Electronic structure, spectroscopic properties, and reactivity of molybdenum and tungsten nitrido and imido complexes with diphosphine coligands: Influence of the trans ligand

Article abstract of DOI:10.1021/ic060141n

A series of molybdenum and tungsten nitrido, [M(N)(X)(diphos)₂], and imido complexes, [M(NH)(X)(diphos)₂)]Y, (M = Mo, W) with diphosphine coligands (diphos = dppe/depe), various trans ligands (X = N ₃^-, Cl^-, NCCH₃) and different counterions (Y^- = Cl^-, BPh₄^-) is investigated. These compounds are studied by infrared and Raman spectroscopies; they are also studied with isotope-substitution and optical-absorption, as well as emission, spectroscopies. In the nitrido complexes with trans-azido and -chloro coligands, the metal-N stretch is found at about 980 cm^-1; upon protonation, it is lowered to about 920 cm^-1. The ¹A₁ → ¹E (n → π*) electronic transition is observed for [Mo(N)(N₃)(depe)₂] at 398 nm and shows a progression in the metal-N stretch of 810 cm^-1. The corresponding ³E → ¹A (π* → n) emission band is observed at 542 nm, exhibiting a progression in the metal-N stretch of 980 cm^-1. In the imido system [Mo(NH)(N₃)(depe)2]BPh ₄, the n → π* transition is shifted to lower energy (518 nm) and markedly decreases in intensity. In the trans-nitrile complex [Mo(N)(NCCH₃)(dppe)₂]BPh₄, the metal-N(nitrido) stretching frequency increases to 1016 cm^-1. The n → π* transition now is found at 450 nm, shifting to 525 nm upon protonation. Most importantly, the reduction of this nitrido trans-nitrile complex is drastically facilitated compared to its counterparts with anionic trans-ligands (E_p^red = -1.5 V vs Fc⁺/Fc). On the other hand, the basicity of the nitrido group is decreased (pK _a{[Mo(NH)(NCCH₃)(dppe)₂](BPh₄) ₂} = 5). The implications of these findings with respect to the Chatt cycle are discussed.

Full text of DOI:10.1021/ic060141n

Inorg. Chem. 2006, 45, 5044−5056
Electronic Structure, Spectroscopic Properties, and Reactivity of
Molybdenum and Tungsten Nitrido and Imido Complexes with
Diphosphine Coligands: Influence of the trans Ligand†
Klaus Mersmann,‡ Andreas Hauser,§ Nicolai Lehnert,‡ and Felix Tuczek*,‡
Institut fu¨r Anorganische Chemie, Christian Albrechts UniVersita¨t, D-24098 Kiel, Germany, and
De´partement de Chimie Physique, Baˆtiment de Science II, UniVersite´ de Gene`Ve,
30 quai Ernest Ansermet, 1211 Gene`Ve 4, Switzerland
Received January 24, 2006
A series of molybdenum and tungsten nitrido, [M(N)(X)(diphos)2], and imido complexes, [M(NH)(X)(diphos)2)]Y, (M
Mo, W) with diphosphine coligands (diphos dppe/depe), various trans ligands (X
N3-, Cl-, NCCH3) and
different counterions (Y- Cl-, BPh4-) is investigated. These compounds are studied by infrared and Raman
spectroscopies; they are also studied with isotope-substitution and optical-absorption, as well as emission,
spectroscopies. In the nitrido complexes with trans-azido and -chloro coligands, the metal N stretch is found at
about 980 cm-1; upon protonation, it is lowered to about 920 cm-1. The 1A1 1E (n f π*) electronic transition
is observed for [Mo(N)(N3)(depe)2] at 398 nm and shows a progression in the metal
N stretch of 810 cm-1. The
corresponding 3E 1A (
n) emission band is observed at 542 nm, exhibiting a progression in the metal
stretch of 980 cm-1. In the imido system [Mo(NH)(N3)(depe)2]BPh4, the n f π* transition is shifted to lower energy
(518 nm) and markedly decreases in intensity. In the trans-nitrile complex [Mo(N)(NCCH3)(dppe)2]BPh4, the metal
)
)
)
)
−
f
−
f
π
*
f
−N
−
N(nitrido) stretching frequency increases to 1016 cm-1. The n f π* transition now is found at 450 nm, shifting to
525 nm upon protonation. Most importantly, the reduction of this nitrido trans-nitrile complex is drastically facilitated
compared to its counterparts with anionic trans-ligands (Epred ) −1.5 V vs Fc+/Fc). On the other hand, the basicity
of the nitrido group is decreased (pKa{[Mo(NH)(NCCH3)(dppe)2](BPh4)2} ) 5). The implications of these findings
with respect to the Chatt cycle are discussed.
1. Introduction
Many years before, Pickett and Talarmin, in a classic
experiment, had already succeeded in producing ammonia
from N2 in a cyclic fashion; however, it was a stepwise
procedure with a much lower yield.4 Their process was
based on the complex [W(NNH2)TsO(dppe)2]+ (TsO )
The transition-metal-mediated conversion of dinitrogen to
ammonia under ambient conditions is a subject of continuing
interest.1 Recently, Yandulov and Schrock succeeded in
catalytically performimg this reaction using a Mo(III)
complex with a sterically shielding triamidoamine ligand.2
With decamethylchromocene as the reductant and lutidinium
BAr′4 as the acid (Ar′ ) 3,5-(CF3)2C6H3), they were able to
generate NH3 from N2 at room temperature and normal
pressure, achieving 6 cycles with an overall yield of 65%.
On the basis of the available experimental information, a
mechanism was proposed for this reaction (Schrock cycle)2
and reproduced by DFT calculations.3
-
p-CH3C6H4SO3 ) which was electrolyzed to give the parent
dinitrogen complex [W(N2)2(dppe)2], ammonia, and H2.
Protonation of the bis(dinitrogen) complex reformed the
starting complex; this procedure was repeated twice. The
(2) (a) Yandulov, D. V.; Schrock, R. R. J. Am. Chem. Soc. 2002, 124,
6252. (b) Yandulov, D. V.; Schrock, R. R.; Rheingold, A. L.;
Ceccarelli, C.; Davis, W. M. Inorg. Chem. 2003, 42, 796. (c) Yandulov,
D. V.; Schrock, R. R. Science 2003, 301, 76. (d) Ritleng, V.; Yandulov,
D. V.; Weare, W. W.; Schrock, R. R.; Hock, A. S.; Davis, W. M. J.
Am. Chem. Soc. 2004, 126, 6150. (e) Yandulov, D. V.; Schrock, R.
R. Inorg. Chem. 2005, 44, 1103. (f) Yandulov, D. V.; Schrock, R. R.
Inorg. Chem. 2005, 44, 5542. (g) Yandulov, D. V.; Schrock, R. R.
Can. J. Chem. 2005, 83, 341.
(3) Studt, F.; Tuczek, F. Angew. Chem., Int. Ed. 2005, 44, 5639.
(4) (a) Pickett, C. J.; Talarmin, J. Nature 1985, 317, 652. (b) Pickett, C.
J.; Ryder, K. S.; Talarmin, J. J. Chem. Soc., Dalton Trans. 1986, 7,
1453.
† Reduction and protonation of end-on terminally coordinated dinitrogen.
Part VI.
* To whom correspondence should be addressed. E-mail: ftuczek@
ac.uni-kiel.de.
‡ Christian Albrechts Universita¨t.
§ Universite´ de Gene`ve.
(1) MacKay, B. A.; Fryzuk, M. D. Chem. ReV. 2004, 104, 385.
5044 Inorganic Chemistry, Vol. 45, No. 13, 2006
10.1021/ic060141n CCC: $33.50
© 2006 American Chemical Society
Published on Web 05/25/2006

