Small-molecule activation by molybdaziridine hydride complexes: Mechanistic sequence of the small-molecule binding and molybdaziridine ring-opening steps

Article abstract of DOI:10.1021/om040045z

The relationship between molybdaziridine hydride Mo(H)(Me ₂C=NAr)(N[i-Pr]Ar)₂ (1, Ar = 3,5-C₆H ₃Me₂) and its three-coordinate isomer Mo(N[i-Pr]Ar) ₃ (2) has been probed via an investigation of the coordination chemistry of 1 and Mo(N[t-Bu]Ar)₃ (3) with 1-adamantyl-isocyanide (AdNC). One or two equivalents of AdNC react with 1 to form the adducts 2-AdNC and 2-(AdNC)₂, respectively. One equivalent of AdNC coordinates to 3, forming 3-AdNC. Similarly, tert-butylisocyanide (t-BuNC) reacts with 1 to form 2-t-BuNC and 2-(t-BuNC)₂ and with 3 to form 3-t-BuNC. An X-ray crystal structure of 2-(AdNC)₂ reveals a trigonal bipyramidal core with a trans disposition of the isocyanide ligands (C-Mo-C, 172.8(4)°; Mo-C, 2.135(11) and 2.083(11) A). The structure of 3-t-BuNC features a bent isocyanide ligand with a C-N-C angle of 137.8(7)°, and the compound has a solution IR stretch of 1759 cm^-1, revealing Mo-C multiple-bond character. The fast reactions of AdNC with 1, 1-d₃, and 3 were studied by stopped-flow spectrophotometry in a wide temperature range (-80 to 25°C). A comparison of the reaction rates and activation parameters indicates that 2 is not an intermediate on the pathway from 1 to 2-AdNC, but rather that the molybdaziridine hydride opens upon AdNC binding in an associative process. The enthalpies of reaction to generate the compounds of interest were measured using solution calorimetry: 2-AdNC, -24.6 kcal·mol^-1. 3-AdNC, -29.1 kcal-mol^-1. The enthalpy of binding of the second equivalent of AdNC to generate 2-(AdNC)2 is -10.2 kcal·mol^-1.

Full text of DOI:10.1021/om040045z

3126
Organometallics 2004, 23, 3126-3138
Sm a ll-Molecu le Activa tion by Molybd a zir id in e Hyd r id e
Com p lexes: Mech a n istic Sequ en ce of th e Sm a ll-Molecu le
Bin d in g a n d Molybd a zir id in e Rin g-Op en in g Step s
Frances H. Stephens, J oshua S. Figueroa, and Christopher C. Cummins*
Department of Chemistry, Massachusetts Institute of Technology, Room 2-227,
77 Massachusetts Avenue, Cambridge, Massachusetts 02139
Olga P. Kryatova, Sergey V. Kryatov, and Elena V. Rybak-Akimova*
Department of Chemistry, Tufts University, 62 Talbot Avenue, Medford, Massachusetts 02155
J . Eric McDonough and Carl D. Hoff*
Department of Chemistry, University of Miami, 1301 Memorial Drive,
Coral Gables, Florida 33146
Received March 23, 2004
The relationship between molybdaziridine hydride Mo(H)(Me2CdNAr)(N[i-Pr]Ar)2 (1, Ar
) 3,5-C6H3Me2) and its three-coordinate isomer Mo(N[i-Pr]Ar)3 (2) has been probed via an
investigation of the coordination chemistry of 1 and Mo(N[t-Bu]Ar)3 (3) with 1-adamantyl-
isocyanide (AdNC). One or two equivalents of AdNC react with 1 to form the adducts 2-AdNC
and 2-(AdNC)2, respectively. One equivalent of AdNC coordinates to 3, forming 3-AdNC.
Similarly, tert-butylisocyanide (t-BuNC) reacts with 1 to form 2-t-BuNC and 2-(t-BuNC)2
and with 3 to form 3-t-BuNC. An X-ray crystal structure of 2-(AdNC)2 reveals a trigonal
bipyramidal core with a trans disposition of the isocyanide ligands (C-Mo-C, 172.8(4)°;
Mo-C, 2.135(11) and 2.083(11) Å). The structure of 3-t-BuNC features a bent isocyanide
ligand with a C-N-C angle of 137.8(7)°, and the compound has a solution IR stretch of
1759 cm-1, revealing Mo-C multiple-bond character. The fast reactions of AdNC with 1,
1-d3, and 3 were studied by stopped-flow spectrophotometry in a wide temperature range
(-80 to 25 °C). A comparison of the reaction rates and activation parameters indicates that
2 is not an intermediate on the pathway from 1 to 2-AdNC, but rather that the
molybdaziridine hydride opens upon AdNC binding in an associative process. The enthal-
pies of reaction to generate the compounds of interest were measured using solution
calorimetry: 2-AdNC, -24.6 kcal‚mol-1; 3-AdNC, -29.1 kcal‚mol-1. The enthalpy of binding
of the second equivalent of AdNC to generate 2-(AdNC)2 is -10.2 kcal‚mol-1
.
In tr od u ction
Dinitrogen cleavage reactions,5-7 when coupled with N
atom transfer reactivity8-14 of the product nitrido
complexes,15-17 may yield valuable methods for the
Elemental nitrogen in the form of N2 gas is an
abundant resource underutilized in synthesis. In addi-
tion to its conversion to ammonia,1,2 there is interest in
incorporation of N2-derived nitrogen atoms into organic
molecules.3,4 To this end, we have been developing
reactive metal complexes capable of dinitrogen scission
according to eq 1.5
(3) (a) Mori, M. J . Heterocycl. Chem. 2000, 37, 623. (b) Hori, M.;
Mori, M. J . Org. Chem. 1995, 60, 1480. (c) Mori, M.; Kawaguchi, M.;
Hori, M.; Hamaoka, S. Heterocycles 1994, 39, 729.
(4) For reviews of dinitrogen-derived organonitrogen compounds, see
the following reviews and references found therein: (a) Fryzuk, M. D.
Chem. Rec. 2003, 3, 2. (b) Fryzuk, M. D.; J ohnson, S. A. Coord. Chem.
Rev. 2000, 200-202, 379. (c) Hidai, M.; Yasushi, M. Top. Organomet.
Chem. 1999, 3, 227. (d) Hidai, M.; Mizobe, Y. Chem. Rev. 1995, 95,
1115. (e) Mizobe, Y.; Ishii, Y.; Hidai, M. Coord. Chem. Rev. 1995, 139,
281. (f) George, T. A.; DeBord, J . R. D. ACS Symp. Ser. 1993, 535,
363.
2 L3Mo + NtN f 2 L3MotN
(1)
(5) (a) Laplaza, C. E.; Cummins, C. C. Science 1995, 268, 861. (b)
Laplaza, C. E.; Odom, A. L.; Davis, W. M.; Cummins, C. C. J . Am.
Chem. Soc. 1995, 117, 4999. (c) Laplaza, C. E.; J ohnson, 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. (d) Tsai, Y.-C.;
Cummins, C. C. Inorg. Chim. Acta 2003, 345, 63.
(6) (a) Mindiola, D. J .; Meyer, K.; Cherry, J .-P. F.; Baker, T. A.;
Cummins, C. C. Organometallics 2000, 19, 1622. (b) Caselli, A.; Solari,
E.; Scopelliti, R.; Floriani, C.; Re, N.; Rizzoli, C.; Chiesi-Villa, A. J .
Am. Chem. Soc. 2000, 122, 3652.
(7) (a) Clentsmith, G. K. B.; Bates, V. M. E.; Hitchcock, P. B.; Cloke,
F. G. N. J . Am. Chem. Soc. 1999, 121, 10444. (b) Fryzuk, M. D.; Kozak,
C. M.; Bowdridge, M. R.; Patrick, B. O.; Rettig, S. J . J . Am. Chem.
Soc. 2002, 124, 8389.
(1) (a) Yandulov, D. V.; Schrock, R. R. J . Am. Chem. Soc. 2002, 124,
6252. (b) Yandulov, D. V.; Schrock, R. R.; Rheingold, A. L.; Ceccarelli,
C.; Davis, W. M. Inorg. Chem. 2003, 42, 796. (c) Yandulov, D. V.;
Schrock, R. R. Science 2003, 301, 76. (d) Schrock, R. R.; Kolodziej, R.
M.; Liu, A. H.; Davis, W. M.; Vale, M. G. J . Am. Chem. Soc. 1990, 112,
4338.
(2) (a) Nishibayashi, Y.; Iwai, S.; Hidai, M. Science 1998, 279, 540.
(b) Nishibayashi, Y.; Takemoto, S.; Iwai, S.; Hidai, M. Inorg. Chem.
2000, 39, 5946. (c) Ishino, H.; Nagano, T.; Kuwata, S.; Yokobayashi,
Y.; Ishii, Y.; Hidai, M.; Mizobe, Y. Organometallics 2001, 20, 188. (d)
Nishibayashi, Y.; Wakiji, I.; Hirata, K.; Du Bois, M. R.; Hidai, M. Inorg.
Chem. 2001, 40, 578. (e) Hidai, M. Coord. Chem. Rev. 1999, 185-186,
99.
10.1021/om040045z CCC: $27.50 © 2004 American Chemical Society
Publication on Web 05/14/2004

