Triamidoamine complexes of molybdenum and tungsten that contain Metal-E (E = N, P, and as) single, double, or triple bonds

Article abstract of DOI:10.1021/ja971727z

Addition of two or more equivalents of LiPPhH to [N₃N]MCl ([N₃N]³ = [(Me₃SiNCH₂CH₂)₃N]^3-; M = Mo or W) produced [N₃N]M≡P complexes via intermediate [N₃N]M(PPhH) complexes. The reaction between [N₃N]-MoCl and 2 equiv of LiAsPhH in the absence of light gave a mixture of [N₃N]Mo≡As (~30% yield) and [N₃N]-MoPh. [N₃N]Mo≡N and [N₃N]W≡N were both prepared via decomposition of intermediate azide complexes. Tungsten nitrido, phosphido, or arsenide complexes react readily with methyl triflate in toluene to give the cationic methyl imido, methyl phosphinidene, and methyl arsinidene complexes, respectively. Addition of methyl triflate or trimethylsilyl triflate to [N₃N]Mo≡N yields the cationic imido complexes {N₃N]Mo=NMe}OTf and {[N₃N]- Mo=NSiMe₃}OTf, respectively, but ([N₃N]Mo=PMe}OTf is not stable in solution at room temperature for more than 1-2 h. The reaction between '[Rh(CO)₂(CH₃CN)₂]PF₆' and 2 equiv of [N₃N]Mo≡P or [N₃N]W≡P gave red. crystalline adducts that contain two [N₃N]M≡P 'ligands', e.g., [Rh{[N₃NW≡P}₂(CO)(CH₃CN)]⁺, while red, crystalline [Rh{[N₃N]W≡As)₂(CO)(CH₃CN)]PF₆ could be prepared by an analogous route. {[N₃N]Mo=NSiMe₃}-OTf could be reduced to '19-electron' [N₃N]Mo=NSiMe₃, while addition of MeMgCl to {[N₃N]Mo=NSiMe₃}OTf or {[N₃N]Mo=NMe}OTf yielded complexes of the type [N₃N]Mo(NR)(Me). The complex in which R=Me was unstable with respect to loss of methane and formation of the iminato complex, [N₃N]Mo(N=CH₂). Both [N₃N(F)]W-(PPhH) and [N₃N(F)]Mo(PPhH) ([N₃N(F)³-=[(C₆F₅NCH₂CH₂)₃N]^3-) could be prepared readily, but all attempts to prepare [N₃N(F)]W≡P failed. X-ray studies of [N₃N]W≡P, [N₃N]Mo(PPhH), [N₃N]Mo≡As, {[N₃N]W=AsMe}OTf, [Rh{[N₃N]W≡P}₂(CO)(CH₃CN)]⁺, and [N₃N]Mo=NSiMe₃ are presented and discussed.

Full text of DOI:10.1021/ja971727z

J. Am. Chem. Soc. 1997, 119, 11037-11048
11037
Triamidoamine Complexes of Molybdenum and Tungsten That
Contain Metal-E (E ) N, P, and As) Single, Double, or Triple
Bonds
Nadia C. Mo1sch-Zanetti, Richard R. Schrock,* William M. Davis,
Klaus Wanninger, Scott W. Seidel, and Myra B. O’Donoghue
Contribution from the Department of Chemistry 6-331, Massachusetts Institute of Technology,
Cambridge, Massachusetts 02139
ReceiVed May 27, 1997X
Abstract: Addition of two or more equivalents of LiPPhH to [N3N]MCl ([N3N]3- ) [(Me3SiNCH2CH2)3N]3-; M )
Mo or W) produced [N3N]MtP complexes via intermediate [N3N]M(PPhH) complexes. The reaction between [N3N]-
MoCl and 2 equiv of LiAsPhH in the absence of light gave a mixture of [N3N]MotAs (∼30% yield) and [N3N]-
MoPh. [N3N]MotN and [N3N]WtN were both prepared via decomposition of intermediate azide complexes.
Tungsten nitrido, phosphido, or arsenido complexes react readily with methyl triflate in toluene to give the cationic
methyl imido, methyl phosphinidene, and methyl arsinidene complexes, respectively. Addition of methyl triflate or
trimethylsilyl triflate to [N3N]MotN yields the cationic imido complexes {[N3N]ModNMe}OTf and {[N3N]-
ModNSiMe3}OTf, respectively, but {[N3N]Mo)PMe}OTf is not stable in solution at room temperature for more
than 1-2 h. The reaction between “[Rh(CO)2(CH3CN)2]PF6” and 2 equiv of [N3N]MotP or [N3N]WtP gave red,
crystalline adducts that contain two [N3N]MtP “ligands”, e.g., [Rh{[N3N]WtP}2(CO)(CH3CN)]+, while red,
crystalline [Rh{[N3N]WtAs}2(CO)(CH3CN)]PF6 could be prepared by an analogous route. {[N3N]ModNSiMe3}-
OTf could be reduced to “19-electron” [N3N]ModNSiMe3, while addition of MeMgCl to {[N3N]ModNSiMe3}OTf
or {[N3N]ModNMe}OTf yielded complexes of the type [N3N]Mo(NR)(Me). The complex in which R ) Me was
unstable with respect to loss of methane and formation of the iminato complex, [N3N]Mo(NdCH2). Both [N3NF]W-
(PPhH) and [N3NF]Mo(PPhH) ([N3NF]3- ) [(C6F5NCH2CH2)3N]3-) could be prepared readily, but all attempts to
prepare [N3NF]WtP failed. X-ray studies of [N3N]WtP, [N3N]Mo(PPhH), [N3N]MotAs, {[N3N]WdAsMe}OTf,
[Rh{[N3N]WtP}2(CO)(CH3CN)]+, and [N3N]Mo)NSiMe3 are presented and discussed.
Introduction
ized.12 Tantalum arsinidene complexes of the type [N3N]-
TadAsR were believed to be the ultimate products of reactions
between [N3N]Ta(C2H4) and arsines such as Me3SiAsH2, but
these species were not isolated in pure form.14 Finally, although
complexes that are believed to contain a triple bond between
antimony and a metal, namely [(CO)n-1MtSb]- (M ) Cr, Mo,
and W), had been detected in the gas phase,15 and compounds
in which a terminal phosphido complex is a ligand had been
prepared,6 no monomeric transition metal complex that con-
tained a metal-pnictogen bond other than the well-known
nitrido complexes16 had been isolated and characterized.
There has been much speculation in the literature concerning
monomeric transition metal compounds that contain a multiple
metal-phosphorus bond.1-7 In the last few years several
mononuclear phosphinidene complexes in fact have been
characterized crystallographically.8-13 Monomeric transition
metal complexes that contain a double bond between the metal
and the higher pnictogens are much rarer. For example, only
one metal arsinidene complex, (silox)3TadAsPh (silox ) t-Bu3-
SiO), has been synthesized and crystallographically character-
X Abstract published in AdVance ACS Abstracts, October 15, 1997.
(1) Binger, P. In Multiple Bonds and Low Coordination in Phosphorus
Chemistry; Regitz, M., Scherer, O. J., Eds.; Thieme Medical Publishers,
Inc.: New York, 1990.
(2) Nixon, J. F. Chem. ReV. 1988, 88, 1327.
(3) Chisholm, M. H.; Folting, K.; Pasterczyk, J. W. Inorg. Chem. 1988,
27, 3057.
(4) Becker, G.; Becker, W.; Knebl, R.; Schmidt, H.; Hildenbrand, U.;
Westerhausen, M. Phosphorus & Sulfur 1987, 30, 349.
(5) Scherer, O. J. Angew. Chem., Int. Ed. Engl. 1990, 29, 1104.
(6) Scheer, M.; Schuster, K.; Budzichowski, T. A.; Chisholm, M. H.;
Streib, W. E. J. Chem. Soc., Chem. Commun. 1995, 1671.
(7) Cowley, A. H.; Barron, A. R. Acc. Chem. Res. 1988, 21, 81.
(8) Bohra, P.; Hitchcock, P. B.; Lappert, M. F.; Leung, W.-P. Polyhedron
1989, 8, 1884.
(9) Cowley, A. H.; Pellerin, B.; Atwood, J. L.; Bott, S. G. J. Am. Chem.
Soc. 1990, 112, 6734.
(10) Cummins, C. C.; Schrock, R. R.; Davis, W. M. Angew. Chem., Int.
Ed. Engl. 1993, 32, 756.
Transition metal-phosphorus triple bonds became a reality
with the synthesis and structural characterization of two related
terminal phosphido complexes of Mo and W.17,18 In each
complex three metal-amido linkages are present. One type,
trigonal bipyramidal [(Me3SiNCH2CH2)3N]MtP (M ) Mo or
W),18 contains a trimethylsilyl-substituted triamidoamine ligand
([N3N]3-),19 while the other, trigonal monopyramidal [N(t-Bu)-
(3,5-C6H3)]3MotP, contains monodentate tert-butylarylamido
ligands and no ligand trans to the phosphido ligand.17 [N3N]MtP
compounds are especially interesting in view of the stability of
(14) Freundlich, J. S.; Schrock, R. R.; Davis, W. M. J. Am. Chem. Soc.
1996, 118, 3643.
(15) Arnold, F. P. J.; Ridge, D. P.; Rheingold, A. L. J. Am. Chem. Soc.
1995, 117, 4428.
(16) Nugent, W. A.; Mayer, J. M. Metal-Ligand Multiple Bonds;
Wiley: New York, 1988.
(17) Laplaza, C. E.; Davis, W. M.; Cummins, C. C. Angew. Chem., Int.
Ed. Engl. 1995, 34, 2041.
(18) Zanetti, N. C.; Schrock, R. R.; Davis, W. M. Angew. Chem., Int.
Ed. Engl. 1995, 34, 2044.
(11) Hou, Z.; Breen, T. L.; Stephan, D. W. Organometallics 1993, 12,
3158.
(12) Bonanno, J. B.; Wolczanski, P. T.; Lobkovsky, E. B. J. Am. Chem.
Soc. 1994, 116, 11159.
(13) Arney, D. S. J.; Schnabel, R. C.; Scott, B. C.; Burns, C. J. J. Am.
Chem. Soc. 1996, 118, 6780.
(19) Verkade, J. G. Acc. Chem. Res. 1993, 26, 483.
S0002-7863(97)01727-7 CCC: $14.00 © 1997 American Chemical Society

