Spectroscopic characterization of molybdenum dinitrogen complexes containing a combination of di- and triphosphine coligands: 31P NMR analysis of five-spin systems

Article abstract of DOI:10.1021/ic800389g

The three molybdenum-N₂ complexes [Mo(N₂)(dpepp) (depe)] (1), [Mo(N₂)(dpepp)(dppe)] (2), and [Mo(N₂)(dpepp) (1,2-dppp)] (3), all of which contain a combination of a bi- and a tridentate phosphine ligand, were prepared and investigated by vibrational and ³¹P NMR spectroscopy. As a tridentate ligand bis(2- diphenylphosphinoethyl)phenylphosphine (dpepp) has been employed. The three different bidentate ligands are 1,2-bis(diethylphosphino)ethane (depe), 1,2-bis(diphenylphosphino)ethane (dppe), and R-(+)-1,2-bis(diphenylphosphino) propane (1,2-dppp). N-N as well as metal-N vibrations of 1-3 are identified and interpreted in terms of the geometric and electronic structures of the complexes. ³¹P NMR spectra are recorded and fully analyzed. Moreover, correlation spectroscopy (COSY)-45 measurements are performed to determine the relative signs of coupling constants. Special attention is directed to a detection of different isomers and their ³¹P NMR, as well as vibrational spectroscopic properties. The implications of the results for the area of synthetic nitrogen fixation with phosphine complexes are discussed.

Full text of DOI:10.1021/ic800389g

Inorg. Chem. 2008, 47, 6541-6550
Spectroscopic Characterization of Molybdenum Dinitrogen Complexes
31
Containing a Combination of Di- and Triphosphine Coligands: P NMR
Analysis of Five-Spin Systems
Kristina Klatt, Gerald Stephan, Gerhard Peters, and Felix Tuczek*
Institut fu¨r Anorganische Chemie, Christian Albrechts UniVersita¨t Kiel, Otto Hahn Platz 6/7,
D-24098 Kiel, Germany
Received February 29, 2008
The three molybdenum-N2 complexes [Mo(N2)(dpepp)(depe)] (1), [Mo(N2)(dpepp)(dppe)] (2), and [Mo(N2)(dpepp)(1,2-
dppp)] (3), all of which contain a combination of a bi- and a tridentate phosphine ligand, were prepared and
investigated by vibrational and 31P NMR spectroscopy. As a tridentate ligand bis(2-diphenylphosphinoethyl)phe-
nylphosphine (dpepp) has been employed. The three different bidentate ligands are 1,2-bis(diethylphosphino)ethane
(depe), 1,2-bis(diphenylphosphino)ethane (dppe), and R-(+)-1,2-bis(diphenylphosphino)propane (1,2-dppp). N-N
as well as metal-N vibrations of 1-3 are identified and interpreted in terms of the geometric and electronic
structures of the complexes. 31P NMR spectra are recorded and fully analyzed. Moreover, correlation spectroscopy
(COSY)-45 measurements are performed to determine the relative signs of coupling constants. Special attention
is directed to a detection of different isomers and their 31P NMR, as well as vibrational spectroscopic properties.
The implications of the results for the area of synthetic nitrogen fixation with phosphine complexes are discussed.