Molybdenum and Tungsten Nitrido and Imido Complexes
Scheme 1
corresponding chemistry proceeds within the Chatt cycle,
which provided the first complete scheme for the transition-
metal-mediated conversion of N2 to ammonia (Scheme 1).5
The first two steps of this cycle involve the protonation of
molybdenum and tungsten dinitrogen complexes with diphos-
phine coligands to give the corresponding NNH2 complexes.
We have investigated the molybdenum and tungsten N2,
NNH, and NNH2 intermediates appearing in this chemistry
by a variety of spectroscopic methods coupled to DFT
calculations.6 In the presence of strong acids, an additional
proton can be added giving NNH3 (hydrazidium) species.7
At this stage, the N-N triple bond of N2 has been reduced
to a single bond and can be cleaved by adding two electrons
from an external reductant or electrochemically. Alterna-
tively, two-electron reduction can be performed at the NNH2
stage, leading to the elimination of the trans ligand and
formation of five-coordinate Mo(II) intermediates. This
process has been studied on the basis of alkylated deriva-
tives.8 If the doubly reduced, five-coordinate intermediates,
[M(II)(dppe)2NNR2], are protonated, the N-N bond cleaves,
and secondary amines, as well as M(IV) nitrido or imido
complexes, are formed.9 We spectroscopically and kinetically
followed the splitting of the N-N bond and explored various
mechanistic scenarios to account for the experimental
observations.10
valent ammine complexes that are able to exchange NH3
against N2. One of the problems associated with this
chemistry are the strongly negative potentials which, in
particular, are needed to regenerate the +I and zerovalent
Mo and W species.11 The presence of such negative potentials
accounts for the large fraction of H2 that is formed from the
acids employed in these experiments (vide supra). One of
the possibilities to shift the reduction potentials to less
negative values is to use a weakly donating or even labile
ligand in the trans position to the nitrido/imido group.
Replacing, for example, halide by acetonitrile (i) increases
the total charge of the complex because of the exchange of
a negatively charged by a neutral ligand, (ii) increases the
effective nuclear charge, Zeff, of the metal center because of
a decreased ligand σ/π donor strength, and in particular (iii),
lowers the dxz/dyz orbitals of the metal centers because of
the lack of π-donor capability. All of these effects should
act to make the nitrido complex easier to reduce (i.e., shift
the reduction potential to less negative values).
Herein, we investigate the influence of the trans ligand
on the electronic structure, the spectroscopic properties and
reactivity of the molybdenum and tungsten nitrido com-
plexes. To this end, a series of molybdenum and tungsten
nitrido and imido complexes with dppe and depe coligands
and trans ligands of different σ/π-donor strengths (azide,
chloride, acetonitrile) is investigated by UV-vis, IR, and
Raman spectroscopies and electrochemistry. The results are
interpreted by means of DFT calculations, and the general
implications with respect to a catalytic ammonia synthesis
on the basis of the Chatt cycle are discussed.
In the second part of the Chatt cycle, Mo and W nitrido
complexes have to be converted to the corresponding low-
(5) (a) Henderson, R. A.; Leigh, G. J.; Pickett, C. J. AdV. Inorg. Chem.
Radiochem. 1983, 27, 197. (b) Pickett, C. J. J. Biol. Inorg. Chem.
1996, 1, 601.
(6) (a) Lehnert, N.; Tuczek, F. Inorg. Chem. 1999, 38, 1659. (b) Lehnert,
N.; Tuczek, F. Inorg. Chem. 1999, 38, 1671.
(7) (a) Galindo, A.; Hills, A.; Hughes, D. L.; Richards, R. L. J. Chem.
Soc., Chem. Commun. 1987, 1815. (b) Horn, K. H.; Lehnert, N.;
Tuczek, F. Inorg. Chem. 2003, 42, 1076.
(8) Hussain, W.; Leigh, G. J.; Pickett, C. J. J. Chem. Soc., Chem. Commun.
1982, 747.
2. Experimental Section
Sample Preparation and Isotopic Substitution. A set of
molybdenum and tungsten nitrido complexes with dppe as the
coligand was synthesized (i.e., the trans azido complexes [Mo(N)-
(9) Henderson, R. A.; Leigh, G. J.; Pickett, C. J. J. Chem. Soc., Dalton
Trans. 1989, 425.
(10) Mersmann, K.; Horn, K. H.; Bo¨res, N.; Lehnert, N.; Studt, F.; Paulat,
F.; Peters, G.; Tuczek, F. Inorg. Chem. 2005, 44, 3031.
(11) Alias, Y.; Ibrahim, S. K.; Queiros, M. A.; Fonseca, A.; Talarmin, J.;
Volant, F.; Pickett, C. J. J. Chem. Soc., Dalton Trans. 1997, 4807.
Inorganic Chemistry, Vol. 45, No. 13, 2006 5045

