Molybdenum Triamidoamine Complexes that Contain Hexa-tert-butylterphenyl, Hexamethylterphenyl, or p-Bromohexaisopropylterphenyl Substituents. An Examination of Some Catalyst Variations for the Catalytic Reduction of Dinitrogen

Article abstract of DOI:10.1021/ja0306415

Three new tetramines, (ArNHCH₂CH₂)₃N, have been synthesized in which Ar = 3,5-(2,4,6-t-Bu₃C ₆H₂)₂C₆H₃ (H ₃[HTBTN₃N]), 3,5-(2,4,6-Me₃C₆H ₂)₂C₆H₃ (H₃[HMTN ₃N]), or 4-Br-3,5-(2,4,6-i-Pr₃C₆H ₂)₂C₆H₂ (H₃[pBrHIPTN ₃N]). The diarylated tetramine, {3,5-(2,4,6-t-Bu₃C ₆H₂)₂C₆H₃NHCH ₂CH₂}₂NCH₂CH₂NH ₂, has also been isolated, and the "hybrid" tetramine {3,5-(2,4,6-t-Bu₃C₆H₂)₂C ₆H₃NHCH₂CH₂}₂NCH ₂CH₂NH(4-t-BuC₆H₄) has been prepared from it. Monochloride complexes, [(TerNCH₂CH ₂)₃N]MoCl, have been prepared, as well as a selection of intermediates that would be expected in a catalytic dinitrogen reduction such as [(TerNCH₂CH₂)₃N]Mo≡N and {[(TerNCH₂CH₂)₃N]Mo(NH₃)}{BAr′ ₄} (Ter = HTBT, HMT, or pBrHIPT and Ar′ = 3,5-(CF ₃)₂C₆H₃)). Intermediates that contain the new terphenyl-substituted ligands are then evaluated for their efficiency for the catalytic reduction of dinitrogen under conditions where analogous [HIPTN₃N]Mo species give four turnovers to ammonia under "standard" conditions with an efficiency of ～65%. Only [pBrHIPTN₃N]Mo compounds are efficient catalysts for dinitrogen reduction. The reasons are explored and discussed.

Full text of DOI:10.1021/ja0306415

Published on Web 04/27/2004
Molybdenum Triamidoamine Complexes that Contain
Hexa-tert-butylterphenyl, Hexamethylterphenyl, or
p-Bromohexaisopropylterphenyl Substituents. An
Examination of Some Catalyst Variations for the Catalytic
Reduction of Dinitrogen
Vincent Ritleng, Dmitry V. Yandulov, Walter W. Weare, Richard R. Schrock,*
Adam S. Hock, and William M. Davis
Contribution from the Department of Chemistry, Massachusetts Institute of Technology,
Cambridge, Massachusetts 02139
Received December 1, 2003; E-mail: rrs@mit.edu
Abstract: Three new tetramines, (ArNHCH2CH2)3N, have been synthesized in which Ar ) 3,5-(2,4,6-t-
Bu3C6H2)2C6H3 (H3[HTBTN3N]), 3,5-(2,4,6-Me3C6H2)2C6H3 (H3[HMTN3N]), or 4-Br-3,5-(2,4,6-i-Pr3C6H2)2C6H2
(H3[pBrHIPTN3N]). The diarylated tetramine, {3,5-(2,4,6-t-Bu3C6H2)2C6H3NHCH2CH2}2NCH2CH2NH2, has
also been isolated, and the “hybrid” tetramine {3,5-(2,4,6-t-Bu3C6H2)2C6H3NHCH2CH2}2NCH2CH2NH(4-t-
BuC6H4) has been prepared from it. Monochloride complexes, [(TerNCH2CH2)3N]MoCl, have been prepared,
as well as a selection of intermediates that would be expected in a catalytic dinitrogen reduction such as
[(TerNCH2CH2)3N]MotN and {[(TerNCH2CH2)3N]Mo(NH3)}{BAr′4} (Ter ) HTBT, HMT, or pBrHIPT and
Ar′ ) 3,5-(CF3)2C6H3)). Intermediates that contain the new terphenyl-substituted ligands are then evaluated
for their efficiency for the catalytic reduction of dinitrogen under conditions where analogous [HIPTN3N]Mo
species give four turnovers to ammonia under “standard” conditions with an efficiency of ∼65%. Only
[pBrHIPTN3N]Mo compounds are efficient catalysts for dinitrogen reduction. The reasons are explored
and discussed.
molybdenum nitrogenase16-21 in the last dozen years have settled
important structural questions, but the debate concerning exactly
where and how dinitrogen is reduced continues today.
Introduction
The first example of a dinitrogen complex, [Ru(NH3)5(N2)]2+
,
was published in 1965.1 Since then, tremendous advances have
been made in the synthesis and chemistry of dinitrogen
complexes.2-11 The main goals have been to understand how
dinitrogen might be reduced to ammonia catalytically at a metal
center or how it might be incorporated into organic molecules.
The interest in reduction of dinitrogen to ammonia arises from
the fact that dinitrogen is reduced to ammonia in nature by
various nitrogenase enzymes,12-15 all of which contain Fe and,
in the most active enzymes, either Mo or V. It is generally
agreed that dinitrogen is reduced at one or more transition metal
centers, although details are still sparse. X-ray studies of a
We have been interested in complexes that contain a
[(RNCH2CH2)3N]3- (triamidoamine) ligand22,23 for several
reasons: the triamidoamine ligand is multidentate, the apical
site that is created when it binds to a metal is sterically protected
by the three amido substituents, and the metal to which such a
ligand is coordinated is likely to be in an oxidation state of 3+
or greater. In view of the presence of a molybdenum (along
with seven irons) in one nitrogenase,16-21 we have focused on
the synthesis and study of molybdenum complexes that contain
triamidoamine ligands for the reduction of dinitrogen. Most
recently, we have invested in meta-terphenyl-substituted ligands,
and in particular a ligand that is substituted with 3,5-(2,4,6-i-
Pr3C6H2)2C6H3 (or HIPT ) hexaisopropylterphenyl) groups.24,25
Starting with [(HIPTNCH2CH2)3N]MoCl ([HIPTN3N]MoCl),
(1) Allen, A. D.; Senoff, C. V. J. Chem. Soc., Chem. Commun. 1965, 621.
(2) Chatt, J.; Dilworth, J. R.; Richards, R. L. Chem. ReV. 1978, 78, 589.
(3) Hidai, M.; Mizobe, Y. Chem. ReV. 1995, 95, 1115.
(4) Hidai, M. Coord. Chem. ReV. 1999, 185-186, 99.
(5) Fryzuk, M. D.; Johnson, S. A. Coord. Chem. ReV. 2000, 200-202, 379.
(6) Bazhenova, T. A.; Shilov, A. E. Coord. Chem. ReV. 1995, 144, 69.
(7) Barriere, F. Coord. Chem. ReV. 2003, 236, 71.
(8) Richards, R. L. Coord. Chem. ReV. 1996, 154, 83.
(9) Henderson, R. A.; Leigh, G. J.; Pickett, C. J. AdV. Inorg. Chem. Radiochem.
1983, 27, 197.
(16) Chan, M. K.; Kim, J. S.; Rees, D. C. Science 1993, 260, 792.
(17) Chen, J.; Christiansen, J.; Campobasso, N.; Bolin, J. T.; Tittsworth, R. C.;
Hales, B. J.; Rehr, J. J.; Cramer, S. P. Angew. Chem., Int. Ed. Engl. 1993,
32, 1592.
(18) Einsle, O.; Tezcan, F. A.; Andrade, S. L. A.; Schmid, B.; Yoshida, M.;
Howard, J. B.; Rees, D. C. Science 2002, 297, 1696.
(10) Shaver, M. P.; Fryzuk, M. D. AdV. Synth. Catal. 2003, 345, 1061.
(11) MacKay, B. A.; Fryzuk, M. D. Chem. ReV. 2004, 104, 385.
(12) Burgess, B. K. Chem. ReV. 1990, 90, 1377.
(19) Rees, D. C.; Howard, J. B. Curr. Opin. Chem. Biol. 2000, 4, 559.
(20) Rees, D. C.; Chan, M. K.; Kim, J. AdV. Inorg. Chem. 1996, 40, 89.
(21) Bolin, J. T.; Ronco, A. E.; Morgan, T. V.; Mortenson, L. E.; Xuong, L. E.
Proc. Natl. Acad. Sci. U.S.A. 1993, 90, 1078.
(13) Burgess, B. K.; Lowe, D. J. Chem. ReV. 1996, 96, 2983.
(14) Eady, R. R. Chem. ReV. 1996, 96, 3013.
(22) Schrock, R. R. Acc. Chem. Res. 1997, 30, 9.
(15) Dos Santos, P. C.; Dean, D. R.; Hu, Y.; Ribbe, M. W. Chem. ReV. 2004,
104, 1159.
(23) Schrock, R. R. Pure Appl. Chem. 1997, 69, 2197.
(24) Yandulov, D. V.; Schrock, R. R. J. Am. Chem. Soc. 2002, 124, 6252.
9
6150
J. AM. CHEM. SOC. 2004, 126, 6150-6163
10.1021/ja0306415 CCC: $27.50 © 2004 American Chemical Society

Molybdenum Triamidoamine Complexes
A R T I C L E S
Scheme 1. Proposed Intermediates in the Reduction of Dinitrogen
at a [HIPTN3N]Mo (Mo) Center through the Stepwise Addition of
Protons and Electrons
formation of ammonia in ∼65% yield. To our knowledge, this
is the only reported catalytic reduction of dinitrogen in a
relatively well-defined manner. It should be compared with the
catalytic reduction of dinitrogen in a protic environment
(methanol is the solvent) to a 10:1 mixture of hydrazine and
ammonia reported by Shilov.6 In addition to Mo(III), the Shilov
system contains Mg(OH)2, some strong reducing agent (Ti(OH)3,
Hg electrode, Na amalgam, etc.), additives such as various
phosphines, and in some cases phosphatidylcholine that has been
proposed to function (inter alia) as a “detergent” at a surface of
sodium amalgam. As many as 170 equiv of ammonia (relative
to Mo) reportedly have been produced at 110 °C and 1 atm
pressure. The turnover rate also is said to approach that of the
Fe/Mo nitrogenase itself at room temperature (∼1 s-1 Mo-1
)
in the best cases. Relatively powerful reducing agents are
required, that is, those with E° < -1.75 V vs SCE. No details
as to how dinitrogen is reduced in the Shilov systems have been
established. Shilov has favored the idea that at least two metals
(presumably two molybdenum centers) are required to bind and
reduce dinitrogen to hydrazine and ammonia, although convinc-
ing data in support of that proposal have not been forthcoming.
We naturally are interested in variations of the [HIPTN3N]-
Mo complexes and their efficacies for catalytic dinitrogen
reduction. In this paper, we report three systems that are closely
related to the [HIPTN3N]3- system. One that is more protective
of the metal contains the 3,5-(2,4,6-t-Bu3C6H2)2C6H3 (HTBT)
group. A second that is less protective of the metal contains
the 3,5-(2,4,6-Me3C6H2)2C6H3 (HMT) group. A third that is
essentially the same sterically as the [HIPTN3N]3- ligand, but
slightly different electronically, contains the 4-bromo-3,5-(2,4,6-
i-Pr3C6H2)2C6H2 (pBrHIPT) group. We compare potential
catalysts that contain one of these three ligands in terms of their
ability to reduce dinitrogen catalytically under conditions that
have been reported for complexes that contain the [HIPTN3N]3-
ligand.26
we showed that we could prepare many intermediates in
a hypothetical reduction of dinitrogen (Scheme 1). These
intermediates include paramagnetic [HIPTN3N]Mo(N2) (A),
diamagnetic [HIPTN3N]Mo-NdN-H (B), diamagnetic
{[HIPTN3N]ModNNH2}{BAr′4} (C; Ar′ ) 3,5-(CF3)2C6H3),
diamagnetic [HIPTN3N]MotN (D), diamagnetic {[HIPTN3N]-
ModNH}{BAr′4} (E), and paramagnetic {[HIPTN3N]Mo-
(NH3)}{BAr′4} (F) and [HIPTN3N]Mo(NH3) (G). Extensive 15
N
labeling studies, NMR studies, and X-ray studies of A, D, and
F25 all reveal a trigonal pocket in which N2 and its reduced
products are protected to a dramatic degree by three 2,4,6-i-
Pr3C6H2 rings clustered around it (see drawing of A). The
intermediates shown in Scheme 1, in which the oxidation state
of the metal varies between Mo(III) and Mo(VI), are analogous
to those proposed originally by Chatt for low oxidation state
(initially Mo(0) and W(0)) phosphine complexes.2 The “naked”
complex, [HIPTN3N]Mo, has not been observed to date.
Results and Discussion
Ligand Syntheses. Syntheses of the terphenyl bromides,
HTBTBr, HMTBr, and pBrHIPTBr, are analogous to synthesis
of the isopropyl-substituted terphenyl bromide,25 the first steps
of which are shown in eq 1. This method follows that originally
reported by Hart27,28 with one important difference. The
Grignard reagent is prepared with 2 equiv of magnesium. The
Grignard reagents are then treated with 2,4,6-tribromoiodoben-
zene,27 which yields the Grignard reagent of the desired
m-terphenyl derivative and 2,4,6-trialkyliodobenzene. The 2,4,6-
trialkyliodobenzene then reacts in situ with the second equivalent
of magnesium to give the corresponding Grignard reagent,
(2,4,6-R3C6H2)MgX (eq 1). Acid hydrolysis of this mixture then
yields the desired 3,5-bis(2,4,6-trialkylphenyl)bromobenzene and
1,3,5-trialkylbenzene. The 1,3,5-trialkylbenzene is easily sepa-
rated from the 3,5-bis(2,4,6-trialkylphenyl)bromobenzene. Be-
cause 1,3,5-tri-tert-butylbenzene is relatively expensive, it
therefore can be recovered and used to remake 1-bromo-2,4,6-
tri-tert-butylbenzene. 3,5-Bis(2,4,6-tri-tert-butylphenyl)bro-
mobenzene (HTBTBr) and 3,5-bis(2,4,6-trimethylphenyl)-
bromobenzene (HMTBr) were obtained as white solids in 41%
and 26% yields, respectively. If N-bromosuccinimide is added
Recently, we have shown that dinitrogen can be reduced
catalytically in heptane to ammonia at room temperature and 1
atm in the presence of A, B, D, or F.26 Addition of the reductant
(decamethylchromocene) over a period of 6 h in the presence
of a poorly soluble proton source ({2,6-lutidinium}{BAr′4})
leads to the consumption of ∼4 equiv of dinitrogen and the
(25) Yandulov, D. V.; Schrock, R. R.; Rheingold, A. L.; Ceccarelli, C.; Davis,
W. M. Inorg. Chem. 2003, 42, 796.
(26) Yandulov, D. V.; Schrock, R. R. Science 2003, 76, 301.
(27) Du, C. J. F.; Hart, H.; Ng, K. K. D. J. Org. Chem. 1986, 51, 3162.
(28) Vinod, T. K.; Hart, H. J. Org. Chem. 1991, 56, 5630.
9
J. AM. CHEM. SOC. VOL. 126, NO. 19, 2004 6151

