Dinitrogen cleavage by three-coordinate molybdenum(III) complexes: Mechanistic and structural data

Article abstract of DOI:10.1021/ja960574x

The synthesis and characterization of the complexes Mo[N(R)Ar]₃ (R = C(CD₃)₂CH₃, Ar = 3,5-C₆H₃Me₂), (μ-N₂){Mo[N(R)Ar]₃}₂, (μ-¹⁵N₂){Mo[N(R)Ar]₃}₂, NMo[N(R)Ar]₃, ¹⁵NMo[N(R)Ar]₃, Mo[N(t-Bu)Ph]₃, (μ-N₂){Mo[N(t-Bu)Ph]₃}₂, and NMo[N(t-Bu)Ph]₃ are described. Temperature-dependent magnetic susceptibility data indicate a quartet ground state for Mo[N(R)Ar]₃. Single-crystal X-ray diffraction studies for Mo[N(R)Ar]₃ and NMo[N(t-Bu)Ph]₃ are described. Extended X-ray absorption fine structure (EXAFS) structural studies for Mo[N(R)AR]₃, (μN₂){Mo[N(R)Ar]₃}₂, and NMo[N(R)AR]₃ are reported. Temperature-dependent kinetic data are given for the unimolecular fragmentation of (μ-N₂){Mo[N(R)Ar]₃}₂ to 2 equiv of NMo[N(R)Ar]₃ and for the fragmentation of (μ-¹⁵N₂){Mo[N(R)AR]₃}₂ to 2 equiv of ¹⁵NMo[N(R)AR]₃. The temperature dependence of the ¹⁵N₂ isotope effect for the latter N₂ cleavage process was fitted to a simple harmonic model, leading to a prediction for the difference in NN stretching frequencies for the two isotopomers. The latter prediction was consistent with the Raman spectroscopic data for (μ-N₂){Mo[N(R)Ar]₃}₂ and (μ-¹⁵N₂){Mo[N(R)Ar]₃}₂. The Raman spectroscopic data and EXAFS results are both consistent with an NN bond order of approximately 2 in (μ-N₂){Mo[N(R)Ar]₃}₂. Temperature-dependent magnetic susceptibility data consistent with a triplet ground state are given for (μ-N₂){Mo[N(t-Bu)Ph]₃}₂.

Full text of DOI:10.1021/ja960574x

J. Am. Chem. Soc. 1996, 118, 8623-8638
8623
Dinitrogen Cleavage by Three-Coordinate Molybdenum(III)
Complexes: Mechanistic and Structural Data1
Catalina E. Laplaza,† Marc J. A. Johnson,† Jonas C. Peters,† Aaron L. Odom,†
Esther Kim,† Christopher C. Cummins,*,† Graham N. George,*,‡ and
Ingrid J. Pickering‡
Contribution from the Department of Chemistry, Massachusetts Institute of Technology,
77 Massachusetts AVenue, Cambridge, Massachusetts 02139-4307, and Stanford
Synchrotron Radiation Laboratory, SLAC, Stanford UniVersity, P.O. Box 4349, MS 69,
Stanford, California 94309-0210
ReceiVed February 22, 1996X
Abstract: The synthesis and characterization of the complexes Mo[N(R)Ar]3 (R ) C(CD3)2CH3, Ar ) 3,5-C6H3Me2),
(µ-N2){Mo[N(R)Ar]3}2, (µ-15N2){Mo[N(R)Ar]3}2, NMo[N(R)Ar]3, 15NMo[N(R)Ar]3, Mo[N(t-Bu)Ph]3, (µ-N2){Mo-
[N(t-Bu)Ph]3}2, and NMo[N(t-Bu)Ph]3 are described. Temperature-dependent magnetic susceptibility data indicate
a quartet ground state for Mo[N(R)Ar]3. Single-crystal X-ray diffraction studies for Mo[N(R)Ar]3 and NMo[N(t-
Bu)Ph]3 are described. Extended X-ray absorption fine structure (EXAFS) structural studies for Mo[N(R)Ar]3, (µ-
N2){Mo[N(R)Ar]3}2, and NMo[N(R)Ar]3 are reported. Temperature-dependent kinetic data are given for the
unimolecular fragmentation of (µ-N2){Mo[N(R)Ar]3}2 to 2 equiv of NMo[N(R)Ar]3 and for the fragmentation of
(µ-15N2){Mo[N(R)Ar]3}2 to 2 equiv of 15NMo[N(R)Ar]3. The temperature dependence of the 15N2 isotope effect for
the latter N2 cleavage process was fitted to a simple harmonic model, leading to a prediction for the difference in
NN stretching frequencies for the two isotopomers. The latter prediction was consistent with the Raman spectroscopic
data for (µ-N2){Mo[N(R)Ar]3}2 and (µ-15N2){Mo[N(R)Ar]3}2. The Raman spectroscopic data and EXAFS results
are both consistent with an NN bond order of approximately 2 in (µ-N2){Mo[N(R)Ar]3}2. Temperature-dependent
magnetic susceptibility data consistent with a triplet ground state are given for (µ-N2){Mo[N(t-Bu)Ph]3}2.
Introduction
details relevant to the NN bond cleavage step. Synthetic,
structural, and magnetic data are given for three-coordinate
Mo[N(R)Ar]3, along with synthetic and characterization data
for a new analog, Mo[N(t-Bu)Ph]3. Structural characterization
by extended X-ray absorption fine structure (EXAFS)10 of the
thermally-unstable purple intermediate, (µ-N2){Mo[N(R)Ar]3}2
is reported. Also given are EXAFS data for three-coordinate
Mo[N(R)Ar]3 and the nitrido complex NMo[N(R)Ar]3. The
synthesis and single-crystal X-ray structure of the new terminal
nitrido complex NMo[N(t-Bu)Ph]3 are described. Temperature-
dependent kinetic studies bearing on the conversion of purple
(µ-N2){Mo[N(R)Ar]3}2 to 2 equiv of NMo[N(R)Ar]3 are
reported and interpreted. Temperature-dependent kinetic studies
It was shown recently that the three-coordinate molyb-
denum(III) complex Mo[N(R)Ar]3 [R ) C(CD3)2CH3, Ar ) 3,5-
C6H3Me2]2 reacts readily with dinitrogen (1 atm, -35 to 30
°C) to provide the terminal nitrido complex NMo[N(R)Ar]3,3
via an intermediate formulated as (µ-N2){Mo[N(R)Ar]3}2
(Scheme 1). Experiments utilizing 15N2 established dinitrogen
as the origin of the terminal nitrido nitrogen atom. The reaction
(Scheme 1) is without precedent in its operational simplicity,4
and the details of its mechanism are of inherent interest.5 This
molybdenum-based N2 cleavage system is amenable to molec-
ular-level scrutiny of both the N2-binding and NN bond cleavage
processes, rendering it unique among those nitrogenase models
that synthetic chemists have yet developed.6-8 Given the intense
scrutiny to which dinuclear µ-dinitrogen complexes have been
subjected,9 it is gratifying that an N2-cleavage reaction clearly
stemming from such a species is now available for study.
This paper is concerned with those details of N2 cleavage by
Mo[N(R)Ar]3 that are currently in hand, focusing primarily on
of the conversion of (µ-15N2){Mo[N(R)Ar]3}2 to 2 equiv of 15
-
NMo[N(R)Ar]3, carried out to determine the heavy-atom isotope
effect associated with the NN-cleavage step, are reported. Also
described is a Raman spectroscopic study of purple (µ-N2){Mo-
[N(R)Ar]3}2 and its isotopomer (µ-15N2){Mo[N(R)Ar]3}2, car-
ried out to address the issue of the NN bond order in these
species independent of the EXAFS information. The earlier
assertion (based on 2H NMR data) that purple (µ-N2){Mo[N(R)-
Ar]3}2 is paramagnetic with a triplet ground state3 is here
substantiated via a temperature-dependent study of the magnetic
susceptibility of (µ-N2){Mo[N(t-Bu)Ph]3}2.
† Massachusetts Institute of Technology.
‡ Stanford University.
X Abstract published in AdVance ACS Abstracts, August 1, 1996.
(1) Dedicated to Professor Alan Davison of the MIT Department of
Chemistry on the occasion of his 60th birthday.
(2) Laplaza, C. E.; Odom, A. L.; Davis, W. M.; Cummins, C. C.;
Protasiewicz, J. D. J. Am. Chem. Soc. 1995, 117, 4999. Laplaza, C. E.;
Davis, W. M.; Cummins, C. C. Angew. Chem., Int. Ed. Engl. 1995, 34,
2042.
(3) Laplaza, C. E.; Cummins, C. C. Science 1995, 268, 861. See also:
Laplaza, C. E.; Johnson, A. R.; Cummins, C. C. J. Am. Chem. Soc. 1996,
118, 709.
(6) Elegant structural models for the nitrogenase active site have been
synthesized: Nordlander, E.; Lee, S. C.; Cen, W.; Wu, Z. Y.; Natoli, C.
R.; Di Cicco, A.; Filipponi, A.; Hedman, B.; Hodgson, K. O.; Holm, R. H.
J. Am. Chem. Soc. 1993, 115, 5549. For general reviews of transition metal
dinitrogen complexes, see: Allen, A. D.; Harris, R. O.; Loescher, B. R.;
Stevens, J. R.; Whiteley, R. N. Chem. ReV. 1973, 73, 11. Chatt, J.; Dilworth,
J. R.; Richards, R. L. Chem. ReV. 1978, 78, 589. Henderson, R. A.; Leigh,
G. J.; Pickett, C. J. AdV. Inorg. Chem. Radiochem. 1983, 27, 197. Leigh,
G. J. Acc. Chem. Res. 1992, 25, 177. Leigh, G. J. New J. Chem. 1994, 18,
157. Hidai, M.; Mizobe, Y. Chem. ReV. 1995, 95, 1115.
(4) Leigh, G. J. Science 1995, 268, 827.
(5) The mechanism of dinitrogen cleavage by d3 molybdenum and
tungsten complexes was the subject of a recent theoretical study: Cui, Q.;
Musaev, D. G.; Svensson, M.; Sieber, S.; Morokuma, K. J. Am. Chem.
Soc. 1995, 117, 12366.
S0002-7863(96)00574-4 CCC: $12.00 © 1996 American Chemical Society