Mersmann et al.
(N3)(dppe)2] (1) and [W(N)(N3)(dppe)2] (2), as well as the trans
chloro complexes [Mo(N)(Cl)(dppe)2] (3) and [W(N)(Cl)(dppe)2]
(4)). These complexes were prepared by following literature
procedures.12 Another reaction route,8,9,13 starting from [W(15N2)2-
(dppe)2], was employed to prepare the 15N isotopomer of complex
4 (4(15N)). Furthermore we synthesized the depe complex [Mo(N)-
(N3)(depe)2] (5)14 and the trans acetonitrile complex [Mo(N)-
(NCCH3)(dppe)2]BPh4 (6).9
observed. A cross-check with a drop of rubber cement without
compound 5 gave a vanishing intensity at 11 K, proving that the
luminescence in fact derives from [Mo(N)(N3)(depe)2]. The lumi-
nescence intensity is strongly temperature dependent and at 77 K
already quite weak.
Electrochemical Investigations. Cyclic voltammograms were
recorded with an EG and G PAR model 273A potentiostat
controlled by EG and G PAR M270 software using a Pt knob
working electrode, a Pt counter electrode, and a silver reference
electrode. A solution of [MoN(NCCH3)(dppe)2]BPh4 (0.5 mM) and
[NBu4][PF6] (0.1M) in acetonitrile was recorded at 25 °C. The scan
rate was set to 100 mVs-1. The potentials are referenced against
ferrocenium-ferrocene.
Normal Coordinate Analysis. Normal coordinate calculations
were performed using the QCPE computer program by Peterson
and McIntosh. It involves solutions of the secular equation GFL )
ΛL by the diagonalization procedure of Miyazawa.16 The calcula-
tions are based on a general valence force field, and the force
constants are refined using the nonlinear optimization of the simplex
algorithm according to Nelder and Mead.17 Normal coordinate
analysis is based on the QCB-NCA procedure which involves
generation of an initial force field by DFT methods.6
DFT Calculations. Spin-restricted DFT calculations were per-
formed for the nitrido systems, [Mo(N)(N3)(diphos)2] (I), [Mo(N)-
(Cl)(diphos)2] (III), and [Mo(N)(NCCH3)(diphos)2]+ (VI), and for
the corresponding imido systems, [Mo(NH)(N3)(diphos)2]+ (I*H+),
[Mo(NH)(Cl)(diphos)2]+ (III*H+), and [Mo(NH)(NCCH3)(diphos)2]2+
(VI*H+), using Becke’s three parameter hybrid functional with the
correlation functional of Lee, Yang, and Parr (B3LYP).18 All di-
phosphine ligands (dppe, depe) are simplified to H2PCH2CH2PH2
(“diphos”). I, III, and VI are models of compounds 1/2/5, 3/4, and
6, respectively, whereas I*H+, III*H+, and VI*H+ are models of
compounds 1*HBPh4/2*HBPh4/5*HBPh4, 3*HBPh4/4*HBPh4,
and 6*HBPh4, respectively. The LANL2DZ basis set was used for
the calculations. It applies the Dunning/Huzinaga full double-ú19
basis functions on the first row and the Los Alamos effective core
potentials plus DZ functions on all other atoms.20 Charges were
analyzed using the natural bond orbital (NBO) formalism (natural
population analysis, NPA).21 All computational procedures were
used as implemented in the Gaussian 98 package.22 Wave functions
were plotted with GaussView. The f matrix in internal coordinates
was extracted from the Gaussian output using the program
REDONG.23
Protonation of the nitrido complexes, No (No ) 1, 2, ..., 6), was
performed with various acids HX, leading to the imido derivatives
No*HX. Condensation of dry HCl gas on frozen dichloromethane
solutions of the No complexes gave the chloro-imido complexes
No*HCl (No ) 2, 3, 4). The products were precipitated by adding
hexane to the solutions. DCl gas was used to generate the deuterated
counterparts, No*DCl (No ) 2, 3, 4). Imido complexes with
tetraphenylborate as counterion (No*HBPh4) were prepared using
lutidinium tetraphenylborate (HLutBPh4) as acid in thf (No ) 1,
2, 5, 6). The products were obtained upon addition of methanol
and diethyl ether to the solutions. Treatment with methanol-d1 was
used to convert the imido complexes to their deuterated counterparts,
No*DBPh4 (No ) 1, 2, 5).
All reactions and sample preparations were performed under a
nitrogen or argon atmosphere using Schlenk techniques. All solvents
were dried under argon. Sample manipulations for vibrational and
optical spectroscopy were carried out in a glovebox.
Elemental Analyses. The elemental analyses were performed
on a CHN-O-RAPID (Heraeus) instrument in little tin containers
(Elementar). Observed C/H/N values (calcd values):151 65.7/6.4/
5.1 (65.8/5.1/5.9), 2 59.5/5.2/5.9 (60.2/4.7/5.4), 4 59.3/4.9/1.4 (60.6/
4.7/1.4), 4(15N) 58.9/4.7/1.5 (60.6/4.7/1.4), 5 41.3/8.3/8.5 (42.6/
8.6/9.9), 6 74.4/7.2/2.3 (73.9/5.7/2.2), 2*HCl 56.5/5.2/4.4 (58.2/
4.6/5.2), 3*HCl 63.1/6.0/1.5 (63.8/5.1/1.4), 4*HCl 56.6/5.1/1.4
(58.6/4.6/1.3), 1*HBPh4 70.9/6.2/3.9 (71.9/5.5/4.4), 2*HBPh4 67.9/
6.0/3.5 (67.3/5.1/4.1), 5* HBPh4 59.8/8.6/6.0 (59.7/7.9/6.3).
Spectroscopic Instrumentation. FT-IR spectra were recorded
in KBr pellets on a Mattson Genesis type I spectrometer. Optical
absorption spectra of solutions were recorded on a Analytic Jena
Specord S100 spectrometer. UV-vis spectra on neat compounds
pressed between sapphire windows were recorded at 10 K using a
Varian Cary 5 UV-vis-NIR spectrometer equipped with a CTI
cryocooler. Luminescence spectra were measured with a SPEX
Fluorolog System equipped with an optical helium cryostat (λex
)
400 nm).
Luminescence Spectra of [Mo(N)(N3)(depe)2] (5). A little drop
of rubber cement was put on a small copper plate to hold a dense
layer of polycrystalline material. At 11 K, strong luminescence was
(16) Miyazawa, T. J. Chem. Phys. 1958, 29, 246.
(17) Nelder, J. A.; Mead, R. Comput. J. 1965, 7, 308.
(18) Becke, A. D. J. Chem. Phys. 1993, 98, 5648.
(19) Dunning, T. H., Jr.; Hay, P. J. In Modern Theoretical Chemistry;
Schaefer, H. F., III, Ed.; Plenum: New York, 1976.
(12) (a) Dilworth, J. R.; Richards, R. L. Inorg. Synth. 1980, 20, 119. (b)
George, T. A.; Noble, M. E. Inorg. Chem. 1978, 17, 1678. (c) Chatt,
J.; Dilworth, J. R. J. Chem. Soc., Chem. Commun. 1975, 983. (d) Chatt,
J.; Dilworth, J. R. J. Indian Chem. Soc. 1977, 54, 13. (e) Bevan, P.
C.; Chatt, J.; Dilworth, J. R.; Henderson, R. A.; Leigh, G. J. J. Chem.
Soc., Dalton Trans. 1982, 821.
(13) Chatt, J.; Hussain, W.; Leigh, G. J.; Terreros, F. P. J. Chem. Soc.,
Dalton Trans. 1980, 1408.
(14) (a) Hussain, W.; Leigh, G. J.; Ali, H. M.; Pickett, C. J.; Rankin, D.
A. J. Chem. Soc., Dalton Trans. 1984, 1703. (b) Hughes, D. L.;
Mohammed, M. Y.; Pickett, C. J. J. Chem. Soc., Dalton Trans. 1990,
2013.
(15) Larger deviations between the calculated and observed values are
observed for compounds 1 and 6. As evident from NMR, the deviations
in compound 1 are the result of trimethylsilyl impurities. Trimethylsilyl
azide was added to convert [Mo(N2)2(dppe)2] to [MoN(N3)(dppe)2]
(1). In compound 6, the deviation is caused by an admixture of
HNEt3BPh4; triethylamine was added as a base to deprotonate
[Mo(NH)Cl(dppe)2]Cl and NaBPh4 was added to precipitate [Mo(N)-
(CH3CN)(dppe)2]BPh4 (6).
(20) (a) Hay, P. J.; Wadt, W. R. J. Chem. Phys. 1985, 82, 270, 299. (b)
Wadt, W. R.; Hay, P. J. J. Chem. Phys. 1985, 82, 284.
(21) (a) Foster, J. P.; Weinhold, F. J. Am. Chem. Soc. 1980, 102, 7211. (b)
Rives, A. B.; Weinhold F. Int. J. Quantum Chem. Symp. 1980, 14,
201. (c) Reed, A. E.; Weinstock, R. B.; Weinhold, F. J. Chem. Phys.
1985, 83, 735. (d) Reed, A. E.; Curtiss, L. A.; Weinhold, F. Chem.
ReV. 1988, 88, 899.
(22) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
M. A.; Cheeseman, J. R.; Zakrzewski, V. G.; Montgomery, J. A., Jr.;
Stratmann, R. E.; Burant, J. C.; Dapprich, S.; Millam, J. M.; Daniels,
A. D.; Kudin, K. N.; Strain, M. C.; Farkas, O.; Tomasi, J.; Barone,
V.; Cossi, M.; Cammi, R.; Mennucci, B.; Pomelli, C.; Adamo, C.;
Clifford, S.; Ochterski, J.; Petersson, G. A.; Ayala, P. Y.; Cui, Q.;
Morokuma, K.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.;
Foresman, J. B.; Cioslowski, J.; Ortiz, J. V.; Stefanov, B. B.; Liu, G.;
Liashenko, A.; Piskorz, P.; Komaromi, I.; Gomperts, R.; Martin, R.
L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara,
A.; Gonzalez, C.; Challacombe, M.; Gill, P. M. W.; Johnson, B. G.;
Chen, W.; Wong, M. W.; Andres, J. L.; Head-Gordon, M.; Replogle,
E. S.; Pople, J. A. Gaussian 98; Gaussian, Inc.: Pittsburgh, PA, 1998.
5046 Inorganic Chemistry, Vol. 45, No. 13, 2006

Molybdenum and Tungsten Nitrido and Imido Complexes
Figure 1. IR spectra of nitrido complexes [W(N)(N3)(dppe)2] (2), [W(N)(Cl)(dppe)2] (4), and [W(15N)(Cl)(dppe)2] (4(15N)).
3. Results and Analysis
the antisymmetric combination is found at 2029 cm-1. In
the spectrum of the imido complex, these peaks are shifted
to 1376 and 2080 cm-1, respectively. The metal-N stretch
shifts to 922 cm-1. Upon deuteration, this signal shifts to
892 cm-1. The decrease of the metal-(NH) stretching
frequency upon deuteration is due to an increasing mass of
the NH moiety by one, and therefore, it has an effect
comparable to the 15N substitution in the nitrido complex.
At 3250 cm-1, a broad band can be identified, which in the
deuterated complex is shifted to 2430 cm-1. This signal
derives from the imido N-H stretching vibration and an
O-H stretch which is associated with methanol being present
as a crystal solvent. Unfortunately, attempts to isolate the
imido complex in the absence of methanol failed. In addition
to the nitrido complex, the spectra of the imido system also
contain the signals of the counterion, tetraphenylborate. The
Raman spectra of these compounds, which are given in
Figure S1 (Supporting Information), exhibit bands at similar
positions. The results of the IR and Raman experiments are
collected in Table 1.
A DFT frequency analysis of the optimized model
[Mo(N)(N3)(diphos)2] (I) of 1/2/5 predicts a frequency of
1049 cm-1 for the metal-N stretch, slightly higher than the
observed value. This corresponds to the finding that the
metal-N distance is calculated slightly too short (see below).
The frequency analysis further provides an f matrix which
is used as an initial guess for a normal coordinate analysis
(quantum chemistry-based normal coordinate analysis, QCB-
NCA).6 The results of the fitting procedure are collected in
Table 2. In particular, the fitting of the metal-N stretch to
the experimental value gives a corresponding force constant
of 7.065 mdyn/Å.
Vibrational Spectroscopy and DFT Frequency Analy-
sis. The infrared spectra of the tungsten-dppe nitrido
complexes, [W(N)(N3)(dppe)2] (2) and [W(N)(Cl)(dppe)2]
(4), are presented in Figure 1, along with the IR spectrum
of the 15N isotopomer of 4, 4(15N). On the right-hand side
of Figure 1, the region between 850 cm-1 and 1050 cm-1 is
presented on a larger scale. The spectrum of the trans-azido-
nitrido complex 2 is dominated by the antisymmetric N-N
vibration of the azide ligand, νas(N3 ), at 2062 cm-1. Its
-
-
symmetric counterpart, νs(N3 ), can be found at 1354 cm-1.
The metal-nitrido stretching vibration is located at 979 cm-1.
The remaining bands are mostly caused by vibrations of the
dppe ligand. The IR spectrum of the molybdenum analogue
of 2, [Mo(N)(N3)(dppe)2] (1, not shown), is very similar (i.e.,
the antisymmetric azide vibration is found at 2051 cm-1,
the corresponding symmetric mode is at 1348 cm-1, and the
metal-nitrido stretching vibration appears at 975 cm-1). As
is evident from the spectrum of the nitrido-chloro complex
[W(N)(Cl)(dppe)2] (4), exchange of the trans ligand from
azide to chloride has almost no effect on the position of the
metal-nitrido vibration, which appears in the spectrum of
4 at 978 cm-1. In the 15N isotopomer [W(15N)(Cl)(dppe)2],
4(15N), this vibration shifts by about 30 cm-1 to 947 cm-1.
In Figure 2, the infrared spectra of [Mo(NH)(N3)(depe)2]-
BPh4 (5*HBPh4) and [Mo(ND)(N3)(depe)2]BPh4 (5*DBPh4)
are presented along with the spectrum of the parent nitrido
complex [Mo(N)(N3)(depe)] (5). In the latter system, the
metal-N vibration is easily identified at 980 cm-1; the
symmetric azide stretch is located at 1328 cm-1, whereas
(23) Allouche, A.; Pourcin, J. Spectrochim. Acta 1993, 49A, 571.
Inorganic Chemistry, Vol. 45, No. 13, 2006 5047

