, SiMe₃). ¹³C{¹H} NMR (C₆D₆, 75.4 MHz,

293 K): δ 153.9 (o-NC₅H₅), 147.9 (i-C₆H₅), 138.7 (p-NC₅H₅), 129.3

(m-C₆H₅), 124.0 (m-NC₅H₅), 122.2 (p-C₆H₅), 119.1 (o-C₆H₅), 63.6

(CH₂NMe), 50.7 (NMe), 48.2 (CH₂NSiMe₃), 1.9 (SiMe₃). IR (NaCl

plates, Nujol mull, cm^-1): ν 1600 (m), 1586 (s), 1486 (s), 1445

(s), 1308 (w), 1298 (w), 1240 (s), 1211 (w), 1168 (w), 1078 (m),

1064 (m), 1041 (m), 1012 (w), 989 (w), 919 (s), 841 (s), 799 (m),

745 (s), 697 (s), 636 (w), 621 (w). EI-MS: m/z 489 (16%) [M -

py]⁺, 168 (73%) [NPh₂]⁺. Anal. Found (calcd for C₂₈H₄₄N₆Si₂Ti):

C, 59.0 (59.1); H, 7.7 (7.8); N, 14.7 (14.8).

Ti(NNPh₂)(N₂N^py)(py) (12). To

a

stirred slurry of

Ti(NNPh₂)Cl₂(py)₃(1.67 g, 3.10 mmol) in toluene (50 mL), cooled

to -78 °C, was added Li₂N₂N^py(1.00 g, 3.10 mmol) in toluene

(10 mL) dropwise over 15 min. The green-yellow slurry was

allowed to warm to room temperature and stirred for 14 h, resulting

in a dark brown solution. Volatiles were removed under reduced

pressure, and the dark brown solid was extracted into Et₂O (3 ×

30 mL) and ﬁltered. Volatiles were removed under reduced pressure

to yield a brown solid, which was stirred with pentane (2 × 50

mL) and ﬁltered. The dark green solid product was dried in Vacuo.

Yield: 1.40 g (73%). Diffraction-quality crystals were grown from

slow evaporation of a diethyl ether solution.

Ti(NNPh₂)(N₂N^{C ,SiMe})(py) (10). To a stirred mixture of

2

3

[Ti(NNPh₂)Cl₂(py)₂]₂(1.00 g, 2.18 mmol) and Li₂N₂N^{C ,SiMe}(0.72

g, 2.18 mmol) was added toluene (60 mL), all at -78 °C. The

green suspension was allowed to warm to room temperature and

stirred for a further 90 min, resulting in a dark brown solution.

Volatiles were removed under reduced pressure, and the dark

oily solid was extracted into Et₂O (2 × 30 mL) and ﬁltered.

Volatiles were removed under reduced pressure to yield a brown

solid, which was washed with pentane (3 × 5 mL) cooled to

-78 °C. The green solid product was dried in Vacuo. Yield:

2

3

¹H NMR (C₆D₆, 299.9 MHz, 293 K): δ 9.44 (1H, d, py-H⁶, J

3

) 4.1 Hz), 9.00 (2H, d, o-NC₅H₄, J ) 4.1 Hz), 7.82 (4 H, d, J

3

) 7.6 Hz, o-C₆H₅), 7.29 (4 H, app. t, app. J )7.6 and 8.3 Hz,

Titanium Hydrazides

Organometallics, Vol. 27, No. 24, 2008 6493

4

m-C₆H₅), 7.03 (1 H, ddd, py-H⁴, ³J ) 7.6 Hz, ⁴J ) 1.8 Hz)