11038 J. Am. Chem. Soc., Vol. 119, No. 45, 1997
Mo¨sch-Zanetti et al.
Table 1. Bond Distances (Å) and Angles (deg) for [N3N]WtP, [N3N]MotAs, [N3N]Mo(PPhH), and {[N3N]WdAsMe}+
compd
[N3N]WtPa
M-Neq
M-Nax
Neq-M-Neq
M-Neq-Si
Nax-M-N-Sid
other
2.162(4) (WtP)
1.975(6)
2.34(1)
115.8(1)
125.4(3)
178
101.9(2) (P-W-Neq)
2.252(3) (MotAs)
[N3N]MotAs
1.964(8)
2.33(2)
116.3(2)
127.0(5)
178
101.2(2) (As-Mo-Neq)
2.2494(14) (WdAs)
173.3(6) (W-As-C)
{[N3N]WdAsMe}+
1.940(12)b
1.969(11)
1.949(12)
1.992(5)b
2.014(5)
2.218(11)
114.8(5)c
116.0(5)
116.8(5)
115.0(2)c
124.1(2)
110.7(2)
125.5(6)b
127.8(6)
126.1(7)
127.3(2)b
128.7(2)
135.6(2)
169b
166
169
167b
179
152
101.0(3)-102.5(3) (As-W-Neq)
2.227(3) (Mo-P)
[N3N]Mo(PPhH)a
2.251(5)
2.468(3)
129.7(2) (Mo-P-C)
134(1) (Mo-P-H)
2.010(5)
100.3(2)-101.7(2) (P-Mo-Neq)
1.771(4) (Mo-N(5))
[N3N]ModNSiMe3
2.038(3)b
1.969(3)
1.994(3)
104.89(12)c
124.56(13)
110.48(13)
131.4(2)b
128.2(2)
124.6(2)
135b
152
163
172.4(2) (Mo-N(5)-Si)
172.78(12) (N(4)-Mo-N(5))
a See ref 18. b Measurements concern N(1), then N(2), then N(3). c Measurements concern N(1) and N(2), then N(1) and N(3), then N(2) and
N(3). d Dihedral angles were obtained from a Chem 3D structure.
related complexes that contain metal-carbon triple bonds20-22
and the “diagonal” relationship between C and P.5 Here we
report the full details of the synthesis of complexes of the type
[N3N]MtE (M ) Mo or W) in which E ) N, P, or As, along
with the imido, phosphinidene, or arsinidene products of
electrophilic attack on them, some examples of metal complexes
that contain phosphido or arsenido [N3N]MtE “ligands”, and
some chemistry of amido and phosphido complexes. Portions
of this work have appeared in preliminary form.18,23
[N3N]M complexes in which there is little steric interaction
between the trimethylsilyl groups and the ligand in the trigonal
pocket.25 Most other distances and angles are similar to what
is found in [N3N]WCl24 (W-Neq ) 1.985(11) Å and Neq-W-
Neq ) 117.66(6)°, for example). The exception is the W-Nax
distance in [N3N]WtP (2.34(1) Å) versus the W-Nax distance
in [N3N]WCl (2.182(6) Å). The relatively long W-Nax distance
of 2.34(1) Å is consistent with the expected greater trans
influence of a triply bound element versus a chloride.
When [N3N]WCl is treated with only 1 equiv of LiPPhH,
then the crude product mixture contained small quantities of
what we proposed to be diamagnetic [N3N]W(PPhH) on the
Results
Preparation of Terminal Nitrido, Phosphido, and Ars-
enido Complexes. The attempted synthesis of a phenylphos-
phido complex by treating [N3N]WCl20 with 2 equiv of LiPPhH
in a mixture of toluene and tetrahydrofuran (4:1) at 80 °C led
instead (over a period of 48 h) to [N3N]WtP (eq 1) as yellow
cubes in ∼50% yield. The phosphorus resonance in [N3N]WtP
basis of its proton and 31P NMR spectra (δP ) 91 ppm, JPW
)
719 Hz; δH ) 18.68 ppm, JHP ) 300 Hz, JHW ) 36 Hz).
[N3N]W(PPhH) could not be separated from [N3N]WtP readily.
However, [N3N]W(PPhH) could be prepared and isolated in
∼50% yield by adding PPhH2 to [N3N]W(C6H5)26 in toluene
at room temperature. On the basis of the similarity of the proton
and 31P NMR spectrum of [N3N]W(PPhH) to that of crystal-
lographically characterized [N3N]Mo(PPhH) (see below), we
presume that [N3N]W(PPhH) is structurally analogous to [N3N]-
Mo(PPhH). Its diamagnetism can be ascribed to donation of
the phosphorus lone pair to the metal and consequent pairing
of the two d electrons in the remaining π orbital. The proton
NMR spectrum suggests that [N3N]W(PPhH) is C3V symmetric
on the NMR time scale, i.e., the phenylphosphido ligand
“rotates” readily with respect to the [N3N] ligand, even at -60
°C.
is found at 1080 ppm in the 31P NMR spectrum with JPW
)
1
138 Hz. When 1.3 equiv of LiPPhH and [N3N]WCl were heated
in toluene-d8 at 110 °C over a period of 2 days, then [N3N]WtP
was formed in 88% yield versus an internal standard. [N3N]WtP
also has been prepared by treating [N3N]WCl with 2 equiv of
LiP(SiMe3)2.24 In all such reactions, and as has been docu-
mented more extensively in the analogous Mo system (see
below), it is believed that complexes of the type “[N3N]W-
[P(R)(Li)]” are unobservable intermediates that decompose to
give LiR (R ) phenyl or SiMe324) and [N3N]WtP.
The reaction between [N3N]MoCl27,28 and 1 equiv of LiPPhH
at 50 °C in a mixture of toluene and THF yields orange,
diamagnetic [N3N]Mo(PPhH) in good yield (eq 2). Like the
An X-ray study18 showed [N3N]WtP to be a monomeric
species in which the WtP bond length is 2.162(4) Å. (Bond
distances and angles are compared in Table 1 for [N3N]WtP,18
[N3N]Mo(PPhH),18 and three other crystallographically char-
acterized compounds reported here.) The W-Neq-Si angle
(125°) and Nax-W-N-Si dihedral angle (178°) are typical of
tungsten analog, [N3N]Mo(PPhH) is diamagnetic and has
apparent C3V symmetry on the NMR time scale. The proton
NMR spectrum shows the phosphido proton resonance at 12.16
(20) Shih, K.-Y.; Totland, K.; Seidel, S. W.; Schrock, R. R. J. Am. Chem.
Soc. 1994, 116, 12103.
(21) Schrock, R. R.; Shih, K.-Y.; Dobbs, D.; Davis, W. M. J. Am. Chem.
Soc. 1995, 117, 6609.
(22) Murdzek, J. S.; Schrock, R. R. In Carbyne Complexes; VCH: New
York, 1988.
(23) Johnson-Carr, J. A.; Zanetti, N. C.; Schrock, R. R.; Hopkins, M.
D. J. Am. Chem. Soc. 1996, 118, 11305.
(25) Schrock, R. R. Acc. Chem. Res. 1997, 30, 9.
(26) Schrock, R. R.; Seidel, S. W.; Mo¨sch-Zanetti, N. C.; Dobbs, D. A.;
Shih, K.-Y.; Davis, W. M. Organometallics In press.
(27) Schrock, R. R.; Seidel, S. W.; Mo¨sch-Zanetti, N. C.; Shih, K.-Y.;
O’Donoghue, M. B.; Davis, W. M.; Reiff, W. M. J. Am. Chem. Soc. In
press.
(24) Scheer, M.; Mu¨ller, J.; Ha¨ser, M. Angew. Chem., Int. Ed. Engl. 1996,
35, 2492.
(28) Shih, K.-Y.; Schrock, R. R.; Kempe, R. J. Am. Chem. Soc. 1994,
116, 8804.

Triamidoamine Complexes of Molybdenum and Tungsten
J. Am. Chem. Soc., Vol. 119, No. 45, 1997 11039
give an unobserved lithiated (or anionic) phenylphosphinidene
complex (eq 3), which then decomposes to give the terminal
phosphido complex and phenyl lithium (eq 4). The phenyl
Figure 1. A top view of the structure of [(Me3SiNCH2CH2)3N]Mo-
(PPhH) (P-H proton omitted).
lithium then reacts with PPhH2 that is generated in the
deprotonation reaction (eq 3) to yield benzene and LiPPhH.
Consistent with this proposal is the fact that when [N3N]Mo-
(PPhH) was heated with 1.0 equiv of LiPh at 110 °C in toluene-
d8 for 4 days, [N3N]MotP and benzene (90% relative to
[N3N]MotP) were produced. In a similar experiment that
employed 0.5 equiv of LiPh at 65 °C over a period of 4 days,
[N3N]Mo(PPhH) disappeared in a first-order manner to give
benzene and [N3N]MotP, consistent with LiPh-catalyzed
decomposition of [N3N]Mo(PPhH) to [N3N]MotP.
The proposed mechanism of forming the terminal phosphido
complexes would suggest that the reaction between Li2PPh and
[N3N]MCl should be a suitable method of preparing [N3N]MtP.
Indeed, [N3N]WCl reacts with 1 equiv of Li2PPh to give
[N3N]WtP in ∼50% yield, presumably via decomposition of
unobserved intermediate “[N3N]W(PPhLi)” (eq 4). [N3N]WPh
is formed in ∼10% yield in this reaction, we presume as a
consequence of a competitive, secondary reaction between
[N3N]WCl and LiPh.26 In contrast, the reaction between Li2PPh
and [N3N]MoCl did not produce [N3N]MotP in any significant
yield.
The formation of tungsten alkylidyne complexes via R,R-
dehydrogenation of alkyl complexes of the type [N3N]WCH2R20
and formation of (silox)3TadPPh by adding phenylphosphine
to (silox)3Ta12 together suggest that [N3N]WtP might be the
product of the reaction between [N3N]WCl and LiPH2. Indeed,
LiPH2 reacts with [N3N]WCl in THF over a period of 24 h to
give [N3N]WtP as the only observable product. It is not known
whether this type of reaction involves R,R-dehydrogenation of
[N3N]W(PH2) or a catalyzed decomposition of [N3N]W(PH2)
that proceeds via intermediate [N3N]W(PHLi). In contrast, the
reaction between LiPH2 and [N3N]MoCl again did not produce
[N3N]MotP in any significant yield.
The reaction between [N3N]MoCl and 2 equiv of LiAsPhH
in the absence of light yields orange, diamagnetic [N3N]MotAs
in ∼30% yield. Another pentane-soluble, but paramagnetic,
product that was produced in approximately the same yield in
this reaction proved to be [N3N]MoPh.26 We also have seen
no evidence for [N3N]Mo(AsPhH) as an intermediate in this
reaction. An attempt to prepare [N3N]Mo(AsPhH) or [N3N]-
MotAs by adding AsPhH2 to [N3N]MoPh in toluene-d8 in the
absence of light yielded H3[N3N] as the only recognizable
product.
ppm with a coupling to phosphorus of 283 Hz (cf., 18.68 ppm
and 300 Hz in [N3N]W(PPhH)). The phosphido phosphorus
resonance (142 ppm) is found ∼50 ppm further downfield from
where it is found in [N3N]W(PPhH) (91 ppm).
An X-ray structural study of [N3N]Mo(PPhH) (Figure 1)
shows that it is a trigonal bipyramidal species that contains an
essentially planar phenylphosphido ligand (proton located) with
ModP ) 2.227(3) Å, ModP-C ) 129.7(2)°, and ModP-H
) 134(1)°. The phosphido ligand is turned so that the phenyl
ring points between N(1) and N(3). Consequently, the N(1)-
Mo-N(3) angle is opened slightly (to 124.1(2)°) compared to
the other two Neq-Mo-Neq angles. The plane of the phosphido
ligand does not quite coincide with the N(4)-Mo-N(2) plane,
as shown by the N(2)-Mo-P-Cipso dihedral angle of 168°.
Therefore the SiMe3 group attached to N(3) is tipped furthest
away from the phosphido ligand (Mo-N(3)-Si ) 135.6(2)°
versus 127.3(2)° and 128.7(2)°). The overall result is that two
of the Nax-Mo-N-Si dihedral angles are significantly less than
180° (167° and 152°). A “twisting” of the trimethylsilyl groups
and a consequent decrease in a Nax-Mo-N-Si dihedral angle
was shown to be consistent with an increase in steric hindrance
in analogous molybdenum complexes of the type [N3N]MoR,27
and was postulated to be the first and most facile response to
an increase in steric crowding within the trigonal coordination
pocket.25 The Mo-Nax distance in [N3N]Mo(PPhH) is slightly
shorter (2.251(5) Å) than the W-Nax distance in [N3N]WtP
(2.34(1) Å), consistent with a larger trans effect for the W-P
triple bond than for the Mo-P pseudo double bond. In spite
of the preferred orientation of the phenyl ring between N(1)
and N(3), and the generally crowded nature of the coordination
environment in which the phosphido ligand is located, steric
hindrance between the phenyl ring and the TMS groups is not
great enough to restrict rotation of the phosphido ligand on the
NMR time scale, and the molecule therefore has apparent C3V
symmetry.
Addition of 2 equiv of LiPPhH to [N3N]MoCl in a mixture
of toluene and THF, or 1 equiv of LiPPhH to [N3N]Mo(PPhH),
followed by heating the reaction to 70 °C for 12 h, yields yellow,
diamagnetic [N3N]MotP in good yield. The chemical shift of
the terminal phosphido ligand in [N3N]MotP was found to be
1346 ppm, approximately 266 ppm further downfield than the
phosphorus resonance in [N3N]WtP (1080 ppm). No [N3N]-
MotP is formed upon heating a solution of [N3N]Mo(PPhH)
in toluene to 120 °C for 3 days. We speculate that the
phenylphosphido ligand in [N3N]Mo(PPhH) is deprotonated to
[N3N]MoCl + 2LiAsPhH 50 °C, 2 days8 [N3N]MotAs (5)
toluene/THF
A view of the structure of [N3N]MotAs is shown in Figure
2 and bond distances and angles are listed in Table 1. The