Molybdaziridine Hydride Complexes
Organometallics, Vol. 23, No. 13, 2004 3127
synthesis of organonitrogen compounds without requir-
ing the intermediacy of ammonia.
a hydrogen on the coordinated Cp* ligand is enough to
significantly change the bonding and reactivity picture.
Seeking a sterically less crowded system still capable
of dinitrogen scission, we undertook the synthesis of an
isopropyl-substituted variant of 3, Mo(N[i-Pr]Ar)3 (2).
We found that this system exists not as a three-
coordinate molybdenum(III) complex but as the tauto-
meric molybdaziridine hydride, Mo(H)(Me2CdNAr)(N[i-
Pr]Ar)2 (1).22 This process is correctly described as either
cyclometalation or â-H transfer since the anilide ligand
N(i-Pr)Ar provides a hydrogen â to the metal. The
resultant imine ligand Me2CdNAr is bound η2, forming
a three-membered molybdaziridine ring.23 The molyb-
daziridine hydride complex has a single unpaired elec-
tron, in contrast to compound 3, and can be thought of
as either Mo(III) or Mo(V) at the two extremes of
oxidative addition of the η2-imine ligand. Despite ap-
parent deactivation due to cyclometalation, the molyb-
daziridine hydride complex is active for dinitrogen
scission in addition to a host of related small-molecule
activation processes.22,24,25 The molybdaziridine hydride
functional group effectively masks an open coordination
site required for small-molecule activation via a revers-
ible cyclometalation process. Small molecule activa-
tion by the analogous niobaziridine hydride Nb(H)(t-
Bu[H]CdNAr)(N[Np]Ar)2 (Np ) neopentyl) has also
been demonstrated recently.26,27
These and other cyclometalated systems influence
both the steric and electronic environment that incom-
ing small molecules encounter. There are exceedingly
few systems where rates of reaction can be compared
for both cyclometalated and noncyclometalated systems
and where the crystal structures of both complexes are
known. Furthermore, we have recently shown that
addition of coordinating bases can serve to accelerate
N2 uptake and cleavage for 1 and 3.5d While NMR
spectroscopy did not always permit observation of bases
bound to 1 or 3, an accelerating effect was demonstrated
for both. This highlights the importance of understand-
ing the mechanism of binding and release of ligands in
these complexes. This work reports synthetic, kinetic,
thermodynamic, and theoretical investigations aimed at
probing differences in rate and energy for binding of
isocyanide ligands to 1 and 3. Fundamental to this quest
is a proper description of the sequence of the small-
molecule binding and molybdaziridine ring-opening
steps in order to determine whether 2 must be gener-
ated from 1 as a prerequisite for ligand binding.
The search for metal complexes capable of effecting
the complete six-electron reductive cleavage of N2 is
under way on a variety of fronts.4b,5,8 Three-coordinate
molybdenum(III) complexes such as Mo(N[t-Bu]Ar)3 (3,
Ar ) 3,5-C6H3Me2)5 have provided a paradigm for N2
cleavage (eq 1) in which an open coordination site is
available for N2 binding prior to a bimetallic cleav-
age process.5c The product nitrido complex NtMo(N[t-
Bu]Ar)3 (3-N) is known from structural studies to be
extremely crowded in the vicinity of the molybdenum-
nitrogen triple bond,18 a circumstance potentially un-
favorable with respect to subsequent N atom transfer
processes.
The role of steric and electronic factors in determining
reaction chemistry relevant to nitrogen fixation is still
evolving. Recent work by Chirik and co-workers has
demonstrated different coordination modes and reactiv-
ity when (C5Me4R)2ZrCl2 (R ) H or Me) is treated under
nitrogen with sodium amalgam and then with hydro-
gen.19,20 In the case of R ) H, a µ2-η2,η2-N2 complex is
formed which can be reduced at 85 °C with H2 to
generate a zirconium dihydride and ammonia. In con-
trast, when R ) Me, reduction under N2 of the zirconium
dichloride results in Bercaw’s dinitrogen complex21
which does not generate NH3 when treated with H2.
Apparently replacement of a single methyl group with
(8) (a) Laplaza, C. E.; Davis, W. M.; Cummins, C. C. Angew. Chem.,
Int. Ed. Engl. 1995, 34, 2042. (b) Laplaza, C. E.; J ohnson, A. R.;
Cummins, C. C. J . Am. Chem. Soc. 1996, 118, 709. (c) J ohnson, M. J .
A.; Lee, P. M.; Odom, A. L.; Davis, W. M.; Cummins, C. C. Angew.
Chem., Int. Ed. Engl. 1997, 36, 87. (d) Agapie, T.; Odom, A. L.;
Cummins, C. C. Inorg. Chem. 2000, 39, 174.
(9) (a) Woo, L. K.; Goll, J . G. J . Am. Chem. Soc. 1989, 111, 3755. (b)
Woo, L. K.; Goll, J . G.; Czapla, D. J .; Hays, J . A. J . Am. Chem. Soc.
1991, 113, 8478.
(10) (a) Bottomley, L. A.; Neely, F. L. J . Am. Chem. Soc. 1989, 111,
5955. (b) Neely, F. L.; Bottomley, L. A. Inorg. Chim. Acta 1992, 192,
147. (c) Neely, F. L.; Bottomley, L. A. Inorg. Chem. 1997, 36, 5432. (d)
Bottomley, L. A.; Neely, F. L. Inorg. Chem. 1997, 36, 5435.
(11) Chang, C. J .; Low, D. W.; Gray, H. B. Inorg. Chem. 1997, 36,
270.
(12) (a) Demadis, K. D.; Bakir, M.; Klesczewski, B. G.; Williams, D.
S.; White, P. S.; Meyer, T. J . Inorg. Chim. Acta 1998, 270, 511.
(b) Wong, T.-W.; Lau, T.-C.; Wong, W.-T. Inorg. Chem. 1999, 38,
6181.
(13) McCarthy, M. R.; Crevier, T. J .; Bennett, B.; Dehestani, A.;
Meyer, J . M. J . Am. Chem. Soc. 2000, 122, 12391.
(14) Clough, C. R.; Greco, J . B.; Figueroa, J . S.; Cummins, C. C. J .
Am. Chem. Soc. 2004, in press.
(15) (a) Tomooka, C. S.; Carreira, E. M. Helv. Chim. Acta 2002, 85,
3773. (b) Carreira, E. M.; Hong, J .; Du Bois, J .; Tomooka, C. S. Pure
Appl. Chem. 1998, 70, 1097. (c) Du Bois, J .; Tomooka, C. S.; Hong, J .;
Carreira, E. M. Acc. Chem. Res. 1997, 30, 364. (d) Du Bois, J .; Tomooka,
C. S.; Hong, J .; Carreira, E. M. Angew. Chem., Int. Ed. Engl. 1997,
36, 1645. (e) Du Bois, J .; Hong, J .; Carreira, E. M.; Day, M. W. J . Am.
Chem. Soc. 1996, 118, 915.
(16) (a) Nishimura, M.; Minakata, S.; Takahashi, T.; Oderaotoshi,
Y.; Komatsu, M. J . Org. Chem. 2002, 67, 2101. (b) Minakata, S.;
Nishimura, M.; Takahashi, T.; Oderaotoshi, Y.; Komatsu, M. Tetra-
hedron Lett. 2001, 42, 9019. (c) Nishimura, M.; Minakata, S.;
Thongchant, S.; Ryu, I.; Komatsu, M. Tetrahedron Lett. 2000, 41, 7089.
(d) Ho, C.-M.; Lau, T.-C.; Kwong, H.-L.; Wong, W.-T. J . Chem. Soc.,
Dalton Trans. 1999, 2411. (e) Svenstrup, N.; Bogevig, A.; Hazell, R.
G.; J orgensen, K. A. J . Chem. Soc., Perkin Trans. 1 1999, 1559.
(17) Henderickx, H.; Kwakkenbos, G.; Peters, A.; van der Spoel, J .;
de Vries, K. Chem. Commun. 2003, 2050.
(22) Tsai, Y.-C.; J ohnson, M. J . A.; Mindiola, D. J .; Cummins, C.
C.; Klooster, W. T.; Koetzle, T. F. J . Am. Chem. Soc. 1999, 121, 10426.
(23) For other examples of â-H elimination from organoamide
ligands, see the following: (a) Mayer, J . M.; Curtis, C. J .; Bercaw, J .
E. J . Am. Chem. Soc. 1983, 105, 2651. (b) Bryndza, H. E.; Tam, W.
Chem. Rev. 1988, 88, 1163. (c) Fryzuk, M. D.; Montgomery, C. D. Coord.
Chem. Rev. 1989, 95, 1. (d) Berno, P.; Gambarotta, S. Organometallics
1995, 14, 2159. (e) Driver, M. S.; Hartwig, J . F. J . Am. Chem. Soc.
1996, 118, 4206. (f) Hartwig, J . F. J . Am. Chem. Soc. 1996, 118, 7010.
(g) Bazinet, P.; Yap, G. P. A.; Richeson, D. S. Organometallics 2001,
20, 4129. (h) Cai, H.; Chen, T.; Wang, X.; Schultz, A. J .; Koetzle, T. F.;
Xue, Z. Chem. Commun. 2002, 230.
(24) Cherry, J .-P. F.; Stephens, F. H.; J ohnson, M. J . A.; Diaconescu,
P. L.; Cummins, C. C. Inorg. Chem. 2001, 40, 6860.
(18) The X-ray crystal structure of NMo(N[t-Bu]Ph)3 was deter-
mined. See ref 5.
(19) Pool, J . A.; Lobkovsky, E.; Chirik, P. J . Nature 2004, 427, 527.
(20) (a) Pool, J . A.; Lobkovsky, E.; Chirik, P. J . Organometallics
2003, 22, 2797. (b) Pool, J . A.; Lobkovsky, E.; Chirik, P. J . J . Am. Chem.
Soc. 2003, 125, 2241.
(25) (a) Tsai, Y.-C.; Diaconescu, P. L.; Cummins, C. C. Organo-
metallics 2000, 19, 5260. (b) Blackwell, J . B.; Figueroa, J . S.; Stephens,
F. H.; Cummins, C. C. Organometallics 2003, 22, 3351.
(26) (a) Figueroa, J . S.; Cummins, C. C. J . Am. Chem. Soc. 2003,
125, 4020. (b) Mindiola, D. J .; Cummins, C. C. Organometallics 2001,
20, 3626.
(21) Manriquez, J . M.; Bercaw, J . E. J . Am. Chem. Soc. 1974, 96,
6229.
(27) Figueroa, J . S.; Cummins, C. C. Angew. Chem., Int. Ed. 2004,
43, 984.