A R T I C L E S
Ritleng et al.
The fact that 1a′ can be isolated created the possibility that
“hybrid” ligands can be prepared in which the third arm can be
substituted with a different aromatic group, for example, a
smaller or electronically different group. To demonstrate the
principle, we coupled 1a′ with 4-tert-butylbromobenzene to give
(HTBTNHCH2CH2)2(4-t-BuC6H4NHCH2CH2)N (1d; eq 4). The
yield (12%) was not optimized.
to the mixture of Grignard reagents shown in eq 1, then
pBrHIPTBr (eq 2) can be isolated in good yield (49%) after
recrystallization from ether.
Compounds 1a-1c (eq 3: 1a ) H3[HTBTN3N], 1b ) H3-
[HMTN3N], and 1c ) H3[pBrHIPTN3N]) were then synthesized
in a reaction at 90-110 °C between N(CH2CH2NH2)3 and 3
equiv of the terphenyl bromide in the presence of an excess of
sodium tert-butoxide, 3 mol % Pd2(dba)3, and rac-BINAP (9
mol % vs the terphenylbromide).29,30 A higher temperature and
catalyst loading were required for the synthesis of 1a than for
1b or 1c, which were more typical of conditions employed in
reactions of this general type.31,32 Attempts to synthesize 1a
under milder conditions (80 °C) and with 0.5-2% catalyst failed
to give full conversion. Even at the higher temperatures, the
reaction was not complete and the final mixture always
contained 1a and the disubstituted tetraamine (HTBTNHCH2-
CH2)2(NH2CH2CH2)N (1a′). No disubstituted tetraamine was
observed during the preparation of 1b or 1c. Compounds 1a,
1b, and 1c were isolated as white solids in 65%, 35%, and 61%
yields, respectively. Compound 1a′ is relatively easy to separate
from 1a; the chromatography solvent can be adjusted so that
1a passes through the column, while 1a′ is absorbed strongly
and removed later with a more polar solvent mixture. Compound
1a′ can be isolated as a yellow solid in 25% yield.
Preparation of Monochloride Complexes. When 1a is
added to MoCl4(THF)2 in THF at room temperature, an adduct
forms readily, the nature of which is not known. After 1 h,
addition of 3.1 equiv of LiN(SiMe3)2 at concentrations that were
employed to prepare [HIPTN3N]MoCl25 led to [HTBTN3N]-
MoCl (2a) in 16% yield. Employing CH3MgCl as the base31
led to no improvement. Addition of the trilithium salt of 1a to
either MoCl4(THF)2 or MoCl3(THF)3 under a variety of condi-
tions gave only traces of 2a and/or [HTBTN3N]Mo(N2) (4a;
vide infra).
We suspected that it may be more difficult to form the
appropriate initial adduct when 1a is employed than when H3-
[HIPTN3N] is employed. The precise nature of this initial adduct
is not known, but it almost certainly should be monomeric (not
oligomeric or polymeric) to form 2a upon addition of base. A
1H NMR study in C6D6 showed that primarily two species are
present in samples that contain 1a and MoCl4(THF)2, one of
which is not converted to 2a upon addition of base. Decreasing
the concentration of 1a and MoCl4(THF)2 led to a decrease in
the amount of the unreactive initial adduct. Therefore, more
dilute conditions were employed. When solid MoCl4(THF)2 is
added slowly to a 0.01 M solution of HTBTBr in benzene, which
is at least 10 times more dilute than the reported procedure for
(29) Wolfe, J. P.; Wagaw, S.; Buchwald, S. L. J. Am. Chem. Soc. 1996, 118,
7215.
(30) Hartwig, J. F. Angew. Chem., Int. Ed. 1998, 37, 2046.
(31) Greco, G. E.; Schrock, R. R. Inorg. Chem. 2001, 40, 3850.
(32) Greco, G. E.; Schrock, R. R. Inorg. Chem. 2001, 40, 3860.
9
6152 J. AM. CHEM. SOC. VOL. 126, NO. 19, 2004

Molybdenum Triamidoamine Complexes
A R T I C L E S
synthesizing [HIPTN3N]MoCl,25 and the resulting dark red
solution is stirred for 2 h before adding LiN(SiMe3)2, 2a can
be isolated as an orange solid in 60% yield (eq 5). We later
Figure 1. Thermal ellipsoid plot of the crystal structure of [HTBTN3N]-
MoCl (2a) (50% probability level). Hydrogen atoms are omitted for clarity.
All non-hydrogen atoms were refined anisotropically.
The “hybrid” complex [(HTBTNCH2CH2)2(4-t-BuC6H4NCH2-
CH2)N]MoCl (2d) was prepared in a manner similar to that used
to prepare 2a and was isolated as an orange solid in 39% yield
(not optimized; eq 6). As a consequence of the absence of a C3
found that if a slight excess of MoCl4(THF)2 (0.2 equiv) is
employed in a concentrated reaction mixture (0.05 M), 2a again
can be obtained in good yield (69%). Both methods yield small
quantities (∼4%) of the nitride [HTBTN3N]MotN (3a, vide
infra), in sharp contrast to the analogous reactions involving
other H3[TerN3N] ligands. We suspect that the nitride originates
from competitive substitution of Cl- by N(SiMe3)2 at Mo,
-
followed by loss of the trimethylsilyl groups from the resulting
Mo-N(SiMe3)2 species through unknown pathways.
Compound 2a is paramagnetic and extremely sensitive to air
1
and moisture, as is [HIPTN3N]MoCl.25 Its H NMR spectrum
closely resembles spectra of other [(ArylNCH2CH2)3N]MoCl
complexes25,32 with three paramagnetically shifted resonances
being observed at 14.8, -11.9, and -82.5 ppm. The first of
these is ascribed to the ortho protons on the primary phenyl
ring and the last two to the backbone methylene resonances.
Other resonances within the 0-10 ppm window can be assigned
readily (see Experimental Section).
axis and the presence of a mirror plane, the 1H NMR spectrum
of 2d is more complex than that of 2a or 2b. If the HTBT ring
does not rotate about the Cipso-N bond axis, then the four ortho-
tert-butyl groups on each of the equivalent 3,5-[2,4,6-t-Bu3C6H2]-
C6H4 rings would be inequivalent. Evidently, rotation about the
Cipso-N bond axis is fast on the NMR time scale because only
two ortho-tert-butyl group resonances are observed at 1.73 and
1.65 ppm. For the same reason, only two Hm protons on the
2,4,6-t-Bu3C6H2 rings are observed at 7.88 and 7.85 ppm, instead
of four. Finally, six backbone methylene resonances of relatively
area 2 appear as six broad singlets in the same range of
frequencies as in 2a and 2b at -8.03, -14.00, -16.80 ppm,
and from -75 to -80 ppm (overlapping). The meta and ortho
protons of the 4-C6H4-t-Bu ring are found at 15.49 and 9.48
ppm, respectively, and the ortho protons of the 3,5-C6H3Ar2
groups are detected as a broad singlet at 13.37 ppm.
X-ray quality crystals of 2a were obtained by allowing a
saturated solution of 2a in pentane to stand at room temperature
for several days. A thermal ellipsoid drawing is shown in Figure
1. Crystal data, data collection, and structure refinement are
listed in Table 1, while selected bond distances and bond angles
are listed in Table 2. This structure cannot be compared directly
with the structure of [HIPTN3N]MoCl because an X-ray study
The monochloride complex [HMTN3N]MoCl (2b) can be
prepared by reacting 1b with MoCl4(THF)2 followed by LiN-
(SiMe3)2, either in THF or in benzene under dilute conditions
(eq 5). The reaction works well under either set of conditions
1
as shown by H NMR studies of the crude reaction mixtures.
Rapid workup of a reaction run in benzene allowed brown-red
2b to be isolated in 64% yield. Compound 2b appears to be
more sensitive to traces of oxygen than 2a, with a brown surface
forming on the red crystals. As expected, the 1H NMR spectrum
of 2b shows three paramagnetically shifted signals at 9.7, -16.0,
and -77.6 ppm. The first is ascribed to the ortho protons on
the primary phenyl ring and the latter to the backbone methylene
resonances.
The synthesis of [pBrHIPTN3N]MoCl (2c) followed the
procedure for [HIPTN3N]MoCl itself; dilute conditions were
not necessary. The yield was 55%, and the proton NMR
spectrum showed characteristic shifted resonances at 13.5, -18,
and -84 ppm.
9
J. AM. CHEM. SOC. VOL. 126, NO. 19, 2004 6153

A R T I C L E S
Ritleng et al.
Table 1. Crystal Data and Structure Refinement for [HTBTN3N]MoCl (2a) and [pBrHIPTN3N]MotN (3c)a
empirical formula
formula weight
crystal system
C132H195N4ClMo
1969.42
orthorhombic
C114H156Br3MoN5
1932.11
orthorhombic
Pna2(1)
space group
P2(1)2(1)2(1)
unit cell dimensions
a ) 19.165(2) Å
b ) 20.234(3) Å
c ) 38.069(5) Å
R ) â ) γ ) 90°
4, 14 763(3)
0.886
0.146
4296
a ) 24.5742(13) Å
b ) 43.729(2) Å
c ) 25.7954(14) Å
R ) â ) γ ) 90°
8, 27 720(3)
0.926
Z, volume (Å3)
density (calculated; Mg/m3)
absorption coefficient (mm-1
F(000)
)
0.996
8176
crystal size (mm3)
0.36 × 0.28 × 0.20
θ range for data collection (deg)
index ranges
reflns collected
2.08-21.00
1.24-20.00
-16 e h e 19, -20 e k e 18, -38 e l e 36
47 115
-23 e h e 23, -42 e k e 42, -24 e l e 24
125 318
independent reflns
15 803 [R(int) ) 0.0726]
99.7% (21.00)
15 803/0/1297
1.130
25 869 [R(int) ) 0.1276]
100.0% (20.00)
25 869/2/1733
completeness (to θ ) xx°)
data/restraints/parameters
GOF on F 2
1.028
final R indices [I > 2σ(I)]
R1 ) 0.0849, wR2 ) 0.2144
R1 ) 0.0901, wR2 ) 0.2182
0.06(4)
R1 ) 0.0574, wR2 ) 0.1058
R1 ) 0.0983, wR2 ) 0.1115
0.024(7)
R indices (all data)
absolute structure parameter
largest diff. peak and hole (e Å-3
)
1.095 and -0.777
0.592 and -0.503
a For each structure, the temperature was 193(2) K, the wavelength was 0.71073 Å, and the refinement method was full-matrix least-squares on F 2.
Table 2. Selected Bond Distances [Å] and Angles (deg) for
Table 3. A Summary of IR Dataa
[HTBTN3N]MoCl
compound
ν )
NN or νMoN (cm-1
Mo-N(1)
Mo-N(2)
Mo-N(3)
Mo-N(4)
Mo-Cl(2)
2.003(6)
1.983(6)
1.989(6)
2.193(6)
2.338(2)
C(201)-N(2)-Mo
C(301)-N(3)-Mo
N(2)-Mo-N(4)
N(3)-Mo-N(4)
N(1)-Mo-N(4)
N(2)-Mo-Cl(2)
N(3)-Mo-Cl(2)
N(1)-Mo-Cl(2)
N(4)-Mo-Cl(2)
N(2)-Mo-N(3)
N(2)-Mo-N(1)
N(3)-Mo-N(1)
C(101)-N(1)-Mo
129.7(5)
129.4(4)
80.5(2)
80.8(2)
79.8(2)
97.94(18)
100.22(16)
100.76(17)
178.41(16)
118.5(2)
117.7(2)
115.6(2)
128.7(4)
[HIPTN3N]Mo(N2)
1990 (1924 for 15N2)25
1990 (1924 for 15N2)
1992
[HTBTN3N]Mo(N2) (4a)
[pBrHIPTN3N]Mo(N2) (4c)
[TMSN3N]Mo(N2)
{Bu4N}{[HIPTN3N]Mo(N2)}
{Bu4N}{[HTBTN3N]Mo(N2)} (5a)
{Et4N}{[pBrHIPTN3N]Mo(N2)} (5c)
[HIPTN3N]MotN
193433
185525
1847
1857
1012b (986 for 15N in Nujol)25
1015 (991 for 15N)
1013 (985 for 15N)
[HTBTN3N]MotN (3a)
[pBrHIPTN3N]MotN (3c)
a All spectra were obtained in benzene except where noted. b In Nujol
the value was 1013 cm-1
.
of [HIPTN3N]MoCl was not carried out. However, we can
period of 2 days to give diamagnetic [HMTN3N]MotN (3b),
which was isolated as a bright yellow solid in 62% yield (eq
7). [pBrHIPTN3N]MotN (3c) can be prepared similarly.
compare the structure of 2a with that of [HIPTN3N]MotN.25
The most significant difference is the size of the Mo-N-Cipso
-
Co angle, where Co is the ortho carbon atom that is closest to
the metal. If the “primary” phenyl ring (the one connected to
the amido nitrogen) lies in the plane of the trigonal amido
nitrogen atom, then the Mo-N-Cipso-Co dihedral angle would
be 0°. If the primary phenyl ring lies perpendicular to the plane
of the trigonal amido nitrogen atom, then the Mo-N-Cipso
-
Co angle would be 90°. In 2a, the Mo-N-Cipso-Co angles are
67.0°, 64.6°, and 77.7°, whereas in [HIPTN3N]MotN the Mo-
N-Cipso-Co angles are 35.6°, 35.4°, and 33.5°. Therefore, one
might think that the large difference in the Mo-N-Cipso-Co
angles could be ascribed to steric differences between the i-Pr
system and the t-Bu system. However, Mo-N-Cipso-Co angles
vary significantly within various [HIPTN3N]3- species.25 There-
fore, any additional steric hindrance imposed by the tert-butyl
groups in [HTBTN3N]MoCl versus the steric hindrance in
[HIPTN3N]3- species cannot be judged in terms of Mo-N-
Cipso-Co dihedral angles. Not surprisingly, a space filling model
that contains all protons reveals an impressive degree of steric
congestion in the upper half of 2a, with the chloride barely being
visible from the top.
Slightly harsher conditions (110 °C, 4 equiv of Me3SiN3) were
required to prepare [HTBTN3N]MotN (3a) in good yield
(60%). Compounds 3a-3c are relatively stable in air as solids;
only minor surface decomposition was observed after several
hours. The νMotN stretch in 3a was located at 1015 cm-1, a
value that is close to that in [HIPTN3N]MotN (1012 cm-1
;
Preparation of Nitride Complexes. Compound 2b reacts
smoothly with 2 equiv of Me3SiN3 in toluene at 90 °C over a
Table 3).25 In 3c, the νMotN stretch was found at 1013 cm-1
(985 cm-1 in the 15N analogue). Therefore, any electronic
9
6154 J. AM. CHEM. SOC. VOL. 126, NO. 19, 2004