6.93-6.85 (4 H, overlapping m, p-C₆H₅, p-NC₅H₄, and py-H³),

4-C₆H₂Me₂), 6.62 (1 H, broad, 4-C₅H₅N), 6.56 (2 H, d, J ) 2.1,

3

6-C₆H₂Me₂), 6.38 (1 H, app. t, app. J ) 7.9 Hz, 4-C₅H₄N), 6.34

3

6.61 (2 H, app t, m-NC₅H_4,app. J ) 6.8 Hz), 6.41 (1 H, dd, py-

(2 H, br, 3-C₅H₅N), 6.06 (1 H, app. t, app. ³J ) 6.3 Hz, 5-C₅H₄N),

3

2

H⁵, ³J ) 7.6 Hz, ⁴J ) 1.2 Hz), 3.79 (2 H, d, CHHNSi₃, ²J ) 12.3

5.70 (1 H, d, J ) 7.9 Hz, 3-C₅H₄N), 3.51 (2 H, d, J ) 12.5 Hz,

NCH₂Ar distal to C₅H₄N), 3.20 (2 H, s, NCH₂C₅H₄N), 2.59 (6 H,

s, 3-C₆H₂Me₂), 2.58 (2 H, d, NCH₂Ar proximal to C₅H₄N), 2.25 (6

H, s, 5-C₆H₂Me₂). ¹³C{¹H} NMR (C₆D₆, 75.4 MHz, 293 K): δ

162.9 (2-C₆H₂Me₂), 158.1 (2-C₅H₄N), 152.4 (6-C₅H₄N), 151.2(2-

C₅H₅N), 147.5 (i-C₆H₅), 138.0 (4-C₅H₄N), 132.2 (4-C₅H₂Me₂),

129.5 (m-C₆H₅), 129.2 (6-C₅H₂Me₂), 127.0 (5-C₅H₂Me₂), 124.5 (3-

C₅H₅N), 124.3 (3-C₅H₂Me₂), 123.2 (1-C₅H₂Me₂), 122.0 (p-C₆H₅),

121.9 (5-C₅H₄N), 120.8 (3-C₅H₄N), 119.0 (o-C₆H₅), 63.4 (NCH₂Ar),

58.1 (NCH₂C₅H₄N), 21.1(5 C₅H₂Me₂), 19.0 (3 C₅H₂Me₂). IR (NaCl

plates, Nujol mull, cm^-1): ν 1604 (m), 1595 (s), 1583 (s), 1475

(s), 1444 (s), 1377 (m), 1318 (s), 1272 (s), 1214 (m), 1162 (m),

1102 (w), 1070 (w), 1053 (w), 1039 (m), 1023 (m), 992 (w), 960

(m), 856 (m), 822 (s), 749 (s), 729 (m), 694 (s), 648 (w), 632 (m),

587 (m), 548 (m). EI-MS: m/z 121 [OC₆H₂Me₂]⁺(99%), 105

[NNPh]⁺(26%), 91 [NPh]⁺(43%), 79 [NC₅H₅]⁺(24%). Anal.

Found (calcd for C₄₁H₄₁N₅O₂Ti): C, 71.9 (72.0); H, 6.0 (6.0); N,

10.2 (10.2).

2

Hz), 3.37 (2 H, d, CHHNSi₃, J ) 12.3 Hz), 1.20 (3 H, s, Me),

-0.03 (18 H, s, NSiMe₃). ¹³C{¹H} NMR (C₆D₆, 75.4 MHz, 293

K): δ 161.7(py-C²), 152.4 (o-NC₅H₄), 151.9 (py-C⁶), 148.1 (i-C₆H₅),

138.3 (py-C⁴), 137.6 (p-NC₆H₅) 129.6 (m-C₆H₅), 123.8 (m-NC₅H₄),

122.2 (p-C₆H₅), 121.3 (py-C⁵), 119.7 (py-C³), 119.5 (o-C₆H₅), 64.3

(CH₂NSi₃), 47.3(MeC), 24.6 (MeC), 1.3 (NSiMe₃). IR (NaCl plates,

Nujol mull, cm^-1): ν 1596 (s), 1585 (s), 1559 (w), 1506 (s), 1486

(s), 1467 (m), 1313 (s), 1299 (m), 1282 (s), 1240 (s), 1212 (m),

1159 (w), 1134 (w), 1085 (s), 1063 (s) 1038 (s), 1023 (m), 1011

(w), 997 (w), 989 (w), 897 (s), 880 (s), 831 (w), 782 (w), 754 (s),

738 (m), 695 (m), 673 (w), 618 (w), 589 (w). EI-MS: m/z 168

(90%) [NPh₂]⁺. Anal. Found (calcd for C₃₂H₄₄N₆Si₂Ti): C, 62.2

(62.3); H, 7.2 (7.2); N, 13.7 (13.6).