Mersmann et al.
Figure 2. IR spectra of nitrido and imido complexes with molybdenum and depe and a tetraphenylboron counterion: [Mo(N)(N3)(depe)2] (5), [Mo(NH)-
(N3)(depe)2]BPh4 (5*HBPh4), and [Mo(ND)(N3)(depe)2]BPh4 (5*DBPh4).
-
Table 1. Comparison of Metal-N, N-H, Symmetric (sym) N3-, and Antisymmetric (asym) N3 Vibrations in Different Nitrido and Imido Complexesa
-
-
M-N
N-H
sym N3
asym N3
[Mo(N)(N3)(dppe)2]
[W(N)(N3)(dppe)2]
1
2
4
975
979
978
947
980
1016
950
912
950
910
946
916
905
920
891
924
893
925
1348
1354
2051
2062
[W(N)(Cl)(dppe)2]
[W(15N)(Cl)(dppe)2]
4(15N)
5
[Mo(N)(N3)(depe)2]
1330
2029
[Mo(N)(NCCH3)(dppe)2]BPh4
[W(NH)(N3)(dppe)2]Cl
[W(ND)(N3)(dppe)2]Cl
[Mo(NH)(Cl)(dppe)2]Cl
[Mo(ND)(Cl)(dppe)2]Cl
[W(NH)(Cl)(dppe)2]Cl
[W(ND)(Cl)(dppe)2]Cl
[Mo(NH)(N3)(dppe)2]BPh4
[W(NH)(N3)(dppe)2]BPh4
[W(ND)(N3)(dppe)2]BPh4
[Mo(NH)(N3)(depe)2]BPh4
[Mo(ND)(N3)(depe)2]BPh4
[Mo(NH)(NCCH3)2](BPh4)2
6
2*HCl
2*DCl
3*HCl
3*DCl
4*HCl
4*DCl
1*HBPh4
2*HBPh4
2*DBPh4
5*HBPh4
5*DBPh4
6*HBPh4
2500-3000
1750-2250
2500-3000
1750-2250
2350-3100
1800-2250
3250
3300
2470
3250
2430
1390
1391
2093
2093
1387
1395
1395
1376
1376
2092
2104
2104
2080
2080
3150-3350
a All data given in cm-1
.
Table 2. Vibrational Data of Nitrido and Imido Complexes and
Comparison to Calculated Results and Force Constants
complexes. The N-H stretching vibration is calculated to
be 3586 cm-1, and the N-D stretch is calculated to be 2639
cm-1, also higher than observed. After normal coordinate
analysis, force constants of 6.7 mdyn/Å for the metal-N
and 5.848 mdyn/Å for the N-H bonds are obtained. This
demonstrates that the strength of the metal-N bond is only
slightly (i.e., by 0.365 mdyn/Å) reduced upon protonation.
Table 2 shows the results of this analysis.
exptl
B3LYP
NCA
f(NCA)
(mdyn/Å)
[Mo(N)(N3)(diphos)2]
( cm-1
)
(cm-1
)
(cm-1
)
Mo-N
980
947
810
1049
1019
980 cm-1
951
810
7.065
4.9
Mo-15N
Mo-N (1E exc. st.)
exptl
B3LYP
NCA
f (NCA)
(mdyn/Å)
[Mo(NH)(N3)(diphos)2]+
(cm-1
)
(cm-1
)
(cm-1
)
Counterions which form hydrogen bridges with the imido
proton have an influence on the metal-(NH) and N-H
vibrations. This can be seen in the spectral pattern of the
N-H vibration in [W(NH)(Cl)(dppe)2]Cl (4*HCl) (Figure
S2, Supporting Information), which appears in the range
between 2400 and 2900 cm-1. In this case, the metal-(NH)
vibration is centered at 946 cm-1 and appears to be split. In
the deuterated complex, the signal intensities are markedly
decreased. The N-D vibration appears in the range between
1750 and 2200 cm-1, and the metal-N vibration is shifted
N-H
3250
2430
924
3586
2639
964
3261
2402
924
5.848
6.700
N-D
Mo-NH
Mo-ND
893
932
892
A DFT frequency analysis was also performed for the
imido-azido model system [Mo(NH)(N3)(diphos)2]+ (I*H+)
of compounds 1*HBPh4, 2*HBPh4, and 5*HBPh4. The
calculated metal-N vibrational frequencies of 964 cm-1 for
the protonated and 932 cm-1 for the deuterated complexes
are again somewhat too high, as found for the nitrido
5048 Inorganic Chemistry, Vol. 45, No. 13, 2006

Molybdenum and Tungsten Nitrido and Imido Complexes
Figure 3. IR spectrum of nitrido-trans-nitrile complex [Mo(N)(NCCH3)(dppe)2]BPh4 (6) with IR and Raman spectra of the protonated product, [Mo(NH)-
(NCCH3)(dppe)2](BPh4)2 (6*HBPh4).
Figure 4. Geometry-optimized structures of [Mo(N)(N3)(PH2CH2CH2PH2)2] (I) (left), I (3E exc. st.) (middle), and [Mo(NH)(N3)(PH2CH2CH2PH2)2]+
(I*H+) (right).
to 916 cm-1. The positions of the metal-N vibrations for
the systems with chloride as counterion are reproduced by
the Raman spectra of 2*HCl and 2*DCl (Figure S3) which
exhibit these vibrations at 950 cm-1 for the protonated and
912 cm-1 for the deuterated complexes.
acetonitrile is again observed. The results are also collected
in Table 1.
A DFT frequency analysis of the model [Mo(N)(NCCH3)-
(diphos)2]+ (VI) for the trans nitrilo-nitrido system predicts
a frequency of 1088 cm-1 for the metal-N vibration. On
the basis of a comparison with the other calculated frequen-
cies, this would support our assignment of the shoulder at
1016 cm-1 to the metal-N stretch. The corresponding
analysis for the imido system, VI*H+, leads to frequencies
of 913 cm-1 for the metal-N stretch and 3059 cm-1 for the
N-H stretch. Thus the frequency of the metal-N vibration
is, in this case, predicted too low compared to the experi-
mental value.
DFT Geometry Optimizations. The optimized structure
of the simplified model [Mo(N)(N3)(diphos)2] (I) (diphos )
PH2CH2CH2PH2) is shown in Figure 4. In this case, the azido
ligand forms an almost linear unit together with the metal
center and the trans nitrido ligand ( Mo-N-N ) 174.6°).
This agrees with the crystal structure of the complex
[Mo(N)(N3)(dppe)2] determined by Dilworth et al.24 Apart
from minor discrepancies, the bond lengths resulting from
the optimization also agree with the X-ray structural analysis.
Finally, the IR and Raman spectra of complexes with trans
nitrile ligands were recorded (Figure 3). In these systems,
the IR spectrum of the nitrido complex [Mo(N)(NCCH3)-
(dppe)2]BPh4 (6) exhibits weak bands at 2270 and 2248 cm-1
which are attributed to coordinated acetonitrile. At 1016
cm-1, a shoulder appears which could be associated with
the metal-N stretch. A more definitive assignment is difficult
because of the intense bands at 1000 and 1030 cm-1, which
originate from the aromatic ligands. Unfortunately, no
complementary information with respect to this problem
could be derived from the corresponding Raman spectrum.
Because of the low basicity of this complex (see below),
the protonated species, [Mo(NH)(NCCH3)(dppe)2](BPh4)2
(6*HBPh4), is only obtained as a mixture with the proton-
ating agent lutidinium tetraphenylborate. The corresponding
sample does not exhibit the shoulder at 1016 cm-1. Instead,
new bands at 932 and 920 cm-1 appear, which could be
associated with the metal-N stretch of the imido complex.
At 2245 cm-1, the C-N vibration of the coordinated
(24) Dilworth, J. R.; Dahlstrom, P. L.; Hyde, J. R.; Zubieta, J. Inorg. Chim
Acta 1983, 71, 21.
Inorganic Chemistry, Vol. 45, No. 13, 2006 5049