Molybdenum Triamidoamine Complexes
A R T I C L E S
Table 4. Selected Bond Distances [Å] and Angles (deg) for
[pBrHIPTN3N]MotN
Mo-N(1)
Mo-N(2)
Mo-N(3)
Mo-N(4)
Mo-N(5)
2.047(8)
2.018(10)
1.980(10)
2.324(9)
1.679(9)
C(201)-N(2)-Mo
C(301)-N(3)-Mo
N(2)-Mo-N(4)
N(3)-Mo-N(4)
N(1)-Mo-N(4)
N(2)-Mo-N(5)
N(3)-Mo-N(5)
N(1)-Mo-N(5)
N(4)-Mo-N(5)
N(2)-Mo-N(3)
N(2)-Mo-N(1)
N(3)-Mo-N(1)
C(101)-N(1)-Mo
130.5(10)
128.5(9)
79.3(5)
75.8(4)
78.3(4)
101.1(5)
102.4(5)
103.2(4)
178.1(4)
119.3(4)
114.2(4)
113.5(4)
128.0(9)
be isolated as dark emerald-green crystals from pentane/benzene
in 59% yield. This diamagnetic species shows all of the expected
resonances in the 1H NMR spectrum, and its IR spectrum
features a νNN stretch at 1847 cm-1, a value that is 8 cm-1 lower
than that in {Bu4N}{[HIPTN3N]MoN2} (Table 3).25 According
to this datum, the [HTBTN3N]3- ligand is slightly more electron-
donating than [HIPTN3N]. It is interesting to note that the lower
stretching frequency is not observed in the neutral N2 adducts,
but only in the anionic complexes, in which back-bonding to
N2 is more pronounced. However, the effect is still small.
An attempted reduction of 2b under the same conditions as
reduction of 2a did not lead to an analogous dinitrogen complex
4b. A reaction took place, but the product appeared to be too
unstable, at least under the conditions employed, to allow a
subsequent oxidation to a dinitrogen complex.
Figure 2. Thermal ellipsoid plot of the crystal structure of [pBrHIPTN3N]-
MotN (3c) (50% probability level). Hydrogen atoms are omitted for clarity.
All non-hydrogen atoms were refined anisotropically.
Attempts to prepare and isolate pure [pBrHIPTN3N]Mo(N2)
(4c) also have not been successful, even though the proce-
dure followed was nearly identical to the procedure published
for the synthesis of [HIPTN3N]Mo(N2). However, IR spectra
in C6D6 of the crude mixture obtained upon reduction of 2c
show that it consists largely of [pBrHIPTN3N]Mo(N2) and
[pBrHIPTN3N]Mo-NdN-Mg(THF)xCl, judging from νNN
peaks at 1992 cm-1 and νNN 1788 cm-1, respectively (in
benzene). The low value of 1788 cm-1 for νNN is a consequence
of binding of the magnesium to the â nitrogen atom.25 The
former should be compared with the νNN peak at 1990 cm-1 in
[HIPTN3N]Mo(N2) itself.25 The shift of 2 cm-1 to higher energy
is in the direction expected for a more electron-withdrawing
ligand and consequently less back-bonding to the dinitrogen
ligand. The magnitude of the shift was smaller than we expected,
although it is at least observable in the neutral dinitrogen
complex in this case. A sample of {Et4N}{[pBrHIPTN3N]-
MoN2} could be isolated (in 25% yield) that showed a νNN value
of 1857 cm-1 (vs 1855 cm-1 in the HIPT analogue as the
{Bu4N}+ salt). Again, the shift of 2 cm-1 to higher energy is
in the direction expected for a more electron-withdrawing ligand.
When a C6D6 solution of 4a-15N2 was placed under 1 atm of
14N2, 4a was formed over a period of weeks. The consumption
of 4a-15N2 was followed by IR and was found to proceed with
t1/2 ≈ 750 h at 22 °C. This rate is ∼20 times slower than that
for 15N2/14N2 exchange in the HIPT system (t1/2 ≈ 35 h).25 An
experiment at a pressure of 5 atm of 14N2 yielded essentially
the same result, which suggests that the 15N2/14N2 exchange
mechanism is unimolecular; we propose that “[HTBTN3N]Mo”
is the intermediate (eq 9). The exchange rate in 4a-15N2 is even
slower than that observed in [(Me3SiCH2CH2N)3N}Mo(14N2)
differences between the compounds that contain HIPT, HTBT,
or pBrHITP are reflected minimally in the value for νMotN
.
An X-ray study of 3c reveals it to have a structure similar to
that of [HIPTN3N]MotN25 (Tables 1 and 4, Figure 2). In 3c,
the MotN bond lengths are 1.679(9) Å (vs 1.652(5) Å in
[HIPTN3N]MotN), the Mo-N(4) bond lengths are 2.324(9)
Å (vs 2.395(5) Å), and the Mo-N-Cipso-Co angles are 38.3°,
32.9°, and 22.3° (vs 35.6°, 35.4°, and 33.5°).
Synthesis of Dinitrogen Complexes. Reduction of 2a in THF
with 10 equiv of magnesium powder (activated with 0.5 equiv
of 1,2-dibromoethane) under a dinitrogen atmosphere gave the
{(HTBTNCH2CH2)3N]MoN2}- species as a mixture of mag-
nesium salts,25 as suggested by a color change from red-orange
to dark green. However, the reaction is much slower than the
reduction of [HIPTN3N]MoCl, the dark-green color appearing
only after 4-5 h instead of 1 h. ZnCl2 was added immediately
to a filtered dark-green solution of {[HTBTN3N]MoN2}- to give
zinc metal and [HTBTN3N]Mo(N2) (4a) as a dark-green solid
1
in 64% isolated yield (eq 8). The H NMR spectrum of 4a
features three paramagnetically shifted resonances at 22.85 ppm
(backbone CH2), -7.11 ppm (Ho), and -32.35 ppm (backbone
CH2), and the IR spectrum in C6D6 reveals the NN stretch at
1990 cm-1, which is identical to νNN in [HIPTN3N]Mo(N2).
The 15N2 analogue, 4a-15N2, was prepared in 48% yield by a
procedure under 15N2 that was similar to that used for the
15N15
unlabeled complex. The value of ν(
N) was found to be 1924
cm-1, which is the same as in [HIPTN3N]Mo(15N2).
If the green solution containing {HTBTN3N]MoN2}- formed
upon Mg reduction of 2a under N2 is treated with Bu4NCl and
10 equiv of dioxane, then {Bu4N}{[HTBTN3N]MoN2} (5a) can
9
J. AM. CHEM. SOC. VOL. 126, NO. 19, 2004 6155

A R T I C L E S
Ritleng et al.
Table 5. Electrochemical Data of [TerN3N]Mo Derivatives and
Metallocene Reductantsa
b
E°′, V
[pBrHIPTN3N]Mo(N2)0/-
-1.71
-1.81
-1.84
-1.46
-1.51
-1.61
-1.33
-1.47
-1.84
[HIPTN3N]Mo(N2)0/-
[HTBTN3N]Mo(N2)0/-
[pBrHIPTN3N]Mo(NH3)+/0
[HIPTN3N]Mo(NH3)+/0
[HTBTN3N]Mo(NH3)+/0
+/0
CoCp2
+/0
CrCp*2 +/0
CoCp*2
a Cyclic voltammetry in 0.4 M [Bu4N][PF6] in THF. b Referenced
+/0
+/0
externally and/or internally to FeCp2 or CoCp2 at scan rates of 10-
+/0
25 mV/s; all values are listed relative to FeCp2
.
NH immediately,26 we have postulated (on the basis of IR
evidence) that a proton adds first to an amido nitrogen, the
resulting cation is then reduced and dinitrogen is protonated,
and finally the original proton is lost from the amido nitrogen.
Synthesis of Cationic Ammonia Adducts. Compounds 2a
and 2c react with 4 equiv of ammonia and 1.1 equiv of NaBAr′4
(Ar′ ) 3,5-(CF3)2C6H3)) in CH2Cl2 to yield the respective
{[TerN3N]Mo(NH3)}{BAr′4} salts in good yields (Ter ) HTBT,
6a, 91%; pBrHIPT, 6c, 81%; eq 10). The reaction to give 6a
required 24 h, and that to give 6c required 12 h. (The low overall
yields of the H3[HMTN3N] and “hybrid” ligands limited
synthesis and examination of compounds that contain those
triamidoamine ligands.) These syntheses are analogous to the
synthesis of {[HIPTN3N]Mo(NH3)}{BAr′4}.25
(14N2/15N2 ≈ 2/1 after 1 week under a 15N2 atmosphere).33 The
slower 14N2/15N2 exchange rate observed in [TMSN3N]Mo(14N2)
as compared to [HIPTN3N]Mo(15N2) (assuming that any primary
isotope is small) can be explained in terms of the stronger back-
bonding to dinitrogen in [TMSN3N]Mo(14N2) (νNN ) 1934 cm-1
in pentane; Table 3) versus [HIPTN3N]Mo(14N2) (νNN ) 1990
cm-1). Because the [HTBTN3N]3- ligand appears to be only
slightly more electron-donating than [HIPTN3N]3-, we propose
that the 20-fold decrease in the exchange rate originates
primarily from increased steric crowding in 4a on the periphery
of the binding pocket, which hinders loss of N2 after it
dissociates from the metal. These findings are relevant to
dinitrogen reduction because loss of ammonia from [TerN3N]-
Mo(NH3), and replacement of that ammonia with dinitrogen,
therefore might also be much more difficult in [HTBTN3N]-
Mo(NH3) than in [HIPTN3N]Mo(NH3) (see later).
All isolated [TerN3N]Mo(N2) derivatives undergo reversible
one-electron electrochemical reductions to the corresponding
{[TerN3N]Mo(N2)}- ions in THF. Electrochemical potentials
determined by cyclic voltammetry in THF are collected in Table
5. The peak current ratio for all couples is near unity, and the
peak separation is about 100 mV for all Mo derivatives and for
the Cp2Fe+/0 couple, as a consequence of the relatively high
solution resistance. The CV of {Bu4N}{[HIPTN3N]Mo(N2)} is
identical to that of [HIPTN3N]Mo(N2) in the reverse scan
direction. The CV of [HTBTN3N]Mo(N2) (4a) reveals that
reduction of [HTBTN3N]Mo(N2) is 30 mV more difficult than
reduction of [HIPTN3N]Mo(N2). The shift in the redox potential
is consistent with a slightly greater donating ability of the
[HTBTN3N]3- ligand versus the [HIPTN3N]3- ligand. The CV
of {Et4N}{[pBrHIPTN3N]Mo(N2)} reveals a reversible couple
for [pBrHIPTN3N]Mo(N2)0/- at -1.71 V, 100 mV more positive
than the couple for [HIPTN3N]Mo(N2)0/-. We conclude that
electrochemistry is the most sensitive and most relevant method
for determining the relative electron-donating ability of the
ligands and that the electron-donating order is what one might
expect, HTBT > HIPT > pBrHIPT. The low potentials for
reduction of the [TerN3N]Mo(N2) species make it unlikely that
{[TerN3N]Mo(N2)}- is formed first upon reduction of [TerN3N]-
Mo(N2) by CoCp2 (E°′ ) -1.33 V; Table 5) followed by
addition of a proton, although that possibility cannot yet be ruled
out rigorously. Because addition of CoCp2 and 2,6-lutidinium
to [HIPTN3N]Mo(N2) in benzene yields [HIPTN3N]Mo-Nd
Proton NMR spectra of both 6a and 6c feature patterns
characteristic of a high-spin Mo(IV) environment, with back-
bone methylene resonances paramagnetically shifted to -12
and -106 ppm in 6a, and to -16 and -110 ppm in 6c. All
cationic ammonia adducts undergo reversible one-electron
electrochemical reductions to the neutral derivatives. The
[TerN3N]Mo(NH3)+/0 potentials summarized in Table 5 suggest
an order of electron-donating ability of the [TerN3N]3- ligands
similar to that suggested by the [TerN3N]Mo(N2)0/- potentials,
but with a 100 mV difference between [HTBTN3N]Mo(NH3)+/0
and [HIPTN3N]Mo(NH3)+/0, and a 50 mV difference between
[HIPTN3N]Mo(NH3)+/0 and [pBrHIPTN3N]Mo(NH3)+/0. These
results suggest that, while CrCp*2 is able to reduce {[HIPTN3N]-
Mo(NH3)}{BAr′4} (consistent with the successful catalytic
reduction of dinitrogen by [HIPTN3N]Mo compounds with
CrCp*2 as the reductant26), CrCp*2 is not likely to reduce
{[HTBTN3N]Mo(NH3)}{BAr′4} as efficiently.
Both {[HIPTN3N]Mo(NH3)}{BAr′4} and 6a react with [Cp2-
Fe][PF6] and Et3N in benzene or CH2Cl2 to yield the respective
(33) O’Donoghue, M. B.; Davis, W. M.; Schrock, R. R. Inorg. Chem. 1998,
37, 5149.
9
6156 J. AM. CHEM. SOC. VOL. 126, NO. 19, 2004