1. Introduction
tional models of the enzyme nitrogenase.3 Despite an early
success with respect to a cyclic conversion of N2 to NH3,4
however, no truly catalytic action could be achieved on the
basis of this chemistry (“Chatt cycle”) so far. This is in
contrast to the recently synthesized Mo(III) triamidoamine
complex containing sterically demanding residues for which
a catalytic generation of ammonia from N2 has been
demonstrated (Schrock cycle).5
To obtain information on the energetics of the Chatt cycle,
we have recently treated all involved intermediates, the
employed acid as well as the reductant, with density-
functional theory (DFT) and have determined the free
31P NMR spectroscopy has proven to be a valuable tool
for the elucidation of the solution structure of transition-
metal phosphine complexes.1 This class of compounds has
also been of significant interest with respect to synthetic
nitrogen fixation. As demonstrated by the groups of Chatt,
Hidai, and others, molybdenum and tungsten dinitrogen
complexes of the type [M(N2)2(monophos)4] or [M(N2)2-
(diphos)2] can be protonated at the N2 ligand to ultimately
give NH3 (monophos ) PR3, R ) Me, Ph; diphos )
Ph2PCH2CH2PPh2 (dppe) and Et2PCH2CH2PEt2 (depe)).2 For
the diphos systems, many complex-bound intermediates of
the reduction and protonation of N2 have been isolated and
characterized, respectively, making these compounds func-
(3) (a) Prokaryotic Nitrogen Fixation: A Model System for the Analysis
of a Biological Process; Triplett, E. W., Ed.; Horizon Scientific Press:
Norfolk, U.K., 2000. (b) Burgess, B. K.; Lowe, D. J. Chem. ReV. 1996,
96, 2983.
* To whom correspondence should be addressed. E-mail: ftuczek@
ac.uni-kiel.de.
(4) (a) Pickett, C. J.; Talarmin, J. Nature 1985, 317, 652. (b) Pickett, C. J.;
Ryder, K. S.; Talarmin, J. J. C. S. Dalton Trans., Inorg. Chem. 1986,
7, 1453.
(1) (a) Henry, R. M.; Shoemaker, R. K.; Newell, R. H.; Jacobsen, G. M.;
DuBois, D. L.; Rakowski DuBois, M. Organometallics 2005, 24, 2481.
(b) Pregosin, P. S.; Kunz, R. W., NMR Basic Principles and Progress;
Springer Verlag: New York, 1979; Vol. 16. (c) Verkade, J. G. Coord.
Chem. ReV. 1972, 9, 1.
(5) (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.
(2) (a) MacKay, B. A.; Fryzuk, M. D. Chem. ReV. 2004, 104, 385. (b)
Hidai, M.; Mizobe, Y. Chem. ReV. 1995, 95, 1115. (c) Henderson,
R. A.; Leigh, G. J.; Pickett, C. J. AdV. Inorg. Chem. Radiochem. 1983,
27, 197. (d) Chatt, J.; Dilworth, J. R.; Richards, R. L. Chem. ReV.
1978, 78, 589.
10.1021/ic800389g CCC: $40.75  2008 American Chemical Society
Inorganic Chemistry, Vol. 47, No. 14, 2008 6541
Published on Web 06/14/2008

Klatt et al.
reaction enthalpies for all protonation and reduction reac-
tions.6 As in our previous DFT study of the Schrock cycle,7
decamethylchromocene was employed as reductant. For the
protonation reactions, two acids were considered, HBF4/
diethylether and a lutidinium salt. For all protonation and
reduction steps, the corresponding free reaction enthalpy
changes were calculated. The derived energy profile and
corresponding reaction mechanism bear strong similarities
to the Schrock cycle. For HBF4/diethylether as acid and
Cp*2Cr as reductant, a catalytic cycle consisting of thermally
allowed reactions was predicted. However, this cycle in-
volves a Mo(I) fluoro complex as dinitrogen intermediate
which is unstable toward disproportionation. If a truly
catalytic action of the Chatt system is intended, strategies
have to be developed to avoid disproportionation of the Mo(I)
dinitrogen complex to a Mo(II) complex carrying two anionic
ligands. Important points in this respect are (i) avoiding the
presence of strongly Lewis-basic species (such as F-) in
solution and (ii) employing multidentate ligands with more
than two P-donors that also occupy the trans-position of
coordinated N2, in contrast to the conventional Chatt-type
Mo and W diphosphine systems.