11040 J. Am. Chem. Soc., Vol. 119, No. 45, 1997
Mo¨sch-Zanetti et al.
Table 2. Stretching Frequencies (cm-1) for Compounds of the
Type [N3N]MtEa
MtE
νME
MtE
νME
WtN
Wt15N
WtP
1015 (R), 1012 (IR)
987 (R)
516 (R), 516 (IR)
343 (R)
MotN
1001 (IR)
MotP
MotAs
521 (R)
374 (R)
WtAs
a (R) implies a determination by Raman spectroscopy on a solid
sample.23 (IR) implies a determination by Infrared spectroscopy on a
solid sample in Nujol.
complex. However, the Raman spectra are considerably sharper,
and the MtE stretch could be clearly identified for all
compounds except [N3N]MotN, which had not yet been
prepared at the time the Raman study was carried out.23 Once
the locations of the MtE stretches were known with certainty,
it was possible to locate the WtN and WtP absorptions in
the IR spectra (Nujol) at 1012 and 516 cm-1, respectively. An
IR spectrum of [N3N]MotN (Nujol) shows a strong stretch at
1001 cm-1 analogous to that found in the spectrum of
Figure 2. A view of the structure of [(Me3SiNCH2CH2)3N]MotAs.
MotAs bond length was found to be 2.25 Å, ∼0.1 Å longer
than the WtP bond in [N3N]WtP, and consistent with the
Mo-As bond being a triple bond. The bond distance of
2.252(3) Å should be compared with a W-As distance of
2.2903(11) Å in recently structurally characterized [N3N]WtAs.24
The values for M-Neq, M-Nax, and Neq-M-Neq are also
similar to analogous bond distances and angles in [N3N]WtAs24
(W-Neq ) 1.989(4) Å, W-Nax ) 2.336(6) Å, and Neq-W-
Neq ) 115.69(8)°), as one might expect. Other distances and
angles are close to what they are in [N3N]WtP and deserve
no further comment. It should be noted that [N3N]WCl,24
[N3N]WtP,18 and [N3N]WtAs24 all crystallize in the space
group Pa3h, so to ensure that bond distances to P or As are
accurate, it is crucial that the sample of [N3N]WtP or
[N3N]WtAs used in an X-ray study not be contaminated with
cocrystallized [N3N]WCl.
An attempt to prepare [N3N]WtAs by adding 2 equiv of
LiAsPhH to [N3N]WCl (2 days in toluene/THF, 22 °C, absence
of light) yielded [N3N]WPh as the only significant product (by
NMR); only traces of [N3N]WtAs were found. However,
[N3N]WtAs can be prepared in good yield by allowing a slight
excess of phenylarsine to react with [N3N]WPh for 48 h at room
temperature in the dark (eq 6). Approximately 2 equiv of
[N3N]WtN. Therefore we assume that νMoN ) 1001 cm-1
.
Observed stretching frequencies from both Raman and IR
studies are listed in Table 2.
Electrophilic Attack on Nitrido, Phosphido, and Arsenido
Complexes. Tungsten nitrido, phosphido, or arsenido com-
plexes react readily with methyl triflate in toluene to give the
imido, phosphinidene, and arsinidene complexes, respectively,
in quantitative yield (eq 8). The phosphorus resonance in
toluene
[N3N]WtE + MeOTf E ) N, P, or As8
{[N3N]WdEMe}+OTf- (8)
{[N3N]WdPMe}+ is found at 339 ppm in the 31P NMR
spectrum with JPW ) 748 Hz. These values should be compared
with 1080 ppm and JPW ) 138 Hz in [N3N]WtP. Changes in
the magnitude of coupling to tungsten are more dramatic than
what is found in [N3N]Wt15N and {[N3N]Wd15NMe}+. In
[N3N]Wt15N the nitrido 15N resonance is found at 551.2 ppm
relative to ammonia with JNW
) 48 Hz, while in
PhAsH2
{[N3N]Wd15NMe}+ the 15N resonance is found at 145.3 ppm
with JNW ) 116 Hz. The increase in JPW from 138 Hz in
[N3N]WtP to 748 Hz in {[N3N]WdPMe}+ (a 5.4-fold
increase) is consistent with a more significant degree of
rehybridization of the σ portion of the WtN bond upon
alkylation, assuming that the WE coupling constants in any set
of [N3N]WtE and {[N3N]WdER}+ compounds are of the same
sign and are dominated by the Fermi contact interaction. On
the basis of the small value of JWP alone the σ portion of the
W-P bond would appear to have largely p character, and the
lone pair therefore would reside in an orbital with largely s
character.
[N3N]WPh
8 [N3N]WtAs
(6)
-2C6H6
1
benzene are also formed, according to H NMR spectra. The
object of this reaction was to form intermediate [N3N]W-
(AsPhH) via proton transfer from As to the phenyl ligand in
“[N3N]WPh(AsPhH2)”. However, we found no evidence that
[N3N]W(AsPhH) is an intermediate in this reaction. [N3N]WtAs
has also been prepared by treating [N3N]WCl with 2 equiv of
LiAs(SiMe3)2, although higher temperatures are required than
for synthesis of [N3N]WtP from [N3N]WCl and LiP(SiMe3)2,
and there was no evidence for formation of intermediate
[N3N]W[As(TMS)2].24
An X-ray structural study of {[N3N]WdAsMe}OTf (Figure
3) showed it to be overall similar to other molecules of this
general type (Table 1). The most important features are the
essentially linear arsinidene ligand (W-As-C ) 173.3(6)°) and
a WdAs bond length of 2.2494(14) Å, which is virtually the
same as the MotAs bond length in [N3N]MotAs. The Mo-
Nax distance (2.218(11) Å) is similar to what it is in [N3N]Mo-
(PPhH) (2.251(5) Å) and [N3N]WCl (2.182(6) Å24), consistent
with the relatively poor trans effect of the arsenic atom in the
linear arsinidene ligand compared to (especially) phosphorus
and arsenic in the terminal phosphido and arsenido complexes,
respectively. The linearity of the arsinidene ligand in this
species contrasts strongly with the dramatically bent phenyl
arsinidene ligand (107.2(4)°) and longer TadAs bond length
The tungsten nitrido complex, [N3N]WtN, can be synthe-
sized as shown in eq 7, while the analogous molybdenum
complex is best prepared from [N3N]MoCl and Me3SiN3 in
[N3N]WCl NaN , CH CN/THF8 [N3N]WtN
(7)
3
3
-N2, -NaCl
toluene at 90 °C over a period of 24 h. Intermediate azide
complexes are proposed to form in each case. Decomposition
of azide complexes (from sodium azide or trimethylsilyl azide)
is a relatively common method of preparing nitrido complexes,
and a large number of nitrido complexes are known.16
Infrared spectra of the [N3N]MtE (E ) N, P, and As)
complexes in the region below 1500 cm-1 are relatively