3128 Organometallics, Vol. 23, No. 13, 2004
Stephens et al.
F igu r e 1. Reaction of 3-d18 with n equivalents of AdNC
(n ) 0, 1, 10; bottom to top) as monitored by 2H NMR. The
peak at δ ) 9.9 ppm (∆ν1/2 ) 12.8 Hz) is assigned to
3-AdNC. No indication of 3-(AdNC)2 is spectroscopically
detectable.
F igu r e 2. X-ray crystal structure of 3-t-BuNC. Thermal
ellipsoids are at the 50% probability level. Hydrogen atoms
have been omitted for clarity. Selected bond lengths and
angles are listed in Table 1.
Ta ble 1. Selected Bon d Len gth s a n d An gles for
3-t-Bu NC (Ca lcu la ted Bon d Len gth s a n d An gles for
(η1-HNC)Mo(NHMe)(NH2)2 in P a r en th eses)
Resu lts a n d Discu ssion
Rea ction of Mo(N[t-Bu ]Ar )3 (3) w ith Bu lk y Iso-
cya n id es. To encourage η1 binding of the isocyanide to
the molybdenum-containing complexes detailed here-
in,28,29 1-adamantylisocyanide (AdNC) was selected due
to its steric bulk. Reaction of Mo(N[t-Bu]Ar)3 (3) with
one or more equivalents of AdNC resulted in rapid
formation of a 1:1 adduct, 3-AdNC. Compound 3-AdNC
is a brown, crystalline solid with a solution magnetic
moment µeff ) 1.83 µB, corresponding to one unpaired
electron. This species has a strong infrared CN stretch
νCN ) 1762 cm-1. This is similar to Schrock’s [N3N]Mo-
(CN[t-Bu]) (where [N3N]3- ) [(Me3SiNCH2CH2)3N]3-),
Bond Lengths (Å)
Mo(1)-C(41)
Mo(1)-N(1)
Mo(1)-N(2)
Mo(1)-N(3)
C(41)-N(4)
N(4)-C(42)
1.936(6) (1.918)
1.972(5) (1.975)
1.981(5) (1.959)
1.967(5) (1.941)
1.184(8) (1.228)
1.495(9)
Bond Angles (deg)
C(41)-Mo(1)-N(1)
C(41)-Mo(1)-N(2)
C(41)-Mo(1)-N(3)
Mo(1)-C(41)-N(4)
C(41)-N(4)-C(42)
99.9(2)
99.6(2)
104.4(2)
174.2(5) (169.06)
137.8(7) (136.36)
which has an IR band νCN ) 1838 cm-1 30
These
.
perpendicular to the C41-N4-C42 plane is expected to
contain the unpaired electron, while a 2e back-bond
takes residence in a molybdenum d orbital lying in the
C41-N4-C42 plane. This back-bond is reflected in νCN
for 3-t-BuNC (1759 cm-1), at much lower frequency than
free t-BuNC (2145 cm-1). Furthermore, the Mo-C
distance 1.936(6) Å is in the range of a double bond.33
The coordinated isocyanide CR-N distance is 1.184(8)
Å. This is only slightly lengthened from free tert-
butylisocyanide.34
compounds have multiple bonding character between
the molybdenum and the carbon of the coordinated
isocyanide due to a strong π-back-bond from the π-basic
molybdenum, explaining the low-energy IR band.
Binding of a second equivalent of AdNC to 3 was not
2
observed even in the presence of excess isocyanide. H
NMR was used to monitor the production of para-
5
magnetic products in the reaction of Mo(N[t-Bu-d6]Ar)3
(3-d18) with n equivalents of AdNC (n ) 0, 1, 10) in C6H6
(Figure 1).
Rea ction of Mo(H)(Me2CdNAr )(N[i-P r ]Ar )2 (1)
w ith Bu lk y Isocya n id es. Reaction of Mo(H)(Me2Cd
NAr)(N[i-Pr]Ar)2 (1) with 1 equiv of AdNC results in
rapid, irreversible formation of a 1:1 adduct. This
compound is a brown oily solid with an infrared band
νCN ) 1718 cm-1. This strong IR band is at slightly lower
frequency than that observed for 3-AdNC (vide supra),
possibly signifying a more significant back-bonding
interaction in the less sterically crowded complex.
Reaction of 3 with tert-butylisocyanide (t-BuNC) led
to products analogous to those formed in the AdNC
system. X-ray structure determination of 3-t-BuNC
revealed a bent conformation of the isocyanide ligand
(Figure 2): C(41)-N(4)-C(42), 137.8(7)° (Table 1). This
bend is similar to that in (t-BuCN)Re[N(CH2CH2S)3]
(154.0(10)°)31,32 and (t-BuCN)W[N3NF] ([N3NF]3-
)
[(C6F5NCH2CH2)3N]3-; 132.2(10)°).30 Deviations from
linearity are generally attributed to the degree of back-
bonding from metal to isocyanide ligand. Back-bonding
in the d3 system 3-t-BuNC is such that the plane
In contrast to 3, compound 1 reversibly binds a second
2
equivalent of AdNC. H NMR was again used to study
the isocyanide binding phenomenon. Reaction of the
(28) Most substrates bind η1 to Mo(N[t-Bu]Ar)3, but dimethylcyan-
amide is a noteworthy exception. Tsai, Y.-C.; Stephens, F. H.; Meyer,
K.; Mendiratta, A.; Gheorghiu, M. D.; Cummins, C. C. Organometallics
2003, 22, 2902.
(33) Nugent, W. A.; Mayer, J . M. Metal-Ligand Multiple Bonds;
Wiley: New York, 1988.
(34) Free t-BuNC has a NC bond length of between 1.166 and 1.184
Å. Bak, B.; Hansen-Nygaard, L.; Rastrup-Andersen, J . J . Mol. Spec-
trosc. 1958, 2, 54.
(35) Seino, H.; Arita, C.; Nonokawa, D.; Nakamura, G.; Harada, Y.;
Mizobe, Y.; Hidai, M. Organometallics 1999, 18, 4165.
(36) Free MeNC has a NC bond length of 1.167 Å as determined by
microwave absorption spectroscopy. Kessler, M.; Ring, H.; Trambarudo,
R.; Gordy, W. Phys. Rev. 1950, 79, 54.
(29) For examples of η2-bound substrates on the Mo(N[i-Pr]Ar)3
fragment, see refs 22 and 25a.
(30) Greco, G. E.; O’Donoghue, M. B.; Seidel, S. W.; Davis, W. M.;
Schrock, R. R. Organometallics 2000, 19, 1132.
(31) Glaser, M.; Spiers, H.; Lugger, T.; Hahn, F. E. J . Organomet.
Chem. 1995, 503, C32.
(32) Spies, H.; Glaser, M.; Pietzsch, H.-J .; Hahn, F. E.; Lugger, T.
Inorg. Chim. Acta 1995, 240, 465.

Molybdaziridine Hydride Complexes
Organometallics, Vol. 23, No. 13, 2004 3129
F igu r e 4. X-ray crystal structure of 2-(AdNC)2. Thermal
ellipsoids are at the 50% probability level. Hydrogen atoms
have been omitted for clarity. Selected bond lengths and
angles are listed in Table 2.
F igu r e 3. Reaction of 1-d18 with n equivalents of AdNC
2
(n ) 0, 1, 2, 10; bottom to top) as monitored by H NMR.
The peak at δ ) 6.2 ppm (∆ν1/2 ) 12.0 Hz) can be assigned
to 2-AdNC-d18; the peak at δ ) 3.5 ppm (∆ν1/2 ) 11.5 Hz)
nitrogen bonds of the coordinated isocyanides are 1.153-
(12) and 1.153(13) Å, which is the same as the C-N
distance in free aliphatic isocyanides within experimen-
tal error.34,36
can be assigned to 2-(AdNC)2-d18
.
deuterated variant Mo(H)((CD3)2CdNAr)(N[i-Pr-d6]Ar)2
Stop p ed -F low Kin etic Mea su r em en ts. Compound
3 is electronically very similar to 2, as can be seen by a
comparison of the CO stretching frequencies in com-
22
(1-d18
)
with n equivalents (n ) 0, 1, 2, 10) of AdNC in
2
C6H6 followed by observation of the H NMR spectrum
gave evidence indicative of the species 2-AdNC-d18 (δ
) 6.2 ppm, ∆ν1/2 ) 12.0 Hz) and 2-(AdNC)2 (δ ) 3.5 ppm,
∆ν1/2 ) 11.5 Hz) (Figure 3). Addition of more than 2
equiv of AdNC resulted in no further reaction.
pounds 2-CO (1827 cm-1 24 and 3-CO (1840 cm-1).37
)
However, the apparent simplicity of the binding of
isocyanides in these systems is deceptive. The simplest
system to approach is 3 since no gross ligand rearrange-
ments are expected in its ligand binding reactions.
However a spin change occurs during the reaction from
high spin (3, three unpaired electrons) to low spin (3-
AdNC, one unpaired electron). In contrast, reaction of
1 (one unpaired electron) must occur with reverse
cyclometalation at some time along the reaction coor-
dinate (Scheme 1).
Several possibilities exist for the mechanism of con-
version between 1 and 2; two important limiting cases
are shown in Scheme 1. There could exist a preequilib-
rium between the two species such that a small popula-
tion of 2 is present in solution. This step could be a rapid
preequilibrium step or a potentially slow and rate-
determining one. Subsequently, reactants approach and
react with 2 directly (Scheme 1, pathway A). An
alternative mechanism can be proposed involving as-
sociative concerted attack of 1 by isocyanide without
prior formation of 2, where substrate forces the molyb-
daziridine hydride to undergo a reverse â-hydrogen
elimination (Scheme 1, pathway B). An additional
complexity is that 2-AdNC rapidly and reversibly binds
a second mole of AdNC, forming 2-(AdNC)2 (Scheme 1,
reaction 6). Outlined herein is a strategy to probe the
mechanism by which molybdaziridine hydride 1 permits
access to the tris-amide Mo(III) complex 2. In this
strategy, the 1/2 system is investigated by comparison
to three-coordinate Mo(N[t-Bu]Ar)3 (3), which has no
â-hydrogen atoms to eliminate.
Compound 2-(AdNC)2 is a purple, crystalline solid
with an infrared band νCN ) 2035 cm-1. This IR band
agrees well with literature compounds displaying trans
disposed isocyanide ligands. For example, trans-(dppe)2-
Mo(CNPh)(CN[t-Bu]) exhibits an infrared band νCN
)
2010 cm-1 corresponding to the CN stretch of the tert-
butylisocyanide ligand.35 The solution magnetic moment
(µeff) for 2-(AdNC)2 is 1.76 µB, corresponding to one
unpaired electron. Heating of 2-AdNC at 65 °C for 24 h
resulted in no substantive change in the 1H NMR
spectrum. However, heating of 2-(AdNC)2 under the
same conditions resulted in formation of 2-AdNC as
observed by 1H NMR with concomitant color change
from purple to brown. This implies that the second
equivalent of AdNC is more weakly bound than the first.
Indeed, a color change from purple to brown and the
corresponding UV-vis and FTIR spectral changes can
also be observed when dilute solutions of 2-(AdNC)2 are
prepared (vide infra).
An X-ray diffraction study revealed the solid state
structure of 2-(AdNC)2 (Figure 4). The isocyanide ligands
adopt a trans disposition, and the core geometry is
approximately trigonal bipyramidal. The N(i-Pr)Ar
ligands are arranged in a pseudo-Cs configuration, thus
optimizing lone pair donation from two of the amide
ligands. The AdNC ligands are now close to linear (C-
N-C: 173.9(11)°, 167.1(11)°), implying a reduction in
the degree of back-bonding vis-a`-vis 3-t-BuNC. The
distances between the molybdenum atom and the
coordinated isocyanide ligands’ carbons are 2.135(11)
and 2.083(11) Å. This is similar to previously reported
molybdenum-isocyanide single bonds.30,35 Further-
more, this Mo-C distance is longer by 0.17 Å than
the analogous bond in 3-t-BuNC, indicating a reduction
of Mo-C multiple bonding character. The carbon-
Reactions between coordinatively unsaturated com-
pounds 1 and 3 with AdNC proved to be rapid (color
changes were immediately observed upon mixing the
metal species with isocyanide solutions); therefore reac-
(37) Greco, J . B.; Peters, J . C.; Baker, T. A.; Davis, W. M.; Cummins,
C. C.; Wu, G. J . Am. Chem. Soc. 2001, 123, 5003.