The only existing phosphine-coordinated molybdenum
dinitrogen complexes which meet the second criterion have
been prepared by George et al. using a combination of a
diphosphine and a triphosphine ligand.8 In analogy to the
Mo triamidoamine complexes, these systems exhibit only
one binding site for N2 and can be protonated to give NNH2
complexes. In a previous study, we investigated the electronic
structure and the spectroscopic properties of the complex
[Mo(N2)(dpepp)(dppm)] (dpepp ) bis(diphenylphosphino-
ethyl)phenylphosphine, PhP(CH2CH2PPh2)2; dppm ) bis-
(diphenylphosphino)methane, Ph2PCH2PPh2) and its hydrazi-
do(2-) derivative.9 15N- and 31P NMR spectroscopic data were
presented and analyzed with the help of spectral simulations.
Moreover, infrared (IR) and Raman spectra were recorded
and evaluated using a quantum chemistry based normal
coordinate analysis (QCB-NCA). The spectroscopic results
were interpreted with the help of DFT calculations. These
studies showed that the activation of the N2 ligand of
the parent N2 complex is lower than in the trans-acetonitrile
complex [Mo(N2)(NCCH3)(dppe)2], but higher than in the
bis(dinitrogen) complex [Mo(N2)2(dppe)2]. This was inter-
preted in terms of the relative σ/π-donor and π-acceptor
capabilities of the respective trans-ligands. More recently,
we also applied our DFT analysis to the cyclic reduction
and protonation of N2 mediated by this complex and found
that a catalytic ammonia synthesis from N2 should in fact
be thermodynamically feasible if the phosphine ligands
remain coordinated to the metal center.10
Herein we present spectroscopic investigations on molyb-
denum dinitrogen complexes containing a combination of
dpepp with diphos ligands exhibiting C2 bridges, that is, depe,
dppe, and R-(+)-1,2-dppp (Ph2PC*H(CH3)CH2PPh2). Specif-
ically, the three compounds [Mo(N2)(dpepp)(depe)] (1),
[Mo(N2)(dpepp)(dppe)] (2), and [Mo(N2)(dpepp)(1,2-dppp)]
(3) were prepared and investigated by vibrational and 31P
NMR spectroscopy. Complex 1 was found to exist in only
one form. Compound 2 had already been synthesized by
George et al..8 In agreement with these authors, we identified
two isomers, [Mo(N2)(dpepp)(dppe)] (2a) and iso-
[Mo(N2)(dpepp)(dppe)] (2b). For compound 3 evidence for
the existence of two diastereomers (isomer A and B, 3a and
3b) was obtained as well. In this case, however, the
isomerism is not induced by the coordination geometry of
the complex but by an optically active ligand. In the present
paper particular emphasis is put on the one- and two-
dimensional 31P NMR spectroscopic characterization of
complexes 1-3 in solution, as well as on a full analysis of
the corresponding five-spin systems. Moreover, IR and
Raman spectra of these complexes are recorded and cor-
related with the NMR-spectroscopic information. The im-
plications of the results with respect to synthetic nitrogen
fixation are discussed.
2. Experimental Section
Synthesis and Sample Preparation. All reactions were carried
out under an N2-atmosphere by using Schlenk-techniques. Solvents
were dried and freshly destilled under argon. All other reagents
were used in the available qualities. The phosphine ligands were
obtained from Strem Chemicals. The preparations were performed
according to the literature8 with slight modifications.
[Mo(N2)(dpepp)(depe)] (1). This synthesis was carried out as
described for 2 (see below) by using sodium amalgam (200 mg
Na, 30.0 g Hg), 450 mg of (0.58 mmol) [MoCl3(dpepp)], 140 mg
of (0.68 mmol) depe, and 20 mL of THF. Elemental Anal. Calcd:
C, 61.1; H, 6.6; N, 3.2. Found: C, 61.0; H, 6.9; N, 2.6. 31P{1H} )
2
2/3
δ31P ) 87.25 ppm (dddd, Jea/eb ) -15.5 Hz, |
J
| ) 4.7 Hz,
ed/ec
2
2
2
Pe), 62.12 ppm (dddd, Jcd ) -9.5 Hz, Jac/bd ) -18.6 Hz, Jad/bc
) 98.3 Hz, Pc/d), 48.80 ppm (dddd,
2/3
J ) -4.9 Hz, Pa/b).