8624 J. Am. Chem. Soc., Vol. 118, No. 36, 1996
Laplaza et al.
Results and Discussion
mer-MoCl3(THF)321 with 2 equiv of Li[N(R)Ar](OEt2)22 in ether
(see the Experimental Section for preparative details). To avoid
contamination with the dinuclear dinitrogen complex (µ-N2)-
{Mo[N(R)Ar]3}2 or with NMo[N(R)Ar]3, Mo[N(R)Ar]3 is
purified by recrystallization under an argon atmosphere at -35
°C. A typical preparation yields ca. 1.5 g of Mo[N(R)Ar]3. The
compound exhibits a diagnostic H NMR resonance at ca. 64
ppm for its deuterated tert-butyl groups. The proton NMR
spectrum of Mo[N(R)Ar]3 exhibits the expected four broad
(i) Design, Synthesis, and Characterization of Mo[N(R)-
Ar]3. The choice of the N[R]Ar ligand for the synthesis of an
isolable, three-coordinate molybdenum(III) complex was pre-
mised on a number of considerations. N-tert-Butylarylamido
ligands such as N[R]Ar are sterically11 and electronically12
comparable to the bis(trimethylsilyl)amido ligand. The bis-
(trimethylsilyl)amido ligand is noteworthy for the large number
of three-coordinate M[N(SiMe3)2]3 complexes it stabilizes both
for the first-row transition metals13 and for uranium.14 Nonethe-
less, the bis(trimethylsilyl)amido ligand has well-documented
shortcomings as an ancillary ligand, including its susceptibility
to cyclometalation15 and SiN bond cleavage16 processes. The
latter Achilles’ heels are presumably responsible, at least in part,
for the conspicuous absence of any second- or third-row
transition metal M[N(SiMe3)2]3 complexes. Radical ligand
degradation processes should also be considered in this context.17
It was our hope that N[R]Ar and related ligands would emulate
the beneficial aspects accordant to N(SiMe3)2 without manifest-
ing its shortcomings. A potential further benefit of N-tert-
butylarylamido ligands is the wide range of substituted anilines
available for their preparation,18 which should permit electronic
tuning of reactivity.19 Finally, the ready incorporation of a
deuterium NMR handle into the N[R]Ar ligand has facilitated
the preparation of paramagnetic complexes including Mo[N(R)-
Ar]3, by virtue of the fact that, for paramagnetic complexes, 2H
2
(9) Dititanium µ-dinitrogen complexes: (a) Berry, D. H.; Procopio, L.
J.; Carroll, P. J. Organometallics 1988, 7, 570. (b) Duchateau, R.;
Gambarotta, S.; Beydoun, N.; Bensimon, C. J. Am. Chem. Soc. 1991, 113,
8986. (c) Beydoun, N.; Duchateau, R.; Gambarotta, S. J. Chem. Soc., Chem.
Commun. 1992, 244. (d) Sanner, R. D.; Duggan, D. M.; McKenzie, T. C.;
Marsh, R. E.; Bercaw, J. E. J. Am. Chem. Soc. 1976, 98, 8358. (e) Zeinstra,
J. D.; Teuben, J. H.; Jellinek, F. J. Organomet. Chem. 1979, 170, 39.
Dizirconium µ-dinitrogen complexes: (f) Sanner, R. D.; Manriquez, J. M.;
Marsh, R. E.; Bercaw, J. E. J. Am. Chem. Soc. 1976, 98, 8351. (g) Fryzuk,
M. D.; Haddad, T. S.; Mylvaganam, M.; McConville, D. H.; Rettig, S. J.
J. Am. Chem. Soc. 1993, 115, 2782. (h) Cohen, J. D.; Mylvaganam, M.;
Fryzuk, M. D.; Loehr, T. M. J. Am. Chem. Soc. 1994, 116, 9529.
Divanadium µ-dinitrogen complexes: (i) Edema, J. J. H.; Meetsma, A.;
Gambarotta, S. J. Am. Chem. Soc. 1989, 111, 6878. (j) Song, J.-I.; Berno,
P.; Gambarotta, S. J. Am. Chem. Soc. 1994, 116, 6927. (k) Berno, P.; Hao,
S.; Minhas, R.; Gambarotta, S. J. Am. Chem. Soc. 1994, 116, 7417. (l)
Buijink, J.-K. F.; Meetsma, A.; Teuben, J. H. Organometallics 1993, 12,
2004. (m) Ferguson, R.; Solari, E.; Floriani, C.; Chiesi-Villa, A.; Rizzoli,
C. Angew. Chem., Int. Ed. Engl. 1993, 32, 396. Diniobium µ-dinitrogen
complexes: (n) Dilworth, J. R.; Henderson, R. A.; Hills, A.; Hughes, D.
L.; Macdonald, C.; Stephens, A. N.; Walton, D. R. M. J. Chem. Soc., Dalton
Trans. 1990, 1077. (o) Berno, P.; Gambarotta, S. Organometallics 1995,
14, 2159. Ditantalum µ-dinitrogen complexes: (p) Turner, H. W.; Fellmann,
J. D.; Rocklage, S. M.; Schrock, R. R.; Churchill, M. R.; Wasserman, H.
J. J. Am. Chem. Soc. 1980, 102, 7809. (q) Rocklage, S. M.; Turner, H. W.;
Fellmann, J. D.; Schrock, R. R. Organometallics 1982, 1, 703. (r) Rocklage,
S. M.; Schrock, R. R. J. Am. Chem. Soc. 1982, 104, 3077. (s) Churchill,
M. R.; Wasserman, H. J. Inorg. Chem. 1981, 20, 2899. (t) Churchill, M.
R.; Wasserman, H. J. Inorg. Chem. 1982, 21, 218. (u) Schrock, R. R.;
Wesolek, M.; Liu, A. H.; Wallace, K. C.; Dewan, J. C. Inorg. Chem. 1988,
27, 2050. Dichromium µ-dinitrogen complexes: (v) Denholm, S.; Hunter,
G.; Weakley, T. J. R. J. Chem. Soc., Dalton Trans. 1987, 2789. Dimolyb-
denum µ-dinitrogen complexes: (w) Forder, R. A.; Prout, K. Acta
Crystallogr. 1974, B30, 2778. (x) Schrock, R. R.; Kolodziej, R. M.; Liu,
A. H.; Davis, W. M.; Vale, M. G. J. Am. Chem. Soc. 1990, 112, 4338. (y)
Kol, M.; Schrock, R. R.; Kempe, R.; Davis, W. M. J. Am. Chem. Soc.
1994, 116, 4382. (z) Shih, K.-Y.; Schrock, R. R.; Kempe, R. J. Am. Chem.
Soc. 1994, 116, 8804. (aa) Luo, X.-L.; Kubas, G. J.; Burns, C. J.; Butcher,
R. J.; Bryan, J. C. Inorg. Chem. 1995, 34, 6538. Ditungsten µ-dinitrogen
complexes: (bb) Anderson, S. N.; Richards, R. L.; Hughes, D. L. J. Chem.
Soc., Chem. Commun. 1982, 1291. (cc) Anderson, S. N.; Richards, R. L.;
Hughes, D. L. J. Chem. Soc., Dalton Trans. 1986, 245. (dd) Churchill, M.
R.; Li, Y.-J.; Theopold, K. H.; Schrock, R. R. Inorg. Chem. 1984, 23, 4472.
(ee) Murray, R. C.; Schrock, R. R. J. Am. Chem. Soc. 1985, 107, 4557. (ff)
Churchill, M. R.; Li, Y.-J. J. Organomet. Chem. 1986, 301, 49. (gg)
O’Regan, M. B.; Liu, A. H.; Finch, W. C.; Schrock, R. R.; Davis, W. M.
J. Am. Chem. Soc. 1990, 112, 4331. (hh) Harlan, C. J.; Jones, R. A.;
Koschmieder, S. U.; Nunn, C. M. Polyhedron 1990, 9, 669. Dimanganese
and ditechnetium µ-dinitrogen complexes: (ii) Weidenhammer, K.; Herr-
mann, W. A.; Ziegler, M. L. Z. Anorg. Allg. Chem. 1979, 457, 183. (jj)
Joachim, J. E.; Apostolidis, C.; Kanellakopulos, B.; Maier, R.; Meyer, D.;
Rebizant, J.; Ziegler, M. L. J. Organomet. Chem. 1993, 455, 137. Dinuclear
µ-dinitrogen complexes of iron, ruthenium, or osmium: (kk) Berke, H.;
Bankhardt, W.; Huttner, G.; von Seyerl, J.; Zsolnai, L. Chem. Ber. 1981,
114, 2754. (ll) Che, C.-M.; Lam, H.-W.; Tong, W.-F.; Lai, T.-F.; Lau, T.-
C. J. Chem. Soc., Chem. Commun. 1989, 1883. (mm) Sellmann, D.;
Friedrich, H.; Knoch, F.; Moll, M. Z. Naturforsch. 1994, 49b, 76. Dinuclear
µ-dinitrogen complexes of cobalt, rhodium, or iridium: (nn) Cecconi, F.;
Ghilardi, C. A.; Midollini, S.; Moneti, S.; Orlandini, A.; Bacci, M. J. Chem.
Soc., Chem. Commun. 1985, 731. (oo) Klein, H.-F.; Beck, H.; Hammer-
schmitt, B.; Koch, U.; Koppert, S.; Cordier, G. Z. Naturforsch. 1991, B46,
147. (pp) Yoshida, T.; Okano, T.; Thorn, D. L.; Tulip, T. H.; Otsuka, S.;
Ibers, J. A. J. Organomet. Chem. 1979, 181, 183. (qq) Gutie´rrez, E.; Monge,
A.; Nicasio, M. C.; Poveda, M. L.; Carmona, E. J. Am. Chem. Soc. 1994,
116, 791. Dinickel µ-dinitrogen complexes: (rr) Jolly, P. W.; Jonas, K.;
Kru¨ger, C.; Tsay, Y.-H. J. Organomet. Chem. 1971, 33, 109. Mixed-metal
µ-dinitrogen complexes: (ss) Mizobe, Y.; Yokobayashi, Y.; Oshita, H.;
Takahashi, T.; Hidai, M. Organometallics 1994, 13, 3764. (tt) Mercer, M.;
Crabtree, R. H.; Richards, R. L. J. Chem. Soc., Chem. Commun. 1973, 808.
(uu) Mercer, M. J. Chem. Soc., Dalton Trans. 1974, 1637; see also ref 9x
above.
1
NMR lines are substantially narrower than corresponding H
NMR signals.20 Complexes containing the N[R]Ar ligand
exhibit one 2H NMR signal for each unique ligand environment,
2
making H NMR a rapid and convenient method for assaying
paramagnetic reaction mixtures.
As communicated previously,2 Mo[N(R)Ar]3 is obtained as
an orange-red crystalline solid in 70% yield upon treatment of
(7) Recent reports involving mononuclear molybdenum and tungsten
dinitrogen complexes: Oshita, H.; Mizobe, Y.; Hidai, M. J. Organomet.
Chem. 1993, 456, 213. Seino, H.; Ishii, Y.; Hidai, M. J. Am. Chem. Soc.
1994, 116, 7433. Ishii, Y.; Kawaguchi, M.; Ishino, Y.; Aoki, T.; Hidai, M.
Organometallics 1994, 13, 5062. Baker, R. T.; Calabrese, J. C.; Harlow,
R. L.; Williams, I. D. Organometallics 1993, 12, 830. Ferna´ndez-Trujillo,
M. J.; Basallote, M. G.; Valerga, P.; Puerta, M. C. J. Chem. Soc., Dalton
Trans. 1993, 923. Jime´nez-Tenorio, M.; Puerta, M. C.; Valerga, P.; Hughes,
D. L. J. Chem. Soc., Dalton Trans. 1994, 2431. Hughes, D. L.; Ibrahim, S.
K.; Pickett, C. J.; Querne, G.; Laouenan, A.; Talarmin, J.; Queiros, A.;
Fonseca, A. Polyhedron 1994, 13, 3341. George, T. A.; Kaul, B. B.; Chen,
Q.; Zubieta, J. Inorg. Chem. 1993, 32, 1706. Dzie¸gielewski, J. O.; Małecki,
J. G.; Grzybek, R. Polyhedron 1994, 13, 241. Luo, X.-L.; Kubas, G. J.;
Burns, C. J.; Eckert, J. Inorg. Chem. 1994, 33, 5219. Recent reports
involving mononuclear rhenium dinitrogen complexes: Orth, S. D.; Barrera,
J.; Sabat, M.; Harman, W. D. Inorg. Chem. 1994, 33, 3026. Dilworth, J.
R.; Hu, J.; Miller, J. R.; Hughes, D. L.; Zubieta, J. A.; Chen, Q. J. Chem.
Soc., Dalton Trans. 1995, 3153. Cusanelli, A.; Sutton, D. Organometallics
1995, 14, 4651. Richards, T. C.; Bard, A. J.; Cusanelli, A.; Sutton, D.
Organometallics 1994, 13, 757. Wang, Y.; Da Silva, J. J. R. F.; Pombeiro,
A. J. L.; Pellinghelli, M. A.; Tiripicchio, A. J. Organomet. Chem. 1993,
454, 211. Recent reports involving mononuclear iron dinitrogen com-
plexes: Hills, A.; Hughes, D. L.; Jime´nez-Tenorio, M.; Leigh, G. J.; Rowley,
A. T. J. Chem. Soc., Dalton Trans. 1993, 3041. Hughes, D. L.; Leigh, G.
J.; Jime´nez-Tenorio, M.; Rowley, A. T. J. Chem. Soc., Dalton Trans. 1993,
75. Perthuisot, C.; Jones, W. D. New J. Chem. 1994, 18, 621. Komiya, S.;
Akita, M.; Yoza, A.; Kasuga, N.; Fukuoka, A.; Kai, Y. J. Chem. Soc., Chem.
Commun. 1993, 787. de la Jara Leal, A.; Jime´nez-Tenorio, M.; Puerta, M.
C.; Valerga, P. Organometallics 1995, 14, 3839. Recent reports involving
mononuclear ruthenium and osmium dinitrogen complexes: Li, Z.-W.;
Harman, W. D.; Lay, P. A.; Taube, H. Inorg. Chem. 1994, 33, 3635. Filipek,
K. Inorg. Chim. Acta 1995, 231, 237. Bianchini, C.; Linn, K.; Masi, D.;
Peruzzini, M.; Polo, A.; Vacca, A.; Zanobini, F. Inorg. Chem. 1993, 32,
2366. Rao, N. N. J. Mol. Catal. 1994, 93, 23. Takahashi, T.; Hiratani, K.;
Kimura, E. Chem. Lett. 1993, 1329.
(8) Catalytic nitrogen-fixing systems: Antipin, M.; Struchkov, Y.; Shilov,
A.; Shilova, A. Gazz. Chim. Ital. 1993, 123, 265. Luneva, N. P.; Mironova,
S. A.; Shilov, A. E.; Antipin, M. Y.; Struchkov, Y. T. Angew. Chem., Int.
Ed. Engl. 1993, 32, 1178. Shilov, A. E. Pure Appl. Chem. 1992, 64, 1409.

Dinitrogen CleaVage by Molybdenum(III) Complexes
J. Am. Chem. Soc., Vol. 118, No. 36, 1996 8625
Scheme 1
resonances, the sharpest of which is found at -9.6 ppm (C6D6,
ca. 25 °C), which we assign to the meta methyl groups on the
Ar residues. Solution magnetic susceptibility measurements
obtained by the method of Evans23 are consistent with a quartet
ground state for Mo[N(R)Ar]3 [µeff ) 3.56 µB].2 SQUID
magnetic susceptibility measurements for solid Mo[N(R)Ar]3
yield a magnetic moment very close to the spin-only value for
(10) Cramer, S. P.; Hodgson, K. O. Prog. Inorg. Chem. 1979, 25, 1.
(11) Bradley has articulated the concept of steric saturation of the primary
coordination sphere: Bradley, D. C. Chem. Br. 1975, 11, 393. Eller, P. G.;
Bradley, D. C.; Hursthouse, M. B.; Meek, D. W. Coord. Chem. ReV. 1977,
24, 1.
(12) The NH pKa for HN(SiMe3)2 has been given as 25.8 in tetrahydro-
furan as ascertained using 13C NMR methods; see: Fraser, R. R.; Mansour,
T. S. J. Org. Chem. 1984, 49, 5284. Fraser, R. R.; Mansour, T. S. J. Org.
Chem. 1984, 49, 3442.
(13) (a) Lappert, M. F.; Power, P. P.; Sanger, A. R.; Srivastava, R. C.
Metal and Metalloid Amides; Ellis Horwood: Chichester, 1980. (b) Bradley,
D. C.; Copperthwaite, R. G. Inorg. Synth. 1978, 18, 112. (c) Bradley, D.
C.; Hursthouse, M. B.; Newing, C. W.; Welch, A. J. J. Chem. Soc., Chem.
Commun. 1972, 567. (d) Alyea, E. C.; Bradley, D. C.; Copperthwaite, R.
G. J. Chem. Soc., Dalton Trans. 1972, 1580. (e) Bradley, D. C.; Newing,
C. W. J. Chem. Soc., Chem. Commun. 1970, 219. (f) Bradley, D. C.;
Hursthouse, M. B.; Rodesiler, P. F. J. Chem. Soc., Chem. Commun. 1969,
14. (g) Ellison, J. J.; Power, P. P.; Shoner, S. C. J. Am. Chem. Soc. 1989,
111, 8044. (h) Bu¨rger, H.; Wannagat, U. Monatsh. Chem. 1964, 95, 1099.
(i) Bu¨rger, H.; Wannagat, U. Monatsh. Chem. 1963, 94, 1007.
(14) (a) Andersen, R. A. Inorg. Chem. 1979, 18, 1507. (b) Van Der Sluys,
W. G.; Burns, C. J.; Huffman, J. C.; Sattelberger, A. P. J. Am. Chem. Soc.
1988, 110, 5924. (c) Avens, L. R.; Bott, S. G.; Clark, D. L.; Sattelberger,
A. P.; Watkin, J. G.; Zwick, B. D. Inorg. Chem. 1994, 33, 2248.
(15) Simpson, S. J.; Andersen, R. A. Inorg. Chem. 1981, 20, 3627.
Simpson, S. J.; Andersen, R. A. Inorg. Chem. 1981, 20, 2991. Berno, P.;
Gambarotta, S. Angew. Chem., Int. Ed. Engl. 1995, 34, 822. Berno, P.;
Minhas, R.; Hao, S.; Gambarotta, S. Organometallics 1994, 13, 1052.
(16) Bu¨rger, H.; Wannagat, U. Monatsh. Chem. 1963, 94, 761.
(17) We have encountered the following radical bond cleavage proc-
esses: MoVOR f MVIO + [•R] and MoVNRAr f MVINAr + [•R], to be
reported in due course. Peters, J.; Wanandi, P. W.; Odom, A. L.; Davis,
W. M.; Cummins, C. C. J. Am. Chem. Soc., in press. Johnson, A. R.; Odom,
A. L.; Cummins, C. C. Manuscript in preparation.
Figure 1. SQUID magnetic susceptibility data for solid Mo(N[R]Ar)3
from 5 to 300 K fit to the Curie-Weiss law (µ ) 3.82 µB; see
Experimental Section for details).
three unpaired electrons (µs ) 3.87 µB): the best least-squares
fit of the data to the Curie-Weiss law gave µ ) 3.82 µB over
the temperature range 5-300 K (Figure 1; see Experimental
Section for details).
The new three-coordinate molybdenum(III) complex Mo[N(t-
Bu)Ph]3, obtained in ca. 55% yield as burgundy crystals, was
prepared in a fashion analogous to Mo[N(R)Ar]3 (see Experi-
mental Section for details). 1H NMR, magnetic susceptibility,
and chemical reactivity data (Vide infra) are consistent with
formulation of Mo[N(t-Bu)Ph]3 as a three-coordinate molyb-
denum(III) monomer.
(ii) Solid-State and Solution Structure of Mo[N(R)Ar]3.
The X-ray crystal structure of Mo[N(R)Ar]3 was communicated
previously in conjunction with a study on its cleavage of the
nitrous oxide NN bond.2 Salient features of the structure that
should be borne in mind when considering reactions of this
(18) The 1992-1993 edition of the Aldrich Structure Index lists ca. 100
variously substituted anilines that conceivably could be employed in the
synthesis of N-tert-butyl arylamido ligands.
(19) Schrock and co-workers have demonstrated exquisite control over
the reactivity of molybdenum metathesis catalysts, of general formula Mo-
(NAr)(CHR)(OR′)2, by varying the alkoxide substituent. See: Schrock, R.
R. Pure Appl. Chem. 1994, 66, 1447 and references therein.
(20) La Mar, G. N.; Horrocks, W. DeW., Jr.; Holm, R. H. NMR of
Paramagnetic Molecules; Academic Press: New York, 1973. Johnson, A.;
Everett, G. W., Jr. J. Am. Chem. Soc. 1972, 94, 1419. Wheeler, W. D.;
Kaizaki, S.; Legg, J. I. Inorg. Chem. 1982, 21, 3248. Hill, D. H.; Parvez,
M. A.; Sen, A. J. Am. Chem. Soc. 1994, 116, 2889. Hill, D. H.; Sen, A. J.
Am. Chem. Soc. 1988, 110, 1650. Li, Z.; Goff, H. M. Inorg. Chem. 1992,
31, 1547.
(21) We prepare orange mer-MoCl3(THF)3 according to Dilworth and
Zubieta: Dilworth, J. R.; Zubieta, J. Inorg. Synth. 1986, 24, 193. For a
paramagnetic H NMR study of the solution stability of monomeric mer-
MoCl3(THF)3, see: Poli, R.; Mui, H. D. J. Am. Chem. Soc. 1990, 112,
2446. For the single-crystal X-ray structure of mer-MoCl3(THF)3, see:
Hofacker, P.; Friebel, C.; Dehnicke, K.; Ba¨uml, P.; Hiller, W.; Stra¨hle, J.
Z. Naturforsch. 1989, 44b, 1161.
(22) Laplaza, C. E.; Davis, W. M.; Cummins, C. C. Organometallics
1995, 14, 577.
1
(23) Sur, S. K. J. Magn. Reson. 1989, 82, 169. Evans, D. F. J. Chem.
Soc. 1959, 2003.