3130 Organometallics, Vol. 23, No. 13, 2004
Stephens et al.
Sch em e 1. (Top ) Lim itin g Ca se Scen a r ios for
Rea ction of Mo(H)(Me2CdNAr )(N[i-P r ]Ar )2 (1) w ith
Su bstr a te 1-Ad a m a n tylisocya n id e (Ad NC); Ad )
1-a d a m a n tyl, Ar ) 3,5-d im eth ylp h en yl, L )
N[i-P r ]Ar . (Bottom ) Rea ction 7 Is th e Rea ction of
Ad NC w ith th e Bu lk ier Mo(N[t-Bu ]Ar )3 (3); L2
N[t-Bu ]Ar
)
F igu r e 5. Time-resolved (2.5 s intervals) spectral changes
and kinetic trace upon mixing 0.3 mM toluene solutions of
3 and AdNC in a 1:1 ratio at -80 °C observed by stopped-
flow spectrophotometry.
F igu r e 6. Dependence of the observed pseudo-first-order
rate constant on [AdNC] for the reaction between Mo(N[t-
Bu]Ar)3 (3) and AdNC in toluene at -60 °C. The initial
concentration of 3 was 0.15 mM (after mixing). A least-
squares fit of the data gives an intercept of 0.86 s-1 and a
slope (corresponding to a second-order rate constant) of
3270 M-1 s-1 with R2 ) 0.997.
tion rates were studied by the stopped-flow method with
spectrophotometric registration. The spectral changes
observed in the stopped-flow experiments were identical
to the spectral changes for these systems observed in
quartz cells with a conventional spectrophotometer. In
most cases, the reactions at room temperature were too
rapid even for the stopped-flow methodology, and ac-
curate measurements of the reaction rates were possible
only upon cooling.
Reaction of Mo(N[t-Bu ]Ar )3 (3) with 1-Adam an tyl-
isocya n id e. Reaction of 3 with AdNC was studied by
rapid scanning spectrometry in toluene solution over the
temperature range -80 to 25 °C using a broad concen-
tration range of isocyanide (0.3-8 mM). Only one
process was observed under all studied conditions
(Figure 5), accompanied by growth of an absorption
band with λmax ) 445 nm (ꢀ ) 5800 M-1cm-1). These
spectral changes are consistent with formation of a 1:1
adduct between 3 and AdNC.
The system containing equal concentrations of 3 and
AdNC gave kinetic traces that were fitted to a second-
order kinetic equation (with mean standard deviation
within 3%) (Figure 5, inset). In a separate series of
experiments, a 10-fold excess of AdNC was used, and
kinetic traces (more than 5 half-lives) were fitted with
a first-order kinetic equation, suggesting that the reac-
tion is first-order in 3. Indeed, varying the concentration
of 3 while keeping the concentration of AdNC constant
yielded identical values of pseudo-first-order rate con-
stant kobs3, confirming that kobs3 is independent of the
concentration of 3, and the reaction rate depends on the
first power of [3]. A plot of kobs3 versus initial concentra-
tion of AdNC is a straight line with an intercept near
zero (Figure 6), showing that the reaction is practically
irreversible and is first-order in AdNC. Consequently,
the reaction between 3 and AdNC is a second-order
process (first-order in each reactant) in the broad
temperature range from -80 to 25 °C (eq 2).
v ) kobs3[3] ) k2[3][AdNC]
(2)
The rate constants measured at different tempera-
tures under otherwise identical conditions gave linear
Arrhenius and Eyring plots (Figures S1 and S2, Sup-
porting Information). The values of activation param-
eters calculated from the data obtained under second-
order conditions (equal concentrations of both reactants)
and pseudo-first-order conditions (10-fold excess of
AdNC) are in a good agreement: ∆Hq ) 5.5 ( 0.5
kcal‚mol-1, ∆Sq ) -15 ( 4 eu (Table 4). In summary,
the formation of a monoadduct of 3 with AdNC was
found to be a rapid, low-barrier, second-order reaction
(first-order in both reagents).
Rea ction of Mo(H)(Me2CdNAr )(N[i-P r ]Ar )2 (1)
w ith 1-Ad a m a n tylisocya n id e. The reaction between

Molybdaziridine Hydride Complexes
Organometallics, Vol. 23, No. 13, 2004 3131
Ta ble 2. Selected Bon d Len gth s a n d An gles for
2-(Ad NC)2 (Ca lcu la ted Bon d Len gth s a n d An gles
for (η1-HNC)2Mo(NHMe)(NH2)2 in P a r en th eses)
Bond Lengths (Å)
Mo(1)-N(1)
Mo(1)-N(2)
Mo(1)-N(3)
Mo(1)-C(41)
Mo(1)-C(51)
C(41)-N(4)
N(4)-C(42)
C(51)-N(5)
N(5)-C(52)
2.112(8) (1.988)
1.983(8) (1.984)
1.990(8) (1.987)
2.135(11) (2.079)
2.083(11) (2.050)
1.153(12) (1.201)
1.422(13)
1.153(13) (1.209)
1.454(13)
Bond Angles (deg)
N(1)-Mo(1)-C(41)
N(2)-Mo(1)-C(41)
N(3)-Mo(1)-C(41)
N(1)-Mo(1)-C(51)
N(2)-Mo(1)-C(51)
N(3)-Mo(1)-C(51)
Mo(1)-C(41)-N(4)
C(41)-N(4)-C(42)
Mo(1)-C(51)-N(5)
C(51)-N(5)-C(52)
C(51)-Mo(1)-C(41)
82.7(3)
90.7(4)
91.6(4)
90.7(4)
94.7(4)
90.8(4)
175.9(9) (171.43)
173.9(11)
176.7(10) (175.36)
167.1(11)
F igu r e 7. Time-resolved (2.5 s intervals) spectral changes
and a kinetic trace obtained upon mixing toluene solutions
of 1 (0.3 mM) and AdNC (3 mM) in a 1:1 ratio in the
stopped-flow apparatus at -80 °C.
changed (within experimental error) upon further in-
crease in concentration of AdNC. This result allowed
us to estimate the stoichiometry of AdNC addition to 1
and indicated that a bis-adduct, 2-(AdNC)2, formed at
low temperatures (-80 to -20 °C) and/or high concen-
tration of isocyanide. Spectral differences observed at
various temperatures for the toluene solution containing
0.15 mM of 1 and 0.3 mM of AdNC suggested revers-
ibility of binding of the second molecule of AdNC to 1.
The equilibrium of AdNC binding to 1 in toluene was
further characterized by conventional spectrophoto-
metric titrations at room temperature, which showed
complete formation of a 1:1 adduct (Scheme 1, reaction
1) followed by reversible addition of a second AdNC
(Scheme 1, reaction 6). The Kassoc for the first, nearly
irreversible process (Scheme 1, reaction 1) could be
estimated as K1 > 3 × 106 M-1. The equilibrium
constant for reversible reaction 6 (Scheme 1) was
calculated: Keq ) 4500 M-1. In concentrated solutions
(ca. 0.1 M), stoichiometric AdNC (2 equiv) is sufficient
for quantitative formation of 2-(AdNC)2.
Even though the reaction between 1 and AdNC
involves two processes, the observed kinetics are fairly
similar to the simple 1:1 adduct formation described
above for 3-AdNC. Formation of the adduct from 1 and
a 10-fold excess of AdNC can be clearly seen as a pseudo-
first-order process in the spectral window from 500 to
575 nm (Figure 7, inset; small absorbance changes
below 500 nm gave rise to noisy kinetic traces).
The adduct formation rate increased substantially
with an increase in AdNC concentration, indicating that
the reaction is first-order in 1. At low temperature (-80
or -40 °C) the plot of the pseudo-first-order rate
constant, kobs1, versus the initial concentration of AdNC
is a straight line with a small intercept (0.014 at -80
°C (∼1.5% of rate scale); 0.086 at -40 °C (∼1% of rate
scale)). Similar results were obtained at 24 °C in single-
wavelength (λ ) 550 nm) experiments (Figure S3,
Supporting Information). These experiments confirmed
that the rate-limiting step is nearly irreversible and
first-order in AdNC.
172.8(4) (178.10)
Ta ble 3. Activa tion P a r a m eter s a n d Rea ction
En th a lp y for Rea ction s of 1 Equ iv of Ad NC w ith
Molybd a zir id in e Hyd r id e (1, r ea ction 1) a n d
Mo(N[t-Bu ]Ar )3 (3, r ea ction 7)
reaction 1
reaction 7
∆Hq (kcal‚mol-1
∆Sq (eu)
)
4.5 ( 0.5
-24 ( 4
-24.6 ( 0.5
5.5 ( 0.5
-15 ( 4
-29.1 ( 0.5
∆Hrxn (kcal‚mol-1
)
Ta ble 4. Cr ysta llogr a p h ic Da ta for 2-(Ad NC)2 a n d
3-t-Bu NC
2-(AdNC)2
3-t-BuNC
formula
C
60H90N5Mo
977.31
C41H63N4Mo
707.89
P21/n
15.1350(14)
13.5623(12)
19.6178(18)
90
91.1720(10)
90
4026.0(6)
4
orange prism
1.168
0.357
1516
fw
space group
a, Å
b, Å
c, Å
P1h
10.6383(11)
10.9340(11)
24.980(3)
91.598(2)
101.624(2)
101.257(2)
2784.4(5)
2
purple cube
1.166
0.276
1054
1.219
R, deg
â, deg
γ, deg
V, Å3
Z
cryst description
D
calcd, g‚cm-3
µ(Mo KR), mm-1
F(000)
GOF on F2
R(F), %a
1.143
0.0647
0.1563
0.1077
0.2495
Rw(F), %a
Quantity minimized ) Rw(F2) ) ∑[w(Fo2 - Fc2)2]/∑[(wFo2)2]1/2
;
a
R ) ∑∆/∑(Fo), ∆ ) |(Fo - Fc)|, w ) 1/[σ2(Fo2) + (aP)2 + bP], P )
2
[2Fc + Max(Fo,0)]/3.
1 and AdNC is more complex than analogous adduct
formation with 3. The observed spectral changes de-
pended upon the concentration of AdNC and on tem-
perature. At low temperature (-80 °C), an intense
absorption band with λmax ) 524 nm (ꢀ ) 6600 M-1cm-1
)
)
and a shoulder at ca. 575 nm (ꢀ ) 5100 M-1cm-1
appeared (Figure 7). This spectrum is substantially
different from the spectrum of 3-AdNC, where a single
maximum at 445 nm was observed (Figure 5). Stopped-
flow “titration” of 1 with AdNC at -40 °C showed that
absorbance from 525 to 575 nm doubles upon increasing
the 1:AdNC ratio from 1:1 to 1:2 and remains un-
The kinetic data suggest a second-order rate law for
the rate-limiting step of the reaction (eq 3). The observed
rate constants increased with temperature, yielding
linear Arrhenius and Eyring plots. The measurements
were performed using various concentrations: a 1:1