ab
[Mo(N2)(dpepp)(dppe)] (2). A suspension of 570 mg of (0.74
mmol) [MoCl3(dpepp)] and 350 mg of (0.87 mmol) dppe in 20
mL of THF was added to sodium amalgam (200 mg Na, 30.0 g
Hg) and stirred for 3 h at 0 °C and 16 h at ambient temperature
under N2. The solution was decanted, filtered and reduced in vacuo
to 6 mL. Six milliliters of methanol were added, and the solution
was reduced again. After addition again of 6 mL of methanol, the
formed precipitate was filtered off, washed with 4 × 4 mL
methanol, and dried in vacuo. Elemental Anal. Calcd: C, 68.1; H,
5.4; N, 2.6. Found: C, 67.9; H, 5.6; N, 2.2. 2a: 31P{1H} ) δ31P
73.80 (quin; Pe), 63.40 (d; Pa,b,c,d) ppm. 2b: 31P{1H} ) δ31P ) 90.94
)
(6) Stephan, G.; Sivasankar, Ch.; Tuczek, F. Chem. Eur. J. 2008, 14, 644–
652.
2
2
2/3
2/3
(ddd, Jea) 106.7 Hz, Jeb) -16.6 Hz, Jec) 7.2 Hz, Jed) 1.5
(7) Studt, F.; Tuczek, F. Angew. Chem., Int. Ed. 2005, 44, 5639–5642.
(8) (a) George, T. A.; Tisdale, R. C. Inorg. Chem. 1988, 27, 2909. (b)
George, T. A.; Tisdale, R. C. J. Am. Chem. Soc. 1985, 107, 5157. (c)
George, T. A.; Ma, L.; Shailh, S. N.; Tisdale, R. C.; Zubieta, J. Inorg.
Chem 1990, 29, 4789.
2
2
2
Hz; Pe), 72.30 (dddd, Jcb ) 101.6 Hz, Jca ) -21.2 Hz, Jcd
)
(10) Stephan, G.; Tuczek, F. In ActiVating UnreactiVe Substrates; Bolm,
C. , Hahn, E., Eds.; Wiley VCH: New York, 2008;. submitted for
publication.
(9) Stephan, G. C.; Peters, G.; Lehnert, N.; Habeck, C. M.; Na¨ther, C.;
Tuczek, F. Can. J. Chem. 2005, 83, 385.
6542 Inorganic Chemistry, Vol. 47, No. 14, 2008

Molybdenum Dinitrogen Complexes
Figure 2. IR spectra of 1 (top), 2 (middle), and 3 (bottom) measured in
KBr.
Figure 1. Synthesized complexes.
2/3
-13.1 Hz; Pc), 69.10 (dddd, 2Jbc ) 101.6 Hz,
J
) 5.1 Hz, 2Jbd
ba
) -9.7 Hz; Pb), 67.00 (ddt; Pa), 61.10 (ddd; Pd) ppm.