8626 J. Am. Chem. Soc., Vol. 118, No. 36, 1996
Laplaza et al.
Table 1. Selected Bond Distances (Å) and Angles (deg) for Mo(N[R]Ar)3
Mo-N(1)
1.960(7)
1.964(7)
1.977(7)
1.42(1)
1.50(1)
1.44(1)
1.49(1)
1.43(1)
1.49(1)
Mo(A)-N(1A)
Mo(A)-N(2A)
Mo(A)-N(3A)
N(1A)-C(11A)
N(1A)-C(17A)
N(2A)-C(21A)
N(2A)-C(27A)
N(3A)-C(31A)
N(3A)-C(37A)
1.970(7)
Mo-N(2)
1.968(7)
1.956(7)
1.42(1)
1.48(1)
1.43(1)
1.49(1)
1.43(1)
1.48(1)
Mo-N(3)
N(1)-C(11)
N(1)-C(17)
N(2)-C(21)
N(2)-C(27)
N(3)-C(31)
N(3)-C(37)
N(1)-Mo-N(2)
N(1)-Mo-N(3)
N(2)-Mo-N(3)
Mo-N(1)-C(11)
Mo-N(1)-C(17)
C(11)-N(1)-C(17)
Mo-N(2)-C(21)
Mo-N(2)-C(27)
C(21)-N(2)-C(27)
Mo-N(3)-C(31)
Mo-N(3)-C(37)
C(31)-N(3)-C(37)
120.4(3)
118.0(3)
119.3(3)
110.2(5)
131.7(6)
118.1(7)
111.5(5)
130.7(6)
117.7(7)
108.9(5)
132.0(6)
119.0(7)
N(1A)-Mo(A)-N(2A)
N(1A)-Mo(A)-N(3A)
N(2A)-Mo(A)-N(3A)
Mo(A)-N(1A)-C(11A)
Mo(A)-N(1A)-C(17A)
C(11A)-N(1A)-C(17A)
Mo(A)-N(2A)-C(21A)
Mo(A)-N(2A)-C(27A)
C(21A)-N(2A)-C(27A)
Mo(A)-N(3A)-C(31A)
Mo(A)-N(3A)-C(37A)
C(31A)-N(3A)-C(37A)
120.3(3)
120.0(3)
118.9(3)
108.6(6)
132.1(6)
118.8(8)
110.2(6)
130.7(6)
119.1(7)
110.4(6)
131.8(6)
117.8(7)
compound with small molecules include (i) a trigonal planar
MoN3 unit, (ii) amido NC2 planes roughly perpendicular to the
MoN3 plane, (iii) tert-butyl groups occupying one hemisphere
and aryl groups the other, and (iv) aryl residues oriented roughly
perpendicular to their contiguous MoNC2 planes (See Table 1
for selected bond lengths and angles). An ORTEP diagram and
space-filling models, depicted in Figure 2, suggest that the most
accessible approach to the Mo atom is a shallow channel
bounded by the three tert-butyl groups (Figure 2c). A second
channel, bounded by the three aryl groups (Figure 2d), presents
a much deeper path to the Mo atom and should thus appear
less enticing to an incoming small molecule. Three axial clefts
(Figure 2b), which appear impossibly narrow, offer the only
other access to the Mo atom. An expansion of the tert-butyl-
bounded cavity upon penetration by a small molecule (e.g., N2
or N2O) appears to be required, on the basis of van der Waals
radii considerations. Such an expansion should be accompanied
by contraction of the aryl-bounded cavity on the opposing
hemisphere, as discussed below in connection with the X-ray
structure of NMo[N(t-Bu)Ph]3. NMR (Vide supra) and EXAFS
(Vide infra) data for Mo[N(R)Ar]3 are consistent with an average
3-fold symmetric structure for the molecule in solution, as
observed in the solid state. We have no evidence for a “two-
up, one-down” structure as has been observed for trigonal
bipyramidal tris(arylthiolate)ML2 complexes (M ) Tc, Re; L
) e.g., acetonitrile).24 Edge-to-face aryl interactions could
contribute to the apparent preference for the observed “three-
up” structure: metrical parameters for the lowest energy gas-
phase structure of the benzene trimer25 are strikingly similar to
those for three aromatic rings oriented as in Mo[N(R)Ar]3. A
preliminary X-ray structure carried out for Mo[N(t-Bu)Ph]3 has
revealed that it is quite similar in shape to Mo[N(R)Ar]3;26 thus,
no special structural significance should be attributed to the aryl
methyl substituents in the latter complex.
(iii) Dinitrogen Binding by Mo[N(R)Ar]3. The uptake of
dinitrogen (at ca. 1 atm) by Mo[N(R)Ar]3 (benzene, 28 °C, 3.5
mM) is a slow process, with e5% conversion to NMo[N(R)-
Ar]3 observed in 12 h. The intermediate purple complex (µ-
N2){Mo[N(R)Ar]3}2 is not observed when the reaction is carried
out at ca. 28 °C, presumably because at that temperature it
converts rapidly to NMo[N(R)Ar]3. A typical procedure for
generating samples rich in purple (µ-N2){Mo[N(R)Ar]3}2 in-
volves storage of a ca. 0.05 M toluene solution of Mo[N(R)-
Ar]3 in a glovebox refrigerator at ca. -35 °C for several (3-8)
days. During this incubation period, the solution turns royal
2
purple and, according to H NMR monitoring, a single new
species proposed to be (µ-N2){Mo[N(R)Ar]3}2 (δ ) 14 ppm)
appears. Upon warming to 28 °C the purple color soon gives
way to the amber color of NMo[N(R)Ar]3, which according to
2
1H and H NMR analysis forms quantitatively, and which has
been isolated from such reaction mixtures in 76% recrystallized
yield. Solutions of Mo[N(R)Ar]3 do not acquire a purple color,
2
and do not exhibit the characteristic H NMR resonance at 14
ppm, when cooled under argon instead of dinitrogen. The use
of 1 atm of 99%-15N2 led to smooth production of the labeled
complex 15NMo[N(R)Ar]3, which has been characterized by
infrared spectroscopy [ν(MoN) ) 1014 cm-1 versus 1042 cm-1
for its unlabeled counterpart] and 15N NMR spectroscopy (δ )
+840 ppm27). Essentially identical spectroscopic properties
were observed for 15NMo[N(R)Ar]3 prepared independently
from Mo[N(R)Ar]3 and selectively-labeled 15NNN(p-C6H4Me).2
The putative mononuclear dinitrogen complex (η1-N2)Mo[N(R)-
Ar]3 (Scheme 1), the logical immediate precursor to (µ-N2)-
{Mo[N(R)Ar]3}2 has not been observed. Apparently the
equilibrium constant for dinitrogen capture by Mo[N(R)Ar]3 is
small under the conditions utilized.
(24) de Vries, N.; Dewan, J. C.; Jones, A. G.; Davison, A. Inorg. Chem.
1988, 27, 1574. de Vries, N.; Jones, A. G.; Davison, A. Inorg. Chem. 1989,
28, 3728. Blower, P. J.; Dilworth, J. R.; Hutchinson, J. P.; Zubieta, J. A. J.
Chem. Soc., Dalton Trans. 1985, 1533. Bishop, P. T.; Dilworth, J. R.;
Hutchinson, J.; Zubieta, J. J. Chem. Soc., Dalton Trans. 1986, 967. Dilworth,
J. R.; Hutchinson, J.; Zubieta, J. A. J. Chem. Soc., Chem. Commun. 1983,
1034. Koch, S. A.; Millar, M. J. Am. Chem. Soc. 1983, 105, 3362. For a
“three-up” tris(arylthiolate) complex with acetonitrile lodged in the pocket
bounded by aromatic residues, see: Dilworth, J. R.; Neaves, B. D.;
Hutchinson, J. P.; Zubieta, J. A. Inorg. Chim. Acta 1982, 65, L223.
(25) The following papers deal with aspects of the structure and energetics
of the benzene trimer: Kratzschmar, O.; Selzle, H. L.; Schlag, E. W. J.
Phys. Chem. 1994, 98, 3501. Xiao, Y. L.; Williams, D. E. Chem. Phys.
Lett. 1993, 215, 17. Henson, B. F.; Venturo, V. A.; Hartland, G. V.; Felker,
P. M. J. Chem. Phys. 1993, 98, 8361. Krause, H.; Ernstberger, B.; Neusser,
H. J. Ber. Bunsen-Ges. Phys. Chem. 1992, 96, 1183.
(iv) X-ray Absorption Near-Edge Spectra and EXAFS of
Mo[N(R)Ar]3, (µ-N2){Mo[N(R)Ar]3}2, and NMo[N(R)Ar]3.
The X-ray absorption near-edge spectra of Mo[N(R)Ar]3, (µ-
N2){Mo[N(R)Ar]3}2, and NMo[N(R)Ar]3 are shown in Figure
3. All of the spectra exhibit a relatively broad edge, as expected
(26) Johnson, M. J. A.; Odom, A. L.; Cummins, C. C. Manuscript in
preparation.
(27) The 15N NMR shift is quoted with reference to external neat
nitromethane, whose chemical shift is +380.2 ppm with respect to liquid
NH3 (taken as 0 ppm): von Philipsborn, W.; Mu¨ller, R. Angew. Chem.,
Int. Ed. Engl. 1986, 25, 383.

Dinitrogen CleaVage by Molybdenum(III) Complexes
J. Am. Chem. Soc., Vol. 118, No. 36, 1996 8627
a
b
c
d
Figure 2. (a) Structural diagram and atom-labeling scheme for Mo(N[R]Ar)3. Thermal ellipsoids are at the 35% probability level. Space-filling
representations of Mo(N[R]Ar)3 viewed (b) perpendicular to the N3 plane normal, (c) along the N3 plane normal facing the hemisphere protected
by tert-butyl-d6 groups, and (d) along the N3 plane normal facing the hemisphere protected by 3,5-dimethylphenyl groups.
observed as a well-defined peak at 20 005.9 eV. The pre-edge
feature in (µ-N2){Mo[N(R)Ar]3}2 and Mo[N(R)Ar]3 may be
assigned to a 1s f 4d transition, which is formally dipole-
forbidden and weakly quadrupole-allowed but gains intensity
in a non-centrosymmetric molybdenum center by admixture of
metal p-orbitals. This transition is analogous to the well-known
“oxo-edge” feature of molybdenum enzymes, said to be
characteristic of a species possessing terminal MoO groups.28
The strong appearance of the 1s f 4d feature in the spectrum
of NMo[N(R)Ar]3 is highly consistent with the occurrence of a
very short MoN triple bond. The intermediate appearance of
the spectrum for (µ-N2){Mo[N(R)Ar]3}2 is therefore consistent
with an additional MoN bond somewhat longer than the triple
bond of NMo[N(R)Ar]3, but shorter than the MoN bonds to
the N[R]Ar ligands.
Figure 4 (top) shows the k3-weighted EXAFS of the three
compounds and Figure 4 (bottom) the corresponding Fourier
Figure 3. Mo K-edge X-ray absorption near-edge spectra of the three
compounds (a) Mo(N[R]Ar)3, (b) (µ-N2)[Mo(N[R]Ar)3]2, and (c) NMo-
(N[R]Ar)3.
transforms. The data are of outstanding quality for all three
compounds, extending out to k ) 21 Å-1 for all spectra. The
EXAFS of Mo[N(R)Ar]3 and NMo[N(R)Ar]3 are broadly similar
for the short core-hole lifetime of the molybdenum K-edge. For
in frequency; however, unlike Mo[N(R)Ar]3, NMo[N(R)Ar]3
Mo[N(R)Ar]3, a weak transition in the edge at about 20 007.0
exhibits a beat in its amplitude envelope which is maximal at
eV, visible as a shoulder about halfway up the rising edge, is
around k ) 9 Å-1 and again close to the end of the spectrum,
observed. For (µ-N2){Mo[N(R)Ar]3}2, a more pronounced and
lower-energy transition is seen at about 20 005.5 eV, and for
NMo[N(R)Ar]3, the most pronounced pre-edge feature is
(28) Kutzler, F. W.; Natoli, C. R.; Misemer, D. K.; Doniach, S.; Hodgson,
K. O. J. Chem. Phys 1980, 73, 3274.