Mersmann et al.
Table 3. Comparison of Important Internal Coordinates of Optimized Structuresa
I
I
(3E exc. st.)
I*H+
III
III*H+
VI
VI*H+
N-H
1.021
1.763
2.059
1.233
1.181
2.590
179.7
179.8
179.1
179.9
94.5
1.024
1.742
2.570
1.024
1.729
2.353
1.174
1.462
2.593
179.6
179.7
179.8
179.9
95.8
Mo-N
1.711
2.295
1.230
1.195
2.560
1.792
2.403
1.233
1.199
1.701
2.781
1.690
2.629
1.174
1.464
2.572
Mo-R
R-â
â-γ
Ø Mo-P
H-N-Mo
N-Mo-R
Mo-R-â
R-â-γ
2.56 (fixed)
2.551
179.9
2.573
179.9
179.9
179.6
174.6
179.9
96.0
79.8
99.0
173.8
127.1
179.1
96.0 (fixed)
80.6
178.8
178.1
179.7
94.2
79.4
100.0
Ø N-Mo-P
Øring P-Mo-P
Øfree P-Mo-P
99.3
79.8
97.2
98.8
79.8
97.5
79.6
99.8
79.4
99.5
98.2
a R, â, and γ are the atoms of the trans ligand; the distances are given in Å and the angles in deg.
The metal-nitrido distance is calculated to be 1.711 Å, which
is somewhat too short with respect to the experimentally
determined value (1.79(2) Å). The metal-N(azido) distance,
in turn, is calculated to be 2.295 Å, which is somewhat too
long with respect to the observed value (2.20(2) Å).
In additon to the optimization of the singlet ground state,
the triplet excited state 3E of I was geometry optimized. As
an unrestricted optimization was found to lead to dissociation
of one phosphine ligand, the N-metal-P angles and the
metal-P bond distances were fixed. This way, the triplet
state, which results from a one-electron HOMO-LUMO
excitation, could be optimized. The spin density in the triplet
excited state was found to be 1.48 for the metal and 0.45
for the nitrido-N. The metal-N bond is elongated by 0.081
Å because of the population of the π* orbital. Moreover,
the slight bending of the azido ligand in the ground state is
drastically enhanced. The metal-azide bond length increases
Figure 5. MO diagram of [Mo(N)(N3)(PH2CH2CH2PH2)2] (I) with contour
plots of some important orbitals.
from 2.295 to 2.403 Å, which corresponds to a weakening
of this bond in the excited state.
(III*H+) give comparable results. The metal-N distances
are 1.701 and 1.742 Å, respectively, and hence, they are
somewhat shorter than those in I and I*H+. The metal center
is markedly displaced from the phosphine plane, as can be
inferred from the average nitride-metal-P angles (Table
3). The geometries of the models for the trans nitrile systems
[Mo(N)(NCCH3)(diphos)2]+ (VI) and [Mo(NH)(NCCH3)-
(diphos)2]2+ (VI*H+) can be inferred from Table 3 as well.
In these cases, the metal-N bond is, with distances of 1.69
and 1.729 Å, respectively, markedly decreased. On the other
hand, the distances from the metal center to the trans ligands
(2.629 and 2.353 Å) are longer than in the analogous trans-
azido systems.
MO Schemes. Figure 5 presents the MO scheme and some
important orbitals of I; charge decompositions of important
molecular orbitals are collected in Table S1 (Supporting
Information). The HOMO of I corresponds to the metal-
ligand nonbonding dxy orbital. At lower energy, the degener-
ate πazide orbitals in the x and y directions are located; they
are nonbonding with respect to the metal. Then, the metal-
ligand bonding MOs follow; they are involved in the metal-
A geometry optimization was also performed for the model
of the imido complexes, [Mo(NH)(N3)(diphos)2]+ (I*H+).
As evident from Figure 4, the azido-metal-N unit is almost
linear ( Mo-N-N ) 179.1°); moreover, the imido-
hydrogen is bound in a linear fashion. The metal-N bond
of 1.763 Å is only elongated by 0.05 Å, in comparison to
the nitrido system, and the metal-azide distance is decreased
by 0.236 Å to a length of 2.059 Å. The structural parameters
agree well with the crystal structure of the imido complex
[MoBr(NH)(dppe)2]Br*MeOH obtained by Dilworth et al.24
If a hydrogen bridge between the crystal solvent and a
putative imido-hydrogen atom is assumed, a linear metal-
N-H bond can be inferred from this structure as well. The
linear arrangement of the imido-hydrogen is also supported
by the crystal structures of alkylated complexes such as
[W(NCH2CH3)Cl(dppe)2]PF6, which was crystallographically
characterized by Seino et al.25 In this case, the R-carbon atom
is bound at a W-N-C angle of 172.80° to the metal center
(i.e., again close to a linear fashion).
The optimized structures of the trans chloro complexes
[Mo(N)(Cl)(diphos)2] (III) and [Mo(NH)(Cl)(diphos)2]+
nitride triple bond. These comprise the σ-bonding orbital
2
pz-dz ≡ σ and the degenerate π-bonding orbitals px-dxz
≡
(25) Seino, H.; Tanabe, Y.; Ishii, Y.; Hidai, M. Inorg. Chim. Acta 1998,
280, 163.
πx and py-dyz ≡ πy. The corresponding metal-ligand
5050 Inorganic Chemistry, Vol. 45, No. 13, 2006

Molybdenum and Tungsten Nitrido and Imido Complexes
Figure 7. UV-vis spectra of [Mo(N)(N3)(depe)2] (5) in thf.
Figure 6. MO diagram of [Mo(NH)(N3)(PH2CH2CH2PH2)2]+ (I*H+) with
contour plots of some important orbitals.
antibonding combinations dxz-px ≡ π/ and dyz-py ≡ π/ form
x
y
the LUMO and the LUMO+1 of the complex. Evidently,
the degeneracy of these orbitals is lifted through the presence
of the bidentate equatorial phosphine ligands. NPA charge
analysis reveals a molybdenum p contribution of 32% for
the π/x orbital and one of 25% for the π/y orbital. The
presence of d-p mixing accounts for the comparatively high
intensity of the HOMO-LUMO transition (see below).
Figure 6 shows the MO scheme and the most important
orbitals of the imido complex I*H+. In analogy to the parent
nitrido system I, the HOMO corresponds to the dxy orbital;
HOMO-1 and HOMO-2 correspond to the πazide orbitals
as discussed above. An important difference, compared to
the nitrido complex, is observed for the metal-N σ-bonding
orbital which shifts below the metal-ligand π-bonding
orbitals πx and πy. Again the corresponding π-antibonding
combinations, πx/ and π/y, form the LUMO and the
LUMO+1 of the complex. The charge decompositions
indicate that these orbitals now only have 5 and 2% Mo p
character (Table S2 in Supporting Information). Because of
the protonation, the energy of these orbitals thus has
decreased so much that the higher-lying Mo p orbitals are
not admixed any more to a significant degree. Correspond-
ingly, the intensity of the n f π* transition in this case is
comparable to that of an ordinary ligand-field band (see
below).
Figure 8. UV-Vis and luminescence spectra of [Mo(N)(N3)(depe)2] (5)
along with fit of the progression.
(π/x), the low-energy component of the 1A1 f 1E transition.
The TD-DFT calculation based on the simplified model
[Mo(N)(N3)(diphos)2] (I) calculates this singlet f singlet
transition to be at 401 nm, in excellent agreement with the
experimental observation. The transition from the dxy orbital
into the dyz-py (π/) orbital (i.e., the higher-energy compo-
y
1
1
nent of the A1 f E transition) is calculated to be at 370
nm but with a much lower intensity (Table S3, Supporting
Information) Another absorption band is found at 330 nm
(Figure S4). This is assigned to a ligand field transition, either
dxy f dx -y or dxy f dz .26 Alternatively, it could be assigned
2
2
2
1
1
to the high-energy component of A1 f E. Nevertheless,
its line width is much smaller than that of dxy f dxz-px, and
it does not exhibit a progression. At lower energy, an
absorption band of much weaker intensity is observed and
assigned to a spin-forbidden transition from the singlet
Electronic Absorption and Emission Spectroscopy. In
Figure 7, optical absorption spectra of [Mo(N)(N3)(depe)2]
(5) in thf solution are presented. The UV region exhibits
two intense absorptions at 300 and 250 nm, and in the visible
region, a weaker absorption band is found at 398 nm. The
latter band is also present in the solid-state spectrum of
compound 5 pressed between two sapphire plates (Figure
S4, Supporting Information). When the sample is cooled from
room temperature to 10 K, a vibrational structure appears
on top of the 398 nm band. This is presented in Figure 8
along with a Gaussian fit of the band shape. The 398 nm
band is assigned to the HOMO-LUMO transition which
corresponds to the transfer of an electron from dxy to dxz-px
3
ground state to the E excited state.
Because the dxy orbital is nonbonding and the dxz/dyz
orbitals are antibonding with respect to the nitrido ligand,
the singlet and triplet transitions have n f π* character, and
the metal-ligand bond length is increased in the excited
states. This excited-state displacement in turn acts to excite
the metal-N vibrations which appear as a progression in
(26) Bendix, J.; Bøgevig, A. Inorg. Chem. 1998, 37, 5992.
Inorganic Chemistry, Vol. 45, No. 13, 2006 5051