[Mo(N2)(dpepp)(1,2-dppp)] (3). This synthesis was performed
in analogy to that of 2 by using sodium amalgam (200 mg Na,
30.0 g Hg), 419 mg of (0.540 mmol) [MoCl3(dpepp)], 265 mg of
(0.640 mmol) 1,2-dppp, and 20 mL of THF. Elemental Anal. Calcd:
C, 69.1; H, 5.4; N, 2.6. Found: C, 68.9; H, 5.7; N, 2.1. 3a: 31P{1H}
2
2
2/3
) δ31P ) 71.90 (dd, Jea ) -10.6 Hz, Jeb ) -15.1 Hz,
J
)
)
ec
2/3
2
2
1.5 Hz,
J
ed
2
) -1.5 Hz; Pe), 69.81 (ddd, Jad ) 99.2 Hz, Jbd
2/3
-25.1 Hz, Jcd ) -4.1 Hz,
J
) -1.5 Hz; Pd), 62.15 (dddd,
de
2Jad ) 99.2 Hz, Jae ) -10.6 Hz, Jac ) -21.0 Hz,
J ) -6.6
2
2
2/3
ab
2
2
2
Hz; Pa), 58.40 (dddd, Jbc ) 98.0 Hz, Jbe ) -15.1 Hz, Jbd
)
2/3
2
2
-25.1 Hz,
J ) -6.6 Hz; Pb), 55.90 (ddd, Jcb ) 98.0 Hz, Jca
ba
) -21.0 Hz, 2Jcd ) -4.1 Hz,
J
ce
) 1.5 Hz; Pc) ppm. 3b: 31P{1H}
2/3
2
2
2
) δ31P ) 77.76 (dddd, Jbc ) 100.3 Hz, Jbd ) -24.4 Hz, Jbe
J ) -8.2 Hz; Pb), 72.50 (ddd, Jda ) 98.7, Jdb
-24.4 Hz, Jdc ) -3.2 Hz,
)
)
Figure 3. Raman spectra of complexes 1(top), 2 (middle), and 3 (bottom)
measured as pure solids.
2/3
2
2
2
-15.8 Hz,
ba
2/3
2
J
) -1.5 Hz; Pd), 71.02 (dd, Jeb
de
2/3
2
2/3
) -15.8 Hz, Jea ) -10.8 Hz,
J
) 1.5 Hz,
J ) -1.5 Hz;
Table 1. Vibrational Data of Complexes 1-3
ec
2
ed
2
2
Pe), 56.60 (ddd, Jcb ) 100.3 Hz, Jca ) -22.8 Hz, Jcd ) -3.2
[Mo(N2)(dpepp)
[Mo(N2)(dpepp)
[Mo(N2)(dpepp)
2/3
Hz,
J
) 1.5 Hz; Pc), 43.17 (dddd, 2Jad ) 98.7 Hz, 2Jac ) -22.8
vibration
(depe)] 1
(dppe)] 2
(1,2-dppp)] 3
ec
2
2/3
Hz, Jae ) -10.8 Hz,
J ) -8.2 Hz; Pa) ppm.
ab
ν(NN)
1952 IR
2007/1953 IR
2020/1960 Raman 1953 Raman
1957 IR
Spectroscopic Characterization. Infrared spectra of the solid
compounds were obtained from KBr pellets using a Bruker IFS
v66/S FT-IR spectrometer. The FT Raman spectra were recorded
with a Bruker IFS 666/CS NIR FT-Raman spectrometer at RT. As
lightsource a NdYAG-Laser with an excitation wavelength of 1064
nm was used. The samples which absorbed at this wavelength were
measured with 15% intensity. NMR spectra were measured with a
Bruker AVANCE 400 Puls Fourier Transform spectrometer operat-
1949 Raman
∼ 450 (IR, Raman)
ν(MoN)
δ(MoNN) 499, 493 (IR)
507, 497 (IR)
510, ∼500 (IR)
of complexes 1-3 are displayed in Figure 2; the correspond-
ing Raman spectra are shown in Figure 3. By comparison
with the spectra of [Mo(N2)(dpepp)(dppm)] for which a 15
N
substitution has been performed it is possible to identify the
N-N- and metal-N vibrations of the three complexes (cf.
Table 1).9 31P NMR spectra of 1-3 are displayed in Figures
4–6; the corresponding parameters are collected in Table 2.
For compounds 2 and 3, 31P-31P-COSY-45{1H CPD}
measurements were performed. Assuming that the trans-
coupling constant is positive, the absolute signs of the
coupling constants were determined. Because of the similar
geometry, the signs obtained from the analysis of the COSY-
45 spectra of compound 3 were also adapted to compound
1. The resulting coupling constants are shown in Scheme 1.
Detailed analyses of the spectra are presented in the
following, with particular emphasis on the identification of
the different isomers.