Triamidoamine Complexes of Molybdenum and Tungsten
J. Am. Chem. Soc., Vol. 119, No. 45, 1997 11041
An attempt to prepare {[N3N]ModAsMe}OTf by a method
analogous to that used to prepare {[N3N]ModPMe}OTf was
even less successful. What we believe is {[N3N]ModAsMe}-
1
OTf could be observed by H NMR in THF-d8 at -70 °C.
However, this belief is based solely on a methyl resonance at
3.72 ppm, as found for {[N3N]ModPMe}OTf, and typical
[N3N]3- resonances. However, this compound could not be
isolated in pure form, and attempts to recrystallize it by ether
diffusion into a THF solution at -35 °C resulted in a
considerable amount of decomposition and no improvement in
purity. The decomposition products could not be identified.
Therefore the existence of {[N3N]ModAsMe}OTf could not
be confirmed by elemental analysis, and {[N3N]ModAsMe}+
therefore must be regarded as the only missing member of the
series of {[N3N]MdEMe}+ (E ) N, P, and As) complexes.
At this stage we do not know why all {[N3N]WdEMe}+
complexes can be prepared, whereas the stabilities of {[N3N]-
ModEMe}+ complexes decline rapidly in the order E ) N >
P > As. The most stable compounds appear to be those in
which the dative portion of the M-E pseudo triple bond is likely
to be the strongest (W > Mo and N > P > As) and E sterically
most protected as a consequence of the short MdE bond.
Terminal Phosphido and Arsenido Complexes as Ligands.
The reaction between “[Rh(CO)2(MeCN)2]PF6” (which is gener-
ated from [Rh(CO)2Cl]2 and TlPF6 in acetonitrile) and 2 equiv
of [N3N]MotP or [N3N]WtP led to soluble red crystalline
complexes in good yield (eq 9). The 31P NMR spectra show
Figure 3. A view of the structure of {[(Me3SiNCH2CH2)3N]WdAsMe}-
OTf (cation only).
(2.428(2) Å) in (silox)3TadAsPh,12 the only other crystallo-
graphically characterized arsinidene complex. At this stage it
is not known to what extent the linearity of the arsinidene ligand
in {[N3N]WdAsMe}OTf is dictated by steric factors. Although
the Neq-W-Neq angles (166-169°) suggest that there is some
“twisting” of the trimethylsilyl groups in response to steric
congestion in the trigonal pocket, the arsinidene ligand would
appear to be able to bend in a manner analogous to what is
observed for the phosphido ligand in {[N3N]Mo(PPhH)}.
Therefore at this stage we believe the arsinidene to be linear
largely for electronic reasons. In that case the “bent” nature of
the arsinidene ligand in (silox)3TadAsPh might be ascribed to
relatively poor AsfTa π bonding, as a result of competitive
OfTa π bonding from the silox ligands in a pseudotetrahedral
environment. In a pseudotrigonal bipyramidal [N3N]3- complex
π bonding to the three equatorial amido nitrogens involves
2[N3N]MtP-CO
“[Rh(CO)2(S)2](PF6)”
8
(S ) MeCN; M ) Mo, W)
trans-[Rh{[N3N]MtP}2(S)(CO)]PF6 (9)
one resonance for the two equivalent phosphorus atoms as a
doublet at 791 (M ) Mo) or 643 ppm (M ) W) with coupling
to rhodium of 67 (M ) Mo) or 79 Hz (M ) W), respectively.
The NMR data for the latter should be compared with that for
trans-{[N3N]WtP}2W(CO)4, in which the terminal phosphido
resonance is found at 680 ppm and JWP ) 426 (PtW) and 151
Hz (P-W).24 IR spectra show absorptions at 2021 (M ) Mo)
and 2014 cm-1 (M ) W) for coordinated CO, and a CtN
stretch for acetonitrile (at 2361 cm-1 when M ) W); no
absorptions corresponding to M-P stretching modes could be
identified.
2
2
primarily the dx -y and dxy orbitals, which are virtually
independent of the dxz and dyz orbitals that are used for π bonding
to an apical ligand.
Addition of methyl triflate or trimethylsilyl triflate to [N3N]-
MotN yields the cationic imido complexes, {[N3N]ModNMe}-
OTf and {[N3N]ModNSiMe3}OTf, respectively, in high yield.
Imido complexes are relatively common,16 and there is nothing
unusual about these. However, addition of methyl triflate to
[N3N]MotP is complex when carried out at room temperature.
Only when [N3N]MotP and 1 equiv of methyl triflate are
allowed to react slowly in ether at -30 °C over a period of
several days can an insoluble orange microcrystalline solid be
prepared in ∼40% yield. The 31P NMR spectrum shows a
phosphorus resonance at 282.7 ppm, which is significantly
upfield of the phosphorus resonance in {[N3N]WdPMe}OTf
(339 ppm). In [N3N]M(PPhH) and [N3N]MtP complexes the
phosphorus resonance in the Mo complex is downfield from
its position in the tungsten analog. Nevertheless, the 31P
chemical shift is at least of approximately the same magnitude
as it is in {[N3N]WdPMe}OTf. Since the compound analyzes
correctly for C, H, and N, we are relatively confident that {[N3N]-
ModPMe}OTf has been prepared. {[N3N]ModPMe}OTf can
be recrystallized from a mixture of THF and diethyl ether, but
is not stable in solution at room temperature. After 1-2 h,
these solutions in dichloromethane or THF darken; 31P and 1H
NMR spectra suggest that {[N3N]ModPMe}OTf has largely
decomposed within 2 h and the only identifiable product (in
∼30% yield) is paramagnetic [N3N]MoOTf.26
An X-ray study of [Rh{[N3N]WtP}2(CO)(CH3CN)]+ re-
vealed the structure shown in Figure 4. (See also Table 3.)
The core geometry of [Rh{[N3N]WtP}2(CO)(CH3CN)]+ is
distorted slightly from square planar in terms of the various
angles at rhodium, the trans [N3N]WtP ligands are slightly
bent so as to produce a “banana-shaped” W-P-Rh-P-W
arrangement, and the Rh-P-W angles are less than 180°. A
cis arrangement of [N3N]WtP ligands would appear to be
precluded on steric grounds. The Rh-P distances are what one
would expect for a Rh-P distance in a rhodium phosphine
complex, while the W-P distances (2.177(5) and 2.173(5) Å)
are not lengthened to a statistically significant degree from the
W-P distance in [N3N]WtP (2.162(4) Å). The observed
distortions from a strictly square planar geometry are not likely
to be costly energetically, and perhaps are dictated to a
significant degree by crystal packing forces. This structure
should be compared with that for trans-{[N3N]WtP}2W-
(CO)4,24 in which the WtP bonds are lengthened to 2.202(2)
Å, while the dative P-W bond lengths are 2.459(2) and 2.460(2)
Å.
Red, crystalline [Rh{[N3N]WtAs}2(CO)(CH3CN)]PF6 was
prepared by adding [N3N]WtAs to “[Rh(CO)2(MeCN)2]PF6”
(generated from [Rh(CO)2Cl]2 and TlPF6 in acetonitrile). The

11042 J. Am. Chem. Soc., Vol. 119, No. 45, 1997
Mo¨sch-Zanetti et al.
confirm that [N3N]ModNSiMe3 is a d1 Curie paramagnet down
to 5 K in the solid state with a magnetic moment of 1.64 µB.
Attempts to prepare the analogous [N3N]ModNMe compound
so far have failed. Isolation of the iminato complex, [N3N]-
Mo(NdCH2) (see below), would suggest that [N3N]ModNMe
might be susceptible to loss of a hydrogen radical from the imido
methyl group at some point during its attempted synthesis.
An X-ray study revealed that [N3N]ModNSiMe3 is a C3-
symmetric species that contains a nearly linear imido ligand
(Mo-Nimido-Si ) 172.4°; Figure 5, Table 1). The ModN bond
length (1.771(4) Å) is also typical of Mo-nitrogen pseudo triple
bonds in amido complexes.16 The imido TMS group is tipped
between N(1) and N(3), which accounts for the larger N(1)-
Mo-N(3) angle. The [N3N]3- ligand backbond is distorted to
a significant degree in response to the bulk of the TMS group
on the imido ligand, as shown by the Nax-Mo-N-Si dihedral
angles being considerably less than 180° (135° for the N(4)-
Mo-N(1)-Si(1) dihedral angle; Table 1). “Tipping” of the
TMS groups away from the “upright” position (dihedral angle
∼180°) appears to be the most facile response of the ligand
backbone to steric congestion in the trigonal coordination
pocket.25,27 The Mo-N(4) bond (2.468(3) Å) is long compared
to other Mo-Nax bonds in the [N3N]Mo complexes reported
here. Therefore it seems possible the “19-electron” [N3N]-
ModNSiMe3 has a pseudo triple Mo-N(5) bond and an
unpaired electron in an orbital with a significant degree of Mo-
Nax σ antibonding character. Alternatively, the imido ligand
may be forced into nearly a linear Mo-N(5)-Si arrangement
for steric reasons, but the amount of Mo-N(5) dative π bonding
is minimal, and either the dxz or the dyz orbital contains the single
electron. The long Mo-N(4) bond would seem to suggest the
former, but the slight bending of the Mo-N-Si linkage away
from 180° also would be consistent with the latter. At this point
it is not possible to choose between these two explanations.
A next logical step toward removing a nitrogen-containing
product would be to add an electrophile to the imido nitrogen
in [N3N]ModNSiMe3 to form a cationic Mo(V) amido complex.
Reactions were attempted with MeOTf, Me3SiOTf, pyridinium
chloride, acetyl chloride, Me3SiCl, or methyl iodide. All
apparently failed. Only oxidation to [N3N]MotN or {[N3N]-
ModNSiMe3}+ was observed, or the [N3N]3- ligand was lost
from the metal as H3[N3N]. Further reduction of [N3N]-
ModNSiMe3 also failed. No reaction occurred with sodium
naphthalenide or sodium amalgam in THF, while an attempted
reduction with sodium sand in THF led only to decomposition.
The failure to reduce “19-electron” [N3N]ModNSiMe3 is not
surprising in view of the likely high energy of the orbital into
which the electron must be placed.
Figure 4. Drawing of the structure of [Rh{[N3N]WtP}2(CO)(CH3-
CN)]+.
Table 3. Selected Bond Distances (Å) and Angles (deg) in
[Rh{[N3N]WtP}2(CO)(CH3CN)]+
Distances (Å)
Rh-C(1) 1.81(3) Rh-P(1) 2.316(5) W(2)-P(1) 2.177(5)
Rh-N(2) 2.09(2) Rh-P(2) 2.305(5) W(1)-P(2) 2.173(5)
Angles (deg)
C(1)-Rh-P(2)
C(1)-Rh-N(1)
C(1)-Rh-P(1)
P(2)-Rh-N(1)
86.8(7)
173.8(8)
89.2(7)
93.0(4)
P(2)-Rh-P(1)
P(1)-Rh-N(1)
W(1)-P(2)-Rh
W(2)-P(1)-Rh
166.7(2)
92.4(4)
172.3(3)
173.3(3)
physical properties, IR data, and NMR data for [Rh{[N3N]-
WtAs}2(CO)(CH3CN)]PF6 are analogous to those for
[Rh{[N3N]WtP}2(CO)(CH3CN)]PF6. Therefore there is no
reason to suspect that the structure of [Rh{[N3N]WtAs}2-
(CO)(CH3CN)]PF6 is not analogous to that of [Rh{[N3N]WtP}2-
(CO)(CH3CN)]PF6.
Further Reactions of Mo Imido Complexes. In a reaction
in which dinitrogen is bound to Mo and reduced in a stepwise
manner, a terminal nitrido complex is a possible intermediate.29-32
To be in a position to reduce dinitrogen catalytically it would
be necessary to remove the nitrido ligand in a well-defined and
controlled manner. Therefore we explored some reactions that
bear on removing the nitrido ligand.
We have shown that the nitrido ligand in [N3N]MotN is
readily attacked by electrophiles, a process that is relatively
common, at least for electrophiles other than a proton.16
A
plausible next step would be reduction by one electron. {[N3N]-
ModNSiMe3}OTf can be reduced cleanly with either Li2C8H8
or sodium naphthalenide to a Mo(V) amido complex. No
reduction is observed over several days when magnesium is
employed as the reducing agent in THF. [N3N]ModNSiMe3
is a blue paramagnetic compound with a proton NMR spectrum
that reveals only three shifted, broad resonances (at room
temperature) for the two types of TMS groups and a resonance
for one of the two backbone methylene groups. Temperature-
dependent magnetic susceptibility (SQUID) measurements
(29) Richards, R. L. Chem. Br. 1988, 24, 133.
(30) Dilworth, J. R.; Richards, R. L. In ComprehensiVe Organometallic
Chemistry; Wilkinson, G., Stone, F. G. A., Abel, E. W., Eds.; Pergamon:
New York, 1982.
(31) Hidai, M.; Mizobe, Y. Chem. ReV. 1995, 95, 1115.
(32) Chatt, J.; Dilworth, J. R.; Richards, R. L. Chem. ReV. 1978, 78,
589.
In view of the existence of [(C6F5NCH2CH2N)3N]Mo-
(NMe2),33 we attempted to prepare [N3N]Mo(NMe2). This
(33) Kol, M.; Schrock, R. R.; Kempe, R.; Davis, W. M. J. Am. Chem.
Soc. 1994, 116, 4382.