Mersmann et al.
the singlet f singlet absorption band. The Gaussian fit of
the line shape gives a spacing of 810 cm-1 which corresponds
to the metal-N frequency in the excited state. The intensities
of the individual transitions from the zero-point level of the
ground state to the nth vibrational level of the excited state
are determined by a Poisson distribution, as expressed by
the equations
I0n
I0n Sn
S
n
)
and
)
(1)
I0n-1
I00 n!
S is the Huang-Rhys factor which gives the excited-state
displacement, ∆Q, in terms of quanta of the corresponding
vibration
Shν ) 1/2k′∆Q2
(2)
Figure 9. UV-vis spectra of [Mo(NH)(N3)(depe)2]BPh4 (5*HBPh4) in
CH2Cl2.
The Gaussian fit gives S ) 2.15. Furthermore, by application
of the excited-state frequency of 810 cm-1 in the normal
coordinate analysis, an excited-state force constant, k′, of
4.9 mdyn/Å is determined. Substitution of these values, in
turn, into eq 2 gives an excited-state displacement, ∆Q, of
0.12 Å, which is in very good agreement with the theoretical
prediction (∆Q ) 0.08 Å, cf Table 3).
d-p mixing which is markedly reduced in the corresponding
imido system because of the increased effective nuclear
charge on the metal (vide supra).
The absorption spectra of [Mo(NH)(Cl)(dppe)2]Cl (3*HCl)
and [W(NH)(Cl)(dppe)2]Cl (4*HCl) have already been
published by Henderson et al.27 Here the n f π* transitions
were found at 525 and 488 nm, respectively. The TD-DFT
calculation, based on model complex III, predicts an energy
of 530 nm for this transition. Unfortunately, the n f π*
transition energy of the parent complex [W(N)(Cl)(dppe)2]
(4) could not be exactly determined because of a new intense
band which appears in complexes with aromatic phosphine
ligands. We attribute this band to a transition from the metal
to the phenyl phosphine group. In the corresponding alkyl
phosphine complexes, this transition also exists but probably
is at a much higher energy.6,28 In the case of complex 4, the
metal-phosphine CT band appears at 360 nm. From the
appearance of a weak shoulder, the much less intense n f
π* transition of 4 may be located around 400 nm, which
would be in agreement with the TD-DFT calculation predict-
ing a transition energy of 392 nm for this system.
Figure 10 presents the UV-vis absorption spectra of the
nitrido nitrile compound [Mo(N)(NCCH3)(dppe)2]BPh4 (6).
The CT transition from the metal to the phosphine ligands
is found at 330 nm. In comparison with complex 5, which
carries an anionic ligand trans to the nitrido group, the n f
π* transition is shifted to lower energy (450 nm). The TD-
DFT calculation based on model complex VI predicts this
transition at 437 nm, in good agreement with the experi-
mental result. The measured energy shift of the HOMO-
LUMO transition upon substitution of an anionic trans ligand
by a neutral acetonitrile reflects two antagonistic effects: on
one hand, the energy of the π* orbitals is increased because
of the strengthening of the metal-N bond, as can be inferred
from the increase of the metal-N vibration frequency (vide
In addition to the absorption spectrum, Figure 8 also shows
the emission spectrum of compound 5. The luminescence is
strongly temperature dependent, and at 11 K, it also shows
a vibrational structure. The corresponding vibrational fre-
quency is connected with the electronic ground state and is
determined to be 980 cm-1, in agreement with the results of
the IR and Raman spectroscopy (vide supra). From the
analysis of the intensity distribution of the luminescence with
a Gaussian fit, a Huang-Rhys factor of 2.25 is derived. The
corresponding horizontal displacement, ∆Q, is determined
to 0.11 Å, in very good agreement with the value obtained
from the absorption profile. Importantly, the 0 T 0 transitions
of the absorption and emission spectrum of 5 do not coincide
but exhibit an energy difference of 3030 cm-1, indicating
that the electronic states involved in absorption and lumi-
nescence are different. If the absorption band is assigned to
the spin-allowed dxy f dxz-px singlet f singlet transition
and the emission band to the spin-forbidden dxz-px f dxy
triplet f singlet transition, the 3030 cm-1 energy difference
corresponds to the energy gap between the (low-symmetry
1
3
split) E and the E excited states.
The UV-vis absorption spectrum of the protonated
complex [Mo(NH)(N3)(depe)2]BPh4 (5*HBPh4) is shown in
Figure 9. Again, two high-energy bands can be identified
(at 236 and 290 nm) which are shifted toward the UV region
with respect to the parent complex 5. In contrast, the low-
energy n f π* transition is shifted toward the IR region
and now is found at 518 nm. This transition was calculated
by TD-DFT applied to the I*H+ model at 523 nm, again in
excellent agreement with the experiment. As already men-
tioned, the intensity of this transition (ꢀ ) 145 M-1 cm-1)
is typical for a ligand-field transition, whereas the intensity
of the corresponding n f π* transition in the parent
compound (ꢀ ) 460 M-1 cm-1) is unusually high. We
attribute this high intensity in the nitrido complex to extensive
(27) (a) Henderson, R. A.; Davies, G.; Dilworth, J. R.; Thorneley, R. N.
F. J. Chem. Soc., Dalton Trans. 1981, 40. (b) Henderson, R. A. J.
Chem. Soc., Dalton Trans. 1983, 51.
(28) (a) George, T. A.; Busby, D. C.; Iske, S. D. A., Jr. Inorg. Organomet.
Photochem. 1978, 147. (b) Bossard, G. E.: Busby, D. C.; Chang, M.;
George, T. A.; Iske, S. D. A., Jr. J. Am. Chem. Soc. 1980, 102, 1001.
5052 Inorganic Chemistry, Vol. 45, No. 13, 2006