Ti(NNPh₂)(N₂NN′)(py) (13). To

a stirred mixture of

[Ti(NNPh₂)Cl₂(py)₂]₂(0.30 g, 0.66 mmol) and Li₂N₂NN′ (0.23 g,

0.66 mmol) was added toluene (30 mL), all at -78 °C. The yellow-

green suspension was allowed to warm to room temperature and

stirred for a further 90 min, resulting in a dark brown solution.

Volatiles were removed under reduced pressure, and the brown solid

was extracted into Et₂O (2 × 5 mL) and ﬁltered. Volatiles were

removed under reduced pressure to yield a green-brown solid, which

was washed with pentane (3 × 5 mL) cooled to -78 °C. The dark

green solid product was dried in Vacuo. Yield: 0.12 g (31%).

Crystal Structure Determinations of Ti(NNPh₂)Cl₂(py)₃

(7), Ti(NNPh₂)Cl₂(py′)₃(8), Ti(NNPh₂)(N₂N^{C ,Me})(py) (2),

2

Ti(NNPh₂)(N₂N^{C ,SiMe})(py) (10), Ti(NNPh₂)(N₂N^{C ,Me})(py)

(11), and Ti(NNPh₂)(N₂N^py)(py) (12). Crystal data collection and

processing parameters are given in Table 6. Crystals were

mounted on glass ﬁbers using perﬂuoropolyether oil and cooled

rapidly in a stream of cold N₂using an Oxford Cryosystems

Cryostream unit. Diffraction data were measured using an Enraf-

Nonius KappaCCD diffractometer. As appropriate, absorption

and decay corrections were applied to the data and equivalent

reﬂections merged.¹⁰²The structures were solved by direct

methods (SIR92¹⁰³), and further reﬁnements and all other

crystallographic calculations were performed using the CRYS-

TALS program suite.¹⁰⁴Other details of the structure solution

and reﬁnements are given in the Supporting Information (CIF

data). A full listing of atomic coordinates, bond lengths and

angles, and displacement parameters for all the structures has

been deposited at the Cambridge Crystallographic Data Centre.

See Notice to Authors, Issue No. 1.

2

3

¹H NMR (C₆D₆, 299.9 MHz, 293 K): δ 8.98 (1 H, d, J ) 6.5

Hz, 6-NC₅H₄), 7.75 (4 H, d, ³J ) 8.3 Hz, o-C₆H₅), 7.24 (4 H, app.

3

t, app. J ) 8.3 and 7.6 Hz, m-C₆H₅), 6.87 (2 H, t, J ) 7.6 Hz,

3

p-C₆H₅), 6.61 (1 H, app. t, app. J ) 7.6 and 7.7 Hz, 4-NC₅H₄),

3

6.12 (1 H, app. t, app. J ) 7.7 Hz, 5-NC₅H₄), 6.06 (1 H, d, J )