8628 J. Am. Chem. Soc., Vol. 118, No. 36, 1996
Laplaza et al.
of the curve-fitting are summarized in Table 2. Excellent
correspondence between data and fit is found for all three
compounds.
The final refined interatomic distances for the frozen solution
of Mo[N(R)Ar]3 were found to be in good agreement with the
-86 °C crystal structure (see above), indicating that the
molecule is not highly constrained by crystal packing forces.
The refined interatomic distances to the ligands for the three
compounds also show very little variation within the estimated
errors, suggesting that their radial configuration is largely
invariant. However, some slight spatial rotation or tilting of
the ligands cannot be ruled out by EXAFS alone, although this
is unlikely due to steric considerations.
The fit of the MoNNMo core of (µ-N2){Mo[N(R)Ar]3}2 was
found to be extremely sensitive to the Mo‚‚‚Mo interatomic
distance, which was always very close to 4.94 Å. The closer
core distances were a little more difficult to determine due in
part to correlations with ligand interactions. Therefore, it was
decided to carry out search profiles to help determine both the
linearity and the dinitrogen NN distance; Figure 5 shows the
results of these searches. In the left-hand panel, the MoNN
angle θ was varied, assuming a symmetrical molecule and a
constant dinitrogen NN distance for both cis and trans geometric
arrangements. New phase, amplitude, and mean free path
functions were calculated for each point, and with fixed core
distances, other parameters were refined to give the best fit.
Figure 5 shows that the fit is very sensitive to θ and that the
minimum in fit-error lies very close to 180° for both conforma-
tions. It is also interesting to note from this figure that the cis
conformation is much more sensitive than the trans to θ; the
small inset Fourier transforms show that for 170° cis the 5-Å
peak has all but disappeared, whereas for 160° trans it is still
substantial. This can be explained by examining the angular
displacement (i.e., the line of sight) of the MoN (dinitrogen)
vectors with respect to the Mo‚‚‚Mo vector which, for these
interatomic distances and the closest MoN, is approximately 4
times greater for the cis arrangement for a given angle θ.30 The
right-hand graph shows the effect of varying the dinitrogen NN
distance for a fixed Mo‚‚‚Mo distance, assuming a linear
molecule, and exhibits a very well-defined minimum at around
1.19 Å.
The final best fit for (µ-N2){Mo[N(R)Ar]3}2 as shown in
Table 2 and Figure 6 was found to be with a linear MoNNMo
arrangement and a dinitrogen NN distance of 1.19(2) Å. As
expected for such an arrangement and as calculated by feff,29
substantial multiple scattering contributions to the outer coor-
dination shells were necessary to account for these components.
Figure 6 shows a deconvolution of all the paths contributing
significantly to the EXAFS of (µ-N2){Mo[N(R)Ar]3}2. The left-
hand panel shows the data and fit at the top and then the
individual waves contributing; the right-hand panel shows
pictorially the corresponding scattering paths with respect to
the molecule. Thus, the ligands contribute five unique waves
(b-f), whereas the MoNNMo unit contributes one single-
scattering wave (a) to the first shell, three unique waves to the
3-Å shell (one each 2-leg, 3-leg, and 4-leg paths, g-i) and 10
unique waves to the 5-Å shell (one 2-leg, two 3-leg, four 4-leg,
two 5-leg, and one 6-leg paths, j-s). However, despite the
numerous paths included in the refinement, only four new
variables were required to fit the Mo‚‚‚N and Mo‚‚‚Mo outer
shells of the MoNNMo unit because the parameters R and σ2
for all single- and multiple-scattering paths of a given distance
were floated as a common parameter.
Figure 4. Mo K-edge EXAFS (upper panel) and EXAFS Fourier
transforms (lower panel) of the three compounds (a) Mo(N[R]Ar)3, (b)
(µ-N2)[Mo(N[R]Ar)3]2, and (c) NMo(N[R]Ar)3. The solid lines indicate
data and the broken lines the best fits. The EXAFS is k3-weighted raw
data. The Fourier transforms have been phase-corrected using the first
shell phase functions.
and minimal at around k ) 15 Å-1. Examination of the Fourier
transforms shows that the origin of this beat is in the well-
resolved split first shell of NMo[N(R)Ar]3, which is observed
to have an extra peak, fitted as a short MoN triple bond, to the
low-R side of the N[R]Ar MoN peak. Both Mo[N(R)Ar]3 and
NMo[N(R)Ar]3 also show low-amplitude scattering at around
3 Å which is attributable to Mo‚‚‚C interactions. In marked
contrast, the EXAFS spectrum of (µ-N2){Mo[N(R)Ar]3}2 is
dominated by high-frequency waves which persist to the end
of the data. The Fourier transform exhibits an increase in
amplitude (compared with Mo[N(R)Ar]3 and NMo[N(R)Ar]3)
at around 3 Å, consistent with the presence of Mo‚‚‚N
backscattering, and a substantial peak at just below 5 Å, which
is attributable to interactions involving Mo‚‚‚Mo backscattering.
The intensity of the peak, given its radial distance and
coordination number (just 1), suggests that considerable en-
hancement from multiple scattering is likely in this outer-shell
EXAFS.
The EXAFS total amplitude, phase-shift, and mean free path
functions used in the curve-fitting were calculated using the
program feff of Rehr and co-workers29 (see Experimental Section
for details). The results of EXAFS curve-fitting are shown as
broken lines in Figure 4, and the corresponding numerical results
(29) (a) Rehr, J. J.; Mustre de Leon, J.; Zabinsky, S. I.; Albers, R. C. J.
Am. Chem. Soc. 1991, 113, 5135. (b) Mustre de Leon, J.; Rehr, J. J.;
Zabinsky, S. I.; Albers, R. C. Phys. ReV. 1991, B44, 4146.
(30) For cis, the displacement φ ) 180 - θ, whereas for trans, sin
θ/RMo-Mo ) sin φ/RNtN
.

Dinitrogen CleaVage by Molybdenum(III) Complexes
J. Am. Chem. Soc., Vol. 118, No. 36, 1996 8629
Table 2. EXAFS Curve-Fitting Resultsa
Mo(N[R]Ar)3
(µ-N2)[Mo(N[R]Ar)3]2
NMo(N[R]Ar)3
interaction
N
R
σ2
R
σ2
R
σ2
MotN
Mo-N
Mo-N
1
1.655(2)
0.0014(2)
1b
3
1.876(5)
1.983(2)
3.153h
0.0019(1)g
0.0021(1)g
0.0071(18)
0.0052
1.982(2)
3.155(8)
2.854(16)
3.29(4)
0.0022(1)
0.0035(6)
0.0097(20)
0.014(5)d
0.014(5)d
1.975(2)
3.151(8)
2.862(7)
3.38(2)
0.0020(1)
0.0031(2)
0.0035(5)
0.0064(15)f
0.0064(15)f
Mo‚‚‚CR
Mo‚‚‚CAr1
Mo‚‚‚CAr2
Mo‚‚‚CAr6
Mo‚‚‚N
3
3
3
2.838(19)
3.34h
0.015(11)e
0.015(11)e
0.0026(9)
0.0025(1)
3
3.76(6)
3.71h
3.66(2)
1b
1b
3.069(12)
4.944(2)
Mo‚‚‚Mo
E0
-12.2(6)
0.150
200
-13.1h
0.270
10
-13.9(9)
0.149
10
errorc
T/K
a
Coordination number N, interatomic distance R (Å), Debye-Waller factor σ2 (Å2) and offset for k ) 0, E0 (eV). Three times the estimated
standard deviations of floated parameters, obtained from the diagonal elements of the covariance matrix, are shown in parentheses after the values.
These coordination numbers were adjusted to account for 75% sample purity (see Experimental Section). Error is given by ∑[(øobsd - øcalcd)2k6]/
b
c
∑[øobsd2k6]. d-f Floated to be equal as a single parameter in the fit. Floated with a constant ratio as a single parameter in the fit. Value fixed at
g
h
the mean of those of the starting material and the end product.
Figure 5. Search profiles for the linearity of the MoNNMo and for the NN distance of (µ-N2)[Mo(N[R]Ar)3]2, illustrating the sensitivity of the
EXAFS fit to the core MoNNMo geometry. Both profiles were calculated using a fixed Mo‚‚‚Mo distance of 4.94 Å. The geometry search profile
(left plot) used a fixed NN distance of 1.19(2) Å, and the NN distance search profile (right plot) assumed a linear MoNNMo arrangement. The other
parameters were refined to obtain the best fits possible. The fit-error in this case is given by ∑[(øobsd - øcalcd)2k6]/n, where n is the number of data
points. The inset Fourier transforms show the fit (filled curve) superimposed on the data (line).
(v) Molecular and Electronic Structure of (µ-N2){Mo-
[N(R)Ar]3}2. According to the EXAFS and NMR data (Vide
supra), and consistent with the Raman and magnetic data (Vide
infra), (µ-N2){Mo[N(R)Ar]3}2 is formulated as a symmetrical
bridging dinitrogen complex9 with a roughly linear MoNNMo
core. Bimetallic complexes with side-on or end-on bridging
dinitrogen ligands have been the subject of theoretical studies.31
The best precedent for our proposed structure is (µ-
N2)[Mo{R′NCH2CH2}3N]2 (R′ ) t-BuMe2Si),9z which contains
molybdenum in the same formal oxidation state as for (µ-N2)-
{Mo[N(R)Ar]3}2, and which is likewise ligated by amido groups.
A point of disparity between the two complexes is the
coordination number 5 at Mo in the former compared with 4 in
the latter. A second point of distinction between the two
complexes is the presence of marginally less electron-releasing
silylated amido substituents in the former. A third, rather
intriguing point of distinction between these two closely related
dinitrogen complexes is that (µ-N2)[Mo{R′NCH2CH2}3N]2 has
not been reported to undergo thermal NN bond cleavage. It
has been predicted, on theoretical grounds, that such an NN
bond cleavage reaction for (µ-N2)[Mo{R′NCH2CH2}3N]2 would
(31) Blomberg, M. R. A.; Siegbahn, P. E. M. J. Am. Chem. Soc. 1993,
115, 6908. Tatsumi, K.; Hoffmann, R. J. Am. Chem. Soc. 1981, 103, 3328.
See also ref 9f above.

8630 J. Am. Chem. Soc., Vol. 118, No. 36, 1996
Laplaza et al.
Figure 6. Scattering interactions contributing to the EXAFS of (µ-N2)[Mo(N[R]Ar)3]2. The left plot shows the experimental data and fit (solid and
broken lines, respectively, at top) and the deconvolution of the individual scattering interactions contributing to the total EXAFS. The right picture
shows, for each of the waves at the left, the corresponding scattering paths in the molecule. Note that paths b-f have a 3-fold degeneracy due to
the geometry of the ligands and paths h, k-m, o, q, and r are 2-fold degenerate because the reverse scattering path also contributes.
be an uphill process.5 On the other hand, note that the nitrido
complex NMo(MeNCH2CH2)3N is known32 and has not been
reported to dimerize.33
and the Mo centers, respectively, due to the substantial energy
mismatch between them.35 Because the molybdenum π-donor
orbitals are energetically well matched with the NN π* levels,
the 1eg and 2eg levels to which they give rise are quite
delocalized. Ten electrons are available for population of the
four doubly degenerate levels in (µ-N2){Mo[N(R)Ar]3}2, leading
to the expected configuration (1eu)4(1eg)4(2eu)2(2eg)0. Such a
configuration is reasonable for an isolable molecule since the
doubly-degenerate HOMO is nonbonding and molybdenum-
based, while the LUMO is everywhere antibonding. Further-
more, this configuration demands that the molecule be para-
magnetic, with two unpaired electrons. Experimental confirmation
of the latter prediction was achieved through SQUID magne-
tometry studies carried out on (µ-N2){Mo[N(t-Bu)Ph]3}2, of
which relatively pure powder samples could be obtained. The
data fit the Curie-Weiss law with good agreement in the
temperature range 29-300 K with µ ) 2.85 µB (Figure 8),
It is instructive to compare the important metrical parameters
for (µ-N2){Mo[N(R)Ar]3}2 with the literature values for (µ-
N2)[Mo{R′NCH2CH2}3N]2; these are, respectively, as follows
(Å): d(NN) ) 1.19(2), 1.20(2); d(Mo-N2) ) 1.876(5),
1.907(8); and d(Mo-NRR′) ) 1.983(2), 2.01(1). An NN bond
order slightly greater than 2 is indicated for both molecules
according to the observed d(NN) values; we elaborate upon this
issue in the section devoted to Raman data (Vide infra). In
addition, both complexes are intensely purple and paramagnetic.
The slightly longer Mo-NRR′ values for the silylated compound
can be attributed either to its higher coordination number or to
a somewhat reduced electron-donor capacity of silylated amido
ligands as compared with the N[R]Ar ligand. Some support
for the latter interpretation rests on the observation of a
significantly shorter d(Mo-N2) for (µ-N2){Mo[N(R)Ar]3}2,
indicative of greater dπ f pπ* back-donation in this case.
Molecules possessing linear MNNM moieties have been the
subject of several molecular orbital studies in the past.31 The
analysis given by Bercaw and co-workers is particularly lucid.9f
Four doubly-degenerate levels comprise the MNNM π-system.
In the point group S6, to which (µ-N2){Mo[N(R)Ar]3}2 likely
belongs,34 these levels bear the Mulliken labels 1eu (NN π),
1eg (MoN π, NN π*), 2eu (Mo nonbonding), and 2eg (MoN π*
and NN π*) and have the indicated characters (see Figure 7,
left). The 1eu and 2eu levels are localized primarily on the N2
consistent with the presence of two unpaired electrons.
A
further issue of interest in the context of the electronic structure
of (µ-N2){Mo[N(R)Ar]3}2 is the origin of its purple color.
While it is tempting to ascribe the intense 547-nm band to a
(2eu)2(2eg)0 f (2eu)1(2eg)1 transition having π f π* character,
Bercaw and co-workers have pointed out that metal-to-metal
charge transfer bands could be important for systems of this
type.9f
(vi) Raman Spectra of (µ-N2){Mo[N(R)Ar]3}2 and (µ-
15N2){Mo[N(R)Ar]3}2. Since a symmetrical linear structure for
(µ-N2){Mo[N(R)Ar]3}2 is indicated by the EXAFS data de-
scribed above, the Raman spectrum of this species was recorded
to garner additional information concerning the NN bond length
and order. Spectra of (µ-N2){Mo[N(R)Ar]3}2 and (µ-15N2){Mo-
[N(R)Ar]3}2 are displayed in Figure 9. Each complex exhibits
one strong Raman band attributable to ν(NN) in the observation
window, located at 1630 cm-1 for (µ-N2){Mo[N(R)Ar]3}2 and
(32) Plass, W.; Verkade, J. G. J. Am. Chem. Soc. 1992, 114, 2275.
(33) An example of µ-N2 bridge formation via coupling of terminal
nitrido ligands has been reported: Buhr, J. D.; Taube, H. Inorg. Chem.
1979, 18, 2208. Lam, H.-W.; Che, C.-M.; Wong, K.-Y. J. Chem. Soc.,
Dalton Trans. 1992, 1411. See also ref 9ll above.
(34) The related molecule (µ-P)[Mo(tBuNPh)3]2 (Ci) nearly belongs to
the point group S6 according to a single-crystal X-ray diffraction study:
Johnson, M. J. A.; Odom, A. L.; Cummins, C. C. Manuscript submitted
elsewhere.
(35) This qualitative notion is supported by extended Hu¨ckel calculations
carried out using the CAChe system.