Molybdenum and Tungsten Nitrido and Imido Complexes
Table 5. Peak Potentials for the Reduction of Nitrido and Imido
Complexes with Anionic and Neutral trans Ligands
Epred (V vs Fc+/Fc)
[Mo(N)(NCCH3)(dppe)2]+
[Mo(N)(Cl)(dppe)2]
-1.53
<-2.4
-2.0911
-2.2111
-1.530
[Mo(NH)(Cl)(dppe)2]+
[Mo(NEt)(Cl)(dppe)2]+
[Mo(NEt)(NCCH3)(depe)2]2+
the nitrido to the imido complex. TD-DFT predicts the n f
π* transition in the imido nitrile model complex VI*H+ at
597 nm, in fair agreement with the experimental result. The
inset in Figure 11 shows the spectrum of a dilute solution
of 6 in the presence of a large excess of acid. In this case an
approximately complete conversion of 6 to its protonated
counterpart, [Mo(NH)(NCCH3)(dppe)2](BPh4)2 (6*HBPh4),
is observed. In the intermediate range, the acid strength of
the protonated complex 6*HBPh4 can be determined by the
equation29
Figure 10. UV-vis spectrum of [Mo(N)(NCCH3)(dppe)2]BPh4 (6) in
CH3CN. The inset shows the UV-vis spectrum of the pure compound
pressed between sapphire plates.
Table 4. Calculated and Experimental dxy f dxz-px (n f π*)
Transition Energiesa
∆pKa ) pKa(6*HBPh4) - pKa(HLutBPh4) )
[6*HBPh4][Lut]
I
I*H+
III
392
III*H+
VI
437
450
(dppe)
VI*H+
TD-DFT
exptl
401
398
(depe)
523
518
(depe)
530
596
525
(dppe)
log
(3)
(
)
∼400
52527
(dppe)
[HLutBPh4][6]
(dppe)
a All data given in nm.
In acetonitrile, the pKa of complex 6*HBPh4 differs from
lutidinium by 1.7 units. On the basis of a pKa of lutidinium
of 6.729 in water, a pKa value of 5 is derived for 6*HBPh4.
This is much lower than that determined for imido complexes
with anionic trans ligands.27 For the complex [Mo(NH)(F)-
(dppe)2]BF4, for example, a pKa value of 12.7 was deter-
mined. This result indicates that the basicity of the nitrido
N drastically decreases upon exchange of an anionic to a
neutral trans ligand. Correspondingly, the conversion of 6
to its protonated counterpart using the weak acid HLutBPh4
is found to be incomplete in solutions containing a higher
concentration of 6. On the other hand, the reduction potential
should increase upon substitution of the trans-halide by a
trans-nitrile group (i.e., become less negative), as shown in
the next section.
Electrochemical Investigation of 6. For Mo/W nitrido
systems with anionic trans ligands no reduction potential
could be determined within the stability range of ordinary
solvents and electrolytes, respectively. In contrast, electro-
chemical data for protonated and alkylated systems have been
obtained (cf., Table 5).11 Figure 12 shows a cyclovoltam-
mogram of the trans nitrido nitrile complex 6, indicating an
irreversible reduction at -1.53 V vs ferrocenium/ferrocene.
Furthermore, two irreversible oxidations are observed at 65
and 540 mV. An inspection of Table 5 shows that the
replacement of a trans halide by acetonitrile shifts the
reduction potential by ∼1 V toward more positive values.
In fact, the reduction potential of nitrido complex 6 is
comparable to that of the trans-alkylimido acetonitrile
complex [Mo(NEt)(NCCH3)(depe)2](OTf)2, which has depe
instead of dppe coligands.30 It thus can be concluded that
Figure 11. UV-vis spectrum of [Mo(N)(NCCH3)(dppe)2]BPh4 (6) in
CH3CN after the addition of different amounts of [HLutBPh4]. The inset
shows the UV-vis spectrum of 6 with a great excess of acid.
supra). On the other hand, the energy of these orbitals is
lowered because π donation of the trans ligand is eliminated.
This effect obviously exceeds the first one in magnitude and
leads to the observed red shift of the n f π* transition. The
inset of Figure 10 shows the solid-state spectrum of
compound 6 recorded at 10 K. Again, a progression is visible,
but this time spectral resolution is not good enough to allow
further analysis. The results of the TD-DFT calculations and
the experimental energies of the n f π* transition are
collected in Table 4.
Spectrometric Determination of pKa. Figure 11 shows
the absorption spectra of a 2 mM solution of 6 before and
after the addition of different amounts of acid (HLutBPh4).
Evidently the signal at 450 nm decreases and the signal at
525 nm increases in intensity, reflecting the conversion of
(29) Kaljurand, I.; Ku¨tt, A.; Soova¨li, L.; Rodima, T.; Ma¨emets, V.; Leito,
I.; Koppel, I. A. J. Org. Chem. 2005, 70, 1019.
(30) Sivasankar, C.; Bo¨res, N.; Peters, G.; Habeck, C. M.; Studt, F.; Tuczek,
F. Organometallics 2005, 24, 5393.
Inorganic Chemistry, Vol. 45, No. 13, 2006 5053

Mersmann et al.
stretch is observed at 925-920 cm-1. On the basis of the
latter value, a force constant of 6.7 mdyn/Å is derived. This
indicates that protonation only causes a minor decrease (0.3
mdyn/Å) of the metal-N force constant. For both systems
with chloride and BPh4 counterions, deuteration causes a
further decrease of the metal-N stretching vibration fre-
quency by 30-40 cm-1, reflecting the increased mass of the
N-H moiety.
The vibrational-spectroscopic investigations of the nitrido
and imido complexes are complemented by measurements
of their electronic absorption and emission spectra. For this
purpose, Mo and W complexes with depe ligands are better
suited than corresponding dppe complexes as the latter
exhibit low-energy metal f phosphine CT transitions which
tend to obscure ligand-field (LF) transitions. Correspond-
ingly, the molybdenum azido-nitrido complex [Mo(N)(N3)-
(depe)2] (5) was found to give a clean LF spectrum which,
Figure 12. Cyclic voltammogram of [Mo(N)(NCCH3)(dppe)2]BPh4 (6)
(0.5 mM) in CH3CN, 0.1 M (NBu4)PF6 at 25 °C and a scan rate of 100 mV
in particular, exhibited the dxy(nb) f dxz-px(π/) transition at
s-1
.
x
398 nm. The relatively high intensity of this spin-allowed n
f π* transition (ꢀ ) 460 M-1 cm-1) is ascribed to extensive
d-p mixing. At lower energy, an absorption band of much
weaker intensity is observed, which is assigned to the
corresponding spin-forbidden transition from the singlet
ground state to the lower-energy component of the 3E (n f
π*) excited state. When the sample had cooled, a vibrational
progression appears on top of the singlet f singlet absorption
band. Its spacing, 810 cm-1, corresponds to the metal-N
molybdenum trans imido- or alkylimido-acetonitrile com-
plexes with dppe coligands should even be reducible at
potentials above -1.5 V vs Fc+/Fc.
4. Discussion
The preceding sections continue our line of theoretical and
spectroscopic studies on the intermediates of the Chatt cycle.
Previous publications have dealt with the protonation of the
coordinated N2 ligand in Mo(0) and W(0) diphosphine
complexes and the subsequent N-N splitting process.6,7,10
Heterolytic cleavage of the N-N bond in Mo(II) and W(II)
systems generates the corresponding Mo(IV) and W(IV)
nitrido and imido systems which have to be protonated and
reduced to the zero-valent state to produce NH3 and rebind
N2. A problem of the Chatt cycle are the strongly negative
potentials (down to -2.3 V vs Fc+/Fc)11 involved in these
processes. It is therefore of considerable interest to investigate
in more detail the electronic structures of these systems and
to explore possibilities to shift these potentials to less
negative values.
In the nitrido complexes with trans-azido and -chloro
coligands, the metal-N stretch is observed at about 980
cm-1, relatively independent of the nature of the trans ligand
and the equatorial phosphine ligation. 15N substitution in
compound 4 causes an isotopic shift of this vibration to ∼950
cm-1. A metal-N force constant of 7.065 mdyn/Å is
determined using the QCB-NCA procedure. When the
corresponding imido systems are generated by protonation,
the metal-N stretching frequency is lowered to about 950-
920 cm-1. In this case, a relatively strong dependence on
the nature of the counterion is observed. As evident from a
highly complex vibrational structure in the N-H stretching
frequency region, extensive hydrogen bonding occurs in
imido complexes with chloro counterions. In this case, the
metal-N stretching frequency is observed at ∼950 cm-1.
In the absence of hydrogen bonding (i.e., in the compounds
with tetraphenylborate counterions), the metal-N(imido)
1
stretching frequency in the E (n f π*) excited state. An
excited-state force constant of 4.9 mdyn/Å is derived by
inserting this frequency into the ground-state normal coor-
dinate analysis. Moreover, a Franck-Condon analysis of this
absorption feature gives a Huang-Rhys factor of 2.15 and
an excited-state displacement of 0.12 Å.
Remarkably, a vibrational progression is found in the
emission spectrum of compound 5 at low temperatures as
well. To the best of our knowledge, this is the first
luminescence spectrum of such a molybdenum-nitrido
system.31 In this case, a spacing of 980 cm-1 is observed
which corresponds to the metal-N stretching vibration in
the electronic ground state. Franck-Condon analysis of the
emission line shape gives a Huang-Rhys factor of 2.25 and
a displacement, ∆Q, of 0.11 Å, in good agreement with the
absorption data. As the emission band is at comparable
energy to the low-intensity spin-forbidden transition in the
absorption spectrum, it is assigned to the phosphorescence
from the lower-energy component of the 3E(n f π*) excited
1
state to the A1 ground state. On the basis of the energy
difference between the 0 T 0 transitions in absorption and
emission, the splitting between the 1E and 3E(n f π*) excited
states is 3030 cm-1. When the fact that this splitting
corresponds to 6B + 2C32 is considered, it becomes clear
that this system is highly covalent.33,34 Upon protonation of
(31) Different combinations of research topics such as molybdenum/
tungsten, nitrido, and luminescence/phosphorescence with SciFinder
resulted in no relevant hits.
(32) Da Re, R. E.; Hopkins, M. D. Inorg. Chem. 2002, 41, 6973.
(33) With a literature value of B ) 682 cm-1 for a Mo4+ ion and C ) 4B,
a splitting of 9548 cm-1 is evaluated.
5054 Inorganic Chemistry, Vol. 45, No. 13, 2006