Triamidoamine Complexes of Molybdenum and Tungsten
J. Am. Chem. Soc., Vol. 119, No. 45, 1997 11043
Figure 7. Temperature-dependent shift of the NMe2 group in
[(Me3SiNCH2CH2)3N]Mo(NMe2) fit to eq 12.
One of the methylene proton resonances also shifts, to a less
significant degree, from near 2 ppm at 180 K to closer to 1
ppm at 304 K, but the other methylene resonance (near 3 ppm)
does not shift appreciably in this temperature range. To explain
the temperature-dependent behavior of the dimethylamido and
backbone methylene resonances we propose that the diamagnetic
form of [N3N]Mo(NMe2) is in rapid equilibrium with a high
spin form and that the dimethylamido resonance is contact
shifted by the paramagnetic form. The high spin form is likely
to follow Curie-Weiss behavior in the temperature range 180-
304 K, as do [N3N]MoCl and [N3N]Mo(alkyl) complexes.27 If
we assume the rate of interconversion of the high spin and low
spin forms is fast at all temperatures here, we can employ the
expression shown in eq 1234-36 to determine the enthalpy and
Figure 5. A drawing of the structure of [(Me3SiNCH2CH2)3N]-
ModNSiMe3.
C
δ ) δdia
+
(12)
∆H° ∆S°
RT
T 1 + exp
-
(
(
))
R
entropy difference between the high spin and low spin states.
(In eq 12 δ is the observed chemical shift at temperature T, δdia
is the chemical shift of the diamagnetic form, which is assumed
to be independent of temperature, and C is a constant. Situations
of this general type also have been analyzed with use of a
slightly different expression.37-44) The temperature dependence
of the chemical shift of the NMe2 group is shown in Figure 7.
A fit to eq 12 yields δdia ) 3.8(4) ppm, ∆H° ) 9.9(1.3) kJ
mol-1, and ∆S° ) -40 J K-1 mol-1. The error of ∆S° is
estimated to be at least ( 20 J K-1 mol-1. A chemical shift of
3.8 ppm is in the expected range for a metal amide complex
(cf. ∼3.9 ppm in [(C6F5NCH2CH2N)3N]Mo(NMe2)33 at 22 °C).
A similar treatment of the chemical shift of the upfield backbone
methylene resonance yielded δdia ) 1.91(3) ppm, ∆H° )
10.2(1.4) kJ mol-1, and ∆S° ) -50 J K-1 mol-1, in agreement
with the previously obtained values for ∆H° and ∆S°. Therefore
Figure 6. Temperature-dependent proton NMR spectrum of
[(Me3SiNCH2CH2)3N]Mo(NMe2).
complex can be prepared virtually quantitatively (according to
proton NMR spectra) as shown in eq 11. [N3N]Mo(NMe2) is
extremely soluble in pentane and usually could be isolated only
as a blue-purple oil. It has been crystallized by diffusing
acetonitrile into an ether solution. The d2 dimethylamido
complex appears to be diamagnetic at first glance, as one would
expect, since the pseudo doubly bound dimethylamido ligand
has local C2V symmetry and the metal has local C3V symmetry.
Therefore the two d electrons are paired up in a dxz or dyz orbital.
(If the axial ligand has cylindrical symmetry then the dxz and
(34) Gu¨tlich, P.; McGarvey, B. R.; Klaeui, W. Inorg. Chem. 1980, 19,
3704.
(35) Klaeui, W.; Eberspach, W.; Gu¨tlich, P. Inorg. Chem. 1987, 26, 3977.
(36) Smith, M. E.; Andersen, R. A. J. Am. Chem. Soc. 1996, 118, 1119.
(37) LaMar, G. N., Horrocks, W. D., Jr.; Holm, R. H., Eds. NMR of
Paramagnetic Molecules; Academic: New York, 1973.
(38) Kriley, C. E.; Fanwick, P. E.; Rothwell, I. P. J. Am. Chem. Soc.
1994, 116, 5225.
(39) Saillant, R.; Wentworth, R. A. D. Inorg. Chem. 1969, 8, 1226.
(40) Cotton, F. A.; Eglin, J. L.; Hong, B.; James, C. A. J. Am. Chem.
Soc. 1992, 114, 4915.
(41) Cotton, F. A.; Chen, H. C.; Daniels, L. M.; Feng, X. J. Am. Chem.
Soc. 1992, 114, 89980.
(42) Fettinger, J. C.; Keogh, D. W.; Kraatz, H.-B.; Poli, R. Organome-
tallics 1996, 15, 5489.
dyz orbitals remain degenerate and the complex is high spin.25,27
)
However, the resonance for the methyl group appears as a broad
1
singlet with unusually large chemical shift in the H NMR
spectrum near room temperature (Figure 6), and the chemical
shift of this resonance, to a larger extent than any other, depends
upon temperature. At 180 K the dimethylamido resonance
appears near 5 ppm, while it appears near 14 ppm at 304 K.
(43) Campbell, G. C.; Reibenspies, J. H.; Haw, J. F. Inorg. Chem. 1991,
30, 171.
(44) Boersma, A. D.; Phillipi, M. A.; Goff, H. M. J. Magn. Reson. 1984,
57, 197.

11044 J. Am. Chem. Soc., Vol. 119, No. 45, 1997
Mo¨sch-Zanetti et al.
of these species suggest that they have mirror symmetry, we
propose that the methyl group has attacked the metal to give a
Mo(VI) species, as shown in eq 13. (The results of an X-ray
study of [N3N]Mo(NTMS)(Me) were of sufficiently high quality
to confirm the pseudooctahedral core and the Mo-CH3 linkage,
but the structure could not be refined to an acceptable level.)
The Mo(CH3) resonance appears at 0.43 ppm in [N3N]Mo-
(NTMS)(Me) and 0.19 ppm in [N3N]Mo(NMe)(Me) in proton
NMR spectra, and 20.7 and 18.8 ppm, respectively, in carbon
NMR spectra. [N3N]Mo(NMe)(CD3) has the same spectrum
as [N3N]Mo(NMe)(Me) except the resonance at 0.19 ppm is
absent. The relatively small chemical shift of the added methyl
group in the proton NMR spectrum is consistent with its being
bound to the metal, not to a nitrogen atom. If we assume that
the imido ligand is bound to the metal through a pseudo triple
bond, then only a single additional π bond can be formed if an
18-electron count in metal-ligand bonding orbitals is to be
maintained. A suitable π bonding orbital of this general type
would be the combination in the MoN3C plane shown below.
Figure 8. Temperature-dependent proton NMR spectrum of
[(C6F5NCH2CH2N)3N]Mo(NMe2).
we feel relatively confident that the high spin form of [N3N]-
Mo(NMe2) is only ∼2.5 kcal mol-1 higher in energy than the
low spin form. At room temperature the equilibrium constant
is of the order of 10-3, i.e., only ∼0.1% of [N3N]Mo(NMe2) is
in the high spin form. We can only speculate as to the nature
of the high spin form. An interesting possibility is that the
amido ligand in the high spin form is not planar, i.e., the electron
count in metal-based orbitals is 16 and an electron pair is present
in an orbital that is largely centered on nitrogen.
The temperature dependence of the proton NMR spectrum
of [N3N]Mo(NMe2) contrasts with the apparently “normal”
spectrum observed for [(C6F5NCH2CH2N)3N]Mo(NMe2).33 In
fact, the spectrum of [(C6F5NCH2CH2N)3N]Mo(NMe2) is also
temperature dependent (Figure 8), with the dimethylamido
resonance shifting from 2.75 ppm at 188 K to 6 ppm at 357 K
(with concomitant broadening almost into the baseline). If the
chemical shift of the dimethylamido ligand is plotted versus
temperature and fit to eq 12 we obtain δdia ) 2.74(1) ppm, ∆H°
) 18.2(0.7) kJ mol-1, and ∆S° ) -20 J K-1 mol-1. The much
higher value for ∆H° is consistent with the much less dramatic
shift for the dimethylamido resonance from ∼2.8 to 6 ppm as
the temperature is raised from 188 to 367 K, and with a complex
in which the π bond is stronger than it is in [N3N]Mo(NMe2)
as a consequence of the electron-withdrawing ability of the C6F5
substituents relative to TMS substituents.
The chemical shifts for the phosphido protons in [N3N]Mo-
(PPhH) (12.16 ppm) and [N3N]W(PPhH) (18.68 ppm) led us
to question as to whether these too depend dramatically upon
temperature. However, variable-temperature proton NMR
spectra of [N3N]Mo(PPhH) between 22 and 80 °C show only
small, “normal” shifts for the phosphido proton. We conclude
that the high spin form of [N3N]Mo(PPhH) is so high in energy
that it cannot be accessed to any significant extent in this
temperature range. Since the precise nature of the high spin
form is not known, we cannot provide any quantifiable
explanation as to why the difference in energy of the high spin
(HS) and low spin (LS) forms of [N3N]Mo(PPhH) is so much
greater than the high and low spin forms of [N3N]Mo(NMe2).
However, in the complexes examined so far, one could correlate
the LS/HS gap with the degree of π bonding in the LS complex,
i.e., π bonding of the phosphido ligand in [N3N]Mo(PPhH) most
efficiently “locks” the complex in the relatively low energy,
low spin state.
[N3N]Mo(NTMS)(Me) is thermally quite stable. After heat-
ing a solution of [N3N]Mo(NTMS)(Me) to 100 °C for 3 days,
it decomposes to yield [N3N]MotN as the major product.
[N3N]Mo(NMe)(Me) is much less thermally stable than
[N3N]Mo(NTMS)(Me), decomposing in solution slowly even
at room temperature to give methane and what we propose to
be the diamagnetic iminato complex, [N3N]Mo(NdCH2) (eq
14). After 48 h at 65 °C [N3N]Mo(NdCH2) is formed in
quantitative yield, according to NMR spectra. However, its high
solubility limits its isolated yield to 34% on a small scale (∼100
mg). The methylene group of the iminato ligand appears as a
singlet at 5.19 ppm.45
Attempts to Prepare [C6F5NCH2CH2)3N]3- ([N3NF]3-
)
The failure to attack [N3N]ModN(SiMe3) with electrophiles
led us to attempt to treat {[N3N]ModNR}+ complexes with
carbon nucleophiles. For example, the reaction between {[N3N]-
ModNMe}+ and MeMgCl might yield known [N3N]Mo-
(NMe2). However, both {[N3N]ModNR}+ complexes (R )
Me and SiMe3) react with MeMgCl to yield dark purple
diamagnetic crystalline products in virtually quantitative yield
that analyze as [N3N]Mo(NR)(Me) species. Since NMR spectra
Complexes. The next most common type of triamidoamine
ligand is the [C6F5NCH2CH2)3N]3- ([N3NF]3-) ligand.25 There-
fore we became interested in whether we could prepare
[N3NF]MtP complexes by the same techniques used to prepare
[N3N]MtP complexes.
(45) For other examples of iminato complexes see: Shapley, P. A.;
Shusta, J. M.; Hunt, J. L. Organometallics 1996, 15, 1622, and references
therein.