3132 Organometallics, Vol. 23, No. 13, 2004
Stephens et al.
F igu r e 8. Dependence of the observed pseudo-first-order
constants on [AdNC] for the reaction between Mo(H)-
(Me2CdNAr)(N[i-Pr]Ar)2 (1) and AdNC in toluene at dif-
ferent temperatures. In all cases the initial concentration
of 1 was 0.15 mM (after mixing). A least-squares fit of the
data generated the following intercepts and slopes (corre-
sponding to a second-order rate constant): at -80 °C, 0.014
(240 M-1 s-1) (R2 ) 0.998); at -40 °C, -0.086 (2080 M-1
s-1) (R2 ) 0.998).
F igu r e 9. Eyring plot for the reaction of Mo(H)(Me2Cd
NAr)(N[i-Pr]Ar)2 (1) (squares) and Mo(D)(Me2CdNAr)(N[i-
Pr]Ar)2 (1-d3) (triangles) with AdNC in toluene. The data
were obtained under pseudo-first-order conditions: c(1,
1-d3) ) 0.15 mM, c(AdNC) ) 1.5 mM. Calculated activation
parameters for Mo(H) (∆Hq ) 4.5 ( 0.5 kcal‚mol-1, ∆Sq )
-24 ( 4 eu) agree within experimental error with those
calculated for Mo(D) (∆Hq ) 4.6 ( 0.5 kcal‚mol-1, ∆Sq
-23 ( 4 eu).
)
molar ratio of 1 to AdNC (second-order conditions) and
a large excess of AdNC (pseudo-first-order conditions).
1 are depicted in Scheme 1. For complex 3 bearing tert-
butyl alkyl substituents, molybdaziridine hydride can-
not form, and the reaction scheme simplifies to reaction
7 (Scheme 1). The experimentally observed second-order
rate law and activation parameters for this reaction
(∆Hq ) 5.5 ( 0.5 kcal‚mol-1, ∆Sq ) -15 ( 4 eu) are
consistent with a bimolecular transition state involving
a simple associative addition at a crowded center. These
data provide a basis for comparison to the reaction of 1
with AdNC.
v ) kobs1[1] ) k[1][AdNC]
(3)
Virtually identical values for the activation parameters
were obtained under all conditions studied: ∆Hq ) +4.5
( 0.5 kcal‚mol-1, ∆Sq ) -24 ( 4 eu.
The kinetic data, taken together, indicate that reac-
tion 1 is the rate-limiting step, with reaction 6 being
faster (Scheme 1). The rate-limiting step, an addition
of the first AdNC ligand to 1, is practically irreversible,
as evidenced by a zero intercept of the plots of kobs
versus [AdNC]. The rate law and the kinetic parameters
(second-order rate constants and activation enthalpies
and entropies) do not depend on the stoichiometry of
the reaction mixture, also suggesting that the rate-
limiting step does not change upon increase in AdNC
concentration. The spectral changes, however, are con-
sistent with a bis-adduct formation at high concentra-
tions of AdNC. This observation agrees with the ob-
served kinetics if the addition of a second monodentate
ligand is a rapid process.
Rea ction betw een Deu ter a ted Com p lex 1-d 3 a n d
Ad NC. Use of 1-d3, where deuterium atoms have been
incorporated into the methine position of the ligand
isopropyl group, has permitted an investigation of the
isotope effect of the hydride (deuteride) migration from
molybdenum to imine ligand. If Mo-H(D) bond break-
ing were the rate-limiting step in isocyanide coordina-
tion, then a primary kinetic isotope effect would be
expected when 1-d3 is utilized. The reaction between
1-d3 and excess AdNC was studied by rapid scanning
stopped-flow spectrophotometry as a function of tem-
perature (from -80 to 20 °C). The spectral changes
observed were identical to those observed previously for
the reaction between 1 and AdNC. The rates of the
reactions of AdNC with 1 and 1-d3 at the same temper-
atures were nearly identical (deviations within 8%), and
Eyring plots yielded very similar values for the activa-
tion parameters (Figure 9). Therefore, there is no
detectable isotope effect in this system.
For complex 1, the addition of one molecule of AdNC
could occur, as discussed above (Scheme 1, pathway A),
by initial rate-determining conversion to 2. The second-
order rate law dependence on incoming AdNC rules out
that possibility. A more plausible scenario would be a
rapid preequilibrium between 1 and 2 (reaction 2, Keq)
followed by slower trapping of 2 by AdNC (reaction 3,
k3). However, three additional pieces of evidence argue
2
against pathway A. First, previously reported H NMR
studies22 demonstrated that high temperatures (ca. 70
°C) were necessary for coalescence of deuterium atoms
bound to Mo and those on the isopropyl groups of the
amides (methine position). Rapid equilibration at the
low temperatures utilized in the stopped-flow experi-
ments therefore seems unlikely. Furthermore, such an
equilibrium would be expected to display an inverse
equilibrium isotope effect,38 where 1-d3 reacted at a
faster rate than 1. However, rates of reaction and
activation parameters are identical within experimental
error.
Finally, if rapid preequilibrium kinetics were fol-
lowed, the derived rate law would resemble that shown
in eq 4:
v ) kobs(A)[1][AdNC], kobs(A) ) Keqk3,
∆Hqobs(A) ) ∆H°eq + ∆Hq (4)
3
The observed enthalpy of activation would be a sum of
the reaction enthalpies for the molybdaziridine hydride
rearrangement (∆H°eq) and an activation enthalpy of
Mech a n istic Con sid er a tion s. The major limiting
reaction pathways for adding one molecule of AdNC to
(38) (a) J ones, W. D. Acc. Chem. Res. 2003, 36, 140. (b) Labinger, J .
A.; Bercaw, J . E. Nature 2002, 417, 507.

Molybdaziridine Hydride Complexes
Organometallics, Vol. 23, No. 13, 2004 3133
AdNC addition to the three-coordinate complex 2 (∆Hq ).
Sch em e 2. Th er m och em ica l Cycle for Estim a tion
of th e En th a lp y of Oxid a tive Ad d ition of th e C-H
Bon d of 2 to F or m 1; L ) N[i-P r ]Ar , L2 ) N[t-Bu ]Ar
3
Comparative activation parameters as well as enthal-
pies of reaction for reaction of 1 and 3 with AdNC are
summarized in Table 3. Thermochemical data and
theoretical calculations (vide infra) indicate that (∆H°eq)
is expected to be on the order of 5 kcal‚mol-1 for this
reaction. This implies a near zero value for ∆Hq , which
3
is not reasonable.
On the basis of these observations we find no evidence
that reaction of 1 with AdNC proceeds via complex 2
as shown in pathway A of Scheme 1, and it can be con-
cluded that reaction occurs via the associative pathway
B. This pathway yields the observed second-order rate
law (i.e., for the rate-limiting reaction 4) (eq 5).
kobs(B) ) k4, ∆Hqobs(B) ) ∆Hq
(5)
4
Th er m och em istr y. The enthalpies of binding of the
first mole of AdNC to 1 and to 3 with all species in
toluene solution are ∆H ) -24.6 ( 0.5 kcal‚mol-1 and
∆H ) -29.1 ( 0.4 kcal‚mol-1, respectively (Table 4), as
determined by solution calorimetry. These values cor-
respond to the formal definition of the molybdenum-
isocyanide bond strengths for the two complexes. Dif-
ferences in the enthalpies of binding of AdNC can be
used to derive an estimate for the enthalpy of the 2 to
1 tautomerization. The enthalpy of transfer of the
isocyanide ligand from 2-AdNC to 3, generating 3-AdNC
and 1, is shown at the bottom of Scheme 2 and is
calculated from direct measurements (from solid AdNC)
The lack of kinetic isotope effect and the values of
activation and thermodynamic parameters also favor
pathway B.39
When comparing the activation parameters in Table
3, it can be seen that 1 has a significantly more negative
entropy of activation than does 3. This is consistent with
a more precise orientation requirement of the incoming
monodentate ligand for its coordination at the binding
site in molybdaziridine hydride 1. Somewhat surpris-
ingly the enthalpies of activation are comparable,
although that of 3 is slightly lower than that of 1.
Even though the reactions between 1 or 3 with AdNC
are complete within seconds and their second-order rate
to be -4.5 ( 0.3 kcal‚mol-1 46
This will equal the sum
.
constants are high (1.2 × 104 and 3 × 105 M-1 s-1
,
of the three steps in the upper cycle of Scheme 2:
BDE2-AdNC + ∆H2f1 - BDE3-AdNC ) -4.5 kcal‚mol-1
(BDE ) bond dissociation energy). Provided BDE2-AdNC
is approximately equal to BDE3-AdNC, these two terms
cancel and it is reasonable to assign ∆H ) -4.5
kcal‚mol-1 to the 2 to 1 tautomerization.
respectively, at 25 °C), these processes are several
orders of magnitude slower than the diffusion limit. In
contrast, reactions at sterically unhindered coordination
sites that do not require steric or electronic reorganiza-
tion are much faster. For example, a rate constant of
1013 M-1 s-1 was reported for CO addition to Cr(CO)5
in the gas phase,40 and the rate constant of CO addition
to [IrCl(PPh3)2] at room temperature was found to be
2.7 × 108 M-1 s-1 in benzene solution at room temper-
ature.41 Similarly, the rate constants of small-molecule
binding to vacant coordination sites in hemoproteins
and their models approach the diffusion limit and often
It is possible that oxidative addition of the anilide
ligand N(i-Pr)Ar may be slightly more exothermic than
this first estimate. The lower CN stretching frequency
of 2-AdNC (νCN ) 1718 cm-1) compared to 3-AdNC (νCN
) 1762 cm-1) appears to indicate (presumably for steric
reasons) that the 2-AdNC bonding interaction may be
slightly stronger than that in 3-AdNC. There is no
simple correlation between this vibration and the rela-
tive AdNC-Mo bond strengths; however it is reasonable
to conclude that 2 to 1 tautomerization is exothermic
equal 108-109 M-1 s-1 42-44
The reaction of complex 3
.
with N2O, which ultimately resulted in the N-N bond
cleavage and proceeded through N2O coordination to
Mo(III) at a rate-limiting step, had a higher activation
enthalpy (9.7 kcal‚mol-1) than the addition of AdNC to
3 (5.5 kcal‚mol-1). The activation entropies for the two
reactions (-24 eu for N2O and -15 eu for AdNC)
reflected a more ordered transition state required for
N2O activation.45
by 5 ( 2 kcal‚mol-1
.
That molybdaziridine hydride 1 is thermodynamically
downhill from its tautomer 2 supports kinetic studies
preferring pathway B over A in Scheme 1. The observed
activation energy of 4.5 ( 0.5 kcal‚mol-1 (Table 3)
overlaps within experimental error the 5 ( 2 kcal‚mol-1
derived ground state enthalpy difference for 1 f 2. If
the reaction proceeded via pathway A, it would therefore
require no barrier to oxidative addition of the C-H bond
and no barrier to addition of AdNC to 2. Neither of these
is reasonable, and all data (kinetic as well as thermo-
dynamic) support pathway B of Scheme 1.
(39) We cannot unambiguously distinguish between the associative
mechanism (that involves
a six-coordinate intermediate) and the
associative interchange (Ia) mechanism that proceeds through a six-
coordinate transition state.
(40) Seder, T. A.; Church, S. P.; Ouderkirk, A. J .; Weitz, E. J . Am.
Chem. Soc. 1985, 107, 1432.
(41) Wink, D. A.; Ford, P. C. J . Am. Chem. Soc. 1987, 109, 436.
(42) Wilkins, R. G. Kinetics of formation of biological oxygen carriers.
In Oxygen Complexes and Oxygen Activation by Transition Metals;
Martell, A. E., Sawyer, D. T., Eds.; Plenum Press: New York, 1987;
pp 49-60.
(43) Momenteau, M.; Reed, C. A. Chem. Rev. 1994, 94, 659.
(44) Warburton, P. R.; Busch, D. H. Dynamics of iron(II) and
cobalt(II) dioxygen carriers. In Perspectives on Bioinorganic Chemistry;
J AI Press Ltd.: London, 1993; Vol. 2, pp 1-79.
(45) Cherry, J .-P. F.; J ohnson, A. R.; Baraldo, L. M.; Tsai, Y.-C.;
Cummins, C.; Kryatov, S. V.; Rybak-Akimova, E. V.; Capps, K. B.;
Hoff, C. D.; Haar, C. M.; Nolan, S. P. J . Am. Chem. Soc. 2001, 123,
7271.
(46) Nolan, S. P.; Lopez de la Vega, R.; Hoff, C. D. Organometallics
1986, 5, 2529.