7.6 Hz, 3-NC₅H₄), 3.66 (2 H, m, CH₂NSiMe₃inner protons), 3.46

(2 H, m, CH₂NSiMe₃outer protons), 3.01 (2 H, s, CH₂NC₅H₄),

2.60 (2 H, m, CH₂N inner protons), 2.08 (2 H, m, CH₂N outer

protons), 0.45 (18 H, s, SiMe₃). ¹³C{¹H} NMR (C₆D₆, 75.4 MHz,

293 K): δ 159.3 (2-NC₅H₅), 154.5 (6-NC₅H₅), 147.8 (i-C₆H₅), 140.0

(4-NC₅H₅), 129.5 (m-C₆H₅), 127.8, 123.5 (3-NC₅H₅and 5-NC₅H₅),

122.0 (p-C₆H₅), 120.5 (o-C₆H₅), 57.3 (CH₂NC₅H₅), 56.4 (NCH₂),

49.2 (CH₂NSiMe₃), 3.2 (SiMe₃). IR (NaCl plates, Nujol mull,

cm^-1): ν 1593 (m), 1583 (m), 1487 (s), 1312 (w), 1237 (m), 1168

(w), 1156 (w), 1086 (s), 1024 (m), 946 (m), 934 (m), 866 (m), 833

(s), 793 (m), 756 (w), 740 (m), 701 (w), 692 (w), 630 (w), 584

(w). EI-MS: m/z 168 (88%) [NPh₂]⁺, 73 (88%) [SiMe₃]. Anal.

Found (calcd for C₂₈H₄₂N₆Si₂Ti): C, 59.2 (59.3); H, 7.4 (7.5); N,

14.7 (14.8).

Computational Details. All the calculations have been per-

formed with the Gaussian03 package¹⁰⁵at the B3PW91

level.^106,107The titanium atom was represented by the relativistic

effective core potential (RECP) from the Stuttgart group (12

valence electrons) and its associated basis set,¹⁰⁸augmented by

an f polarization function (R ) 0.869).¹⁰⁹The remaining atoms

(C, H, N) were represented by a 6-31G(d,p) basis set.¹¹⁰Full

optimizations of geometry without any constraint were per-

formed, followed by analytical computation of the Hessian matrix

to conﬁrm the nature of the located extremum as a minimum on

the potential energy surface for the experimental complex 2. For

the model systems Hyd_0, Hyd_90, Hyd_0_rot, Imide_90, and

Imide_00, constraints were imposed in the geometry optimiza-

Alternative Synthesis of Ti(NNPh₂)(Me₄taa) (14, NMR

tube scale). To a solution of [Ti(NNPh₂)Cl₂(py)₂]₂(12 mg, 13.1

µmol) in C₆D₆(0.2 mL) was added a solution of Li₂Me₄taa (4.7

mg, 13.1 µmol) in C₆D₆(0.2 mL). A color change of the initially

cloudy solution to light brown was observed. Analysis by ¹H NMR

indicated that a reaction had occurred immediately and cleanly to

give the known product, Ti(NNPh₂)(Me₄taa) (14).³⁹

Ti(NNPh₂)(O₂NN′)(py) (15). To

a stirred solution of

[Ti(NNPh₂)Cl₂(py)₂]₂(0.500 g, 1.09 mmol) in toluene (20 mL),

cooled to -78 °C, was added a solution of Na₂O₂NN′ · 0.16(THF)

(0.470 g, 1.09 mmol) in toluene (20 mL) at -78 °C. The solution

was allowed to warm to room temperature and then stirred for a

further 16 h. The volatiles were removed under reduced pressure

to give an oily solid, which was extracted into toluene at -78 °C

(2 × 20 mL) and ﬁltered. Volatiles were removed under reduced

pressure, and the product was washed with hexane (2 × 10 mL).

The brown solid product was dried in Vacuo. Yield: 0.336 g (49%).

(102) Otwinowski, Z.; Minor, W. Processing of X-ray Diffraction Data

Collected in Oscillation Mode; Academic Press: New York, 1997.

(103) Altomare, A.; Cascarano, G.; Giacovazzo, G.; Guagliardi, A.;

Burla, M. C.; Polidori, G.; Camalli, M. J. Appl. Crystallogr. 1994, 27, 435.

(104) Betteridge, P. W.; Cooper, J. R.; Cooper, R. I.; Prout, K.; Watkin,

D. J. J. Appl. Crystallogr. 2003, 36, 1487.

(105) Frisch,M. J.; et al. Gaussian 03, ReVision C.02;Gaussian, Inc.:

Wallingford, CT, 2004.