Molybdenum Triamidoamine Complexes
A R T I C L E S
nitrides immediately and quantitatively in 1H NMR scale
reactions (eq 11). This route to nitrides allows fully 15N-labeled
Mo(NH3)}{BAr′4} and [HIPTN3N]Mo(N2) is established, with
{[HIPTN3N]Mo(NH3)}{BAr′4} being the major species present,
which suggests that, although no [HIPTN3N]Mo(NH3) can be
observed upon reduction of {[HIPTN3N]Mo(NH3)}{BAr′4} with
CoCp2, enough must be formed to allow step B. When employed
in 2-fold excess in C6D6, stronger reductants such as CrCp*2
or CoCp*2 rapidly and quantitatively reduce {[HIPTN3N]Mo-
(NH3)}{BAr′4} to [HIPTN3N]Mo(NH3). The outcome of the
CrCp*2 reduction is important in showing that precipitation of
{CrCp*2 }{BAr′4} out of C6D6 increases the reaction driving
force according to the potentials in THF (Table 5) and this effect
should be even more pronounced in the heptane used in catalytic
dinitrogen reduction studies. Once formed, [HIPTN3N]Mo(NH3)
undergoes exchange step B on the time scale of minutes to
hours. For example, an equilibrium, near-equimolar mixture of
[HIPTN3N]Mo(NH3) and [HIPTN3N]Mo(N2) is formed in under
6 h in C6D6, and full conversion to [HIPTN3N]Mo(N2) can be
effected in under 30 min when ammonia is scavenged by added
BPh3.
derivatives to be prepared. Thus, 6a-15N, formed in situ from
2a and 15NH3, afforded 3a-15N in 23% isolated (nonoptimized)
yield. Both the 15N NMR chemical shift (898.3 ppm) and νMo
15N
(991 cm-1) of 3a-15N are similar to those of [HIPTN3N]Mot
15N.25 We propose that triethylamine initially deprotonates
{[TerN3N]Mo(NH3)}+ to give [TerN3N]Mo(NH2), which is then
oxidized to yield {[TerN3N]Mo(NH2)}+. A subsequent depro-
tonation/oxidation/deprotonation sequence then yields [TerN3N]-
MotN. Of course, the reaction shown in eq 11 is the reverse
of the proposed method of converting [TerN3N]MotN into
{[TerN3N]Mo(NH3)}+ in the catalytic cycle.25,26
The HTBT derivative, {[HTBTN3N]Mo(NH3)}{BAr′4} (6a),
undergoes reduction at a potential 100 mV more negative than
that of {[HIPTN3N]Mo(NH3)}+ (Table 5). Accordingly, chemi-
cal reduction of 6a with 2 equiv of CrCp*2 in C6D6 under
dinitrogen now affords only a 40:60 mixture of 6a and
[HTBTN3N]Mo(NH3). The two are present in a fast equilibrium
on the NMR time scale, such that paramagnetically shifted
resonances of the separate species are not observed; only
weighted-average resonances are observed in the diamagnetic
region. (The locations of resonances for [HTBTN3N]Mo(NH3)
were determined by reducing 6a quantitatively with CoCp*2.)
The 40:60 equilibrium mixture of {[HTBTN3N]Mo(NH3)}+ and
[HTBTN3N]Mo(NH3) evolves further upon exchange of am-
monia with dissolved N2 to yield [HTBTN3N]Mo(N2) (4a),
although this takes place much more slowly than in the
[HIPTN3N]3- system; a 50% conversion of all Mo species
present into 4a is observed after 111 hs at 22 °C. Thus, the
more electron-donating [HTBTN3N]3- ligand hinders reduction
step A in the overall conversion shown in eq 12 relative to that
of {[HIPTN3N]Mo(NH3)}+ with the same reductant, CrCp*2.
Furthermore, if exchange step B takes place via a dissociative
mechanism, by analogy to the N2 exchange in 4a (vide supra),
it should be faster with a more electron-rich metal center, other
conditions being equal. The opposite is observed experimentally.
Therefore, it appears that the increased steric crowding of the
[HTBTN3N]3- ligand in the canopy above the binding pocket
impedes the loss of ammonia once it dissociates from Mo. In
sum, these observations demonstrate that 6a is less likely to
undergo the transformation shown in eq 12 under the conditions
used to effect dinitrogen reduction with {[HIPTN3N]Mo(NH3)}+
as the catalyst.
Conversion of {[TerN3N]Mo(NH3)}+ to [TerN3N]Mo(N2).
Two key steps in the catalytic reduction of dinitrogen by
[TerN3N]Mo derivatives of the type described here appear to
be reduction of the cationic ammonia adducts to neutral
ammonia complexes (step A; eq 12) and replacement of
ammonia by dinitrogen (step B; eq 12). Conversion of [TerN3N]-
Mo(NH3) to [TerN3N]Mo(N2) is believed to take place through
formation of intermediate “[TerN3N]Mo”, a type of species that
has not been observed in any [RN3N]Mo system so far (R )
TMS,33 C6F5,34 Aryl,32 or Terphenyl25), and that is the same
intermediate proposed in the replacement of 15N2 with 14N2 in
[TerN3N]Mo(15N2) derivatives (eq 9). Therefore, we turned to
a qualitative exploration of the conversion of {[TerN3N]Mo-
(NH3)}+ to [TerN3N]Mo(N2) derivatives in the presence of
various reducing agents.
Addition of 5 equiv of CoCp2 to {[pBrHIPTN3N]Mo(NH3)}-
{BAr′4} in C6D6 results in complete reduction of the cationic
species to [pBrHIPTN3N]Mo(NH3) in approximately 30 min,
and formation of a mixture of [pBrHIPTN3N]Mo(NH3) and
[pBrHIPTN3N]Mo(N2) over a period of 6-12 h. Therefore,
reduction of {[pBrHIPTN3N]Mo(NH3)}{BAr′4} to [pBrHIPTN3N]-
Mo(NH3) with CoCp2 is more complete than in the [HIPTN3N]3-
system, but the observed ammonia/nitrogen exchange rate
appears to be slower than in the [HIPTN3N]3- system, most
likely because the metal is slightly more electrophilic.
According to the redox potentials in THF (Table 5), cobal-
tocene is not a sufficiently strong reductant to completely reduce
{[HIPTN3N]Mo(NH3)}{BAr′4} in step A; no reaction takes
place between {[HIPTN3N]Mo(NH3)}{BAr′4} and 3 equiv of
CoCp2 in C6D6. However, after 19 h, a mixture of {[HIPTN3N]-
(34) Kol, M.; Schrock, R. R.; Kempe, R.; Davis, W. M. J. Am. Chem. Soc.
1994, 116, 4382.
9
J. AM. CHEM. SOC. VOL. 126, NO. 19, 2004 6157

A R T I C L E S
Ritleng et al.
Table 6. Dinitrogen Reduction with Various [TerN3N]Mo
significant problem is likely to be the low solubility of the initial
complex (3b) and perhaps other intermediates in the hypothetical
catalytic cycle in heptane. An attempt to increase the solubility
of 3b and/or enable its conversion into other, more soluble
derivatives by using a 0.3 mL:0.3 mL mixture of benzene and
heptane resulted in formation of even less ammonia (0.99 equiv
per Mo), with most of the 3b still not being dissolved during
the run. Therefore, other problems may be responsible for the
low yield of ammonia when 3b is employed as the catalyst.
Derivativesa
equiv of NH
yield
3
b
[TerN N]Mo
reductant
runs
per Mo
NH , %
3
3
[HTBTN3N]MotN (3a)
[HTBTN3N]Mo(N2) (4a)
[HMTN3N]MotN (3b)
3b
[pBrHIPTN3N]MotN (3c)
3c
CrCp*2
CrCp*2
CrCp*2
CrCp*2
CrCp*2
CoCp2
CoCp2
CrCp*2
1
1
1
1.06
1.27
1.49
0.99
6.4-7.0
2.6-2.9
3.6
8.8
10.6
12.4
8.3
1c
4d
3e
1
59(2)
23(1)
37
[HIPTN3N]MotN
[HIPTN3N]MoH
3
7.65(3)
65.6(2)f
In contrast, [pBrHIPTN3N]MotN is a catalyst for the
formation of ammonia in yields only slightly less than those
observed employing [HIPTN3N]3- derivatives (∼65%).26 (The
four runs were carried out by two researchers in different
ammonia “reactors.”26) While {[pBrHIPTN3N]Mo(NH3)}+ is
more easily reduced than {[HIPTN3N]Mo(NH3)}+, qualitative
studies in the previous section suggest that ammonia is lost less
readily from [pBrHIPTN3N]Mo(NH3), presumably as a conse-
quence of the more electron-withdrawing nature of the
[pBrHIPTN3N]3- ligand. Therefore, the yield of ammonia is
decreased, but only slightly.
a Unless otherwise indicated, all runs were carried out through dropwise
addition over a period of 6 h of a solution of 36 equiv of CrCp*2 in 10.0
mL of heptane at a rate of 1.7 mL/h to a mixture of 0.6 mL of heptane,
5.85-5.94 µmol of [TerN3N]Mo, and 48 equiv of [2,6-LutH][BAr′4] with
constant stirring at 1250 rpm at 22-26 °C. Stirring was continued for one
additional hour. The apparatus was sealed under 1 atm of N2; it had a
headspace volume of 68 mL. b Yield ) NH3(found, average)/NH3(theory).
For simplicity, the theoretical amount of ammonia is based on electrons
added (one NH3 per three electrons). Any ammonia that might be formed
from nitride without addition of electrons is not included. c CrCp*2 solution
was added to a mixture of [HMTN3N]MotN and [2,6-LutH][BAr′4] in a
mixture of 0.3 mL of C6H6 and 0.3 mL of heptane. d These runs employed
5.85-7.04 µmol of 3c, 31.35-36.42 equiv of CrCp*2, and 48-50 equiv
of [2,6-LutH][BAr′4]; 6.35, 6.58, 6.91, and 7.04 equiv of ammonia per Mo
were formed for an average yield of 59% (σ ) 2). e The reducing agent
was added over a period of 6 h in two experiments and 20 h in a third.
These runs employed 6.7-7.0 µmol of 3c, 36.3-38.1 equiv of CoCp2, and
44-49 equiv of [2,6-LutH][BAr′4]; 2.56 (20 h), 2.55 (6 h), and 2.89 (6 h)
equiv of ammonia per Mo were formed for an average yield of 23% (σ )
1). f 7.67, 7.67, and 7.62 equiv of ammonia out of 11.67 theoretical equiv.
We found that CoCp2 can also be employed as the reductant
using 3c as the catalyst (three runs; Table 6), although the yield
of ammonia (2.9 equiv) is less than half that when CrCp*2 is
the reductant. (Lengthening the time of addition of CoCp2 to
20 hours resulted in little change in this yield of ammonia.) A
similar run in which [HIPTN3N]MotN was employed as the
catalyst yielded 3.6 equiv of ammonia (37%; Table 6). Both
catalysts are less efficient with CoCp2 as the reductant, because
less of the neutral ammonia complex is formed with the weaker
reductant. Interestingly, [HIPTN3N]MotN is still more efficient
than [pBrHIPTN3N]MotN, which is the order found when
CrCp*2 is the reductant (∼64% with [HIPTN3N]MotN vs
∼59% with [pBrHIPTN3N]MotN). Because both CrCp*2 and
CoCp2 are suitable reducing agents (with efficiencies mirroring
their reducing abilities), it seems likely that they behave as
simple reducing agents; that is, they are not involved in any
chemically more intimate manner.
Catalytic Dinitrogen Reduction Studies. Several [TerN3N]-
Mo derivatives (Ter ) HTBT, HMT, or pBrHIPT) were
screened as catalysts for dinitrogen reduction under conditions
identical to those employed for the HIPT derivatives.26 The
results are summarized in Table 6. Reactions that involved 3a
and 4a yielded only 1.06 and 1.27 equiv of ammonia per Mo,
respectively. The first suggests that the nitride (D, Scheme 1)
is reduced to the cationic ammonia complex (F), but there is
little reduction of F and formation of A. Visually, it can be
seen that {[HTBTN3N]Mo(NH3)}+ (6a) is not reduced readily;
that is, the reaction effectively stops at 6a, 1 equiv of ammonia
being recovered upon workup. This result is also consistent with
the electrochemical results (Table 5), which show that
{[HTBTN3N]Mo(NH3)}+ is reduced in THF at a potential 100
mV more negative than {[HIPTN3N]Mo(NH3)}+. Also, as
outlined in the previous section, it is almost certainly true that
ammonia cannot escape as readily from [HTBTN3N]Mo(NH3)
as it does from [HIPTN3N]Mo(NH3) and the conversion of G
to A therefore is slow.35 This combination of problems will
clearly limit conversion of F to A. The second result suggests
that there may be additional problems associated with conversion
of A to D, although the nature of these problems is unknown at
this stage. The primary problem may simply be a slowing of
all steps as a consequence of overall crowding in the
[HTBTN3N]3- systems. In any case, it is startling how dramatic
are the consequences of changing from a HIPT-based to a
HTBT-based ligand.
In all catalytic dinitrogen reduction studies, we presume that
formation of dihydrogen is the alternative to formation of
ammonia, either at the metal center (see below), or through the
direct reduction of protons by the relatively strong reducing
agents employed here. The less efficient formation of ammonia
when CoCp2 is the reducing agent would lead to formation of
more dihydrogen, because the competing reduction of protons
with CoCp2 is still relatively fast. It is desirable to confirm that
dihydrogen is formed and quantitate it in all of the experiments
listed in Table 6 (and analogous experiments reported else-
where26). However, under the present experimental conditions,
as little as 1-2 mL (at 1 atm) of dihydrogen would be formed.
We hope to be able to quantitate such small amounts in future
studies.
It seems possible that [HIPTN3N]MoH, a known compound,25
might be formed under catalytic conditions. Therefore, we tested
whether [HIPTN3N]MoH is an efficient precursor for catalytic
formation of ammonia. As shown in Table 6, it is as efficient
as any catalyst for dinitrogen reduction of this general type.26
This fact suggests that [HIPTN3N]MoH can be protonated and
reduced to yield dihydrogen and [HIPTN3N]Mo(N2) in the
presence of dinitrogen. We have no way of knowing how much
dihydrogen is formed at the metal in this manner (i.e., with the
Derivative 3b is also a poor catalyst for the reduction of
dinitrogen, giving only 1.49 equiv of NH3 under our standard
conditions; that is, the reaction is barely catalytic. Here, a
(35) Measuring the rate of exchange of ammonia for dinitrogen in the Mo(III)
species is not a trivial experiment, although we hope to devise a means of
doing so eventually.
9
6158 J. AM. CHEM. SOC. VOL. 126, NO. 19, 2004