3134 Organometallics, Vol. 23, No. 13, 2004
Stephens et al.
The value of -5 ( 2 kcal‚mol-1 for ∆H2f1 is consistent
with the fact that 2 is not observed at room tempera-
ture.22 The exothermic nature of this cyclometalation
is in accord with an unfavorable entropic component due
to restricted rotation in the molybdaziridine hydride.
Recent theoretical work by Milstein and co-workers47
on cycloaddition reactions of C-H and C-C bonds has
indicated unfavorable -T∆S terms on the order of 2
kcal‚mol-1 that must be overcome for cycloadditions. A
room-temperature estimate of ∆G1f2 ) 3 kcal‚mol-1 is
consistent with the failure to observe 2 directly and with
the aforementioned observation of high-temperature
coalescence in NMR studies.22 On the other hand if the
intramolecular oxidative addition were more exothermic
than 5 kcal‚mol-1, oxidative addition of external H-H
or C-H bonds might be expected to occur if steric factors
did not preclude this. Activation of hydrocarbons by 1,
for example, does not occur. Therefore, the estimate of
the oxidative addition appears reasonable in terms of
these observations.
In contrast to rapid and irreversible binding of the
first mole of AdNC to 1, the second mole of AdNC adds
rapidly but the reaction is reversible. Direct measure-
ment of the enthalpy of binding of AdNC to 2-AdNC
yielded ∆H ) -10.3 ( 0.5 kcal‚mol-1 in toluene solution.
Despite this relatively low value, FTIR data (see Sup-
porting Information) indicate that, at room temperature
and 0.1 to 0.001 M concentration range, there is little
dissociation of 2-(AdNC)2 to 2-AdNC and AdNC. It is
estimated based on these data that, at room tempera-
ture, Keq for binding of AdNC to 2-AdNC is greater than
5000, in reasonable agreement with UV-vis estimates
discussed earlier. These data imply that the unfavorable
T∆S of binding of the second mole of AdNC is greater
than -5.3 kcal‚mol-1 at room temperature. This is less
than the -10 kcal‚mol-1 often observed for ligand
binding reactions. A more disordered structure of
2-(AdNC)2 compared to 2-AdNC may account for this.
The low value for the second enthalpy of binding of
AdNC to 1 also explains the failure of 3 to bind a second
mole of AdNC even in the presence of 10-fold excess.
The binding of the second mole of isocyanide occurs not
only with increased steric pressure at the binding site
but with a concomitant reduction in the Mo-CNAd bond
order of the trans isocyanide.
of binding of a second mole of AdNC to 2-AdNC is ca.
15 kcal‚mol-1 lower than the first mole; thus, the rate
increase can be anticipated. Stopped-flow kinetic studies
(see Supporting Information) of the comproportionation
reaction between 2-AdNC2 and 1 are interpreted in
terms of the mechanism in eq 7, which is a combination
of reactions 1 and 6 (Scheme 1).
k2
9
2-(AdNC)2 [ k ] 2-AdNC + AdNC + 1
8 2 2-AdNC
1
k-1
(7)
This reaction is rapid at room temperature, and its rate
constants (k1, k-1, and k2) could not be unequivocally
resolved even at low temperature. Nevertheless, the fol-
lowing semiquantitative observations could be made: (1)
Temperature dependence studies revealed an overall
enthalpy of activation of ca. 9 kcal‚mol-1, indicating that
dissociation of AdNC from 2-(AdNC)2, the k-1 step in
eq 7, plays a key role in the mechanism; (2) the rate of
binding of AdNC, surprisingly, was roughly equivalent
for 1 and for 2-AdNC (i.e., k2 ≈ k1).
The observation of near quantitative binding of the
second mole of AdNC to 2-AdNC at low temperatures
led to attempts to determine whether 3-t-BuNC would
add a second equivalent of isocyanide at low tempera-
ture in concentrated solution:
3-t-BuNC + t-BuNC / 3-(t-BuNC)2
(8)
However, even when 3 was dissolved in 3 M t-BuNC in
toluene at ca. -50 °C, no formation of a 2:1 adduct was
observed by FT-IR spectroscopy. Isocyanide exchange
reactions are observed for 3-AdNC and 3-t-BuNC (vide
infra), and this is most likely done via generation of the
thermodynamically unstable bis-adduct of 3. Spectro-
scopic data allow the conservative estimate that Keq for
eq 8 is less than 0.033 M-1 at 293 K. The analogous
binding of AdNC by 2-AdNC has a Keq of 4500 M-1. The
difference in binding constants of 5 orders of magnitude
for the second mole of isocyanide corresponds at room
temperature to a free energy difference of g6 kcal‚mol-1
.
This is a reasonable lower bound for the “steric pres-
sure” that would be generated if the ligands’ isopropyl
groups are replaced by tert-butyl groups.
Despite thermodynamic instability, there is kinetic
evidence that suggests 3-(AdNC)(t-BuNC) is a viable
intermediate in isocyanide exchange. Small differences
in the FTIR band shapes of 3-AdNC and 3-t-BuNC and
the free isocyanides AdNC and t-BuNC allowed study
of the equilibrium reaction shown in eq 9:
Isocya n id e Exch a n ge a n d Com p r op or tion a tion
Rea ction s. The enthalpic preference of ca. 5 kcal‚mol-1
for binding of AdNC to 3 compared to 1 prompted
investigation of the exchange reaction (eq 6):
2-AdNC + 3 f 1 + 3-AdNC
(6)
3-t-BuNC + AdNC / 3-AdNC + t-BuNC (9)
While it was anticipated that the reaction shown in eq
6 would proceed as shown, the reverse reaction was also
attempted. As described in the Supporting Information,
reactions monitored after several days at room temper-
ature showed no sign of exchange in either direction.
Since kinetic data showed that trapping of AdNC by 3
is faster than trapping by 1, it can be concluded that
dissociation of isocyanides from the 1:1 adducts occurs
at a negligible rate at room temperature.
At room temperature and in dilute solution, this equi-
librium is established within minutes of mixing of
solutions. As shown in the Supporting Information, the
equilibrium constant is approximately 1, and the reac-
tion can be approached from either side. Because bound
isocyanide dissociates very slowly from 3-AdNC and 3-t-
BuNC (vide supra), a dissociative exchange mechanism
can be excluded, and we assume a mechanism similar
to that shown in Scheme 3.
In contrast, dissociation of AdNC from the bis-adduct
2-(AdNC)2 occurs at a much faster rate. The enthalpy
Quantitative kinetic study of this reaction was not
made, but it is several orders of magnitude slower than
any of the steps in the comproportionation reaction
(47) Rybtchinski, B.; Oevers, S.; Montag, M.; Vigalok, A.; Rozenberg,
H.; Martin, J . M. L.; Milstein, D. J . Am. Chem. Soc. 2001, 123, 9064.

Molybdaziridine Hydride Complexes
Organometallics, Vol. 23, No. 13, 2004 3135
Sch em e 3. P r op osed Mech a n ism for Associa tive
Isocya n id e Exch a n ge in 3-Ad NC
mechanism shown in eq 7. The much slower rate of the
reaction (eq 9) is attributed primarily to slower uptake
of a second isocyanide to form the proposed five-
coordinate intermediate 3-(AdNC)(t-BuNC), which, as
discussed above, is estimated be at least 6 kcal‚mol-1
less stable than 2-(AdNC)2.
Th eor etica l Ca lcu la tion s
Density functional theory calculations have been
effectively used to elucidate bonding in molybdenum
tris-amide complexes of various ligands.21,23,28,37 Rela-
tively complex N(t-Bu)Ar ligands can be effectively
approximated by the much simpler NHMe and NH2
ligands, and the bulky 1-adamantyl (or tert-butyl)
group was simplified to H. Therefore, we investigated
computationally the compounds Mo(NHMe)(NH2)2,
Mo(H)(H2CdNH)(NH2)2, (η1-HNC)Mo(NH2)3, and (η1-
HNC)2Mo(NH2)3. The spin-unrestricted, all-electron
calculations focused on optimizing the geometry of the
complexes compared to X-ray crystallographic determi-
nations and on an investigation of the molecular orbitals
and their relative energies with specific concern paid
to the spin states of the fragments. Binding of AdNC to
2 or 3 involves spin change from a quartet to a doublet
ground electronic state.49-51 In contrast, binding of
AdNC to 1 does not involve spin change since both
reactant and product have been shown to have doublet
ground states. However the latter reaction involves, at
some point, conversion of the molybdaziridine hydride
form (1) to the tris-amide form (2).
Geometry optimization of Mo(H)(H2CdNH)(NH2)2
and Mo(NHMe)(NH2)2 revealed structural parameters
that agree well with X-ray crystallographically deter-
mined structures of 122 and 3.5 For molybdaziridine
hydride, the experimentally (theoretically) determined
parameters are as follows: Mo-H, 1.69(5) (1.675); Mo-
C(imine), 2.18(2) (2.153); Mo-N(imine), 1.93(2) (2.009);
Mo-N(amide), 1.93(2) (1.947) and 2.02(2) (1.951) Å.
Similarly, for the tris-amide species Mo(NHMe)(NH2)2,
the experimentally (theoretically) determined param-
eters are as follows: Mo-N(unique), 1.960(7) (1.965);
F igu r e 10. (Top) Calculated molecular orbital energies of
Mo(H)(H2CdNH)(NH2)2 (Mo(H)), Mo(NHMe)(NH2)2 (MoL3),
(HNC)Mo(NHMe)(NH2)2 ((HNC)MoL3), and (HNC)2Mo-
(NHMe)(NH2)2 ((HNC)2MoL3). Dashed lines connect the
LUMOs (top) and the three highest occupied molecular
orbitals. (Bottom) Calculated total bonding energy com-
parison. Each energy level depicts a system with the same
formula.
Mo-N, 1.964(7) (1.976) and 1.977(7) (1.978) Å. Using
the orbital energy levels calculated from these geometry
optimized structures, an orbital energy level diagram
was prepared (Figure 10).52 Important features of this
diagram include the HOMO-LUMO gaps. For the
molybdaziridine hydride model (Figure 10, far left), the
three highest energy occupied orbitals represent one
unpaired electron (HOMO) and the back-bond from the
metal to the η2-imine functionality (HOMO-1, HOMO-
2). In three-coordinate Mo(NHMe)(NH2)2, the highest
occupied molecular orbitals are three singly occupied d
(48) (a) Deleted in proof. (b) Mindiola, D. J .; Tsai, Y.-C.; Hara, R.;
Chen, Q.; Meyer, K.; Cummins, C. C. Chem. Commun. 2001, 125. (c)
Agapie, T.; Diaconescu, P. L.; Cummins, C. C. J . Am. Chem. Soc. 2002,
124, 2412. (d) Deleted in proof.
(49) (a) Harvey, J . N.; Poli, R.; Smith, K. M. Coord. Chem. Rev. 2003,
238-239, 347. (b) Poli, R.; Harvey, J . N. Chem. Soc. Rev. 2003, 32, 1.
(c) Poli, R. Acc. Chem. Res. 1997, 30, 494. (d) Poli, R. Chem. Rev. 1996,
96, 2135.
(50) (a) Hess, J . S.; Leelasubcharoen, S.; Rheingold, A. L.; Doren,
D. J .; Theopold, K. H. J . Am. Chem. Soc. 2002, 124, 2454. (b) Detrich,
J . L.; Reinaud, O. M.; Rheingold, A. L.; Theopold, K. H. J . Am. Chem.
Soc. 1995, 117, 11745.
(51) Schro¨der, D.; Shaik, S.; Schwarz, H. Acc. Chem. Res. 2000, 33,
139.
2
orbitals (dxz, dyz, and dz ). The HOMO-LUMO gap in
the former compound is slightly smaller than that in
(52) Molecular orbital pictures for Mo(H)(Me2CdNH)(NH2)2, Mo-
(NHMe)(NH2)2, (η1-HNC)Mo(NHMe)(NH2)2, and (η1-HNC)2Mo(NHMe)-
(NH2)2 can be found in the Supporting Information.