Dinitrogen CleaVage by Molybdenum(III) Complexes
J. Am. Chem. Soc., Vol. 118, No. 36, 1996 8631
Figure 7. Qualitative molecular orbital correlation diagram for the conversion of S6 (µ-N2)[Mo(N[R]Ar)3]2 to 2 NMo(N[R]Ar)3 (C2h) via a C2h
transition state. Only the most important σ and π levels of the MoNNMo core are depicted. The numbering scheme for the energy levels takes into
account only those orbitals that are pictured.
Ar]3}2 is indicated by comparison of its ν(NN) and d(NN) with
those for the other NN-containing molecules in Figure 10.
(vii) Activation Parameters for NN Bond Cleavage and
Comments on the Transition State. Conversion of purple (µ-
N2){Mo[N(R)Ar]3}2 to 2 equiv of the nitrido complex NMo-
[N(R)Ar]3 is conveniently monitored by following the decay
of an intense 547-nm band. The product nitrido complex
exhibits negligible absorbance at this wavelength. Clean,
reproducible first-order kinetic behavior over ca. 4 half-lives
was observed for both (µ-N2){Mo[N(R)Ar]3}2 and (µ-15N2)-
{Mo[N(R)Ar]3}2 over the temperature range 25-65 °C. Mul-
tiple (three to six) rate constant determinations were performed
at 2.5 °C intervals over this range. Averages of the observed
rate constants for breakup of (µ-N2){Mo[N(R)Ar]3}2 are plotted
against temperature in Figure 11, where the solid curve
represents the best fit to the exponential form of the Eyring
Figure 8. SQUID magnetic susceptibility data for (µ-N2)[Mo(N[t-Bu]-
Ph)3]2 from 29 to 300 K fit to the Curie-Weiss law (see Experimental
Section for details). The fit yields µ ) 2.85 µB.
(36) (a) Hydrazine d(NN) ) 1.46 Å: Collin, R. L.; Lipscomb, W. N.
Acta Crystallogr. 1951, 4, 10. (b) hydrazine ν(NN) ) 938 cm-1: Durig, J.
R.; Bush, S. F.; Mercer, E. E. J. Chem. Phys. 1966, 44, 4238. (c) (PMe2-
Ph)4ClRe(µ-N2)MoCl4(OMe) d(NN) ) 1.21 Å, ν(NN) ) 1660 cm-1: ref
9tt above. (d) Azobenzene ν(NN) ) 1440 cm-1: Hacker, H. Spectrochim.
Acta 1965, 21, 1989. (e) Azobenzene d(NN) ) 1.248(3) Å: Bouwstra, J.
A.; Schouten, A.; Kroon, A.; Helmholdt, R. B. Acta Crystallogr. 1985, C41,
420. (f) trans-Diazene d(NN) ) 1.252 Å: Whitelegg, D.; Woolley, R. G.
at 1577 cm-1 for its doubly-labeled isotopomer (calcd 1575
cm-1). These data are compared with those for other NN-
containing molecules in Figure 10,36 which correlates [d(NN)]3/2
with ν(NN) according to Fryzuk, Loehr, and co-workers.9h The
best-fit line (R ) 0.94) in Figure 10 follows the equation ν(NN)
) -1840[d(NN)]3/2 + 4130, leading to a prediction of ca. 1.23
Å for the d(NN) of (µ-N2){Mo[N(R)Ar]3}2. Agreement between
the latter number and that obtained by EXAFS [1.19(2) Å, Vide
supra] is fairly good despite the rough nature of the correlation.
An NN bond order of approximately 2 for (µ-N2){Mo[N(R)-
THEOCHEM 1990, 68, 23. (g) trans-Diazene ν(NN) ) 1541 cm-1
:
Ackerman, M. N.; Burdge, J. J.; Craig, N. C. J. Chem. Phys. 1973, 58,
215. (h) (µ-N2)[ZrCp*2(N2)]2 d(NN) ) 1.182(5) Å, ν(NN) ) 1578 cm-1
:
ref 9f above. (i) (µ-N2)[CpZrN(SiMe2CH2PiPr2)2]2 d(NN) ) 1.301(3) Å,
ν(NN) ) 1211 cm-1 and (µ-N2)[ClZrN(SiMe2CH2PiPr2)2]2 d(NN) )
1.548(7) Å, ν(NN) ) 731 cm-1: ref 9h above. (j) (µ-N2)[Ru(NH3)5]2
4+
d(NN) ) 1.124(15) Å, ν(NN) ) 2100 cm-1: Treitel, I. M.; Flood, M. T.;
Marsh, R. E.; Gray, H. B. J. Am. Chem. Soc. 1969, 91, 6512.

8632 J. Am. Chem. Soc., Vol. 118, No. 36, 1996
Laplaza et al.
Figure 11. Exponential least-squares fit of first-order rate constants
(k) versus temperature (T) to the Eyring equation k ) (kBT/h) exp-
(∆Sq/R) exp(-∆Hq/RT) for the reaction (µ-N2)[Mo(N[R]Ar)3]2 f 2
NMo(N[R]Ar)3. Activation parameters obtained from the fit are ∆Hq
Figure 9. (a) Resonance Raman (RR) spectrum of (µ-N2)[Mo(N[R]-
Ar)3]2 as a solution in toluene with toluene spectrum digitally subtracted;
ν(NN) ) 1630 cm-1. (b) RR spectrum of (µ-15N2)[Mo(N[R]Ar)3]2 as
a solution in toluene with toluene spectrum digitally subtracted;
ν(15N15N) ) 1577 cm-1 (calcd 1575 cm-1). (c) Raman spectrum of
toluene solvent. (d) Raw RR spectrum of (µ-15N2)[Mo(N[R]Ar)3]2 in
toluene with no subtraction of solvent spectrum. (e) Raw RR spectrum
of (µ-N2)[Mo(N[R]Ar)3]2 in toluene with no subtraction of solvent
spectrum.
) 23.3 ( 0.3 kcal‚mol-1 and ∆Sq ) 2.8 ( 1.1 cal‚mol-1‚K-1
.
Figure 12. Exponential least-squares fit of first-order rate constants
(k) versus temperature (T) to the Eyring equation k ) (kBT/h) exp-
Figure 10. Plot of [d(NN)]3/2 versus ν(NN) for a number of
NN-containing molecules. See text for more information.
(∆Sq/R) exp(-∆Hq/RT) for the reaction (µ-15N2)[Mo(N[R]Ar)3]2
f
2 15NMo(N[R]Ar)3. Activation parameters obtained from the fit are ∆Hq
equation. The corresponding data for (µ-15N2){Mo[N(R)Ar]3}2
are presented in Figure 12. See the Experimental Section for
details of the data collection and analysis. Activation parameters
obtained from the fit are as follows: ∆Hq ) 23.3 ( 0.3
kcal‚mol-1 and ∆Sq ) 2.9 ( 0.8 cal‚mol-1‚K-1. Morokuma
and co-workers recently calculated an activation energy of 21
kcal‚mol-1 for hypothetical (µ-N2)[Mo(NH2)3]2.5 The agree-
ment with our experimental value is manifest.
) 22.6 ( 0.3 kcal‚mol-1 and ∆Sq ) 0.4 ( 1.1 cal‚mol-1‚K-1
.
view of the electronic structure of (µ-N2){Mo[N(R)Ar]3}2 as
described above. In particular, while (µ-N2){Mo[N(R)Ar]3}2
possesses 10 π electrons and six σ electrons in its MoNNMo
framework, two NMo[N(R)Ar]3 molecules together require eight
π and eight σ electrons for their nitrido linkages. During NN
cleavage, then, two electrons must pair and pass from the
MoNNMo π-system to the MoN...NMo σ-system. A C2h TS
provides a mechanism for this in that the decreased symmetry
renders nondegenerate the two Mo-based nonbonding orbitals
originating from the 2eu level of ground-state (µ-N2){Mo[N(R)-
Ar]3}2 (see Figure 7, a qualitative molecular orbital correlation
diagram for the process). Furthermore, one of these orbitals
(2bu) lies in the MoNNMo plane of the C2h TS and is of
appropriate symmetry to correlate, via avoided crossings, with
Morokuma and co-workers have suggested that the lowest-
energy transition state for a corresponding NN cleavage process
involving hypothetical (µ-N2)[Mo(NH2)3]2 would involve a
zigzag (C2h) transition structure (TS) with an elongated NN bond
(1.544 Å, bond order of ca. 1) and shortened Mo-N2 bonds
(1.759 Å).5 Such a TS is consistent with our activation
parameters and is characterized by a significant degree of NN
bond cleavage. A zigzag-shaped TS is eminently plausible in

Dinitrogen CleaVage by Molybdenum(III) Complexes
J. Am. Chem. Soc., Vol. 118, No. 36, 1996 8633
the σ symmetry orbital (designated 1bu) that is doubly occupied
in the products but high in energy and unoccupied in linear
(µ-N2){Mo[N(R)Ar]3}2. Any postulated mechanism for simple
dissociation of (µ-N2){Mo[N(R)Ar]3}2 should encompass these
essential symmetry considerations. NN bond cleavage along a
linear trajectory is symmetry forbidden. The postulate of a C2h
TS is reasonable on electronic grounds, but importantly, it is
also reasonable on steric grounds. Whereas the two Mo centers
are separated by 4.944(2) Å in linear (µ-N2){Mo[N(R)Ar]3}2
(Vide supra), the intermolybdenum separation in the bent TS
calculated for (µ-N2)[Mo(NH2)3]2 is ca. 4.81 Å. There exists
independent evidence that two Mo[N(R)Ar]3 moieties can
accommodate an intermolybdenum separation of 4.49 Å with a
single-atom bridge.34 Therefore, the calculated C2h TS cannot
be construed to pose an insurmountable steric problem.
(viii) The 15N2 Isotope Effect for NN Bond Cleavage.
Based on assumptions of (i) a linear transition state, (ii) the
attribution of observed rate differences solely to zero-point
energy differences in the isotopomers, and (iii) a diatomic
system, primary kinetic isotope effects (KIEs) may be calculated
from k1/k2 ) exp[(hc/2kBT)(ν1-ν2)].37 In the case of dinitrogen
itself, the value of ν1 is 2329.9 cm-1 (from the Raman spectrum
of 14N2), and the value of ν2 is 2252.1 cm-1 (from the Raman
spectrum of 15N2).38 Because the force constant (kf) is dependent
only on electronic factors and is therefore the same for 14N2 as
for 15N2,39 and given that the reduced masses µ1 and µ2 for 14N2
and 15N2 are ca. 7.0 and 7.5, respectively,40 the classical
harmonic relationship for these isotopomers is 1.035ν2 ) ν1, in
close agreement with the experimental value. The theoretical
KIE, k(14N2)/k(15N2), for thermal dissociation of dinitrogen to
two N atoms is 1.21 at 298 K and 1.18 at 338 K. As another
example, consider a typical symmetrical alkyl diazo compound
RNdNR having an NN stretching frequency (ν1) of ca. 1575
Figure 13. Plot of rate constant ratios for the reactions (µ-N2)[Mo-
(N[R]Ar)3]2 f 2 NMo(N[R]Ar)3 and (µ-15N2)[Mo(N[R]Ar)3]2
f
2 15NMo(N[R]Ar)3. The unweighted least-squares fit is to the equation
k1/k2 ) exp[(hc/2kBT)(ν1-ν2)], see text, with ν1 - ν2 as the sole
parameter. Given the attendant assumptions, the fit yields the prediction
that the Raman NN stretching frequency for (µ-N2)[Mo(N[R]Ar)3]2
should lie 52 ( 9 cm-1 higher in energy than that for (µ-15N2)[Mo-
(N[R]Ar)3]2.
NN bond cleavage.42 A more sophisticated analysis would take
into account such effects as possible nuclear spin contributions
to the isotope effect,43 and coupling of the NN vibrational mode
with other fundamental modes of the (µ-N2){Mo[N(R)Ar]3}2
molecule.44 15N2 isotope effects have evidently not been
observed for N2 cleavage by nitrogenase enzymes, where
decomplexation of oxidized Fe protein from the FeMo protein
is thought to be rate-limiting.45
(ix) X-ray Structure of NMo[N(t-Bu)Ph]3. The new nitrido
complex NMo[N(t-Bu)Ph]3 is included in the present work in
part because it was possible to characterize this compound
structurally by single-crystal X-ray diffraction, whereas we have
been unable to obtain suitable crystals of NMo[N(R)Ar]3 for
such a study. From a chemical standpoint the two systems are
equivalent in every important respect, and it stands to reason
that the solid-state structure of the latter should resemble closely
that of the former. We have suggested that NMo[N(R)Ar]3 is
a monomeric, pseudotetrahedral species possessed of a short
cm-1 41 the calculated ν2 for R15Nd15NR is 1522 cm-1, so the
;
theoretical KIE for dissociation to two alkylnitrene moieties is
1.14 at 298 K and 1.12 at 338 K. The two preceding examples,
with their attendant assumptions, permit the following prediction
concerning a primary KIE for scission of an NN double-to-
triple bond: k(14N2)/k(15N2) should be somewhere between 1.12
and 1.21 and will be essentially invariant over the temperature
range 298-338 K, the temperature range employed in our
kinetic study. The observed temperature dependence of the
isotope effect for scission of (µ-N2){Mo[N(R)Ar]3}2 to 2 NMo-
[N(R)Ar]3 can be fitted to the equation k1/k2 ) exp[(hc/2kBT)-
(ν1-ν2)] to yield a prediction for ν1 - ν2; such a fit is displayed
in Figure 13. The fit yields a prediction for ν1 - ν2 of 52 ( 9
cm-1, which is (perhaps fortuitously) very close to the observed
value of 53 cm-1 obtained (Vide supra) by Raman spectroscopy
for (µ-N2){Mo[N(R)Ar]3}2 and (µ-15N2){Mo[N(R)Ar]3}2. Within
the confines of the present simplistic analysis, the observed
isotope effect is of the expected magnitude for rate-determining
NMo triple bond, consistent with its ν(MoN) of 1042 cm-1
.
The known,46 crystallographically-characterized molecule NMo-
(NPh2)3 exhibits such features and served as close precedent,
buttressing our assertions concerning the structure of NMo-
[N(R)Ar]3. Related molybdenum nitrido complexes include
t
NMo(OtBu)3,47 NMo(2,4,6-C6H2Me3)3,48 and NMo(CH2 Bu)3.49
The X-ray structure of NMo[N(t-Bu)Ph]3 presented no surprises
(Figure 14; see Experimental Section for a complete list of bond
lengths and angles). The molecule exhibits crystallographic
3-fold symmetry (point group C3), with MoN bond lengths of
1.658(5) and 1.979(2) Å to the nitrido and amido nitrogens,
(37) Collecting the fundamental constants together leads to a user-friendly
form of the equation: k1/k2 ) exp(0.7193916{ν1 - ν2}/T). This equation
leads to the familiar value for kH/kD of 6.9 at 298 K taking ν1 and ν2 as
2900 and 2100 cm-1, respectively, for the breaking of a CH versus a CD
bond. Westheimer, F. H. Chem. ReV. 1961, 61, 265. Carpenter, B. K.
Determination of Organic Reaction Mechanisms; Wiley: New York, 1984.
(38) Bendtsen, J. J. Raman Spectrosc. 1974, 2, 133. Gilson, T. R.; Beattie,
I. R.; Black, J. D.; Greenhalgh, D. A.; Jenny, S. N. J. Raman Spectrosc.
1980, 9, 361. From these references, the experimentally-determined
(42) For isotope effects associated with O-O bond cleavage in acyl
peroxides, see: Goldstein, M. J. Science 1966, 154, 1616. Goldstein, M.
J.; Judson, H. A. J. Am. Chem. Soc. 1970, 92, 4122.
(43) Turro, N. J.; Kraeutler, B. Acc. Chem. Res. 1980, 13, 369.
(44) For an elegant “whole-molecule” deuterium equilibrium isotope
effect study on the binding of ethylene to a transition metal complex, see:
Bender, B. R. J. Am. Chem. Soc. 1995, 117, 11239.
(45) Thorneley, R. N. F.; Lowe, D. J. Biochem. J. 1983, 215, 393.
(46) Gebeyehu, Z.; Weller, F.; Neumu¨ller, B.; Dehnicke, K. Z. Anorg.
Allg. Chem. 1991, 593, 99.
(47) Chan, D. M.-T.; Chisholm, M. H.; Folting, K.; Huffman, J. C.;
Marchant, N. S. Inorg. Chem. 1986, 25, 4170.
vibrational frequency for 14N15N is 2291.3 cm-1
.
(39) Herzberg, G. Molecular Spectra and Molecular Structure; D. Van
Nostrand: New York, 1945.
(40) Calculated using masses for 14N and 15N found in the following:
Weast, R. C. CRC Handbook of Chemistry and Physics, 66th ed.; Weast,
R. C., Ed.; CRC Press: Boca Raton, 1985.
(48) Caulton, K. G.; Chisholm, M. H.; Doherty, S.; Folting, K. Organo-
metallics 1995, 14, 2585.
(41) Dean, J. A. Lange’s Handbook of Chemistry, 14th ed.; McGraw-
Hill: New York, 1992.
(49) Herrmann, W. A.; Bogdanovic, S.; Poli, R.; Priermeier, T. J. Am.
Chem. Soc. 1994, 116, 4989.