Molybdenum Triamidoamine Complexes
A R T I C L E S
prepared according to published methods. All Mo complexes were
stored under N2 at -35 °C.
metal behaving as a hydrogenase) and how much is formed by
direct electron transfer to protons.
Catalytic dinitrogen reduction runs were performed using reagents,
equipment, and procedures described previously.26
Conclusions
NMR spectra were recorded at 298 K on Varian Mercury 300 or
Varian Unity 300 spectrometers operating at 300 MHz for H and at
It is clear that relatively subtle steric and electronic variations
of the [TerN3N]3- ligand system produce relatively profound
changes in the efficacy of the catalytic reduction of dinitrogen
to ammonia, at least for the three variations that we have
examined so far under our “standard” conditions. The steps that
are most sensitive to change in the [TerN3N]3- ligand system
appear to be reduction of {[TerN3N]Mo(NH3)}+ to [TerN3N]-
Mo(NH3), the equilibrium involving loss of ammonia from
[TerN3N]Mo(NH3) to give unobservable [TerN3N]Mo, and
binding of dinitrogen to [TerN3N]Mo to give [TerN3N]Mo(N2)
(eq 12). The slightly more electron-withdrawing [pBrHIPTN3N]3-
ligand allows a weaker reducing agent to be employed, but the
tradeoff is that ammonia is lost less easily from the Mo(III)
center. Insolubility of catalytic intermediates is a problem that
plagues the [HMTN3N]3- system, while too much steric
congestion plagues the [HTBTN3N]3- system, probably through
reducing the rate of virtually all reactions. In a system in which
all intermediates are in solution, it is likely that several could
themselves behave as acids or reducing agents toward others;
whether this happens rapidly under the “standard” reduction
conditions, or is an advantage or a disadvantage, is not known.
Because the buildup of gaseous ammonia would retard loss of
ammonia from [TerN3N]Mo(NH3), it may prove beneficial to
remove any gaseous ammonia efficiently during the reduction
to push the overall efficiency to higher levels. As we strive
toward the goal of using weaker acids and weaker reductants,
and to achieving higher turnover numbers, we expect to
encounter other complex issues in a reaction that may involve
a dozen or more discrete intermediates.
1
75 MHz for 13C {1H}. Chemical shifts (δ) are referenced to the residual
deuterated solvent peaks and are reported in ppm. Coupling constants
(J) are in hertz. IR spectra were recorded on a Nicolet Avatar 360 FT-
IR spectrometer in a demountable solution cell (0.2 mm Teflon spacer,
KBr windows). Elemental analyses were performed by H. Kolbe
Mikroanalytisches Laboratorium, Mu¨lheim an der Ruhr, Germany.
Electrochemical measurements were carried out with BAS CV-50W
potentiostat. Platinum disk (1.6 mm dia), platinum wire, and a silver
wire submerged in a 10 mM solution of AgOTf in 0.4 M [Bu4N][PF6]
in THF were used as working, auxiliary, and reference electrode,
respectively. All measurements were done in 0.4 M [Bu4N][PF6] in
THF and referenced externally and/or internally to Cp2Fe or CoCp2.
3,5-Bis(2,4,6-tri-tert-butylphenyl)bromobenzene (HTBTBr). A
three-neck round-bottom flask set with a N2-inlet, a condenser, and an
addition funnel was charged with Mg turnings (6.6 g, 271 mmol) and
THF (60 mL). Previously degassed 2,4,6-tri-tert-butylbromobenzene
(44.4 g, 136 mmol) was dissolved in THF (215 mL) in a Schlenk flask,
and the solution was added to the addition funnel with a cannula. The
round-bottom flask was heated with a heat gun until THF started to
reflux, and ∼5 mL of the 2,4,6-tri-tert-butylbromobenzene solution was
added to initiate the Grignard reaction. After initiation, the remainder
of the bromide solution was added over a period of 1 h. When
approximately half the solution had been added, the reaction flask was
placed in an oil bath at 85 °C, and the reaction was refluxed overnight.
THF (175 mL) was added to the addition funnel and added to the
refluxing reaction mixture. 2,4,6-Tribromoiodobenzene (17.1 g, 38.9
mmol) was dissolved in THF (70 mL) in a Schlenk flask and added to
the addition funnel with a cannula. The solution was added dropwise
over a period of 45 min, and the reaction mixture was refluxed for 4
h. The resulting orange reaction mixture was cooled to room temper-
ature. The mixture was then decanted from excess Mg and quenched
by pouring it into cold HCl (60 mL, 10%). The aqueous phase was
diluted with 100 mL of water, and some Na2SO3 was added. When
THF was removed on the rotary evaporator, a large amount of solid
formed in the flask. The aqueous phase was extracted with diethyl ether
(3 × 100 mL). The combined organic layers were washed with water
(2 × 100 mL), dried over MgSO4, and concentrated until a white solid
(8.29 g) precipitated. This solid was filtered off and washed with a
small portion of ether. Further concentration yielded 0.85 g more. The
remaining mother liquors were reduced to a thick orange oil, and 1,3,5-
tri-tert-butylbenzene was removed by vacuum distillation. (This is a
tedious process because it crystallizes in the column.) Column
chromatography on silica (25 × 7 cm) with hexanes as eluent yielded
an additional 1.19 g; total yield 10.33 g (16 mmol, 41%): 1H NMR
(CDCl3, 300 MHz) δ 7.53 (s, 4H, C6H2-t-Bu3), 7.52 (t, 1H, C6H3Ar2-
Br, 4J ) 1.5), 7.48 (d 2H, C6H3Ar2Br, 4J ) 1.5), 1.37 (s, 18H, p-t-Bu),
2.07 (s, 36H, o-t-Bu); 13C {1H} NMR (CDCl3, 75 MHz) δ 148.5, 148.4,
144.0, 137.4, 136.3, 133.9, 123.0, and 118.3 (CAr), 38.6 (o-C(CH3)3),
35.2 (p-C(CH3)3), 34.7 (o-C(CH3)3), 31.7 (p-C(CH3)3). Anal. Calcd for
C42H61Br: C, 78.21; H, 9.54; Br, 12.25. Found: C, 78.26; H, 9.61; Br,
12.22. HRMS calcd [M]+: 644.3957. Found (EI): 644.3965.
Experimental Section
General. All manipulations of air- and moisture-sensitive compounds
were carried out with standard Schlenk and glovebox techniques under
an atmosphere of nitrogen using oven-dried glassware. THF and
benzene were purged with dinitrogen and were passed through columns
of activated alumina and supported copper catalyst (Q5; benzene only).36
Toluene and pentane were purged with nitrogen and passed through a
column of activated alumina. For reactions involving molybdenum
compounds, pentane was additionally degassed (freeze-pump-thaw)
three times, and THF, benzene, and toluene were vacuum-transferred
from dark purple Na/benzophenone ketyl solutions into storage Schlenk
flasks and degassed three times using a freeze-pump-thaw method.
C6D6 was dried over Na/benzophenone, vacuum-transferred into a
storage Schlenk flask, and freeze-pump-thaw degassed three times.
CDCl3 was used as received. All dried and deoxygenated solvents were
stored over molecular sieves (4 Å) in a nitrogen-filled glovebox.
1,3,5-Tri-tert-butylbenzene (Fluka), 2,4,6-tribromoaniline, 2-bro-
momesitylene, 4-tert-butylbromobenzene, (Me3Si)2NLi (sublimed),
N(CH2CH2NH2)3, NaO-t-Bu, anhydrous ZnCl2, Mg powder (-50 mesh;
washed with HCl (3.7%), water, THF, diethyl ether, and dried under
vacuum; Aldrich), Pd2(dba)3, rac-BINAP, Mg turnings, CoCp2 (sub-
limed), CrCp*2 (sublimed), MoCl5 (Strem), Me3SiN3 (Acros), and 15N2
(CIL) were used as received unless otherwise indicated. 2,4,6-Tri-tert-
butylbromobenzene,37 2,4,6-tribromoiodobenzene,27 [2,6-lutidinium]-
[BAr′4],25 bromo-2,4,6 triisopropylbenzene,38 and MoCl4(THF)239 were
3,5-Bis(2,4,6-trimethylphenyl)bromobenzene (HMTBr). The Grig-
nard reagent was synthesized in a manner similar to that used to prepare
1a starting from Mg turnings (4.8 g, 200 mmol) in 40 mL of THF and
2-bromomesitylene (19.91 g, 100 mmol) in 160 mL of THF. The
reaction mixture was then refluxed for 6 h. THF (130 mL) was added
to the addition funnel and then to the refluxing reaction mixture. 2,4,6-
(36) Pangborn, A. B.; Giardello, M. A.; Grubbs, R. H.; Rosen, R. K.; Timmers,
F. J. Organometallics 1996, 15, 1518.
(37) Yoshifujii, M. Synth. Organomet. Inorg. Chem. 1996, 3, 120.
(38) Miller, A. R.; Curtin, D. Y. J. Am. Chem. Soc. 1976, 98, 1860.
(39) Stoffelbach, F.; Saurenz, D.; Poli, R. Eur. J. Inorg. Chem. 2001, 2699.
9
J. AM. CHEM. SOC. VOL. 126, NO. 19, 2004 6159