3136 Organometallics, Vol. 23, No. 13, 2004
Stephens et al.
the latter, but the overall bonding energies indicate that
Mo(H)(H2CdNH)(NH2)2 is nearly isoenergetic with
Mo(NHMe)(NH2)2 (Figure 10, bottom). Consequently,
the opening of the molybdaziridine hydride should be a
relatively low-energy process, consistent with the ex-
perimental results (vide supra).
hydride 1 to three-coordinate Mo(N[i-Pr]Ar)3 (2) is
identified with a small reaction enthalpy of ca. 5
kcal‚mol-1. Nonetheless, failure to observe a kinetic
isotope effect (either primary or inverse) in reaction of
molybdaziridine deuteride 1-d3 with AdNC, the rate law,
and the activation parameters all support associative
attack of AdNC on 1 rather than any requirement of
prior conversion to 2.
Utilizing the same amido ligand set as discussed
above, geometry optimization of (η1-HNC)Mo(NHMe)-
(NH2)2 was carried out with no symmetry constraints
(Table 1). The Mo-C distance was computed to be 1.918
Å, just slightly longer than a typical Mo-C double bond
distance (ca. 1.88 Å) in Schrock-type carbene complexes
of molybdenum.33,53 The calculated value is slightly
shorter than the experimentally determined value for
the Mo-C distance in 3-t-BuNC, 1.936(6) Å (Figure 2).
This is likely due to the steric constraints imposed by
the N(t-Bu)Ar ligands in the synthetic system compared
to the reduced steric bulk of NH2 ligands in the
computational system. Approximate C-N double-bond
character is indicated by the C-N-H bond angle of 139°
(experimental, 137.8(7)°), as well as by the C-N bond
distance of 1.223 Å (experimental, 1.184(8) Å). Subse-
quently, a spin unrestricted calculation was carried out
on the d1 complex and the orbitals were examined.52 An
informative way to describe the conformation of the
three amide ligands in the complex is to say that two of
the lone pairs are perpendicular to the Mo-C vector,
while one is parallel. That the computed structure
corresponds to a minimum on the potential energy
surface was confirmed via a frequencies calculation;
there were no imaginary frequencies, and the most
intense vibrational band (νCN) was predicted to occur
at 1802 cm-1. This value agrees well with the experi-
mental values reported above for 2-AdNC (1718 cm-1),
3-AdNC (1762 cm-1), and the t-BuNC analogues.
The unpaired electron of (η1-HNC)Mo(NHMe)(NH2)2
lies in the same plane as the nitrogen lone pair p orbital
on the unique amide ligand.52 The unpaired electron is,
therefore, antibonding with respect to the lone pair of
electrons on the unique amide ligand. At the same time,
it is weakly back-bonding with respect to the isocyanide
LUMO. A two-electron back-bond is in place involving
the other two formal d electrons and the second com-
ponent of the isocyanide LUMO. The two-electron back-
bond is at lower energy than the one-electron back-bond
because it experiences minimal repulsive effects from
amide lone pairs.
The important role that ancillary ligands can play in
determining reactivity was also demonstrated in the fact
that 1 reversibly binds a second equivalent of isocyanide
whereas 3 could not be observed to do so even in
concentrated solution at low temperature. The crystal
structures of 3-t-BuNC and 2-(AdNC)2 reported here,
as well as data on enthalpy and entropy of binding of
the second mole of AdNC, provide a good entry to
further understanding of ligand binding in these steri-
cally congested but low coordination number systems.
Isolable coordinatively unsaturated metal complexes
such as 3 have afforded remarkable new small-molecule
transformations.5,6,8,22,24-29,37,45,55 Comparative reactivity
and mechanistic studies of 1 are making increasingly
clear the fact that coordinative unsaturation in a
starting metal complex is not an absolute requirement
for the type of small-molecule activation processes
exemplified by dinitrogen binding and subsequent scis-
sion. On the contrary, reversible intramolecular ele-
ments of stabilization, from agostic stabilization56 to
near-thermoneutral cyclometalation as seen here,57 may
be employed as design elements in the synthesis of new
reactive motifs.26
Exp er im en ta l Section
Gen er a l Con sid er a tion s: Syn th esis a n d Ch a r a cter iza -
tion . Unless stated otherwise, all operations were performed
in a Vacuum Atmospheres glovebox under an atmosphere of
purified nitrogen or using Schlenk techniques under an argon
atmosphere. Diethyl ether, toluene, benzene, pentane, and
n-hexane were dried and deoxygenated by the method of
Grubbs.58 THF was distilled under nitrogen from purple
sodium benzophenone ketyl. Benzene-d6 was purchased from
Cambridge Isotope Laboratory, degassed, and dried over 4 Å
sieves. Alumina, Celite, and 4 Å sieves were dried in vacuo
overnight at a temperature above 200 °C. Mo(H)(η2-Me2Cd
NAr)(N[i-Pr]Ar)2,22 Mo(D)(η2-Me2CdNAr)(N[i-Pr-d1]Ar)2,22
Mo(H)(η2-(CD3)2CdNAr)(N[i-Pr-d6]Ar)2,22 Mo(N[t-Bu]Ar)3,5,59
Mo(N[t-Bu-d6]Ar)3,5 and 1-adamantylisocyanide60 were syn-
thesized according to literature procedures. 1-Adamantyliso-
cyanide was sublimed before use. Other chemicals were used
as received. Solution infrared spectra were recorded on a
Perkin-Elmer 1600 Series FTIR using KBr plates. 1H and 13C
NMR spectra were recorded on Varian XL-300, Varian Mercury-
Geometry optimization with no symmetry constraints
and a spin-unrestricted calculation were completed on
(η1-HNC)2Mo(NHMe)(NH2)2.52 The calculated bond dis-
tances agreed well with the experimentally determined
distances (Table 2). For example, the Mo-C distances
were computed to be 2.079 and 2.050 Å, while the
distances from the X-ray crystal structure are 2.135(11)
and 2.083(11) Å. Like in the previously discussed 1:1
adduct, the HOMO is a back-bond to the isocyanide
ligands.
(55) (a) Veige, A. S.; Slaughter, L. M.; Lobkovsky, E. B.; Wolczanski,
P. T.; Matsunaga, N.; Decker, S. A.; Cundari, T. R. Inorg. Chem. 2003,
42, 6204. (b) Veige, A. S.; Slaughter, L. M.; Wolczanski, P. T.;
Matsunaga, N.; Decker, S. A.; Cundari, T. R. J . Am. Chem. Soc. 2001,
123, 6419. (c) Wolczanski, P. T. Polyhedron 1995, 14, 3335.
(56) (a) King, W. A.; Scott, B. L.; Eckert, J .; Kubas, G. J . Inorg.
Chem. 1999, 38, 1069. (b) King, W. A.; Luo, X.-L.; Scott, B. L.; Kubas,
G. J .; Zilm, K. W. J . Am. Chem. Soc. 1996, 118, 6782. (c) Luo, X.-L.;
Kubas, G. J .; Bryan, J . C.; Burns, C. J .; Unkefer, C. J . J . Am. Chem.
Soc. 1994, 116, 10312.
Con clu sion s
(57) Bercaw, J . E. J . Am. Chem. Soc. 1974, 96, 5087.
(58) Pangborn, A. B.; Giardello, M. A.; Grubbs, R. H.; Rosen, R. K.;
Timmers, F. J . Organometallics 1996, 15, 1518.
(59) N-tert-Butyl-3,5-dimethylanilide ligands are synthesized via an
improved procedure. See ref 28.
Thermochemical cycles, together with DFT calcula-
tions, suggest that tautomerization of molybdaziridine
(53) Schrock, R. R. Chem. Rev. 2002, 102, 145.
(54) Deleted in proof.
(60) Sasaki, T.; Nakanishi, A.; Ohno, M. J . Org. Chem. 1981, 46,
5445.