1
ing with a H frequency of 400.13 MHz and a 31P frequency of
161.98 MHz using a 5 mm inverse triple-resonance probe head.
Referencing was done with TMS δ1H ) 0 ppm and 85% pure H3PO4
δ31P ) 0 ppm as substitutive standards. All samples were measured
in THF-d8 at 300 K. All 31P NMR spectra have been recorded with
1H composite pulse decoupling. Simulated spectra were created with
the program MESTREC.11
3. Results and Analysis
Schematic structures of the complexes synthesized and
investigated in this study are given in Figure 1. IR spectra
(11) Mestrec, Mestrelab Research, Santiago de Compostela.
Inorganic Chemistry, Vol. 47, No. 14, 2008 6543

Klatt et al.
Figure 4. (a) 31P{1H}-NMR (161.98 MHz, in THF-d8) of [Mo(N2)(dpepp)(depe)] (1), measured (top) and simulated (bottom); (b) M-Spectrum of 1, expansion
of the signal at 48.80 ppm; (c) X-Spectrum of 1, expansion of the signal at 62.12 ppm.
6544 Inorganic Chemistry, Vol. 47, No. 14, 2008

Molybdenum Dinitrogen Complexes
Figure 5. 31P{1H}-NMR (161.98 MHz, in THF-d8) of [Mo(N2)(dpepp)(dppe) (2), measured and simulated.
Figure 6. Expansion of the crosspeak Pa/Pb (67.00 ppm/69.10 ppm) of [Mo(N2)(dpepp)(dppe)] (2, COSY-45, 161.98 MHz, in THF-d8).
Inorganic Chemistry, Vol. 47, No. 14, 2008 6545

Klatt et al.
Table 2. 31P NMR Data of Complexes 1-3
[Mo(N2)(dpepp)(depe)] 1
[Mo(N2)(dpepp)(dppe)] 2a
Pe δ ) 72.80 n.a.
Pa,b,c,d δ ) 63.40
Iso-[Mo(N2)(dpepp)(dppe)] 2b
2/3
Pa/b δ ) 48.80 ppm
Pc/d δ ) 62.12 ppm
Pe δ ) 87.25 ppm
2Jad/bc ) 98.3 Hz
Pa δ ) 67.00 ppm
Pb δ ) 69.10 ppm
Pc δ ) 72.30 ppm
Pd δ ) 61.10 ppm
Pe δ ) 90.94 ppm
J ) 5.1 Hz
ab
2Jcd ) -9.5 Hz
2Jac ) -21.2 Hz
2Jad ) -22.0 Hz
2Jae ) 106.7 Hz
2Jbc ) 101.6 Hz
2Jbd ) -9.7 Hz
2Jbe ) -16.6 Hz
2/3
J
) -4.9 Hz
ab
2Jac/bd ) -18.6 Hz
2/3
|
J
| ) 4.7 Hz
ed/ec
2Jea/eb ) -15.5 Hz
2Jcd ) -13.1 Hz
2/3
J
J
) 7.2 Hz
) 1.5 Hz
ce
2/3
de
[Mo(N2)(dpepp)(1,2-dppp)] isomer A 3a
[Mo(N2)(dpepp)(1,2-dppp)] isomer B 3b
2/3
2/3
Pa δ ) 62.15 ppm
J
) -6.6 Hz
Pa δ ) 43.17 ppm
Pb δ ) 77.76 ppm
Pc δ ) 56.60 ppm
Pd δ ) 72.50 ppm
Pe δ ) 71.02 ppm
J ) -8.2 Hz
ab
ab
Pb δ ) 58.40 ppm
Pc δ ) 55.90 ppm
Pd δ ) 69.81 ppm
Pe δ ) 71.90 ppm
2Jac ) -21.0 Hz
2Jad ) 99.2 Hz
2Jae ) -10.6 Hz
2Jbc ) 98.0 Hz
2Jbd ) -25.1 Hz
2Jbe ) -15.1 Hz
2Jac ) -22.8 Hz
2Jad ) 98.7 Hz
2Jae ) -10.8 Hz
2Jbc ) 100.3 Hz
2Jbd ) -24.4 Hz
2Jbe ) -15.8 Hz
2Jcd ) -4.1 Hz
2Jcd ) -3.2 Hz
2/3
2/3
J
J
) 1.5 Hz
) -1.5 Hz
J
J
) 1.5 Hz
) -1.5 Hz
ce
ce
2/3
2/3
de
de
Scheme 1
[Mo(N2)(dpepp)(depe)] (1). As evident from Figures 2
and 3, the complex [Mo(N2)(dpepp)(depe)] (1) exhibits the
IR and Raman spectra with the sharpest vibrational features.