A R T I C L E S
Ritleng et al.
Tribromoiodobenzene (12.6 g, 28.6 mmol) was dissolved in THF (50
mL) in a Schlenk flask and added to the addition funnel with a cannula.
The solution was added dropwise over a period of 45 min, and the
reaction mixture was refluxed overnight. The resulting orange reaction
mixture was cooled to room temperature, decanted from excess Mg,
and quenched by addition of cold HCl (40 mL, 10%). The aqueous
phase was diluted with 60 mL of water, and some Na2SO3 was added.
THF was removed on the rotary evaporator, and the aqueous phase
was extracted with diethyl ether (3 × 100 mL). The combined organic
layers were washed with water (2 × 100 mL), dried over MgSO4, and
concentrated to a thick yellow oil. Mesitylene was removed by vacuum
distillation. Dissolution of the resulting brownish mushy solid in
minimum hexane led to precipitation of some pure product (white
powder, 1.01 g). Column chromatography on silica (26 × 7 cm) with
hexane as the eluent afforded an additional 1.88 g; total yield 2.89 g
(7.35 mmol, 26%): 1H NMR (CDCl3, 300 MHz) δ 7.28 (d, 2H, C6H3-
recrystallized by dissolving it in CH3CN followed by addition of diethyl
ether to afford a white product (1a, 8.21 g, 4.46 mmol, 65%). Elution
with 8/1/1 to 7/2/1 mixtures of hexane/Et2O/NEt3 produced a yellow
fraction that gave an orange solid after solvent removal (1a′, 2.17 g,
1.70 mmol, 25%).
1H NMR of 1a (CDCl3, 300 MHz) δ 7.51 (s, 12H, C6H2-t-Bu3),
7.00 (t, 3H, C6H3Ar2), 6.50 (d, 6H, C6H3Ar2, 4J ) 1.2), 3.67 (br s, 3H,
3
3
NH), 3.08 (br t, 6H, CH2, J ) 5.7), 2.68 (br t, 6H, CH2, J ) 5.7),
1.37 (s, 54H, p-t-Bu), 1.25 (s, 108H, o-t-Bu); 13C {1H} NMR (CDCl3,
75 MHz) δ 148.6, 147.8, 143.4, 142.8, 138.5, 126.1, 122.9, and 120.7
(CAryl), 54.1 (CH2), 42.2 (CH2), 38.6 (o-C(CH3)3), 35.0 (p-C(CH3)3),
34.1 (o-C(CH3)3), 31.6 (p-C(CH3)3). Anal. Calcd for C132H198N4: C,
86.12; H, 10.84; N, 3.04. Found: C, 85.97; H, 10.76; N, 2.95. HRMS
calcd [M + H]+: 1840.5689. Found (ESI): 1840.5655.
1H NMR of 1a′ (CDCl3, 300 MHz) δ 7.51 (s, 8H, C6H2-t-Bu3), 6.99
(t, 2H, C6H3Ar2), 6.52 (d, 4H, C6H3Ar2, 4J ) 1.2), 3.64 (br t, 2H, CH2),
3.33 (br t, 2H, CH2), 3.10 (br t, 4H, CH2), 2.69 (bt, 4H, CH2), 1.36 (s,
36H, p-t-Bu), 1.25 (s, 72H, o-t-Bu). (The NH resonances could not be
located.) 13C {1H} NMR (CDCl3, 75 MHz) δ 148.4, 147.6, 144.1, 142.6,
138.5, 125.9, 122.8, and 120.6 (CAryl), 59.5 (C2H4NH2), 54.6 (C2H4-
NHTer), 46.4 (C2H4NH2), 42.6 (C2H4NHTer), 38.7 (o-C(CH3)3), 35.1
(p-C(CH3)3), 34.8 (o-C(CH3)3), 31.7 (p-C(CH3)3). Mass calcd for
C90H138N4 [M + H]+: 1276.10. Found (ESI): 1276.12.
{3,5-(2,4,6-Me3C6H2)C6H3NHCH2CH2}3N (1b). This compound
was synthesized in a manner similar to that used to prepare 1a and
from HMTBr (5.30 g, 13.5 mmol), toluene (50 mL), N(CH2CH2NH2)3
(657 mg, 4.49 mmol), and sodium tert-butoxide (1.82 g, 18.9 mmol)
on one hand, and from rac-BINAP (756 mg, 1.21 mmol), toluene (100
mL), and Pd2(dba)3 (371 mg, 0.405 mmol) on the other. The resulting
red-orange suspension was cooled to room temperature after 24 h of
reflux at 110 °C, and NaBr was removed by flitration. Toluene was
removed on the rotary evaporator, and the resulting oil was purified
by filtration on silica (CH2Cl2/hexanes varying from 2/1 to 9/1). A pale
yellow solid was obtained after removing the solvent in vacuo.
Recrystallization from CH2Cl2/hexanes afforded a white solid (1.71 g,
1.58 mmol, 35%): 1H NMR (CDCl3, 300 MHz) δ 6.88 (s, 12H, C6H2-
Me3), 6.32 (d, 6H, C6H3Mes2, 4J ) 1.2), 6.26 (t, 3H, C6H3Mes2), 4.00
(br s, 3H, NH), 3.21 (br t, 6H, CH2, 3J ) 6.0), 2.83 (br t, 6H, CH2, 3J
) 6.0), 2.33 (s, 18H, p-CH3), 2.01 (s, 36H, o-CH3); 13C {1H} NMR
(CDCl3, 75 MHz) δ 148.0, 142.3, 139.4, 136.2, 135.9, 128.0, 120.1,
and 112.1 (CAr), 53.0 (CH2), 41.5 (CH2), 21.4 (p-CH3), 21.0 (o-CH3).
Anal. Calcd for C78H90N4: C, 86.46; H, 8.37; N, 5.17. Found: C, 86.54;
H, 8.42; N, 5.08. HRMS calcd [M + H]+: 1083.7238. Found (ESI):
1083.7216.
4
Mes2Br, J ) 1.5), 6.94 (s, 4H, C6H2Me3), 6.88 (t, 1H, C6H3Mes2Br,
4J ) 1.5), 2.34 (s, 6H, p-CH3), 2.07 (s, 12H, o-CH3); 13C {1H} NMR
(CDCl3, 75 MHz) δ 143.4, 137.7, 137.2, 135.9, 130.7, 129.5, 128.4,
and 122.6 (CAr), 21.2 (p-CH3), 20.9 (o-CH3). Anal. Calcd for C24H25-
Br: C, 73.28; H, 6.41. Found: C, 73.20; H, 6.49. HRMS calcd [M]+:
392.1134. Found (EI): 392.1131.
4-Bromo-3,5-bis(2,4,6-triisopropylphenyl)bromobenzene (pBr-
HIPTBr). The procedure was nearly identical to that published for
HIPTBr. Briefly, magnesium turnings (2.0 g, 82 mmol) and 1,2-
dibromoethane (0.5 mL, in 75 mL of THF) were added to 200 mL of
THF to initiate Grignard formation. After initiation, bromo-2,4,6-
triisopropylbenzene (10.6 g, 37 mmol) in 75 mL of THF was added
dropwise. This mixture was refluxed for 90 min, during which time
the solution became brown and cloudy. Iodo-2,4,6-tribromobenzene
(5.0 g, 11 mmol) dissolved in 75 mL of THF was then added dropwise,
and the reaction was refluxed for another 3.5 h, at which time the
solution was cooled to 0 °C. This Grignard solution was slowly added
to an ice cold slurry of N-bromosuccinimide (NBS) (12.1 g, 68 mmol)
in 400 mL of THF, and the mixture was allowed to stir for 10 h. A
saturated aqueous solution of sodium nitrite was then added. The
mixture was stirred for 2 h and then extracted with ether. The combined
organic phases were washed with water and reduced in volume in vacuo.
Methanol was then added to the resulting orange-colored slurry, and
an off-white solid was filtered off. This solid was further purified by
recrystallization from ether; yield 3.47 g (5.4 mmol, 49% yield): 1H
NMR (CDCl3, 20 °C) δ 7.37 (s, 2H, 4′6′-H), 7.08 (s, 4H, 3,5,3′′,5′′-
H), 2.98 (septet, JHH ) 6.9 Hz, 2H, 4,4′′-CHMe2), 2.56 (septet, JHH
)
6.9 Hz, 4H, 2,6,2′′,6′′-CHMe2), 1.33 (d, JHH ) 6.9 Hz, 12H, 4,4′′-CH-
(CH3)2), 1.19 (d, JHH ) 6.9 Hz, 12H, 2,6,2′′,6′′-CH(CH3)2), 1.17 (d,
JHH ) 6.9 Hz, 12H, 2,6,2′′,6′′-CH(CH3)2); 13C NMR (CDCl3, 20 °C) δ
148.85, 145.82, 144.32, 135.29, 132.33, 127.69, 120.95, 120.61, 34.56,
31.26, 24.96, 24.36, 23.87.
{4-Br-3,5-(2,4,6-Me3C6H2)C6H3NHCH2CH2}3N (1c). A mixture of
pBrHIPTBr (6.92 g, 11 mmol), NaO-t-Bu (1.378 g, 14 mmol), and
triethylamidoamine (0.54 g, 3.5 mmol) in 100 mL of toluene was
prepared in the glovebox. A catalyst prepared from Pd2(dba)3 (0.082
g, 0.09 mmol) and rac-BINAP (0.166 g, 0.26 mmol) in 100 mL of
toluene was filtered through Celite and added to the pBrHIPTBr
solution. The flask was sealed and stirred at 90 °C for 48 h. The solids
were removed by filtration through Celite, and the solution was filtered
through a flash chromotography column (SiO2). Volatiles were removed
from the filtrate under vacuum, and the resulting solid was purified
through chromatography (pentane/toluene) on a silica column. The
resulting off-white foamy solid was extensively dried at 65 °C under
vacuum to yield 3.96 g (2.1 mmol, 61% yield): 1H NMR (C6D6, 20
{3,5-(2,4,6-t-Bu3C6H2)C6H3NHCH2CH2}3N (1a) and {3,5-(2,4,6-
t-Bu3C6H2)C6H3NHCH2CH2}2NCH2CH2NH2 (1a′). A 500 mL Schlenk
flask was charged in the glovebox with HTBTBr (13.25 g, 20.5 mmol),
toluene (60 mL), N(CH2CH2NH2)3 (1.0 g, 6.84 mmol), and sodium
tert-butoxide (2.76 g, 28.7 mmol). The Pd catalyst solution was prepared
by vigorously stirring rac-BINAP (1.15 g, 1.85 mmol) and Pd2(dba)3
(563 mg, 0.615 mmol) in toluene (140 mL) for 16 h at room
temperature. It was filtered and added to the main reaction mixture. A
condenser was then added, and the whole assembly was taken out of
the glovebox and set up on the Schlenk line under N2. The reaction
was then refluxed at 110 °C. The reaction mixture turned dark red within
a few minutes and became clear orange again after a couple of hours
as NaBr precipitated. After 48 h, the reaction mixture was cooled to
room temperature, and NaBr was removed by flitration. Toluene was
removed on the rotary evaporator, and the resulting dark oil was purified
by filtration through a bed of silica on a 300 mL frit. Elution with
hexane/Et2O (9/1) yielded a very pale yellow solution that left a
yellowish solid behind after all solvent was removed in vacuo. It was
°C) δ 7.26 (s, 12H, 3,5,3′′,5′′-H), 6.44 (s, 6H, 4′,6′-H), 3.52 (t, JHH
)
5.0 Hz, 3H, NH), 3.05 (septet, JHH ) 6.9 Hz, 2H, 4,4′′-CHMe2), 2.90
(septet, JHH ) 6.9 Hz, 4H, 2,6,2′′,6′′-CHMe2), 2.70 (br m, 6H, NHCH2-
CH2), 2.16 (approximately t, JHH ) 5.2 Hz, 6H, NHCH2CH2), 1.47 (d,
JHH ) 6.9, 12H, 4,4′′-CH(CH3)2), 1.29 (m, 72H, 2,6,2′6′-CH(CH3)2);
13C NMR (C6D6, 20 °C) δ 149.2, 147.0, 146.6, 144.0, 138.0, 121.5,
116.3, 114.7, 53.0, 41.7, 35.4, 31.9, 25.8, 25.0, 24.7. MS (ESI):
1822.0231 ([M + H]+ calcd 1822.0188). Anal. Calcd for C114H159
-
9
6160 J. AM. CHEM. SOC. VOL. 126, NO. 19, 2004