Molybdaziridine Hydride Complexes
Organometallics, Vol. 23, No. 13, 2004 3137
300, or Varian INOVA-501 spectrometers at room tempera-
ture. Chemical shifts are reported with respect to internal
residual benzene (7.15 and 128.38 (t) ppm). C, H, and N
analyses were performed by H. Kolbe Mikroanalytisches
Laboratorium (Mu¨lheim, Germany).
Cr ysta llogr a p h ic Str u ctu r e Deter m in a tion s. The X-ray
data collections were carried out on a Siemens Platform three-
circle diffractometer (Mo KR, λ ) 0.71073 Å) mounted with a
CCD detector and outfitted with a low-temperature, nitrogen-
stream aperture (189 K). The structures were solved by direct
methods, in conjunction with standard difference Fourier
techniques, and refined by full-matrix least-squares proce-
dures. Selected bond distances and angles are supplied in
Table 2 and Table 1. A summary of crystallographic data is
given in Table 5, with full details found in the Supporting
Information. The systematic absences in the diffraction data
are uniquely consistent with the assigned space group of P21/n
for 3-t-BuNC. No space group symmetry higher than triclinic
was indicated in the diffraction data for 2-(AdNC)2. These
choices led to chemically sensible and computationally stable
refinements. An empirical absorption correction (ψ-scans) was
applied to both data sets. All software for diffraction data
processing and crystal-structure solution and refinement are
contained in the SHELXTL (v5.10) program suite (G. Sheld-
rick, Siemens XRD, Madison, WI).
Syn th esis of (1-Ad NC)Mo(N[i-P r ]Ar )3 (2-Ad NC). To an
orange-brown solution of 1.334 g of Mo(H)(η2-Me2CdNAr)(N[i-
Pr]Ar)2 (2.289 mmol) in 5 mL of Et2O was added 3 mL of a
colorless solution of AdNC (0.369 g, 2.292 mmol, 1 equiv). Upon
mixing of the two solutions at 20 °C, the reaction mixture
became green-brown. After stirring for 1 h, the solvent was
removed in vacuo to yield an oily brown product. 1H NMR (500
MHz, C6D6): δ 18.14, 12.12 (i-Pr methyl), 6.65, 1.96, 1.66, 1.42,
2
-8.68 ppm. H NMR (500 MHz, Et2O): δ 6.21 ppm. IR (Et2O,
KBr): 1718 cm-1. The oily nature of this compound precluded
elemental analysis.
Syn th esis of (1-Ad NC)2Mo(N[i-P r ]Ar )3 (2-(Ad NC)2). To
an orange-brown solution of 2.775 g of Mo(H)(η2-Me2CdNAr)-
(N[i-Pr]Ar)2 (4.759 mmol) in 20 mL of Et2O was added 5 mL
of a colorless solution of AdNC (1.530 g, 9.501 mmol, 2 equiv).
Upon mixing of the two solutions at 20 °C, the reaction mixture
became purple. After 1 h, the solvent was removed in vacuo
and the resulting solid was recrystallized from cold Et2O to
yield 3.427 g of purple crystals (79.7%). 1H NMR (500 MHz,
C6D6): δ 18.00, 6.9, 4.15, 2.07, 1.74, -0.12, -2.23, -8.59,
-11.76 ppm. 2H NMR (500 MHz, Et2O): δ 3.5 ppm (ν1/2 ) 11.5
Hz). µeff ) 1.76 µB (Evans’ method, C6D6, 20 °C). IR (Et2O,
KBr): 2035 cm-1. Anal. Calcd for C55H78N5Mo: C, 72.98; H,
8.68; N, 7.73. Found: C, 72.73; H, 8.82; N, 7.65.
Gen er a l Con sid er a tion s: Stop p ed -F low Kin etics. Tolu-
ene (anhydrous, 99.8%, Acros) was purified further by the
method of Grubbs.58 Toluene solutions of 1, 1-d3, 3, and AdNC
were prepared in a Vacuum Atmospheres glovebox filled with
argon. Kinetic measurements were performed at temperatures
from -80 to 25 °C using a Hi-Tech Scientific (Salisbury,
Wiltshire, UK) SF-43 Multi-Mixing CryoStopped-Flow instru-
ment in diode array and single-wavelength modes. The
stopped-flow instrument was equipped with stainless steel
plumbing, a stainless steel mixing cell with sapphire windows,
and an anaerobic gas-flashing kit. The instrument was con-
nected to an IBM computer with IS-2 Rapid Kinetic software
(Hi-Tech Scientific). The temperature in the mixing cell was
maintained to (0.1 K, and the mixing time was 2 ms. The
driving syringe compartment and the cooling bath filled with
heptane (Fisher) were flushed with argon before and during
the experiments, using anaerobic kit flush lines. All flow lines
of the SF-43 instrument were extensively washed with de-
gassed, anhydrous toluene before charging the driving syringes
with reactant solutions. The experiments were performed in
a single-mixing mode of the instrument, with solutions of two
reagents mixed in a 1:1 (v/v) ratio.
Syn th esis of (1-Ad NC)Mo(N[t-Bu ]Ar )3 (3-Ad NC). To an
orange solution of 1.566 g of Mo(N[t-Bu]Ar)3 (2.505 mmol) in
10 mL of Et2O was added 5 mL of a colorless Et2O solution of
AdNC (0.404 g, 2.507 mmol, 1 equiv). The brown-orange
reaction mixture was stirred at 20 °C for 90 min, and then
the solvent was removed in vacuo. Crystallization from cold
Et2O yielded red-brown needles (1.229 g, 62.4%). 1H NMR (500
MHz, C6D6): δ 28.42, 16.75, 10.42, 3.40, 1.63, -7.64, -7.93,
-18.50 ppm. 2H NMR (500 MHz, Et2O): δ 9.9 ppm (ν1/2 ) 12.8
Hz). IR (Et2O, KBr): 1762 cm-1. µeff ) 1.825 µB (Evans’ method,
C6D6, 20 °C). Anal. Calcd for C47H69N4Mo: C, 71.81; H, 8.84;
N, 7.12. Found: C, 70.94; H, 8.68; N, 7.00.
Syn th esis of (t-Bu NC)Mo(N[t-Bu ]Ar )3 (3-t-Bu NC). To an
orange-brown solution of 2.732 g of Mo(N[t-Bu]Ar)3 (4.371
mmol) in 10 mL of pentane was added a solution of t-BuNC
(0.375 g, 4.371 mmol, 1.0 equiv) in 5 mL of pentane. The
solution was stirred for 1 h at room temperature, at which
time the reaction mixture was filtered through a fine frit. From
the initial filtration, 1.926 g of brown powder was isolated.
The solvent was removed from the filtrate in vacuo, redissolved
in approximately 10 mL of pentane, and a second crop of
orange-brown, crystalline product was recrystallized at -35
°C (2.676 g (overall), 3.780 mmol, 86.4%). IR (KBr, pentane):
1759, 1727 cm-1. Anal. Calcd for C41H63N4Mo: C, 69.56; H,
8.97; N, 7.91. Found: C, 69.10; H, 8.98; N, 7.99.
Gen er a l Con sid er a tion s: Th er m och em istr y. Infrared
data were recorded on a Perkin-Elmer System 2000 FTIR in
sealed solution cells obtained from Harrick Scientific. Solvents
used, toluene, C6D6, and heptane, were all distilled from Na/
benzophenone under argon prior to use. Additional details
regarding infrared spectroscopic experiments are provided in
the Supporting Information.
Enthalpies of reaction were measured in a Setaram C-80
Calvet calorimeter that was loaded in an argon-filled glovebox.
In a typical procedure, a stock solution of approximately 0.2 g
of 1 was dissolved in 7 mL of toluene that had been freshly
distilled from purple sodium benzophenone ketyl. The solution
was filtered, and 5 mL was loaded into the calorimeter
containing an ampule of AdNC (0.0176 g). (The remaining
solution was used to obtain an FTIR spectrum of the stock
solution.) The calorimeter cell was subsequently sealed, taken
from the glovebox, loaded into the calorimeter, and allowed
to thermally equilibrate for 2 h. Rotation of the calorimeter
prior to breaking the ampule gave a small exothermic reaction
due to trace contaminants inside the cell. A second such
rotation showed no thermal signal, and the reaction was
initiated by breaking the ampule. Upon completion of the
reaction and thermal equilibration, the cell was taken back
into the glovebox. An aliquot of the reaction mixture was
loaded into an FT-IR cell for product observation, and the
entire apparatus was cleaned with dried, degassed toluene.
The cell was stored in an evacuated container inside the
glovebox between calorimetric runs.
Syn th esis of (t-Bu NC)Mo(N[i-P r ]Ar )3-d 18 (2-t-Bu NC-
d
18). To a brown solution of 1.519 g of Mo(H)(η2-Me2CdNAr)-
(N[i-Pr]Ar)2-d18 (2.527 mmol) was added a pentane solution
of t-BuNC (0.209 g, 1.0 equiv, 2.524 mmol). The solution was
stirred for 1 h at room temperature, at which time the solvent
was removed in vacuo. The orange-brown residue was oily, and
pure solid was not obtained by recrystallization from cold
pentane (-35 °C). IR (KBr, pentane): 1751, 1718 cm-1
NMR (500 MHz, Et2O): δ 7.1 ppm (ν1/2 ) 9.2 Hz).
.
2H
Syn th esis of (t-Bu NC)2Mo(N[i-P r ]Ar )3 (2-(t-Bu NC)2). To
a brown solution of 0.127 g of Mo(H)(η2-Me2CdNAr)(N[i-Pr]-
Ar)2-d18 (0.211 mmol) was added a diethyl ether solution of
t-BuNC (0.036 g, 2.05 equiv, 0.433 mmol). The reaction
mixture became purple upon mixing and was allowed to stir
for 1 h at room temperature. IR (KBr, pentane): 2019 cm-1
.
2H NMR (500 MHz, Et2O): δ 3.9 ppm (ν1/2 ) 14.6 Hz).

3138 Organometallics, Vol. 23, No. 13, 2004
Stephens et al.
DF T Ca lcu la tion s. Calculations were performed with
version 2002.02 of the Amsterdam DFT program package,
ADF.61-63 Atomic coordinates from the X-ray diffraction study
of 1,22 3,5 3-t-BuNC, and 2-(AdNC)2 were employed. Calcula-
tions were carried out with all the electrons considered
explicitly; that is, no frozen core approximation was utilized.
The atomic basis sets employed were the ZORA/TZ2P bases
supplied with the program, and accordingly, scalar relativistic
effects were included with the ZORA treatment.64 The calcula-
tions were carried out in spin-unrestricted mode. The local
density approximation functional used was that of Vosko, Wilk,
and Nusair, while the functionals for the generalized gradient
approximations took the form of Becke (exchange) and Perdew
(correlation).65-67
Su p p or tin g In for m a tion Ava ila ble: Tables with bond
lengths, bond angles, atomic coordinates, and anisotropic
displacement parameters for 2-(AdNC)2 and 3-t-BuNC; details
of the DFT calculations; details of comproportionation and
exchange reactions; pictorial representations of the frontier
molecular orbitals of Mo(H)(H2CdNH)(NH2)2, Mo(NHMe)-
(NH2)2, (HNC)Mo(NHMe)(NH2)2, and (HNC)2Mo(NHMe)(NH2)2
as obtained from the DFT calculations; stopped-flow kinetic
plots; FTIR information for 3-(AdNC), 2-(AdNC), and 2-(AdNC)2.
This information is available free of charge via the Internet
at http://pubs.acs.org.
OM040045Z
Ack n ow led gm en t. For support of this work, the
authors are grateful to the National Science Foundation
(CRC-0209977).
(64) (a) van Lenthe, E.; Baerends, E. J .; Snijders, J . G. J . Chem.
Phys. 1993, 99, 4597. (b) van Leeuwen, R.; van Lenthe, E.; Baerends,
E. J .; Snijders, J . G. J . Chem. Phys. 1994, 101, 1272. (c) van Lenthe,
E.; Baerends, E. J .; Snijders, J . G. J . Chem. Phys. 1994, 101, 9783. (d)
Sadlej, A. J .; Snijders, J . G.; van Lenthe, E.; Baerends, E. J . J . Chem.
Phys. 1995, 102, 1758.
(61) te Velde, G.; Bickelhaupt, F. M.; van Gisbergen, S. J . A.; Fonseca
Guerra, C.; Baerends, E. J .; Snijders, J . G.; Ziegler, T. J . Comput.
Chem. 2001, 22, 931.
(62) Fonseca Guerra, C.; Snijders, J . G.; te Velde, G.; Baerends, E.
J . J . Theor. Chem. Acc. 1998, 99, 391.
(63) ADF2002.02, SCM, Theoretical Chemistry, Vrije Universiteit,
Amsterdam, The Netherlands, http://www.scm.com.
(65) Baerends, E. J .; Ellis, D. E.; Ros, P. Chem. Phys. 1973, 2,
41.
(66) Versluis, L.; Ziegler, T. J . Chem. Phys. 1988, 88, 322.
(67) te Velde, G.; Baerends, E. J . J . Comput. Phys. 1992, 99, 84.