(106) Becke, A. D. J. Chem. Phys. 1993, 98, 5648.

(107) Perdew, J. P.; Wang, Y. Phys. ReV. B 1992, 45, 13244.

(108) Andrae, D.; Haussermann, U.; Dolg, M.; Stoll, H.; Preuss, H.

Theor. Chim. Acta 1990, 77, 123.

¹H NMR (C₆D₆, 299.9 MHz, 293 K): δ 9.21 (2 H, br, 2-C₅H₅N),

(109) Ehlers, A. W.; Bohme, M.; Dapprich, S.; Gobbi, A.; Hollwarth,

A.; Jonas, V.; Kohler, K. F.; Stegmann, R.; Veldkamp, A.; Frenking, G.

Chem. Phys. Lett. 1993, 208, 111.

3

9.04 (1 H, d, J ) 6.3 Hz, 6-C₅H₄N), 8.31 (4 H, d, J ) 8.8 Hz,

3

o-C₆H₅), 7.35 (4 H, app. t, app. J ) 7.5 and 8.8 Hz, m-C₆H₅),

3

4

6.94 (2 H, t, J ) 7.5 Hz, p-C₆H₅), 6.91 (2 H, d, J ) 2.1 Hz,

(110) Hariharan, P. C.; Pople, J. A. Theor. Chim. Acta 1973, 28, 213.

6494 Organometallics, Vol. 27, No. 24, 2008

Selby et al.

2

tion procedure. For all systems the amido groups NMe₂and the

N_ꢀMe₂group (hydrazido models) were constrained to be planar,

the pyridine rings were eclipsing the TidN_Rbond. The difference

between the various systems lies in the orientation of the amido

and/or the hydrazido N_ꢀMe₂groups, and these various constraints

were imposed upon dihedral angles. In the label of the model

systems “0” refers to amido groups lying in the equatorial plane,

while “90” refers to amido groups perpendicular to this plane.

The label “rot” in Hyd_0_rot implies that the N_ꢀMe₂group of

the hydrazido is perpendicular to the equatorial plane. For all

these systems NBO analyses were performed with NBO 5.0

interfaced with Gaussian.⁹²

(7), Ti(NNPh₂)Cl₂(py′)₃(8), Ti(NNPh₂)(N₂N^{C ,Me})(py) (2),

Ti(NNPh₂)(N₂N^{C ,SiMe})(py) (10), Ti(NNPh₂)(N₂N^{C ,Me})(py) (11),

2

3

Ti(NNPh₂)(N₂N^py)(py) Ti₂(µ-η²η¹-

(12), and

NNMe₂)₂(O₂NN′)₂· 0.5(C₅H₁₂) (16 · 0.5(C₅H₁₂)). Results and discus-

sion for the attempted synthesis of other titanium hydrazides

containing dianionic O₂N₂-donor ligands (16, eq S1, Figure S1,

Tables S1-S2), comparison between experimental and calculated

geometrical parameters for Ti(NNPh₂)(N₂N^{C ,Me})(py), 2 (Table S3),

2

NPA charges for the model compounds Ti(NR)(NMe₂)₂(py)₂(Table

S4), views of selected NLMOs for Hyd_0 (Figure S2), Cartesian

coordinates and electronic energy for the optimized structures,

complete list of authors for the reference of Gaussian03. This

information is available free of charge via the Internet at

http://pubs.acs.org.

Acknowledgment. We thank the EPSRC for support and

Dr. A. R. Cowley for the X-ray structure of 8.

Supporting Information Available: X-ray crystallographic data

in CIF format for the structure determinations of Ti(NNPh₂)Cl₂(py)₃

OM8007597

Article Doi

Titanium hydrazides supported by diamide-amine and related ligands: A combined experimental and DFT study

DOI: 10.1021/om8007597

Source and publish data:

Authors:

Article abstract of DOI:10.1021/om8007597

Full text of DOI:10.1021/om8007597

Products guided by the article

R&D Labs maybe for 1085417-59-9

Relevant to this article

Hot Product