Triamidoamine Complexes of Molybdenum and Tungsten
J. Am. Chem. Soc., Vol. 119, No. 45, 1997 11045
Lithium phenylphosphide reacts readily with [N3NF]WCl33
in toluene to give brown, crystalline, diamagnetic [N3NF]W-
(PPhH) in high yield. The phosphido proton resonance was
located at 17.50 ppm in the proton NMR spectrum (JHP ) 295
Hz, JHW ) 22 Hz) and the phosphorus resonance at 118.1 ppm
(JPW ) 742 Hz). Brown, crystalline [N3NF]Mo(PPhH) was
would have a polymeric structure analogous to that of [(t-
BuO)3WtN]x.49
The mechanism of formation of terminal phosphido and
arsenido complexes from substituted phosphides or arsenides
remains relatively obscure. Observations concerning the forma-
tion of terminal phosphido complexes are consistent with
formation and decomposition of intermediate “lithiophosphido”
complexes, M[P(R)(Li)] (where R ) Ph or SiMe324). In the
case of terminal arsenido complexes, however, there is no
evidence that M(IV) arsenido complexes are in fact intermedi-
ates. The most straightforward formation of a terminal phos-
phido complex is the reaction between Mo[N(t-Bu)(aryl)]3 and
P4,17 although there are important questions that remain to be
answered even in that circumstance. The synthesis of other
types of triamidoamine or other triamido/donor Mo and W
complexes should help answer many of these questions, as the
only two types of [(RNCH2CH2)3N]M complexes for which
general syntheses have been reported are where R ) SiMe3 or
C6F5.25 A molybdenum nitrido complex in which R ) Me has
been reported,50 but no direct entry into [(MeNCH2CH2)3N]M
complexes has yet appeared.
There are now several complexes known, in addition to those
reported here, that contain terminal phosphido complexes as
ligands. [(Me3SiNCH2CH2)3N]MtP complexes would appear
to have some potential as “ancillary” phosphorus ligands, i.e.,
alternatives to phosphine ligands, should they become more
readily available, since even a casual inspection suggests that
the cone angle of a [N3N]MtP “ligand” is well over 180°. The
electronic characteristics of a [N3N]MtP “ligand” also could
prove to be beneficial toward chemistry at the metal center to
which phosphorus is bound, although exactly what that benefit
might be is not predictable at this stage.
synthesized by the same method (δHP ) 11.20 ppm; JHP
)
281 Hz; δP ) 189.6 ppm). These data are analogous to those
found for [N3N]W(PPhH) and [N3N]Mo(PPhH). Many attempts
were made to convert [N3NF]W(PPhH) into [N3NF]WtP
employing various lithium reagents (LiPPhH, LiBu, LiMe, and
LiPh) under a variety of conditions, but in all cases either no
reaction occurred or no products could be identified. Chemical
oxidation ([Cp2Fe]OTf) followed by addition of triethylamine
also did not yield a clean product. Attempts to cleave the
phosphorus-phenyl linkage photochemically or with lithium
metal also were unsuccessful, as was the reaction between
[N3NF]WCl and Li2PPh. We conclude that if [N3NF]WtP is
formed in some of these reactions it is unstable under the
reaction conditions. More likely [N3NF]WtP simply is not
formed as readily as [N3N]WtP. The electron withdrawing
nature of the C6F5 rings renders the metal more electron poor
and therefore less susceptible to oxidation to the 6+ oxidation
state. The C6F5 rings might also be subject to electron transfer
from or nucleophilic attack by a lithium reagent. Other results
also suggest that [N3NF]3- complexes are more difficult to
oxidize than [N3N]3- analogs. For example, [N3NF]W(CH2-
SiMe3) loses hydrogen relatively slowly while the reaction
between [N3N]WCl and LiCH2SiMe3 yields [N3N]WtCSiMe3
rapidly,46 presumably via intermediate [N3N]W(CH2SiMe3), as
has been observed for other [N3N]W(CH2R) complexes.20 Since
[N3NF]M complexes are also less sterically crowded in the
trigonal pocket than are [N3N]M complexes,25 there also would
be less steric pressure on phosphorus substituents to be lost to
give [N3NF]MtP species. Of course, none of these results
constitutes evidence that [N3NF]MtP species are not stable,
only that they cannot be made by the methods employed so
far. It should be noted that [N3NF]MtN (M ) Mo or W)
complexes have been prepared by treating [N3NF]MCl com-
plexes with sodium azide.33
Some of the chemistry reported here provides a glimpse into
the possibility of employing a [N3N]M unit as a relatively inert,
nonlabile “platform” for carrying out chemistry at E in
[N3N]MtE complexes. At this stage we are most interested
in the possibility of removing a nitrido ligand, for example as
an amine, and forming triamidoamine complexes that could be
employed for binding and reducing dinitrogen.51 The develop-
ment of triamido donor ligands that are more stable to a variety
of reaction conditions we believe also will be a key to the
ultimate success of employing a [N3N]M unit as an inert
“platform”, and work toward these goals is currently under way.
Discussion
It is perhaps not surprising that Mo and W terminal phosphido
and arsenido complexes can be prepared when three amido
ligands are present as supporting ligands, since 18-electron
complexes of this general type appear to form strong MtE
bonds (E ) CR or N, for example).25 Although no thermody-
namic data are yet available, some evidence for strong MtE
bonds consists of the homolytic cleavage of dinitrogen by Mo-
[N(t-Bu)(aryl)]3 complexes,47,48 and the spontaneous loss of
molecular hydrogen from [N3N]W(CH2R) complexes20 (and
[N3N]Mo(CH2CMe3)27), to give [N3N]MtCR complexes. It
is also almost certainly important that Mo[N(t-Bu)(aryl)]3 and
[N3N]M complexes are sterically protected against bimetallic
reactions by the trigonal arrangement of amido ligand substit-
uents surrounding the terminal phosphido or arsenido ligand.
The simplest consequence of a bimetallic reaction would be
binding of the phosphido ligand to another metal center to form
a polymer. Although (t-BuO)3WtP has been observed to
function as a ligand,6 it seems plausible that free (t-BuO)3WtP
Experimental Section
General Procedures. All experiments were conducted under
nitrogen in a Vacuum Atmospheres drybox, using standard Schlenk
techniques, or on a high-vacuum line (<10-4 Torr). Pentane was
washed with HNO3/H2SO4 (5/95 v/v), sodium bicarbonate, and H2O,
stored over CaCl2, and then distilled from sodium benzophenone under
nitrogen. Regent grade ether, tetrahydrofuran, and benzene were
distilled from sodium benzophenone under nitrogen. Toluene was
distilled from molten sodium. Methylene chloride was distilled from
CaH2. All solvents were stored in the drybox over activated 4 Å
molecular sieves. Deuterated solvents were freeze-pump-thaw de-
gassed and vacuum transferred from an appropriate drying agent, or
sparged with argon and stored over sieves. NMR spectra are recorded
in C6D6 unless noted otherwise. 1H and 13C data are listed in parts per
million downfield from tetramethylsilane and were referenced using
the residual protonated solvent peak. 19F NMR are listed in parts per
million downfield of CFCl3 as an external standard. 15 NMR chemical
shifts are quoted versus ammonia and 31P NMR chemical shifts versus
(46) Seidel, S. W. Unpublished observations.
(49) Chisholm, M. H.; Hoffman, D. M.; Huffman, J. C. Inorg. Chem.
1986, 22, 2903.
(50) Plass, W.; Verkade, J. G. J. Am. Chem. Soc. 1992, 114, 2275.
(51) O’Donoghue, M. B.; Zanetti, N. C.; Schrock, R. R.; Davis, W. M.
J. Am. Chem. Soc. 1997, 119, 2753.
(47) Laplaza, C. E.; Cummins, C. C. Science 1995, 268, 861.
(48) Laplaza, C. E.; Johnson, M. J. A.; Peters, J. C.; Odom, A. L.; Kim,
E.; Cummins, C. C.; George, G. N.; Pickering, I. J. J. Am. Chem. Soc.
1996, 118, 8623.