Molybdenum and Tungsten Nitrido and Imido Complexes
the nitrido complexes the (n f π*) transition shifts to lower
energy (518 nm); moreover, its intensity decreases consider-
ably. The low-energy shift can be explained by a weaker
π-antibonding interaction between the metal dxz/dyz orbitals
and the p orbitals of the imido ligand, as compared to the
nitrido system. The intensity decrease is attributed to a
decrease in d-p mixing as a consequence of the increase in
the effective nuclear charge of the M(IV) system.
Further insight into the electronic structure and spectro-
scopic properties of the nitrido and imido systems investi-
gated in this study has been obtained by DFT calculations
performed on the model systems [Mo(N)(N3)(diphos)2] (I),
[Mo(N)(Cl)(diphos)2] (III), [Mo(N)(NCCH3)(diphos)2] (VI),
[Mo(NH)(N3)(diphos)2]+ (I*H+), [Mo(NH)(Cl)(diphos)2]+
(III*H+), and [Mo(NH)(NCCH3)(diphos)2]2+ (VI*H+). In
the corresponding geometry-optimized structures, the metal-N
distances are generally found somewhat too short. This is
also reflected by the fact that the metal-N frequencies are
calculated somewhat too high in energy. TD-DFT-calcula-
tions give n f π* transition energies which are in very good
agreement with the experimental data. For the n f π* excited
state of [Mo(N)(N3)(diphos)2] (I), a partial geometry opti-
mization has been performed. In good agreement with the
experimental value, an elongation of the metal-N(nitrido)
distance by 0.08 Å has been found. However, the metal-
N(azido) distance increases as well (by 0.11 Å), and the
bending of the azido ligand increases.
TD-DFT. Otherwise the absorption bands of the Mo-O
systems have properties which are very similar to those of
Mo-nitrido complex 5 (i.e., they show a vibronic spacing
of ∼800 cm-1, which is considerably lower than the ground-
state Mo-O stretching frequency of ∼960 cm-1), and
Franck-Condon analysis gives an excited-state displacement,
∆Q, of ∼0.1 Å.
A major change in the spectroscopic properties and the
reactivity of the nitrido complexes is observed upon exchange
of the anionic trans ligands by acetonitrile. Most conspicu-
ously, the metal-N(nitride) stretching frequency increases
to 1016 cm-1, reflecting a strengthening of the metal-nitrido
bond. The n f π* transition now is found at 450 nm, shifting
to 525 nm upon protonation. Furthermore, the reduction
potential is shifted from more negative values to -1.53 V
vs Fc+/Fc which is close to the E° value of decamethyl-
chromocene used as a reductant in the Schrock cycle.2 This
drastically facilitates the reduction of the nitrido-nitrile
system compared to its counterparts with anionic trans
ligands which cannot be reduced at experimentally accessible
potentials. Even less negative reduction potentials are
expected for the corresponding imido systems [Mo(NH)-
(NCCH3)(dppe)]2+ and [Mo(NR)(NCCH3)(dppe)]2+ (for
[Mo(NEt)(CH3CN)(depe)]2+
a reduction potential of
-1.5 V has been determined,30 but the corresponding
dppe analogues should be reducible at less negative poten-
tials).
It is instructive to compare the properties of the Mo and
W nitrido and imido systems to those of the corresponding
oxo compounds which have been studied in detail by several
groups.26,32,35 Da Re and Hopkins have investigated a series
of molybdenum oxo complexes of the type [MoOL4Cl]+ (L
) CNBut, PMe3, and 1/2dmpe).32 The Mo-O stretching
frequency of these ions (948-959 cm-1) was essentially
found to be insensitive to the nature and geometry of the
equatorial ligands. However, the electronic absorption bands
resulting from the dxy f dxz/dyz n f π*(Mo-O) LF transition
show a large dependence on the geometry of the equatorial
ligation. Specifically, the [MoO(CNBut)4Cl] complex ex-
hibits a single 1[n f π*x,y] band, whereas the spectra of both
[MoO(dmpe)2Cl]+ and [MoO(PMe3)4Cl]+ reveal separate 1[n
Exchange of the anionic trans ligands by nitrile, on the
other hand, also leads to a considerable decrease in basicity
of the nitrido group. By titration experiments with HLutBPh4,
a pKa value of 5.0 was determined for complex 6*HBPh4,
which is much lower than that observed for corresponding
systems with trans halide ligands (pKa ≈ 10).27 Importantly,
however, the nitrido group is still basic enough because
protonation of 6, even with the weak acid HLutBPh4,
proceeds with an equilibrium constant Ka of 1/50. Conversion
to a free reaction enthalpy gives a corresponding ∆G° value
of +10 kJ/mol which does not represent an appreciable
barrier in the transformation of coordinated nitride to
ammonia.
In summary, the electronic structures of a series of Mo
and W nitrido and imido complexes have been determined
and their spectroscopic properties have been investigated in
detail. The experimental and theoretical results have been
correlated to the observed reactivities, in particular reduction
potentials and Brønsted basicities. In this way, it has been
found that the replacement of anionic trans ligands by
acetonitrile causes a pronounced shift in the reduction
potentials toward less negative values, thus rendering these
systems promising intermediates for a catalytic realization
of the Chatt cycle. In this respect, it is advantageous that
the activation of N2 in complexes with trans-nitrile ligands
is higher than in corresponding trans-bis(dinitrogen) com-
plexes.36 The trade-off for the easier reducibility is a con-
siderable decrease of the basicity of the nitrido group which,
/
/
1
f πx] and [n f πy] bands. Nitrido complex 5 is chemi-
cally related to these systems and therefore should have split
singlet and triplet n f π* excited states; however, the
magnitude of this splitting cannot be inferred from the
spectroscopic data. It is clear that luminescence occurs from
3
the lower-energy component of the E excited state. From
the almost identical line shapes in the absorption and
emission spectra, we conclude that the absorption band
corresponds to the electronic transition from 1A1 to the lower-
1
energy component of E, in agreement with the prediction
1
from TD-DFT. The higher-energy component of E is not
visible in the absorption spectrum. A possible explanation
is the much lower intensity predicted for this transition by
(34) Griffith, J. S. The Theory of Transition-Metal-Ions; Cambridge
University Press: Cambridge, U.K., 1971.
(35) Isovitsch, R. A.; Beadle, A. S.; Fronczek, F. R.; Maverick, A. W.
Inorg. Chem. 1998, 37, 4258.
(36) Habeck, C. M.; Lehnert, N.; Na¨ther, C.; Tuczek, F. Inorg. Chim. Acta
2002, 737C, 11.
Inorganic Chemistry, Vol. 45, No. 13, 2006 5055

Mersmann et al.
Supporting Information Available: Tables of charge contribu-
tions, n f π* transitions, and NPA charges, a description of the
charge analysis, and figures showing Raman, IR, and UV-vis
spectra. This material is available free of charge via the Internet at
http://pubs.acs.org.
however, should not present a serious obstacle to further
conversion of the nitrido complex along the Chatt cycle.
Acknowledgment. Dedicated to Prof. Dr. Hansgeorg
Schno¨ckel on the occasion of his 65th anniversary. F.T.
thanks DFT Tu58/12-1 and FCI for financial support. K.M.
thanks U. Cornelissen for help with the spectroscopic
measurements.
IC060141N
5056 Inorganic Chemistry, Vol. 45, No. 13, 2006