8634 J. Am. Chem. Soc., Vol. 118, No. 36, 1996
Laplaza et al.
a
b
d
c
Figure 14. (a) Structural diagram and atom-labeling scheme for NMo(N[t-Bu]Ph)3. Thermal ellipsoids are at the 35% probability level. Selected
bond lengths (Å) and angles (deg) for NMo(N[t-Bu]Ph)3 are as follows: Mo(1)-N(1), 1.658(5); Mo(1)-N(2), 1.979(2); C(1)-N(2), 1.444(4);
N(2)-C(7), 1.513(3); N(1)-Mo(1)-N(2), 102.94(7); N(2)-Mo(1)-N(2)′, 115.13(5); C(1)-N(2)-C(7), 117.3(2); C(1)-N(2)-Mo(1), 113.5(2);
C(7)-N(2)-Mo(1), 128.2(2). Space-filling representations of NMo(N[t-Bu]Ph)3 viewed (b) perpendicular to the nitrido bond axis, (c) along the
nitrido bond axis facing the hemisphere protected by tert-butyl groups, and (d) along the nitrido bond axis facing the hemisphere protected by
phenyl groups.
52
respectively. These values are in excellent agreement with those
determined by EXAFS for NMo[N(R)Ar]3 (Vide supra), namely,
1.655(2) and 1.975(2) Å, and also compare reasonably well with
those calculated for hypothetical NMo(NH2)3,5 namely, 1.678
and 1.966 Å. It is noteworthy that the geometry of the Mo[N(t-
Bu)Ph]3 fragment differs little from that in the starting Mo-
[N(R)Ar]3 (Vide supra) and Mo[N(t-Bu)Ph]3 molecules, grossly
speaking. Space-filling models (Figure 14b-d) reveal that the
major difference is that the channel bounded by tert-butyl groups
(Figure 14c) has expanded somewhat to accommodate the
introduction of a nitrogen atom. The channel bounded by aryl
residues, on the other hemisphere of the molecule, is constricted
commensurately (Figure 14d). No short Mo...NMo contacts of
the kind observed for solid-state NMo(OtBu)3, a “shish kebab”
polymer,47 were found for NMo[N(t-Bu)Ph]3. Indeed, none
should be expected in view of the substantially greater steric
demands of N[t-Bu]Ph or N[R]Ar as compared with tert-
butoxide. A growing list of pseudo-C3-symmetric, monomeric,
structurally characterized complexes related to NMo[N(t-Bu)-
Ph]3 includes PMo[N(R)Ar]3,2 PMo[N(t-Bu)Ph]3,50 ClTi[N(R)-
Ar]3,51 ITi[N(t-Bu)Ph]3,17 MeSn[N(R)Ar]3, and ISn[N(R)Ar]3.22
C3 symmetry thus represents a pervasive motif for complexes
built with three N-tert-butylarylamido ligands and a single apical
atom. The related chromium nitrido complexes NCr(NiPr2)3
and NCr(NiPr2)2(CH2SiMe2Ph)53 are likewise monomeric in the
solid state.
Concluding Remarks
Findings presented here corroborate the proposal of a
bimetallic mechanism (Scheme 1) with an observable intermedi-
ate, for dinitrogen cleavage by three-coordinate molybdenum(III)
amide complexes. The purple, thermally-unstable intermediate
(µ-N2){Mo[N(R)Ar]3}2 was found to convert to the nitrido
product, NMo[N(R)Ar]3, according to a first-order kinetic profile
over the temperature range interrogated (25-65 °C). Temper-
ature-dependent kinetic isotope effect data obtained for breakup
of the isotopomer (µ-15N2){Mo[N(R)Ar]3}2 are in line with rate-
determining NN bond cleavage. A Raman spectroscopic
investigation of (µ-N2){Mo[N(R)Ar]3}2 and its 15N2 isotopomer
suggests an NN bond order of ca. 2 for these complexes.
EXAFS characterization of (µ-N2){Mo[N(R)Ar]3}2 indicates a
linear MoNNMo linkage in this µ-dinitrogen complex, an
assignment corroborated by a temperature-dependent magnetic
susceptibility study indicating a triplet ground state for (µ-N2)-
{Mo[N(t-Bu)Ph]3}2. EXAFS structural characterization data
pertaining to Mo[N(R)Ar]3 and NMo[N(R)Ar]3 are in accord
with single-crystal X-ray structural data for Mo[N(R)Ar]3 and
NMo[N(t-Bu)Ph]3.
(50) Johnson, M. J. A.; Odom, A. L.; Cummins, C. C.; Manuscript in
preparation.
(51) Johnson, A. R.; Wanandi, P. W.; Cummins, C. C.; Davis, W. M.
Organometallics 1994, 13, 2907.
(52) Odom, A. L.; Cummins, C. C.; Protasiewicz, J. D. J. Am. Chem.
Soc. 1995, 117, 6613.
(53) Odom, A. L.; Cummins, C. C. Organometallics 1996, 15, 898.

Dinitrogen CleaVage by Molybdenum(III) Complexes
J. Am. Chem. Soc., Vol. 118, No. 36, 1996 8635
under an argon atmosphere to produce orange-red, crystalline Mo[N(R)-
Ar]3 (mp 126-128 °C, yield 1.247 g, 1.940 mmol, 70%). Mo[N(R)-
Ar]3 is extremely oxygen- and moisture-sensitive, but decomposes <5%
at 25 °C in an inert atmosphere over 24 h. 1H NMR (C6D6): 64.0
(C(CD3)2CH3), ca. 23 (br, ortho ArH), -9.63 (ArCH3), -51.67 ppm
(br, para ArH). Anal. Calcd for C36H36D18MoN3: C, 67.26; H, 8.47;
N, 6.54. Found: C, 67.61; H, 8.38; N, 6.40.
SQUID Magnetic Measurements on Mo[N(R)Ar]3. Measurements
were made using a Quantum Design SQUID magnetometer. The
magnetometer uses the MPMSR2 software (Magnetic Property Meas-
urement System Revision 2). Data were recorded at a field strength
of 5000 G. Gel caps (Gelatin Capsule No. 4 Clear) and straws were
ordered from Quantum Design, Inc.
The potential utility of three-coordinate molybdenum(III)
complexes as models for biological nitrogen fixation54 stems
from the ease with which they can be used to probe the
molecular-level details of dinitrogen binding, reduction, and
cleavage. Especially attractive is the fact that sterically and
electronically tunable three-coordinate molybdenum(III) com-
plexes are now available for study in the context of fully
synthetic dinitrogen cleavage systems. Catalytic dinitrogen
cleavage systems based on the chemistry described here are in
the offing if suitable N-atom transfer chemistries can be
developed to convert molybdenum(VI) nitrido complexes into
molybdenum(III) species active for dinitrogen cleavage.3
Samples of Mo[N(R)Ar]3 for magnetic measurements were prepared
in the drybox as described above. The solid compound was placed
into a preweighed gel cap, and a preweighed plug of Parafilm was
inserted above it, to keep it in place. The mass of the Mo[N(R)Ar]3
sample was then ascertained by weighing the loaded gel cap. The
loaded gel cap was mounted in a straw, and the straw was placed in an
N2-filled vessel for transport to the magnetometer. The straw was
placed on the end of a rod for insertion into the magnetometer. The
field and temperature were adjusted to 5000 G and 5 K, respectively.
Once the temperature had equilibrated and the field was stable, the
sample was centered. This was done by running a full-length DC scan,
adjusting the position automatically, and then recentering using a DC
centering scan. During the run, measurements were taken over the
following temperature ranges with the indicated increments: 5-10 K
(one data point every 1 K), 12-20 K (one data point every 2 K), 23-
50 K (one data point every 3 K), 55-100 (one data point every 5 K),
110-200 K (one data point every 10 K), 220-300 K (one data point
every 20 K). The run required approximately 5 h.
The data for Mo[N(R)Ar]3 fit the Curie-Weiss law in the temper-
ature range 5-300 K (R ) 0.999 87). The calculated curve (Figure 1)
is the best least-squares fit of the observed susceptibility data to the
equation øΜ(obs) ) ((µ2)/(7.997584(T - θ))) - C. The value of µ
obtained in this manner was 3.82 µB, to be compared with the spin-
only value for three unpaired electrons, 3.87 µB. In the least-squares
fit, carried out using the General curve-fitting routine included with
the program Kaleidagraph, both θ and C were treated as variable
parameters. The obtained value of θ, the Weiss constant, was -2.8755
K. The value of the constant C, included to represent the sum of all
temperature-independent contributions to the total observed susceptibil-
ity, including the diamagnetic correction and temperature-independent
paramagnetism, was -0.000 632 06. The raw data plotted in Figure 1
are deposited in the supporting information.
X-ray Crystal Structure of Mo[N(R)Ar]3. The crystal and
structure refinement data in this section have been communicated in
part2 and are deposited here for completeness. An orange parallelepiped
crystal of approximate dimensions 0.24 × 0.32 × 0.32 mm was
obtained by cooling an ether solution to -35 °C. The crystal was coated
with Paratone-N oil and mounted on a glass fiber in a stream of cold
N2. Data collection was carried out at -86 °C on an Enraf-Nonius
CAD-4 diffractometer with graphite-monochromated Mo KR radiation.
A total of 9653 reflections were collected to a 2θ value of 44.9°, of
which 9186 were unique (Rint ) 0.051). The structure was solved by
direct methods. Non-hydrogen atoms were refined anisotropically. The
final cycle of full-matrix least-squares refinement was based on 5584
observed reflections (I > 3.00σ(I)) and 722 variable parameters and
converged (largest parameter shift was 0.16 times its esd) with R )
0.066 and Rw ) 0.065. A final difference Fourier map showed no
chemically significant features. Crystal data are a ) 13.736(4) Å, b
) 15.948(4) Å, c ) 16.187(7) Å, R ) 88.93(2)°, â ) 84.46(2)°, γ )
88.35(2)°, V ) 3527(4) Å3, space group P1h, Z ) 4, molecular weight
) 642.90 for C36H36D18N3Mo, and d(calcd) ) 1.210 g/cm3.
Kinetic Studies: Conversion of (µ-N2){Mo[N(R)Ar]3}2 to 2 NMo-
[N(R)Ar]3 and of (µ-15N2){Mo[N(R)Ar]3}2 to 2 15NMo[N(R)Ar]3.
Stock solutions of (µ-N2){Mo[N(R)Ar]3}2 were prepared according to
the following typical procedure. Toluene (5.0 g) was added to 251
mg of Mo[N(R)Ar]3 (0.390 mmol), giving an approximately 0.4 M
solution. The solution, in a 20-mL scintillation vial having a hole in
its cap, was then stored in the glovebox refrigerator under ca. 1 atm of
N2 at -35 °C for ca. 10 days. During this time the reaction mixture
Experimental Section
General Details. Unless stated otherwise, all operations were
performed in a Vacuum Atmospheres drybox under an atmosphere of
purified nitrogen, or using standard Schlenk techniques under an argon
atmosphere. Anhydrous ether and toluene were purchased from
Mallinckrodt; n-pentane and n-hexane were purchased from EM
Science. Ether was distilled under a nitrogen atmosphere from purple
sodium benzophenone ketyl. Aliphatic hydrocarbon solvents were
distilled under a nitrogen atmosphere from very dark blue to purple
sodium benzophenone ketyl solubilized with a small quantity of
tetraglyme. Toluene was refluxed over molten sodium for at least 2
days, prior to its distillation under a nitrogen atmosphere. Distilled
solvents were transferred under vacuum into Teflon-stopcocked glass
vessels and stored, prior to use, in a Vacuum Atmospheres drybox.
Benzene-d6 was degassed and dried over blue sodium benzophenone
ketyl and transferred under vacuum into a storage vessel. Chloroform-d
and toluene-d8 were degassed and dried over 4-Å sieves. Alumina and
4-Å sieves were activated in Vacuo overnight at a temperature above
180 °C. Li[N(R)Ar](OEt2),22 HN(t-Bu)Ph,55 mer-MoCl3(THF)3,21 and
mesityl azide56 were prepared according to published procedures. Other
chemicals were used as received. Infrared spectra were recorded on a
Perkin-Elmer 1600 Series FTIR. UV-visible spectra were recorded
on a Hewlett-Packard 8453 diode-array spectrophotometer. 1H and 13
C
NMR spectra were recorded on Varian XL-500, Varian XL-300, or
Varian Unity-300 spectrometers. 1H and 13C chemical shifts are
reported with reference to solvent resonances (residual C6D5H in C6D6,
7.15 ppm; C6D6, 128.0 ppm; CHCl3 in CDCl3, 7.24 ppm; CDCl3, 77.0
ppm) and were recorded at ca. 25 °C unless otherwise indicated. 2H
chemical shifts are reported with respect to external C6D6 (7.15 ppm).
15N chemical shifts are reported with respect to external nitromethane
at 380.2 ppm (relative to liquid NH3 at 0 ppm). Solution magnetic
1
susceptibilities were determined by H NMR at 300 MHz using the
method of Evans.23 Routine coupling constants are not reported.
Combustion analyses (C, H, and N) were performed by Oneida Research
Services, Whitesboro, NY. Melting points were obtained in sealed glass
capillaries and are uncorrected.
Synthesis of Mo[N(R)Ar]3. In a typical preparation, mer-MoCl3-
(THF)3 (1.743 g, 4.164 mmol) and Li[N(R)Ar](OEt2) (2.190 g, 8.315
mmol) were added to 70 mL of cold (-100 °C) ether, and the mixture
was stirred for 2.5 h after being warmed to 28 °C. The precipitated
LiCl and excess mer-MoCl3(THF)3 were removed by filtration.
Analysis of the filtrate by 2H NMR spectroscopy showed only one major
product, with a relatively sharp (∆ν1/2 ) 35 Hz) signal at 64.6 ppm
corresponding to the 2H-labeled tert-butyl groups in paramagnetic
Mo[N(R)Ar]3. The filtrate was concentrated and cooled to -35 °C
(54) (a) Kim, J.; Rees, D. C. Biochemistry 1994, 33, 389. (b) Kim, J.;
Woo, D.; Rees, D. C. Biochemistry 1993, 32, 7104. (c) Rees, D. C.; Kim,
J.; Georgiadis, M. M.; Komiya, H.; Chirino, A. J.; Wood, D.; Schlessman,
J.; Chan, M. K.; Joshuator, L.; Santillan, G.; Chakrabarti, P.; Hsu, B. T.
ACS Symp. Ser. 1993, 535, 170. (d) Chan, M. K.; Kim, J. S.; Rees, D. C.
Science 1993, 260, 792. (e) Georgiadis, M. M.; Komiya, H.; Chakrabarti,
P.; Woo, D.; Kornuc, J. J.; Rees, D. C. Science 1992, 257, 1653. (f) Kim,
J. S.; Rees, D. C. Science 1992, 257, 1677. (g) Kim, J. S.; Rees, D. C.
Nature 1992, 360, 553. (h) Howard, J. B.; Rees, D. C. AdV. Protein Chem.
1991, 42, 199. (i) Bolin, J. T.; Ronco, A. E.; Morgan, T. V.; Mortenson, L.
E.; Xuing, N.-H. Proc. Natl. Acad. U.S.A. 1993, 90, 1078.
(55) Biehl, E. R.; Smith, S. M.; Reeves, P. C. J. Org. Chem. 1971, 36,
1841.
(56) Ugi, I.; Perlinger, H.; Behringer, L. Chem. Ber. 1958, 91, 2330.