11046 J. Am. Chem. Soc., Vol. 119, No. 45, 1997
Mo¨sch-Zanetti et al.
H3PO4. Coupling constants are given in hertz, but routine couplings
are not listed. Numerical values of integrals of proton NMR resonances
usually are not listed, unless some clarification is required, and should
be assumed to be consistent with the proposed assignment. Elemental
analyses (C, H, N) were performed on a Perkin-Elmer 2400 CHN
analyzer in our own laboratory.
7.76 (m, Ar), 7.22 (t, Ar), 7.03 (m, Ar), 3.38 (t, CH2), 2.35 (t, CH2),
0.28 (s, TMS); 13C{1H} NMR (toluene-d8) δ 133.13 (d, Cipso), 128.33,
128.25, 127.71, 55.86 (s, CH2), 54.82 (s, CH2), 4.15 (s, SiMe3); 31P-
{1H} NMR (toluene-d8) δ 142.03. Anal. Calcd for C21H45N4PSi3Mo:
C, 44.66; H, 8.03; N, 9.92. Found: C, 44.57; H, 8.05; N, 9.90.
[N3N]W(PPhH). Phenylphosphine (40 mg) was added to a solution
of 40 mg of [N3N]WPh in 1 mL of toluene. The orange solution was
stirred for 48 h at room temperature. The solvent was removed in
vacuo, and the solid was extracted with pentane. The solvent was
reduced to 1 mL and stored at -40 °C overnight to yield 20 mg of
orange crystalline product: 1H NMR 18.68 (d, JHP ) 300; JHW ) 36,
PPhH), 3.37 (t, CH2), 2.21 (t, CH2), 0.33 (s, TMS); 31P NMR δ 91 (s,
JPW ) 719).
{[N3N]ModNMe}OTf. [N3N]MotN (104 mg, 0.22 mmol) was
dissolved in 3 mL of toluene, and MeOTf (30 µL, 0.27 mmol) was
added by syringe. The reaction mixture immediately deepened in color
and a yellow solid precipitated. After 1 h the solvent was removed in
vacuo and the resulting solid was dissolved in the minimum volume
of THF. The solution was cooled to -30 °C to give the product as
yellow needles; yield 129 mg (93%): 1H NMR (CD2Cl2) 4.45 (s,
NCH3), 4.00 (t, CH2), 3.24 (t, CH2), 0.30 (s, SiMe3); 13C{1H} NMR
(CD2Cl2) 57.0 (NCH3), 56.2 (CH2), 54.5 (CH2), 3.2 (TMS). Anal. Calcd
for C17H42N5Si3F3SO3Mo: C, 32.22; H, 6.68; N, 11.05. Found: C,
32.07; H, 6.61; N, 11.17.
All organic compounds were received from commercial suppliers
and were used as received. [N3N]MoCl,27 [N3N]WCl,26 [N3NF]MoCl,33
52
[N3NF]WCl,33 and LiPH2 were prepared according to published
procedures. [N3N]MoPh and [N3N]WPh were prepared by adding LiPh
to [N3N]MCl in diethyl ether.26
[N3N]MotN. [N3N]MoCl (106 mg, 0.22 mmol) was dissolved in
10 mL of toluene and placed in a bomb. TMSN3 (120 µL, 0.90 mmol)
was added by syringe and the bomb was sealed. The solution was
heated at 90 °C for 24 h during which time the color of the reaction
mixture changed to brown/yellow. The solvent was removed and the
residue was extracted into pentane and filtered. The filtrate was reduced
in volume and cooled to -30 °C to give the product as a yellow
crystalline compound; yield 89 mg (88%): 1H NMR δ 3.23 (t, CH2),
2.14 (t, CH2), 0.56 (s, SiMe3); 13C{1H} NMR δ 52.2 (CH2), 51.6 (CH2),
3.1 (SiMe3); IR (Nujol) cm-1 1001 (νMoN). Anal. Calcd for C15H39-
N5Si3Mo: C, 38.36; H, 8.37; N, 14.91. Found: C, 37.99; H, 8.17; N,
14.65.
[N3N]MotP. [N3N]MotP was prepared as described in an earlier
publication.18
{[N3N]ModNSiMe3}OTf. [N3N]MotN (75 mg, 0.16 mmol) was
dissolved in 3 mL of toluene and TMSOTf (40 µL, 0.21 mmol) was
added by syringe. The reaction mixture immediately deepened in color
and a yellow solid precipitated. After 2 h the solvent was removed in
vacuo and the resulting solid was dissolved in the minimum volume
of THF. The solution was cooled to -30 °C to give the product as
yellow needles; yield 92 mg (83%): 1H NMR δ (CD2Cl2) 3.85 (t, CH2),
3.11 (t, CH2), 0.58 (s, 9, SiMe3), 0.33 (s, 27, SiMe3); 13C{1H} NMR
(CD2Cl2) δ 58.2 (CH2), 57.1 (CH2), 3.9 (SiMe3), 2.4 (SiMe3). Anal.
Calcd for C19H48N5Si4F3SO3Mo: C, 32.98; H, 6.99; N, 10.12. Found:
C, 32.84; H, 6.47; N, 9.58.
[N3N]MotAs. LiAsPhH (295 mg, 1.84 mmol) was added to a
solution of 431 mg (0.88 mmol) of [N3N]MoCl in a mixture of toluene
and THF (30 mL/3 mL). The clear orange solution was stirred at 50
°C for 40 h. The solvents were removed, and the resulting brown oil
was extracted twice with 60 mL of pentane. Evaporation of the solvent
yielded an orange powder that was recrystallized from cold ether; yield
140 mg (30 %) of orange cubes: 1H NMR 3.39 (t, CH2), 1.66 (t, CH2),
0.79 (s, TMS); 13C{1H} NMR 54.57 (s, CH2), 51.85 (s, CH2), 5.59 (s,
TMS). Anal. Calcd for C15H39N4AsSi3Mo: C, 33.95; H, 7.41; N,
10.56. Found: C, 34.05; H, 7.60; N, 10.32.
[N3N]WtN. A suspension of [N3N]WCl (136 mg, 0.23 mmol) and
NaN3 (38 mg, 0.58 mmol) in a mixture of acetonitrile (10 mL) and
tetrahydrofuran (10 mL) was stirred for 2 days at room temperature.
The color changed from orange to nearly colorless. The solvent was
removed in vacuo, and the residue was taken up in pentane (100 mL).
Insoluble salts were removed by filtration, and the solvent was
evaporated in vacuo. The white residue was recrystallized from ether
at -40 °C; yield 93 mg (71%) of white needles: 1H NMR δ 3.37 (t,
CH2), 2.00 (t, CH2), 0.59 (s, TMS); 13C{1H} NMR δ 51.59 (s, CH2),
50.71 (s, CH2), 3.12 (s, TMS). Anal. Calcd for C15H39N5Si3W: C,
32.31; H, 7.05; N, 12.56. Found: C, 32.68; H, 6.82; N, 12.28.
A 1:1 mixture of [N3N]WtN and [N3N]Wt15N was prepared in
an analogous fashion with Na15NN2: 15N NMR (toluene-d8, versus NH3)
551.2 (JNW ) 48 Hz).
{[N3N]ModPMe}OTf. Methyl triflate (35µL, 51 mg, 0.31 mmol)
was added at -35 °C to a solution of 142 mg (0.29 mmol) of [N3N]-
MotP in 12 mL of ether. This cold solution was stored at -35 °C
for 5 days, during which time orange-red crystals grew at the wall of
the vial. The solvent was decanted and the crystals washed with cold
ether. {[N3N]ModPMe}OTf can be recrystallized from THF by
diffusion of ether into the THF solution at -35 °C; yield 104 mg
(55%): 1H NMR (THF-d8, -40 °C) δ 4.13 (t, CH2), 3.72 (d, JPH
)
26, PCH3), 3.39 (t, CH2), 0.36 (s, SiMe3); 31P-NMR (THF-d8, -40 °C)
δ 282.7. Anal. Calcd for MoSi3N4PF3SO3C17H42: C, 31.38; H, 6.51;
N, 8.61. Found: C, 31.13 ; H, 6.69, N, 8.27.
{[N3N]ModPMe}OTf decomposes slowly in solution at 22 °C.
{[N3N]ModAsMe}OTf (proposed). Methyl triflate (35µL, 33 mg,
0.20 mmol) was added at -35 °C to a solution of 98 mg (0.18 mmol)
of [N3N]MotAs in 12 mL of ether. This cold solution was stored at
-35 °C for 6 days, during which time red crystals grew on the wall of
the vial. The solution was decanted, and the crystals were washed
with cold ether. The compound is relatively unstable in solution or
the solid state: 1H NMR (THF-d8, -70 °C) δ 4.13 (br, 6, CH2), 3.72
(s, 3, AsCH3), 3.36 (br, 6, CH2), 0.36 (s, 27, SiMe3). No pure compound
could be isolated and in any case appeared to be too unstable for
elemental analysis.
[N3N]WtP. [N3N]WtP was prepared as described in an earlier
publication.18 It also could be prepared under similar conditions and
in similar yields employing Li2PPh or LiPH2.
[N3N]WtAs. An orange solution of [N3N]WPh (50 mg, 0.81 mmol)
and 50 mg (3.2 mmol) of phenylarsine in 10 mL of toluene were stirred
for 48 h at room temperature in the dark. The solvent was removed in
vacuo, and the brown solid was extracted with 40 mL of pentane. The
pentane extract was filtered, and the pentane was removed in vacuo to
give a yellow solid that was recrystallized from cold diethyl ether to
yield 30 mg (61%) of yellow crystals: 1H NMR (toluene-d8) δ 3.55 (t,
CH2), 1.61 (t, CH2), 0.75 (s, TMS); 13C{1H} NMR δ 54.84 (CH2), 52.11
(CH2), 6.46 (TMS). Anal. Calcd for C15H39N4AsSi3W: C, 29.13; H,
6.36; N, 9.06. Found: C, 29.00; H, 6.58; N, 8.99.
[N3N]Mo(PPhH). A mixture of [N3N]MoCl (135 mg, 0.27 mmol)
and LiPPhH (38 mg, 0.33 mmol) in a mixture of toluene (15 mL) and
tetrahydrofuran (5 mL) was heated for 12 h at 50 °C. The solvent was
removed and the residue was taken up in pentane (50 mL). The extract
was filtered and the pentane removed from the filtrate in vacuo. The
orange residue was recrystallized from ether by cooling a solution to
-40 °C; yield 109 mg (70%): 1H NMR δ 12.16 (d, 1JHP ) 283, PH),
Reaction of [N3N]MotAs with MeOTf To Yield [N3N]Mo-
(OSO2CF3). To a solution of 30 mg (0.05 mmol) of [N3N]MotAs in
2 mL of toluene was added 14 µL of methyl triflate. The color changed
from orange to deep red. The solution was filtered and upon
evaporation red crystals of [N3N]Mo(OTf) appeared; yield 33 mg
(97%): 1H NMR δ 11.0 (s, TMS), -42 (br s), -60 (br s).26
{[N3N]WdNMe}OTf. Methyl triflate (24 µL, 0.2 mmol) was added
to a solution of 100 mg (0.18 mmol) of [N3N]WtN in 3 mL of diethyl
ether. A white precipitate appeared immediately. After the suspension
was stirred for 2 h the white residue was filtered off and recrystallized
from a mixture of tetrahydrofuran and pentane; yield 102 mg (79%)
of white crystals: 1H NMR (CDCl3) δ 4.54 (s, NCH3), 4.18 (t, CH2),
3.36 (t, CH2), 0.28 (s, TMS); 13C{1H} NMR (CDCl3) δ 54.46 (s, CH2),
53.34 (s, CH2), 52.53 (s, NCH3), 2.99 (s, TMS). Anal. Calcd for
(52) Baudler, M.; Glinka, K. Inorg. Synth. 1990, 27, 228.
(53) Rosenberger, C.; Schrock, R. R.; Davis, W. M. Inorg. Chem. 1997,
36, 123.