Molybdenum Triamidoamine Complexes
A R T I C L E S
Br3N4: C, 75.02; H, 8.78; N, 3.07; Br, 13.13. Found: C, 75.11; H,
8.87; N, 3.02; Br, 12.98.
solvent was removed, and the resulting dark red residue was extracted
with pentane (100 mL). The mixture was filtered through Celite and
concentrated to 1/3 of its initial volume, which induced formation of
a precipitate of the product. The suspension was allowed to stand at
-35 °C for 1 h. The solid was filtered off and washed with cold pentane
to give a dark red solid (585 mg, 0.483 mmol, 64%): 1H NMR (C6D6,
300 MHz) δ 9.72 (br s, 6H, C6H3Mes2), 6.96 (s, 12H, C6H2Me3), 2.40
(s, 18H, p-CH3), 2.21 (br s, 36H, o-CH3), 1.96 (s, 3H, C6H3Mes2),
-16.00 (br s, 6H, NCH2), -77.60 (br s, 6H, NCH2). Anal. Calcd for
C78H87N4MoCl: C, 77.30; H, 7.24; N, 4.62; Cl, 2.93. Found: C, 77.18;
H, 7.31; N, 4.56; Cl, 3.05.
[3,5-(2,4,6-t-Bu3C6H2)2C6H3NHCH2CH2]2(4-t-BuC6H4NHCH2-
CH2)N (1d). This compound was synthesized in a manner similar to
that used to prepare 1a starting from 1a′ (5.56 g, 4.35 mmol), toluene
(14 mL), p-tert-butylbromobenzene (1.02 g, 4.79 mmol), and sodium
tert-butoxide (585 mg, 6.09 mmol), on one hand, and a preformed
catalyst solution prepared from rac-BINAP (263 mg, 0.423 mmol) and
Pd2(dba)3 (129 mg, 0.141 mmol) in toluene (32 mL). The resulting
dark brown suspension was cooled to room temperature after 48 h of
reflux at 110 °C, and NaBr was removed by filtration over Celite.
Toluene was removed on the rotary evaporator, and the resulting oil
was purified by filtration on a 300 mL frit filled with silica. Elution
with a hexane/Et2O/NEt3 (8/1/1) mixture led to a pale yellow solid
that was dried under high vacuum at 70 °C after solvent removal (750
mg, 0.533 mmol, 12%): 1H NMR (CDCl3, 300 MHz) δ 7.52 (s, 8H,
C6H2-t-Bu3), 7.13 (d, 2H, C6H4-t-Bu, 3J ) 8.7), 7.00 (t, 2H, C6H3Ar2),
[pBrHIPTN3N]MoCl (2c). H3[pBrHIPTN3N] (2.5 g, 0.14 mmol)
and MoCl4(THF)2 (0.524 g, 0.14 mmol) were dissolved in THF (200
mL). This mixture was stirred for 1 h, and (Me3Si)2NLi (0.711 g, 0.42
mmol) was added slowly. The mixture was stirred for 2 h, and the
solvent was removed in vacuo. The solid residue was extracted with
pentane (2 × 10 mL) followed by benzene (3 × 20 mL), and all was
filtered through Celite. The filtrate was then reduced to dryness in
vacuo. Crystallization from pentane yielded 1.5 g (55%) of a red-orange
solid in multiple crops: 1H NMR (C6D6, 20 °C) δ 13.5 (br s), 7.34 (s),
3.11 (br s), 2.995 (br m), 1.6 (s), 1.388 (d), -18 (br s), -84 (br s).
Anal. Calcd for C114H156Br3ClMoN4: C, 70.09; H, 8.05; N, 2.87; Br,
12.27; Cl, 1.81. Found: C, 69.85; H, 7.79; N, 2.99; Br, 12.70; Cl, 1.87.
[(HTBTNCH2CH2)2(4-t-BuC6H4NCH2CH2)N]MoCl (2d). A 250
mL round-bottom flask was charged with 1d (714 mg, 0.507 mmol)
and C6H6 (50 mL). Solid MoCl4(THF)2 (194 mg, 0.507 mmol) was
added over 20 min. The resulting very dark red solution was stirred
for 2 h at room temperature. Solid (Me3Si)2NLi (297 mg, 1.77 mmol)
was then added over 10 min, and the resulting reaction mixture was
stirred at room temperaure for 4 h. The solvents were removed in vacuo,
and the solid residue was extracted with benzene (75 mL). The solution
was filtered through Celite, and the dark red filtrate was evaporated to
dryness in vacuo. The resulting dark red residue was suspended in
pentane (50 mL), and the mixture was stirred vigorously for 20 min.
The suspension was then concentrated to 10 mL, and the product was
filtered off and washed with cold pentane (3 × 5 mL); yield 307 mg
(0.20 mmol, 39%): 1H NMR (C6D6, 300 MHz) δ 15.49 (s, 2H, C6H4-
t-Bu), 13.37 (br s, 4H, C6H3Ar2), 9.48 (br s, 2H, C6H4-t-Bu), 7.88 and
7.85 (2s, 2 × 4H, C6H2-t-Bu3), 3.60 (s, 2H, C6H3Ar2), 1.73 and 1.65
(2 br s, 2 × 36H, o-t-Bu), 1.54 (s, 9H, C6H4-t-Bu), 1.47 (s, 36H, p-t-
Bu), -8.03, -14.00, and -16.80 (3 br s, 3 × 2H, NCH2), - 75 to -
80 (3 br s, 3 × 2H, NCH2). Anal. Calcd for C100H147N4MoCl: C, 78.16;
H, 9.64; N, 3.64; Cl, 2.31. Found: C, 78.01; H, 9.65; N, 3.54; Cl,
2.30.
4
3
6.52 (d, 4H, C6H3Ar2, J ) 1.5), 6.49 (d, 2H, C6H4-t-Bu, J ) 8.7),
3.77 (m, 3H, NH), 3.14 (br t, 4H, CH2, 3J ) 5.7), 3.10 (br t, 3H, CH2,
3J ) 5.7), 2.74 (br t, 4H, CH2), 2.72 (br t, 2H, CH2), 1.37 (s, 36H,
p-t-Bu), 1.28 (s, 9H, t-Bu), 1.26 (s, 72H, o-t-Bu); 13C {1H} NMR
(CDCl3, 75 MHz) δ 148.4, 147.7 (CTer), 145.7 (C6H4-t-Bu), 143.9, 142.7
(CTer), 140.2 (C6H4-t-Bu), 138.4, 126.1, 122.8, 120.7 (CTer), 112.8 (C6H4-
t-Bu), 54.4 (C2H4NHTer), 54.4 (C2H4NHAr), 42.5 (C2H4NHTer and
C2H4NHAr), 38.7 (o-C(CH3)3), 35.1 (p-C(CH3)3), 34.8 (o-C(CH3)3), 34.1
(C(CH3)3), 31.9 (C(CH3)3), 31.7 (p-C(CH3)3). Anal. Calcd for
C100H150N4: C, 86.12; H, 10.84; N, 3.04. Found: C, 85.97; H, 10.76;
N, 2.95. HRMS calcd [M + H]+: 1408.1933. Found (ESI): 1408.1927.
[HTBTN3N]MoCl (2a). (A) A 250 mL round-bottom flask was
charged with 1a (1.84 g, 0.995 mmol) and C6H6 (100 mL). Solid MoCl4-
(THF)2 (380 mg, 0.995 mmol) was added over 50 min. The resulting
dark red solution was stirred for 2 h at room temperature. Solid (Me3-
Si)2NLi (583 mg, 3.48 mmol) was then added over a period of 20 min,
and the reaction mixture was stirred for an additional 4 h at 22 °C.
Concentration of the latter to 1/4 of its initial volume and filtration
through Celite gave a dark red filtrate that was evaporated to dryness.
The resulting dark red residue was thoroughly dried in vacuo and
suspended in pentane (60 mL). The mixture was stirred vigorously for
15 min. Solvent was removed from the suspension until only 15 mL
remained, and the suspension was allowed to stand at -35 °C for 1 h.
Filtration and washing with cold pentane (3 × 10 mL) gave an orange
solid (1.17 g, 0.594 mmol, 60%): 1H NMR (C6D6, 300 MHz) δ 14.76
(br s, 6H, C6H3Ar2), 7.90 (s, 12H, C6H2-t-Bu3), 4.43 (s, 3H, C6H3Ar2),
1.62 (br s, 108H, o-t-Bu), 1.48 (s, 54H, p-t-Bu), -11.92 (br s, 6H,
NCH2), -82.46 (br s, 6H, NCH2). Anal. Calcd for C132H195N4MoCl:
C, 80.50; H, 9.98; N, 2.84; Cl, 1.80. Found: C, 80.39; H, 9.93; N,
2.73; Cl, 1.84.
[HTBTN3N]MotN (3a). A toluene (10 mL) solution of 2a (0.600
g, 0.305 mmol) and Me3SiN3 (162 µL, 1.22 mmol) was heated at 110
°C for 4 days, after which the resulting dark brown-yellow solution
was taken to dryness and exposed to high vacuum at 90 °C for 0.5 h.
The solid was extracted with pentane (70 mL), and the mixture was
filtered through Celite to give a brown-yellow solution that was
concentrated to yield a yellow solid. The solid was collected by filtration
in three crops, washed with cold pentane, and dried in vacuo; yield
355 mg (0.182 mmol, 60%): 1H NMR (C6D6, 300 MHz) δ 7.76 (s,
12H, C6H2-t-Bu3), 7.50 (d, 6H, C6H3Ar2, 4J ) 1.2), 7.30 (t, 3H, C6H3-
Ar2, 4J ) 1.2), 3.35 (t, 6H, NCH2), 2.00 (t, 6H, NCH2), 1.46 (s, 108H,
o-t-Bu), 1.45 (s, 54H, p-t-Bu); IR (C6D6) 1015 cm-1 (νMoN). Anal. Calcd
for C132H195N5Mo: C, 81.39; H, 10.09; N, 3.60. Found: C, 81.45; H,
10.14; N, 3.56.
(B) A mixture of MoCl4(THF)2 (747 mg, 1.96 mmol), 1a (2.955 g,
1.61 mmol), and C6H6 (30 mL) was stirred vigorously for 2 h to yield
a homogeneous, dark red solution. Solid (Me3Si)2NLi (833 mg, 4.98
mmol) was then added in portions over 15 min. The reaction mixture
was stirred for 1 h and brought to dryness in vacuo. The solid residue
was dried in vacuo for 8 h at room temperature. The resulting solid
was triturated with 20 mL of pentane overnight, and the mixture was
filtered. The solid was dissolved in 30 mL of benzene, and the mixture
was filtered through Celite. The combined filtrates were taken to
dryness. The resulting solid was dried in vacuo for 6 h at room
temperature and triturated with 20 mL of pentane; two crops were
collected after the mixture was left standing at -25 °C overnight; yield
2.178 g (1.11 mmol, 69%).
[HMTN3N]MotN (3b). A toluene (10 mL) solution of 2b (0.200
g, 0.165 mmol) and Me3SiN3 (44 µL, 0.330 mmol) was heated at 90
°C for 2 days, after which the resulting dark brown-yellow solution
was taken to dryness and the residue was dried at 90 °C under high
vacuum. Extraction with benzene (50 mL) and filtration through Celite
gave a brown-yellow solution that was concentrated to ∼5 mL until a
yellow solid started to precipitate. Pentane (7 mL) was added, and,
after standing at room temperature for 2 h, the precipitate was collected
[HMTN3N]MoCl (2b). A 250 mL round-bottom flask was charged
with 1b (813 mg, 0.750 mmol) and C6H6 (75 mL). Solid MoCl4(THF)2
(286 mg, 0.750 mmol) was added over a period of 30 min. The resulting
dark red solution was stirred for 2 h at room temperature. Solid (Me3-
Si)2NLi (439 mg, 2.62 mmol) was then added over 15 min, and the
reaction mixture was stirred for 4 h more at room temperature. The
9
J. AM. CHEM. SOC. VOL. 126, NO. 19, 2004 6161