8636 J. Am. Chem. Soc., Vol. 118, No. 36, 1996
Laplaza et al.
acquired a rich purple color (λmax ) 547 nm). Stock solutions of (µ-
15N2){Mo[N(R)Ar]3}2 were prepared according to the following typical
procedure. A 0.1-L glass tube containing 1 atm of 15N2 (98+% isotopic
purity) at STP was purchased from Cambridge Isotopes Laboratory,
and to the cylinder was attached, by glass-blowing techniques, an in-
line Chemglass Teflon stopcock suitable for attachment to a vacuum
or gas system via a hose. Into the small space (ca. 5 mL) between the
breakseal and the Teflon stopcock was added a magnetic bar (to break
the breakseal) and a solution of Mo[N(R)Ar]3 (50 mg, 0.078 mmol) in
toluene (1.0 g). The solution was then carefully degassed for ca. 30 s
by exposure to dynamic vacuum. The stopcock was tightly closed to
isolate the Mo[N(R)Ar]3 solution in the evacuated space, and then the
magnetic bar was addressed via an external magnet and manipulated
to break the breakseal, admitting 15N2 into the region containing the
Mo[N(R)Ar]3 solution. The whole assembly was then placed in the
glovebox refrigerator at -35 °C. After a few minutes, and thereafter
periodically, the stopcock was checked for a tight closure, as Teflon
may shrink upon cooling. After at least 6 days the purple solution
was removed from the vessel using a disposable 1-mL syringe fitted
with a long steel needle. The stock solution was stored thereafter in a
scintillation vial at -35 °C.
Aliquots from the stock solutions were diluted for kinetic measure-
ments. Data collection began when the band at 547 nm had decayed
to ca. 1.3 absorbance units. Data were collected on an Hewlett-Packard
HP 8453 diode array spectrophotometer. The reaction temperature was
controlled to within (0.5 °C using an HP 89090A Peltier temperature
control accessory. The decay of purple (µ-N2){Mo[N(R)Ar]3}2 was
monitored at λmax ) 547 nm. The amber product NMo[N(R)Ar]3 does
not absorb in this region. Samples were stirred at 250 rpm by an
internal Teflon-coated magnetic stir bar and allowed to equilibrate at
the desired temperature for 120 s prior to data acquisition. In a typical
run, between 700 and 1000 data points were processed and used to
determine rates. Between 25 and 65 °C, plots of ln(A-A∞) vs time
were linear through at least 4 half-lives. Rate constants used as data
points in the Eyring plot (Figure 11) were determined using the
Kaleidagraph least-squares curve-fitting routine, using the equation A
) A∞ + A0[exp(-kt)]. From three to six runs were carried out at each
temperature, and the reported rate constants are unweighted averages.
The error associated with each rate constant was estimated from the
reproducibility and is reported at the 95% confidence interval.57 Raw
rate data are deposited in the supporting information.
EXAFS Sample Preparation: (µ-N2){Mo[N(R)Ar]3}2. A solution
of Mo[N(R)Ar]3 (1.16 g, 1.80 mmol) in 9.0 g of toluene was stored in
a -35 °C glovebox refrigerator under ca. 1 atm of dinitrogen. After 6
days an aliquot of the deep purple solution was removed for 2H NMR
analysis, revealing a 1:3 ratio of the resonances at 64 and 14 ppm,
which we assign respectively to Mo[N(R)Ar]3 and (µ-N2){Mo[N(R)-
Ar]3}2. The extent of conversion was probably greater than 75% based
on Mo, in view of the fact that some purple solid, presumably (µ-N2)-
{Mo[N(R)Ar]3}2, was observed to have precipitated. Three rectangular
EXAFS cells (Pyrex, 2 × 1 × 1 cm) were filled with the cold purple
solution, degassed, and flame-sealed with a hand torch, the sample being
kept covered with powdered dry ice throughout. Mo[N(R)Ar]3. An
EXAFS cell (Pyrex, 2 × 1 × 1 cm) was loaded with a solution of
Mo[N(R)Ar]3 (0.168 mmol, 108 mg) in ca. 2 mL of toluene, degassed,
and flame-sealed with a hand torch, the sample being kept covered
with powdered dry ice throughout. NMo[N(R)Ar]3. An EXAFS cell
(Pyrex, 2 × 1 × 1 cm) was loaded with a solution of NMo[N(R)Ar]3
(0.152 mmol, 100 mg) in ca. 2 mL of toluene, degassed, and flame-
sealed with a hand torch, the sample being kept covered with powdered
dry ice throughout. The samples were shipped, on dry ice, from MIT
to Stanford’s SLAC via Federal Express.
crystal to approximately 50% off-peak, and no specular optics were
present in the beamline. X-ray absorption was measured in transmit-
tance using argon-filled ionization chambers. A spectrum of molyb-
denum foil was collected simultaneously with that of the sample, and
spectra were calibrated with reference to the lowest energy inflection
point of the foil, assumed to be 20 003.9 eV.
During data collection, the samples of (µ-N2){Mo[N(R)Ar]3}2 and
NMo[N(R)Ar]3 were maintained at a temperature of approximately 10
K, using an Oxford Instruments liquid helium flow cryostat. The
sample Mo[N(R)Ar]3 was discovered to reversibly crystallize from
solution at low temperatures and therefore was maintained at the higher
temperature of 200 K in order to maintain a solution.
EXAFS Data Analysis. The data were analyzed using the EXAFS-
PAK suite of computer programs58 running on Digital Equipment Corp.
VAXstation 4000 and AXP graphics workstations. The EXAFS
oscillations ø(k) were quantitatively analyzed by curve-fitting.59 The
EXAFS total amplitude and phase-shift functions were calculated using
the program feff (version 6.01) of Rehr and co-workers.29 Mo[N(R)-
Ar]3. The EXAFS spectrum of Mo[N(R)Ar]3 was fitted with reference
to the known crystal structure. Coordinates from the -86 °C crystal
structure (see above) were used in feff to calculate accurate phase,
amplitude, and mean free path functions. These functions were then
used to refine interatomic distances in the 200 K solution form of
Mo[N(R)Ar]3. Of the many scattering paths generated by feff, only
five were found to contribute sufficient intensity to merit inclusion in
the refinement, specifically the single-scattering contributions from the
central carbon of the R group and from three carbons of the Ar group,
in addition to the coordinating nitrogen. NMo[N(R)Ar]3. The
spectrum of NMo[N(R)Ar]3 was fitted assuming a very similar ligand
structure to that of Mo[N(R)Ar]3. The phase, amplitude, and mean
free path functions calculated for the ligand interactions of Mo[N(R)-
Ar]3 were also used for NMo[N(R)Ar]3, and the same numbers of shells
were included. The first shell Mo-N and MotN distances were
iteratively refined and new functions calculated until no further change
was observed. (µ-N2){Mo[N(R)Ar]3}2. The spectrum of (µ-N2){Mo-
[N(R)Ar]3}2 was also fitted in part using the calculated phase, amplitude,
and mean free path functions calculated for Mo[N(R)Ar]3. In this case,
some of the parameters were fixed at the mean of values derived from
Mo[N(R)Ar]3 and NMo[N(R)Ar]3. These were the offset to k ) 0, E0,
and the distances to three of the outer carbons (which contribute only
subtly to the total EXAFS); this was done in order to eliminate
correlations in the minimization, and fits in which these variables were
floated independently yielded values which differed only slightly from
the mean values used. In addition, the highly correlated Debye-Waller
factors for the two Mo-N shells were refined as a single parameter
with values proportional to their distances [justification for this
approximation is found in the values of the independently-refined
Debye-Waller factors for the Mo-N interactions of NMo[N(R)Ar]3].
Phase, amplitude, and mean free path functions for the MoNNMo
molecular unit were calculated using feff by postulating a starting
(linear) coordinate model, then iteratively refining the distances, deriving
new coordinates, and recalculating functions until no further difference
was found. In this case, substantial multiple-scattering contributions
to the outer shells of the linear molecule were predicted and were
included in the fit. In the final analysis, all three interatomic distances
Mo-N, Mo‚‚‚N, and Mo‚‚‚Mo were floated and excellent agreement
between them, assuming a linear model, was found. Deuterium NMR
of the (µ-N2){Mo[N(R)Ar]3}2 sample showed a purity of 75% (based
on Mo, see above), the remainder being the starting material Mo[N(R)-
Ar]3, and therefore appropriate coordination numbers for the MoNNMo
(58) The EXAFSPAK program suite was developed by one of the authors
(G.N.G.) and is available from him by application in writing.
EXAFS Data Collection. X-ray absorption spectroscopy measure-
ments were carried out at the Stanford Synchrotron Radiation Labora-
tory with the SPEAR storage ring containing 55-90 mA at 3.0 GeV.
Molybdenum K-edge X-ray absorption spectra were collected on
beamline 7-3 using a Si(220) double crystal monochromator with an
upstream vertical aperture of 1 mm and a wiggler field of 1.8 T.
Harmonic rejection was accomplished by detuning one monochromator
(59) The EXAFS oscillations ø(k) are given by the approximate equation:
n
NiAi(k,Ri)
-2σi2ki2
e-2R /λ (k,Ri)e
sin[2kRi + Ri(k,Ri)]
i
i
∑
kRi2
i)1
where k is the photoelectron wave number and Ai(k,Ri) and Ri(k,Ri) are the
EXAFS total amplitude and total phase-shift functions, respectively, for
scattering path i. Ni is the degeneracy of scattering path i with a total path
length Ri. λi(k,Ri) is the photoelectron mean free path function, and σi2 is
the Debye-Waller factor (the mean square deviation of Ri). The summation
is over all sets of nondegenerate scattering paths.
(57) Wilson, E. B., Jr. An Introduction to Scientific Research; Dover:
New York, 1990. See especially pages 239-242 where a description of
Student’s t distribution can be found.