Triamidoamine Complexes of Molybdenum and Tungsten
J. Am. Chem. Soc., Vol. 119, No. 45, 1997 11047
C17H42N5F3O3SSi3W: C, 28.29; H, 5.87; N, 9.70. Found: C, 28.24;
H, 6.08; N, 9.51.
amount of pentane: 1H NMR δ 3.61 (s, 3, NCH3), 3.54 (m, 4, CH2),
3.36 (t, 2, CH2), 2.82 (dt, 2, CH2), 2.47 (t, 2, CH2), 2.25 (ddd, 2, CH2),
0.29 (s, 18, SiMe3), 0.23 (s, 9, SiMe3), 0.19 (s, 3, MoCH3); 13C NMR
δ 65.1 (CH2), 63.3 (CH2), 58.5 (CH2), 55.3 (CH2), 52.4 (NCH3), 18.8
(MoCH3), 3.8 (SiMe3), 2.4 (SiMe3). Anal. Calcd for MoSi3N5C17H45:
C, 40.86; H, 9.08; N, 14.01. Found: C, 40.85; H, 8.67; N, 14.18.
[N3N]Mo(NdCH2). [N3N]Mo(NMe)Me (91 mg, 0.182 mmol) was
dissolved in 1 mL of C6D6, and the solution was heated in a closed
NMR tube to 65 °C for 40-48 h. The solution turned from purple to
deep red during this time. The solvent was removed in vacuo to give
85 mg (95%) of essentially pure product that can be recrystallized from
a minimal amount of pentane. The yield of recrystallized product (30
mg, 34%) is limited by the high solubility of the compound in
pentane: 1H NMR δ 5.19 (s, 2, NCH2), 3.29 (t, 6, CH2), 2.22 (t, 6,
CH2), 0.47 (s, 27, SiMe3); 13C NMR δ 122.8 (ΝCΗ2), 54.7 (CH2),
53.29 (CH2), 3.8 (SiMe3). Anal. Calcd for MoSi3N5C16H41: C, 39.72;
H, 8.54; N, 14.48. Found: C, 39.37; H, 8.85; N, 14.01.
[N3N]ModNSiMe3. A sample of {[N3N]Mo(NSiMe3)}OTf (203
mg, 0.293 mmol) was dissolved in 8 mL of THF, and the solution was
cooled to -35 °C. A freshly prepared solution of Li2C8H8 in ether
(0.183 M, 1.6 mL) was added in one portion. The solution turned
deep blue. The mixture was stirred for 10 min and allowed to reach
room temperature. The solvent was removed in vacuo, and the residue
was extracted with 10 mL of pentane. The extract was filtered through
a glass wool filter and concentated to 1 mL. The product crystallized
upon storing this solution at -35 °C; yield 125 mg (79%): 1H NMR
δ 5.6 (br, SiMe3), 1.15 (br , SiMe3), -4.5(br, CH2). Anal. Calcd for
MoSi4N5C18H48: C, 39.82; H, 8.91; N, 12.90. Found: C, 40.11; H,
9.39; N, 12.79.
[N3N]Mo(NMe2). A sample of [N3N]MoCl (600 mg, 1.22 mmol)
was dissolved in 30 mL of ether, and the solution was cooled to -35
°C. LiNMe2 (66 mg, 1.3 mmol) was added to this cold solution, and
the mixture was allowed to reach room temperature while being stirred
over a period of a total of 3 h. The solvent was removed from the
dark purple solution in vacuo and the residue was extracted with 5 mL
of pentane. The extract was filtered through a glasswool filter, and
the pentane was removed from the filtrate in vacuo to give a blue-
purple oil (478 mg, 78%). The product was obtained as fine needles
from ether by diffusion of acetonitrile into the solution in a closed,
two vial, system: 1H NMR (toluene-d8, 30 °C) δ 13.91 (s, 6, NMe2),
2.88 (t, 6 , CH2), 1.05 (t, 6, CH2), 0.41 (s, 27, SiMe3). Anal. Calcd
for MoSi3N5C17H45: C, 40.86; H, 9.08; N, 14.01. Found: C, 40.54;
H, 9.22, N, 13.66.
15N-labeled (50%) {[N3N]Wd15NMe}OTf was prepared in an
analogous fashion: 15N NMR (toluene-d8) δ 145.3 ppm (JNW ) 116
Hz).
{[N3N]WdPMe}OTf. A yellow suspension appeared upon addition
of methyl triflate (8.7 µL, 0.08 mmol) to a solution of 40 mg (0.07
mmol) of [N3N]WtP in toluene. After 1 h the solvent was evaporated
and the yellow residue was recrystallized from a mixture of tetrahy-
drofuran and pentane; yield 50 mg (97%) of yellow needles: 1H NMR
δ 3.87 (t, CH2), 3.18 (t, CH2), 2.57 (d, 2JHP ) 25, PMe), 0.2 (s, TMS);
31P NMR δ 338.6 (s, 1JPW ) 748); 13C NMR (CD2Cl2) δ 55.59 (d, 3JCP
1
) 7.4, CH2), 54.91 (s, CH2), 37.82 (d, JCP ) 33.5, PCH3), 5.16 (s,
TMS). Anal. Calcd for C17H42N4F3O3PSSi3W: C, 27.64; H, 5.73; N,
7.58. Found: C, 27.79; H, 5.81; N, 7.57.
{[N3N]WdAsMe}OTf. Methyl triflate (8.0 µL, 0.07 mmol) was
added to a solution of [N3N]WtAs (41 mg, 0.07 mmol) in toluene.
After 1 h the yellow product was isolated as described for
{[N3N]WdPMe}OTf; yield 50 mg (97%) of yellow clusters: 1H NMR
δ 3.95 (t, CH2), 3.25 (t, CH2), 2.73 (s, AsMe), 0.23 (s, TMS); 13C NMR
(CDCl3) δ 54.82 (s, CH2), 54.50 (s, CH2), 43.29 (s, AsCH3), 5.42 (s,
TMS). Anal. Calcd for C17H42N4AsF3O3SSi3W: C, 26.09; H, 5.41;
N, 7.16. Found: C, 26.38; H, 5.37; N, 7.28.
trans-[Rh{[N3N]MotP}2(CO)(CH3CN)]PF6. To a solution of 20
mg (0.05 mmol) of [RhCl(CO)2]2 was added 38 mg (0.11 mmol) of
TlPF6 in 2 mL of acetonitrile. The mixture was stirred for 2 h during
which time a white precipitate was formed. The suspension was filtered
through Celite. A solution of 0.1 g (0.21 mmol) of [N3N]MotP in 2
mL of CH2Cl2 was added to the yellow filtrate. The solution turned
deep red immediately and after 1 h was filtered again through Celite.
The solvent was removed from the filtrate in vacuo and the red residue
was recrystallized at -40 °C from a mixture of CH2Cl2 and diethyl
ether; yield 100 mg (75%) of red needles: 1H NMR (CD2Cl2) δ 3.34
(t, CH2), 2.73 (t, CH2), 2.23 (s, CH3CN), 0.44 (s, TMS); 31P NMR
(CD2Cl2) δ 791.1 ppm (d, JPRh ) 67 Hz); IR (Nujol) cm-1 2021 (νCO).
Anal. Calcd for C33H81N9F6OP3RhSi6Mo2: C, 30.72; H, 6.33; N, 9.77.
Found: C, 30.38; H, 6.33; N, 9.38.
trans-[Rh{[N3N]WtP}2(CO)(CH3CN)]PF6. This compound was
synthesized in an analogous fashion as above using 13.5 mg (0.035
mmol) of [RhCl(CO)2]2, 27 mg (0.077 mmol) of TlPF6, and 80 mg
(0.14 mmol) of [N3N]WtP; yield 75 mg (73%): 1H NMR (CD2Cl2) δ
4.05 (t, CH2), 2.75 (t, CH2), 2.26 (s, CH3CN), 0.47 (s, TMS); 31P NMR
(CD2Cl2) δ 642.57 (d, JPRh ) 79); IR (Nujol) cm-1 2014 (νCO),
2361(νCN). Anal. Calcd for C33H81N9F6OP3RhSi6W2: C, 27.04; H,
5.57; N, 8.60. Found: C, 26.98; H, 5.22; N, 8.21.
trans-[Rh{[N3N]WtAs}2(CO)(CH3CN)]PF6. This compound was
synthesized in an analogous fashion as above by using 7.9 mg (0.02
mmol) of [RhCl(CO)2]2, 14.1 mg (0.04 mmol) of Tl(PF6), and 50 mg
(0.08 mmol) of [N3N]WtAs; yield 50 mg (80%): 1H NMR (CD2Cl2)
δ 4.1 (t, CH2), 3.75 (t, CH2), 2.27 (s, CH3CN), 0.5 (s, TMS); IR (Nujol)
cm-1 2001 (νCO), 2360 (νCN).
[N3N]Mo(NSiMe3)Me. A sample of {[N3N]ModNSiMe3}[CF3SO3]
(151 mg, 0.218 mmol) was dissolved in 5 mL of THF. The solution
was cooled to -35 °C and MeMgCl (72 µL, 0.23 mmol) was added.
The color changed to deep purple. The mixture was stirred for 30
min and allowed to warm to room temperature. The solvent was
removed in vacuo, and the mixture was extracted with pentane. The
extract was filtered through a glasswool filter, and the solvent was
removed to yield 109 mg of dark purple product (90% yield). This
essentially pure sample was recrystallized from a minimal amount of
pentane; yield 0.18 g (45%): 1H NMR δ 3.63 (ddd, 2, CH2), 3.47 (m,
2, CH2), 3.34 (t, 2, CH2), 2.66 (dt, 2, CH2), 2.33 (t, 2, CH2), 2.23 (ddd,
2, CH2), 0.46 (s, 9, SiMe3), 0.45 (s, 3, MoCH3), 0.38 (s, 18, SiMe3),
0.32 (s, 9, SiMe3); 13C NMR δ 64.3 (CH2), 62.6 (CH2), 59.5 (CH2),
56.7 (CH2), 20.7 (MoCH3), 3.1 (SiMe3), 3.2 (SiMe3), 2.8 (SiMe3). Anal.
Calcd for MoSi4N5C19H51: C, 40.90; H, 9.21; N, 12.55. Found: C,
41.28; H, 9.37; N, 12.68.
[N3NF]W(PPhH). To a stirred slurry of [N3NF]WCl (500 mg, 0.581
mmol) in 30 mL of toluene was added solid LiPPhH (67 mg, 0.58
mmol) followed by 20 drops of THF. Addition of THF caused the
reaction to become homogeneous. After 30 min, a second portion of
LiPPhH (20 mg, 0.17 mmol) was added. The reaction was stirred for
another 30 min, at which point 19F NMR showed it was complete. The
solution was filtered through Celite and the volatiles removed in Vacuo.
The brown solid was recrystallized from CH2Cl2 layered with pentane
at -40 °C, and 469 mg (86%) brown crystals were isolated in three
crops: 1H NMR δ 17.5 (dd, JPH ) 295 Hz, 2JWH ) 22, 1, PPhH), 6.97
(t, 4, phenyl), 6.82 (m, 1, phenyl), 3.23 (t, 6, CH2), 2.35 (t, 6, CH2);
19F NMR δ -150.4 (s, 6, Fortho), -164.0 (t, 3, Fpara), -165.1 (s, 6,
F
meta); 31P NMR δ 118.1 (2, JPW ) 742, PHPh). Anal. Calcd for
C30H18N4F15PW: C, 38.57; H, 1.94; N, 6.00. Found: C, 38.57; H,
2.11; N, 5.55.
[N3NF]Mo(PPhH). Solid LiPPhH (45 mg, 0.39 mmol) was added
to a stirred slurry of [N3NF]MoCl (300 mg, 0.388 mmol) in 20 mL of
toluene. Ten drops of THF was then added, which caused the mixture
to become homogeneous. After 1 h, more LiPPhH (14 mg, 0.12 mmol)
was added and the mixture was stirred for 3 h. An 19F NMR spectrum
showed the reaction to be complete. The reaction was filtered through
Celite, and the solvent was removed in Vacuo. The brown residue was
recrystallized from CH2Cl2 layered with pentane at -40 °C. The
product was obtained as a brown microcrystalline solid, 263 mg in
two crops (80%): 1H NMR δ 11.2 (d, JPH ) 281, MoPPhH), 6.99 (m,
Ph), 6.87 (m, Ph), 3.29 (t, CH2), 2.44 (t, CH2); 19F NMR δ -150.1 (d,
6, Fortho), -164.5 (t, 3, Fpara), -165.3 (t, 6, Fmeta); 31P NMR δ 189.6
(d). Anal. Calcd for C30H18N4F15PMo: C, 42.57; H, 2.14; N, 6.62.
Found: C, 42.57; H, 2.32; N, 6.37.
[N3N]Mo(NMe)Me. This compound was prepared in a manner
similar to that used to prepare [N3N]Mo(NSiMe3)Me from
[[N3N]Mo(NMe)]{CF3SO3} (142 mg, 0.224 mmol) and 75 µL of a 3.2
M solution of MeMgCl. The essentially pure product was obtained
quantitatively (110 mg), but could be crystallized from a minimal

11048 J. Am. Chem. Soc., Vol. 119, No. 45, 1997
Mo¨sch-Zanetti et al.
X-ray Structure of [N3N]MotAs. The structure was solved at 188
K with use of a Siemens SMART/CCD diffractometer. (A typical
procedure can be found elsewhere.53) Empirical formula C17H42-
AsF3N4O3SSi3W, FW ) 782.63, space group P212121, a ) 11.647(3)
Å, b ) 14.061(4) Å, c ) 18.876(6) Å, V ) 3091.3(14) Å3, Z ) 4, Dcalc
) 1.682 Mg/m3. Full details can be found in the Supporting
Information.
X-ray Structure of {[N3N]WdAsMe}OTf. The structure was
solved by direct methods at 188 K with use of an Enraf-Nonius CAD-4
diffractometer. (A typical procedure can be found elsewhere.54)
Empirical formula C15H39N4Si3AsMo, FW ) 530.62, space group Pa3h,
a ) 17.197(2) Å, V ) 5085.4(4) Å3, Z ) 8, Dcalc ) 1.386 Mg/m3. Full
details can be found in the Supporting Information.
) 10.8639(4) Å, b ) 11.9376(5) Å, c ) 11.3007(5) Å, â )
90.0910(10)°, V ) 1465.57(10) Å3, Z ) 2, Dcalc ) 1.230 Mg/m3. Full
details can be found in the Supporting Information.
Acknowledgment. R.R.S. is grateful to the National Insti-
tutes of Health (GM 31978) for research support. N.C.Z. thanks
Ciba-Geigy for postdoctoral support through a Ciba-Geigy
Jubila¨ums Stiftung, and K.W. thanks the Alexander von
Humboldt Foundation for a Lynen Fellowship. R.R.S. also
thanks the National Science Foundation for funds to help
purchase a departmental Siemens SMART/CCD diffractometer.
X-ray Structure of trans-[Rh{[N3N]WtP}2(CO)(CH3CN)]PF6.
The structure was solved at 293 K with use of a Siemens SMART/
CCD diffractometer. Empirical formula C33H66F6N9OP3RhSi6W2, FW
) 1450.98, space group P21/n, a ) 17.2988(12) Å, b ) 16.7599(11)
Å, c ) 19.7912(13) Å, â ) 99.8960(10)°, V ) 5652.6(7) Å3, Z ) 4,
Dcalc ) 1.705 Mg/m3. Full details can be found in the Supporting
Information.
X-ray Structure of [N3N]ModNSiMe3. The structure was solved
at 293 K with use of a Siemens SMART/CCD diffractometer.
Empirical formula C18H48N5Si4Mo, FW ) 542.91, space group P21, a
Supporting Information Available: A detailed description
of X-ray data collection, structure solution, and refinement,
labeled ORTEP diagrams, tables of fractional coordinates,
anisotropic thermal parameters, and selected bond distances and
angles for [N3N]MotAs, {[N3N]WdAsMe}+, trans-[Rh-
{[N3N]WtP}2(CO)(NCCH3)]PF6, and [N3N]Mo(NSiMe3) (34
pages). See any current masthead page for ordering and Internet
access instructions. Supporting information for the structural
studies of [N3N]WtP and [N3N]Mo(PPhH) can be found in
an earlier publication.18
(54) Toreki, R.; Schrock, R. R.; Davis, W. M. J. Organomet. Chem.
1996, 520, 69.
JA971727Z