A R T I C L E S
Ritleng et al.
on a frit, washed with cold pentane, and dried in vacuo; yield 122 mg
filled glovebox. The tube was taken out of the box, and the solution
was freeze-pump-thaw degassed five times, exposed to 15N2 (760
mmHg, ∼6.4 equiv relative to 2a), and vigorously stirred for 2 days.
The Schlenk tube was then pumped back in the 14N2-filled glovebox,
and the reaction mixture was filtered through Celite and treated with
ZnCl2 (21 mg, 0.152 mmol). This resulted in a rapid color change to
brown. The mixture was stirred for 30 min at room temperature, the
solvents were removed in vacuo, and the residue was dried at 60 °C
under high vacuum. The resulting gray residue was worked up as
described for the unlabeled 4a to afford a dark-green solid; yield 240
mg (0.122 mmol, 48%): 1H NMR (C6D6, 300 MHz) δ 22.82 (br s,
6H, NCH2), 7.02 (s, 12H, C6H2-t-Bu3), 1.98 (s, 3H, C6H3Ar2), 1.07 (s,
54H, p-t-Bu), 0.58 (br s, 108H, o-t-Bu), -7.12 (br s, 6H, C6H3Ar2),
(0.103 mmol, 62%): 1H NMR (C6D6, 300 MHz) δ 7.39 (d, 6H, C6H3-
4
4
Ar2, J ) 1.5), 6.93 (s, 12H, C6H2Me3), 6.48 (t, 3H, C6H3Ar2, J )
1.5), 3.40 (t, 6H, NCH2), 2.27 (s, 18H, p-Me), 2.14 (s, 36H, o-Me),
2.09 (t, 6H, NCH2). Anal. Calcd for C78H87N5Mo: C, 78.69; H, 7.37;
N, 5.88. Found: C, 78.75; H, 7.31; N, 5.81.
[HTBTN3N]Mot15N (3a-15N). 15N-labeled ammonia (60 mL, 200
Torr, 0.6 mmol) was vacuum-transferred from a bronze-colored liquid
phase containing Na into a mixture of 2a (300 mg, 0.152 mmol) and
NaBAr′4 (148.5 mg, 0.168 mmol) in CH2Cl2 (4 mL). The mixture was
stirred for 2 h, pressurized to 1 atm with N2, and stirred for additional
22 h. Treatment with Et3N (74 mL, 0.532 mmol), followed by solid
[Cp2Fe][PF6] (111 mg, 0.334 mmol), resulted in an instantaneous color
change to brown-red. The resulting solution was stirred for an hour
and brought to dryness in vacuo. The solid residue was dried for 1 h
at 90 °C and extracted with a total of 50 mL of pentane. The pentane
extracts were filtered through Celite, and the filtrate was concentrated
to 5 mL to afford a bright yellow solid after standing at -20 °C for 2
days, which was collected by filtration, washed with cold pentane, and
dried in vacuo; yield 69 mg (0.035 mmol, 23%, not optimized): 15N
-32.39 (br s, 6H, NCH2); IR (C6D6) cm-1 1924 (ν
N).
15
15
N-
14N2/15N2 Exchange Studies. A C6D6 (3 mL) solution of 4a-15N2
(0.025 g, 0.0127 mmol) was stirred at room temperature under 1 atm
of 14N2 in a closed 10 mL Schlenk tube that was kept in a glovebox.
Aliquots were removed regularly and immediately analyzed by IR
spectrometry. The observed 4a-15N2/4a-14N2 ratios were as follows:
94.1/5.9 (6.5 h); 93.2/6.8 (22 h); 90.8/9.2 (47 h); 88.5/11.5 (71 h); 86.9/
13.1 (99 h); 84.0/16.0 (149 h); 79.5/20.5 (196 h); 76.0/24.0 (243 h);
72.6/27.4 (312 h). The total was constant during this period. A first-
order plot of the disappearance of 4a-15N2 gave a value of k ) 2.4 ×
10-7 s-1 (R ) 0.998).
NMR (C6D6, 20 °C) δ 898.3 (Mot15N); IR (C6D6) 991 cm-1 (νMo N).
15
Anal. Calcd for C132H195N415NMo: C, 81.35; H, 10.08; N, 3.64.
Found: C, 81.42; H, 9.95; N, 3.52.
[pBrHIPT]MotN (3c). [pBrHIPTN3N]MoCl (0.25 g, 0.013 mmol)
and TMS azide (0.1 g, 0.087 mmol) were added to benzene (40 mL)
in a Teflon sealed flask, which was heated at 90 °C for 3 days. All
volatiles were then removed in vacuo, and the solid was recrystallized
from pentane to yield 0.167 g (67%) of bright yellow powder in multiple
crops: 1H NMR (C6D6, 20 °C) δ 7.90 (br s, 6H, 4′,6′-H), 7.25 (s, 12H,
3,5,3′′,5′′-H), 3.53 (br t, JHH ) 5 Hz, 6H, ArNCH2CH2), 3.00
(overlapping septets, JHH ) 6.9 Hz, 18H, 4,4′′-CHMe2 and 2,6,2′′,6′′-
CHMe2), 1.90 (br t, JHH ) 5 Hz, 6H, NHCH2CH2), 1.47 (d, JHH ) 6.9
Hz, 36H, 4,4′′-CH(CH3)2), 1.34 (d, JHH ) 6.9 Hz 36H, 2,6,2′6′-CH-
(CH3)2), 1.08 (d, JHH ) 6.6 Hz, 36H, 2,6,2′6′-CH(CH3)2); IR (C6D6)
1013 cm-1 (νMoN). Anal. Calcd for C114H156Br3MoN5: C, 70.87; H,
8.14; N, 3.62; Br, 12.41. Found: C, 70.73; H, 8.20; N, 3.69; Br, 12.37.
[pBrHIPT]Mot15N (50%) (3c-15N (50%)). [pBrHIPTN3N]MoCl
(0.311 g, 0.017 mmol), 15N labeled sodium azide (0.0443 g, 0.068
mmol), and TMEDA (0.1 g) were placed in 75 mL of benzene and
heated at 100-110 °C for 6 days. The reaction was then stripped to
dryness in vacuo, and the bright yellow compound was recrystallized
from pentane; yield 0.048 mg (14%); IR (C6D6) 1013 cm-1 (νMoN),
In a dinitrogen-filled glovebox, a Parr bomb was charged with a
C6D6 (3 mL) solution of 4a-15N2 (0.025 g, 0.0127 mmol). It was then
closed, taken out of the box, and pressurized to 5 atm of 14N2. Every
24 h, it was pumped back into the glovebox, depressurized, opened,
and an aliquot was removed for IR analysis. The observed 4a-15N2/
4a-14N2 ratios were as follows: 90.0/10.0 (24 h); 87.3/12.7 (48 h); 85.0/
15.0 (70.5 h); 83.3/16.7 (93 h). The total was constant during this period.
A first-order plot of the disappearance of 4a-15N2 gave a value of k )
3.1 × 10-7 s-1 (R ) 0.997).
{Bu4N}{[HTBTN3N]Mo(N2)} (5a). A mixture of 2a (1.0 g, 0.508
mmol) and Mg powder (123.4 mg, 5.08 mmol, activated with 0.25
mmol of 1,2-dibromoethane) was stirred with a glass-coated stirbar
under 1 atm of N2 for 8 h and treated with Bu4NCl (155.2 mg, 0.558
mmol) and 1,4-dioxane (433 µL, 5.08 mmol). Stirring was continued
for additonal 14 h, and the mixture was brought to dryness in vacuo.
The solid residue was dried at 65 °C and extracted with a total of 10
mL of C6H6. Benzene extracts were filtered through Celite, concentrated
to 7 mL, and treated with 40 mL of pentane to yield dark emerald-
green crystals after standing at room temperature for 2 h, which were
collected by filtration, washed with pentane, and dried in vacuo; yield
0.661 g (0.300 mmol, 59%): 1H NMR (C6D6) δ 7.75 (s, 12H, 3,5,3′′,5′′-
H), 7.66 (br s, 6H, 4′,6′-H), 6.93 (br s, 3H, 2′-H), 3.68 (br s, 6H, NCH2-
CH2), 2.33 (br s, 8H, NCH2+), 1.91 (br s, 6H, NCH2CH2), 1.65 (br s,
108H, 2,6,2′′,6′′-C(CH3)3), 1.47 (s, 54H, 4,4′′-C(CH3)3), 1.04 (ap-
proximately quintet, JHH ) 6.9 Hz, 8H, NCH2CH2CH2CH3+), 0.91 (br
sextet, 8H, NCH2CH2CH2CH3+), 0.79 (t, JHH ) 7.2 Hz, 12H, NCH2-
CH2CH2CH3+); IR (C6D6) δ 1847 (ν(NN)). Anal. Calcd for C148H231N7-
Mo: C, 80.64; H, 10.56; N, 4.45. Found: C, 80.49; H, 10.47; N, 4.28.
Observation of a Mixture of [pBrHIPTN3N]Mo-NdN-Mg-
(THF)xCl and [pBrHIPTN3N]Mo(N2). The procedure followed was
similar to the procedure published for the [HIPTN3N]3- analogue.25
Briefly, [pBrHIPTN3N]MoCl (0.16 g) and magnesium powder (0.14
g) were added to ∼30 mL of THF and stirred with a glass stirbar. The
magnesium was then activated with dibromoethane, and the reaction
allowed to proceed for 2 h, during which time the reaction turned from
orange to emerald green. This was stripped to dryness in vacuo, and
the resulting solid was taken up in pentane and filtered through Celite.
This deep red/purple pentane solution was then stripped to dryness,
yielding a reddish brown solid that contains both the magnesium salt
and the neutral dinitrogen species. Attempts to synthesize the neutral
species alone have so far proven unsuccessful. Important paramagneti-
cally shifted NMR peaks (that correspond to the neutral dinitrogen
985 cm-1 (νMo N).
15
[HTBTN3N]Mo(N2) (4a). Magnesium powder (62 mg, 2.54 mmol)
was suspended in THF (2 mL) in a 100 mL round-bottom flask
equipped with a glass-coated stirbar, and the magnesium was activated
with 1,2-dibromoethane (11 µL, 0.127 mmol) for 5 min. Addition of
solid 2a (0.5 g, 0.254 mmol) followed by 6 mL of THF gave a red-
orange suspension that turned dark green over 4-5 h. After 6 h, the
reaction mixture was filtered through Celite and treated with ZnCl2
(21 mg, 0.152 mmol). This resulted in an instantaneous color change
to brown. The mixture was stirred for 45 min at room temperature,
and all solvents were removed at 60 °C under high vacuum. The gray
residue was extracted with pentane, and the extract filtered through
Celite. The resulting dark-green solution was concentrated to yield a
dark-green solid after standing at -35 °C for 30 min. The latter was
collected on a frit, washed with cold pentane (5 × 1 mL), and dried
under vacuum; yield 320 mg (0.164 mmol, 64%): 1H NMR (C6D6,
300 MHz) δ 22.85 (br s, 6H, NCH2), 7.02 (s, 12H, C6H2-t-Bu3), 1.97
(s, 3H, C6H3Ar2), 1.06 (s, 54H, p-t-Bu), 0.57 (br s, 108H, o-t-Bu), -7.11
(br s, 6H, C6H3Ar2), -32.35 (br s, 6H, NCH2). IR (C6D6) 1990 cm-1
(νNN). Anal. Calcd for C132H195N6Mo: C, 80.81; H, 10.02; N, 4.28.
Found: C, 80.92; H, 9.93; N, 4.14.
[HTBTN3N]Mo(15N2) (4a-15N2). A 50 mL Schlenk tube equipped
with a glass-coated stirbar was charged with magnesium powder (62
mg, 2.54 mmol), 2a (0.5 g, 0.254 mmol), and THF (10 mL) in a 14N2-
9
6162 J. AM. CHEM. SOC. VOL. 126, NO. 19, 2004

Molybdenum Triamidoamine Complexes
A R T I C L E S
species) are located at 22.8, -6.9, and -33.50 ppm. Peaks in the
diamagnetic region have yet to be assigned due to the presence of
multiple species. The IR spectrum shows both species distinctly, with
the νNN peak at 1992 cm-1 for [pBrHIPTN3N]Mo(N2) and the νNN peak
at 1788 cm-1 for the anion with magnesium bound to the â nitrogen
atom.
{NEt4}{[pBrHIPTN3N]Mo(N2)}. A sample of [pBrHIPTN3N]MoCl
(0.4 g, 0.25 mmol) and magnesium powder (0.5 g, 20 mmol) were
added to ∼30 mL of THF, and a glass coated stirbar was added. A
drop of dibromoethane was added to activate the magnesium, and the
reaction was allowed to proceed for 2 h, during which time the color
turned from orange to emerald green. Tetraethylammonium chloride
(0.035 g, 0.27 mmol) and 0.25 mL of dioxane were then added, and
the solution was allowed to stir for 48 h, at which point all volatiles
were removed in vacuo. The resulting solid was triturated with pentane
and dissolved in benzene. The solution was filtered through Celite,
and the solution was concentrated to yield bright green {NEt4}-
{[pBrHIPTN3N]Mo(N2)}; yield 150 mg, 25%: IR (C6D6) 1857 cm-1
(νNN). Anal. Calcd for C122H176Br3MoN7: C, 70.57; H, 8.54; N, 4.72.
Found: C, 70.43; H, 8.48; N, 4.65.
H), 7.64 (br s, 4H, C6H3-3,5-(CF3)2), 1.41 (s, 54H, 4,4′′-C(CH3)3), 1.15-
1.50 (v br, 108H, 2,6,2′′,6′′-C(CH3)3), -0.21 (s, 3H, 2′-H), -12 (br s,
6H, NCH2), -106 (br s, 6H, NCH2). Anal. Calcd for C164H210BF24N5-
Mo: C, 69.99; H, 7.52; N, 2.49. Found: C, 70.16; H, 7.44; N, 2.37.
{[pBrHIPT]Mo(NH3)}{BAr′4} (6c). A mixture of [pBrHIPT]MoCl
(0.5 g, 0.25 mmol), NaBAr′4 (0.4 g, 0.45 mmol), and NH3 (100 mL,
440 Torr, transferred from a bronze solution prepared with Na, ∼8
equiv relative to pBrMoCl) was allowed to stir in 75 mL of degassed
dichloromethane for 12 h in a Teflon-sealed glass bomb. The reaction
turned from deep orange to bright red, and the resulting solution was
taken to dryness in vacuo. The solid was then dissolved in pentane,
and the solution was filtered through Celite and reduced to ∼1.5 mL
in vacuo. This concentrated solution was cooled to -30 °C for 6 days
to yield bright red crystals of the product; yield 0.58 g (81%): 1H NMR
(C6D6, 20 °C) δ 8.27 (s, BAr4′-H), 7.65 (s, BAr4′-H), 2.91 (br m), 2.71
(br m), 1.31 (br m), 1.09 (br m), -16.3 (br s), -109.5 (br s). Anal.
Calcd for C146H171BBr3F24MoN5: C, 62.66; H, 6.16; N, 2.50. Found:
C, 62.54; H, 6.22; N, 2.43.
Experimental Procedure for the X-ray Diffraction Analysis of
2a and 3c. Reflections were collected on a Bruker Smart diffractometer
equipped with Kappa CCD area detector using Mo KR graphite-
monochromated radiation (λ ) 0.71073 Å). The structure was solved
and refined using Bruker SHELXTL 5.1.
This procedure was also used to make the corresponding parent
{NEt4}{[HIPTN3N]Mo(N2)} compound for comparison of IR spectra
with the known tetrabutylammonium complex. It was found that the
substitution of tetraethylammonium for tetrabutylammonium had no
Acknowledgment. R.R.S. is grateful to the National Institutes
of Health (GM 31978) for research support.
effect on the C6D6 νNN stretch located at 1855 cm-1
.
{[HTBTN3N]Mo(NH3)}{BAr′4} (6a). Ammonia (60 mL, 200 Torr,
0.6 mmol) was vacuum-transferred from a bronze-colored solution of
Na to a mixture of 2a (300 mg, 0.152 mmol) and NaBAr′4 (148.5 mg,
0.168 mmol) in CH2Cl2 (4 mL). The mixture was stirred for 2 h,
pressurized to 1 atm with N2, stirred for additional 22 h, and brought
to dryness in vacuo. The solid residue was dried at 50 °C and extracted
with a total of 10 mL of pentane. The pentane extracts were filtered
through Celite and brought to dryness to afford a dark red-brown solid;
yield 0.391 g (0.139 mmol, 91%): 1H NMR (C6D6, 20 °C) δ 9.5 (br,
6H, 4′,6′-H), 8.29 (br s, 8H, C6H3-3,5-(CF3)2), 7.75 (s, 12H, 3,5,3′′,5′′-
Supporting Information Available: Crystal data and structure
refinement, atomic coordinates and equivalent isotropic dis-
placement parameters, bond lengths and angles, anisotropic
displacement parameters, hydrogen coordinates, and isotropic
displacement parameters for [HTBTN3N]MoCl and [pBrHIPT]-
MotN (PDF and CIF). This material is available free of charge
via the Internet at http://pubs.acs.org.
JA0306415
9
J. AM. CHEM. SOC. VOL. 126, NO. 19, 2004 6163