Dinitrogen CleaVage by Molybdenum(III) Complexes
J. Am. Chem. Soc., Vol. 118, No. 36, 1996 8637
were reduced accordingly (the assumption, borne out by the final fit
results, being that the ligand contribution is very little changed between
Mo[N(R)Ar]3 and (µ-N2){Mo[N(R)Ar]3}2).
additional ether was added (ca. 5 mL), and the mixture was stirred and
allowed to warm to ca. 10 °C. The aqueous layer was removed via
pipet, and then the yellow solution containing 15NNN(p-C6H4Me) was
filtered through a plug of anhydrous sodium sulfate, in a pipet, into a
flask for evaporation of volatile material. The solution was kept chilled
(ice water bath) while volatile matter was removed in Vacuo. The crude
15NNN(p-C6H4Me) was brought directly into the glovebox, where it
was weighed (550 mg, 4.13 mmol, 54%). A stock solution of 550 mg
of 15NNN(p-C6H4Me) in 9.2 g of cold ether was prepared, and stored
Raman Spectra of (µ-N2){Mo[N(R)Ar]3}2 and (µ-15N2){Mo[N(R)-
Ar]3}2. Samples of the two species were generated as described above
for the kinetic measurements. Spectra were obtained at room temper-
ature, as liquid samples at approximately the concentration of the
toluene stock solutions, and were taken quickly in order to minimize
decomposition to the corresponding terminal nitrido species. The
solution Raman spectra were recorded on a Spex 1677 (0.6 mm)
monochromator equipped with a Princeton Instruments, back thinned
512 × 512 liquid nitrogen cooled CCD detector. Excitation was with
a Coherent I-90-K krypton ion laser (647 nm) operating at a power of
20 mW. The plasma lines were cut off with an Applied Photophysics
laser filter monochromator. Laser radiation from the scattered light
was removed with a Kaiser Optical Systems holographic notch filter.
Optimal Raman spectra were obtained with 100-µm slits. Calibration
was obtained with the 1662- and 1404-cm-1 bands of DMF (dimethyl-
formamide). The data for the two samples and for the toluene
background were imported into the program Kaleidagraph. The toluene
spectrum was subtracted using an appropriate scaling factor; no other
manipulations were applied to the data.
1
at -35 °C. The H NMR spectrum (C6D6) of the crude 15NNN(p-
C6H4Me) was recorded: δ ) 6.685 (d, ArH), 6.753 (d, ArH), 1.964 (s,
ArCH3); this showed the crude material to be ca. 90% pure. GCMS
analysis was also consistent with assignment of the major component
of the crude product as 15NNN(p-C6H4Me). A solution of Mo[N(R)-
Ar]3 (480 mg, 0.75 mmol) in ca. 7 mL of Et2O was added dropwise,
via pipet, to a stirred orange solution of 15NNN(p-C6H4Me) (ca. 210
mg, 1.57 mmol, from a stock solution prepared as described above).
During the addition, the color of the reaction mixture was purple. After
ca. 5 more min, the color of the reaction mixture was brown. After the
volume of the reaction mixure was reduced to ca. 4 mL, it was stored
at -35 °C, leading to the formation of dark crystalline solid (160 mg
isolated, 0.243 mmol, 32%). HRMS (M+): calcd 659.44843, obsd
659.44595. 15N NMR (50.659 MHz): 840 ppm (s). 1H NMR: as
reported above for NMo[N(R)Ar]3. IR (KBr solution cells, pentane):
ν(Mo15N) ) 1014 cm-1 (calcd 1011 cm-1); the 1042 cm-1 band
attributed to ν(MoN) for the unlabeled derivative was absent from this
spectrum.
Synthesis of NMo[N(R)Ar]3 from Mo[N(R)Ar]3 and Dinitrogen.
A preparative-scale reaction between Mo[N(R)Ar]3 and dinitrogen (ca.
1 atm in a glovebox) was conducted by storing a solution of Mo[N(R)-
Ar]3 (300 mg, 0.467 mmol) in toluene (6 mL) at -35 °C for ca. 76 h.
The reaction’s progress was monitored occasionally by 2H NMR.
Aliquots used for monitoring progress were returned to the reaction
mixture so as not to diminish the yield. When conversion of Mo-
[N(R)Ar]3 (signal at 64 ppm) to (µ-N2){Mo[N(R)Ar]3}2 (signal at 14
ppm) was judged complete, the reaction mixture was allowed to warm
to room temperature (ca. 28 °C) for 5 h, during which time the purple
color dissipated and the mixture turned amber. 1H NMR analysis of
the crude material obtained upon removing volatile matter revealed
that NMo[N(R)Ar]3 had formed cleanly. The nitrido complex NMo-
[N(R)Ar]3 was isolated in 76% yield (0.233 g, 0.355 mmol) upon
recrystallization (Et2O, -35 °C). The sample of NMo[N(R)Ar]3 so
prepared was identical in its spectroscopic and physical properties to
samples of NMo[N(R)Ar]3 prepared by other means (see following
paragraph). IR and 15N NMR data for samples of 15NMo[N(R)Ar]3
prepared from Mo[N(R)Ar]3 and 1 atm of 99% 15N2 (see procedure
above for generation of (µ-15N2){Mo[N(R)Ar]3}2) were identical to those
reported below for 15NMo[N(R)Ar]3 prepared from Mo[N(R)Ar]3 and
15NNN(p-C6H4Me).
Preparation of Li[N(t-Bu)Ph](OEt2). A solution of HN(t-Bu)Ph,
20.0 g (0.134 mol), in a mixture of ether (40 mL) and pentane (95
mL) in a 500-mL Erlenmeyer flask was frozen solid in the glovebox
cold well. The vessel was removed from the well, and upon partial
thawing, a hexanes solution of n-butyllithium (92 mL, 1.6 M, 0.15
mol, 1.1 equiv) was added via pipet. The reaction mixture was stirred
for 4 h, with warming to ca. 28 °C. The mixture acquired a pale orange
color. The resulting solution was concentrated to ca. 50 mL under
vacuum, eliciting precipitation of the product as a white powder. The
powder was collected on a sintered glass frit, washed with pentane,
and dried under vacuum (21.0 g, 0.0916 mol, 68.3%). The material
thus obtained was >95% pure according to 1H NMR spectroscopy. 1H
NMR (300 MHz, C6D6, 20.8 °C): δ ) 7.198 (“t” [dd], 2H, Ph meta),
6.879 (d, 2H, Ph ortho), 6.474 (t, 1H, Ph para), 3.092 (q, 4H, O(CH2-
CH3)2), 1.552 (s, 9H, C(CH3)3), 0.865 (t, 6H, O(CH2CH3)2). 13C{1H}
NMR (75 MHz, C6D6, 20.8 °C): δ ) 159.24 (Ph ipso), 129.92 (Ph
ortho), 118.03 (Ph meta), 110.83 (Ph para), 65.124 (O(CH2CH3)2),
52.452 (C(CH3)3), 31.528 (C(CH3)3), 14.651 (O(CH2CH3)2).
Synthesis of NMo[N(R)Ar]3 from Mo[N(R)Ar]3 and Mesityl
Azide. A solution of Mo[N(R)Ar]3 (0.686 mmol, 441 mg) in 7 mL of
Et2O at 25 °C was added to a stirred solution of mesityl azide (161
mg, 0.997 mmol) in 3 mL of Et2O. Volatile material was removed in
Vacuo after 30 min at 30 °C. Pure amber NMo[N(R)Ar]3 was obtained
in 73% yield (0.501 mmol, 329 mg) after one recrystallization (Et2O,
-35 °C). 1H NMR (C6D6): 6.61 (s, 3H, para ArH), 5.94 (s, 6H, ortho
ArH), 2.03 (s, 18H, ArCH3), 1.62 (s, 9H, C(CD3)2CH3). 13C NMR
(C6D6): 150.69 (s, aryl ipso), 137.10 (q, aryl meta), 128.84 (d, aryl
ortho or para), 61.04 (s, C(CD3)2CH3), 33.06 (q, C(CD3)2CH3), 32.55
(m, C(CD3)2CH3), 21.34 (q, aryl meta CH3); one missing peak (d, aryl
ortho or para) was not located and is presumed to have been obscured
by the C6D6 signal. IR (KBr windows, pentane solution): 1042 cm-1
ν(MoN). Anal. Calcd for C36H36D18MoN4: C, 65.82; H, 8.29; N, 8.53.
Found: C, 66.14; H, 7.90; N, 8.45. HRMS (M+): calcd 658.45140,
obsd 658.44713.
Synthesis of 15NMo[N(R)Ar]3 from Mo[N(R)Ar]3 and 15NNN(p-
C6H4Me). Terminally-15N-labeled p-tolyl azide was prepared by the
(slightly modified) method60 of Hillhouse, Goeden, and Haymore: to
a 50-mL Erlenmeyer flask containing a magnetic stir bar was added
1.2127 g of p-tolylhydrazine hydrochloride (7.645 mmol), 7 mL of
distilled water, 2 mL of concentrated HCl, and ca. 5 mL of ether. The
mixture was chilled to 0 °C, and Na15NO2 (0.5457 g, 7.796 mmol), as
a solution in ca. 4 mL of chilled distilled water, was added over ca.
0.5 h dropwise via pipet. After all the Na15NO2 solution had been added,
Synthesis of Mo[N(t-Bu)Ph]3. To argon-sparged ether (100 mL)
were added sequentially Li[N(t-Bu)Ph](OEt2) (6.00 g, 26.2 mmol, 3.00
equiv) and MoCl3(THF)3 (3.65 g, 8.72 mmol) as white and orange
powders, respectively. The stirred reaction mixture was sparged with
Ar for ca. 1 min, following which the reaction vessel was fitted with
a rubber septum. The reaction mixture remained orange for ca. 5 min,
presumably due to unreacted MoCl3(THF)3, before it suddenly adopted
a dark mud-brown color. The mixture was stirred under an Ar
atmosphere for 90 min, at which point a 1H NMR spectrum of an aliquot
indicated that the major constituent species was the desired Mo[N(t-
Bu)Ph]3 (see below for spectral data). After being stirred for a total of
2 h, the reaction mixture was vacuum filtered through a bed of Celite
on a sintered glass frit. Volatile material was removed from the filtrate
in Vacuo, providing a dark brown solid. The solid was dissolved in
ether (ca. 15 mL), and the resulting solution was sparged with argon
for several minutes. The container was tightly sealed and stored at
-35 °C overnight, giving a crop of translucent burgundy crystals (1.54
g, 32.7%). The mother liquor was concentrated in Vacuo and stored
again at -35 °C to provide a second crop of burgundy crystals. Yield
of spectroscopically pure Mo[N(t-Bu)Ph]3: 2.61 g (55.4%). 1H NMR
(300 MHz, C6D6, 22 °C): δ ) 66.5 (br, 27H, t-Bu), 22.4 (br, 6H, Ph
meta or ortho), -23.0 (br, 6H, Ph meta or ortho), -49.8 (br, 3H, Ph
para). µeff (300 MHz, C7D8, 23 °C): 3.41 µB. We have not obtained
a satisfactory combustion analysis for this thermally-sensitive com-
pound; the following are typical results. Anal. Calcd for C30H42-
MoN3: C, 66.65; H, 7.83; N, 7.77. Found: C, 67.80; H, 8.63; N, 7.55.
(60) Hillhouse, G. L.; Goeden, G. V.; Haymore, B. L. Inorg. Chem. 1982,
21, 2064.

8638 J. Am. Chem. Soc., Vol. 118, No. 36, 1996
Laplaza et al.
SQUID Measurements on (µ-N2){Mo[N(t-Bu)Ph]3}2. The sample
was obtained in the following way. A saturated solution of Mo[N(t-
Bu)Ph]3 in pentane at -35 °C was stored in a vial capped by a septum
pierced by a needle to permit dinitrogen influx. Over the course of 2
days, the solution acquired a dark purple color, and a fine, intensely
purple-black microcrystalline solid presumed to be (µ-N2){Mo[N(t-
Bu)Ph]3}2 began to precipitate. The supernatant was transferred to
another vial and treated similarly to obtain further crops. The solid
was washed with cold (-35 °C) ether and pumped to dryness before
being stored at -35 °C. Purity was assayed via integration on the 1H
NMR spectrum of a sample dissolved in C6D6. Only (µ-N2){Mo[N(t-
Bu)Ph]3}2 and NMo[N(t-Bu)Ph]3 were observed; no Mo[N(t-Bu)Ph]3
was present. 1H NMR (300 MHz, C6D6, 20 °C): δ ) 12.62 (br s
[∆ν1/2 ) 33 Hz], 9H, C(CH3)3), ca. 9.8 (br s [∆ν1/2 ) 316 Hz], 2H,
phenyl ortho or meta), ca. 3.8 (br s [∆ν1/2 ) 175 Hz], 2H, phenyl ortho
or meta), 2.36 (br s [∆ν1/2 ) 22 Hz], 1H, phenyl para). 13C{1H} and
13C NMR (75.4 MHz, CDCl3, 22 °C): no peaks observed.
Squid magnetic measurements on samples of (µ-N2){Mo[N(t-Bu)-
Ph]3}2 were made as described for Mo[N(R)Ar]3 (Vide supra).
The data for (µ-N2){Mo[N(t-Bu)Ph]3}2 fit the Curie-Weiss law in
the temperature range 29-300 K (R ) 0.999 82). The calculated curve
(Figure 8) is the best least-squares fit of the observed susceptibility
data to the equation øΜ(obs) ) ((µ2)/(7.997584(T - θ))) - C. The
value of µ obtained in this manner was 2.85 µB, to be compared with
the spin-only value for two unpaired electrons, 2.83 µB. In the least-
squares fit, carried out using the General curve-fitting routine included
with the program Kaleidagraph, both θ and C were treated as variable
parameters. The obtained value of θ, the Weiss constant, was -14.626
K. The value of the constant C, included to represent the sum of all
temperature-independent contributions to the total observed susceptibil-
ity, including the diamagnetic correction and temperature-independent
paramagnetism, was -0.000 395 03. The raw data plotted in Figure 8
are deposited in the supporting information.
Synthesis of NMo[N(t-Bu)Ph]3 from Mo[N(t-Bu)Ph]3 and Mesityl
Azide. Neat mesityl azide (199.3 mg, 1.236 mmol, 1.33 equiv) was
added to a stirred orange-brown solution of Mo[N(t-Bu)Ph]3 (501.3
mg, 0.9273 mmol) in ether (10 mL) at -35 °C. The reaction mixture
acquired a dark purple color very rapidly; this color faded substantially
but did not entirely disappear as the mixture was stirred at 29 °C for
3 h. Ether was then removed in Vacuo, and the resulting dark solid
crystallized from ether at -35 °C. Recrystallization from ether
delivered pure NMo[N(t-Bu)Ph]3 as transparent yellow-orange crystals
(374.6 mg [two crops], 0.6754 mmol, 72.83%) suitable for an X-ray
diffraction study. 1H NMR (300 MHz, CDCl3, 22 °C): δ ) 7.031 (tt,
1H, para), 6.931 (“t” [dd], 2H, meta), 5.933 (d, 2H, ortho), 1.397 (s,
9H, C(CH3)3). 13C{1H} NMR (75.4 MHz, CDCl3, 22 °C): δ ) 150.44
(phenyl ipso), 130.42 (phenyl meta), 128.04 (phenyl ortho), 125.02
(phenyl para), 61.43 (C(CH3)3), 32.98 (C(CH3)3). 13C NMR (75.4 MHz,
CDCl3, 22 °C): δ ) 150.40 (s, phenyl ipso), 130.40 (“dt” [ddd], phenyl
meta), 128.04 (dd, phenyl ortho), 125.02 (dt, phenyl para), 61.46 (s,
C(CH3)3), 32.98 (q, C(CH3)3). Anal. Calcd for C30H42MoN4: C, 64.97;
H, 7.63; N, 10.10. Found: C, 64.79; H, 7.86; N, 10.18.
Z ) 16. The space group was found to be I4h3d. The absorption
coefficient was 0.467 mm-1, the calculated density F ) 1.246 g/cm3,
and F(000) ) 4672. Data were collected using a Siemens platform
goniometer with a CCD detector at 208(2) K using molybdenum KR
radiation [λ ) 0.710 73 Å] in the θ range 21.9-23.24° with limiting
indices -13 e h e 25, -18 e k e 25, -24 e l e 17. Of the 12 790
reflections collected, 1415 were independent (Rint ) 0.0372).
A
semiempirical absorption correction, from ψ scans, was applied. The
structure was solved by direct methods (SHELXTL V5.0, Sheldrick,
G. M., and Siemens Industrial Automation, Inc., 1995) in conjunction
with standard difference Fourier techniques. Least-squares refinement
based upon F2 with 1413 data, zero restraints, and 107 parameters
converged with final residuals: R1 ) 0.0247, wR2 ) 0.0491, and GOF
) 1.265 based upon I > 2σ(I). The largest peak and hole in the final
difference Fourier map were respectively +0.149 and -0.191 e Å-3
.
Acknowledgment. This work made use of MRSEC Shared
Facilities supported by the National Science Foundation under
Award No. DMR-9400334. The Stanford Synchrotron Radia-
tion Laboratory is funded by the Department of Energy, Office
of Basic Energy Sciences. The Biotechnology Program is
supported by the National Institutes of Health, Biomedical
Research Technology Program, Division of Research Resources.
Further support is provided by the Department of Energy, Office
of Health and Environmental Research. We are indebted to
Roger C. Prince of Exxon Research and Engineering Company
for his assistance in data collection and Martin J. George of
SSRL for use of his data collection software. We are also
grateful to John Rehr of the University of Washington for
providing us with a special extended k-range version of his
program feff. Raman spectra were collected in the Harrison
Spectroscopy Laboratory at MIT, supported in part by the
National Institutes of Health. C.C.C. thanks Professors Richard
R. Schrock and Barry K. Carpenter for helpful discussions. For
funding, C.C.C. thanks the National Science Foundation
(CAREER Award CHE-9501992), DuPont (Young Professor
Award), the Packard Foundation (Packard Foundation Fellow-
ship), Union Carbide (Innovation Recognition Award), and 3M
(Innovation Fund Award). C.E.L. thanks MIT’s UROP (Under-
graduate Research Opportunities Program) for funding. M.J.A.J.
is grateful for an NSERC graduate research fellowship.
Supporting Information Available: SQUID magnetic data
for Mo[N(R)Ar]3 and (µ-N2){Mo[N(t-Bu)Ph]3}2, kinetic data
for conversion of (µ-N2){Mo[N(R)Ar]3}2 to 2 equiv of NMo-
[N(R)Ar]3, kinetic data for the fragmentation of (µ-
15N2){Mo[N(R)Ar]3}2 to 2 equiv of 15NMo[N(R)Ar]3, and tables
of atom coordinates, anisotropic thermal parameters, and bond
distances and angles for the X-ray structures of Mo[N(R)Ar]3
and NMo[N(t-Bu)Ph]3 (21 pages). See any current masthead
page for ordering and Internet access instructions.
X-ray Crystal Structure of NMo[N(t-Bu)Ph]3. Data were collected
using a crystal of NMo[N(t-Bu)Ph]3 having dimensions 0.5 × 0.4 ×
0.4 mm. The crystal system was cubic, a ) b ) c ) 22.783(2) Å, and
R ) â ) γ ) 90°, leading to a unit cell volume V ) 11826(2) Å3 with
JA960574X