This agrees with the 31P NMR spectroscopic result that 1
exists in only one isomeric form (see below). The N-N
stretch frequency is located at 1952 cm-1 in the IR spectrum
and at 1949 cm-1 in the Raman spectrum (Table 1). The
position of the metal-N stretch can be inferred from a
comparison with the vibrational spectra of [Mo(N2)-
(dpepp)(dppm)]. From that comparison, it also becomes clear
that the peaks at 527, 523, and 522 cm-1 in the Raman
spectra of 1, 2, and 3, respectively (Figure 3), are not
associated with the metal-N stretching vibrations because
in [Mo(N2)(dpepp)(dppm)] there is a corresponding peak (at
526 cm-1) which does not shift upon 15N substitution. For
the latter complex, ν(Mo-N) was rather shown to be located
at 453 cm-1.9 At the same position there is a (split) peak in
the Raman spectrum of 1 which has a counterpart of weak
intensity in the IR spectrum (Supporting Information, Figure
S1). This intensity difference is characteristic of metal-N
stretching vibrations in dinitrogen complexes and has been
employed to distinguish between metal-stretching and metal-
bending vibrations.12 The δ(Mo-N-N) vibrations are more
difficult to identify. By comparison with the spectrum of
[Mo(N2)(dpepp)(dppm)] they should be located around 500
cm-1. In this spectral region complex 1 exhibits compara-
6546 Inorganic Chemistry, Vol. 47, No. 14, 2008

Molybdenum Dinitrogen Complexes
tively weak IR bands (at 499 and 493 cm-1, cf. Figure 2)
which render a definite assignment difficult. Nevertheless,
these features are much more intense for compounds 2 and
3 and in these systems can be attributed to δ(Mo-N-N)
(see below).
vibrations δ(MoNN) (vide supra). On the basis of the IR
and Raman spectroscopic information available for the lower-
energy region, however, no metal-N stretching vibration can
be assigned with safety.
In agreement with the vibrational data, the 31P NMR
The measured 31P NMR spectrum of 1 (Figure 4a top)
shows a typical AMM′XX′-pattern. The signal at 48.80 ppm
is assigned to the phosphorus atoms Pa and Pb (cf. Figure
1), the signal at 62.12 ppm is associated with Pc and Pd, and
the signal at 87.25 ppm belongs to Pe. The simulated
spectrum (Figure 4a bottom) was obtained by an analysis of
the M- and the X-half-spectra (Figure 4b,c) according to the
literature.13 Every half-spectrum consists of ten lines (a-j)
which in turn are split into doublets because of the coupling
with Pe. The coupling constants are obtained by the solution
of following equations:
spectrum of [Mo(N2)(dpepp)(dppe)] (Figure 5, center) reveals
the presence of two isomers which can be separated by
COSY-45 (cf. Supporting Information, Figure S2). The
AA′BB’M spectrum of the first isomer (compound 2a; Figure
5 top) consists of two groups of signals at 73.80 ppm and
63.40 ppm with an intensity ratio of 1:4. The signal at 73.80
ppm belongs to the axial phosphorus atom, Pe, and the signal
at 63.40 ppm is caused by the four phosphorus atoms which
are located in the equatorial plane. No coupling constants
can be derived from this spectrum with safety. The 31P NMR
spectrum of the other isomer (iso-[Mo(N2)(dpepp)(dppe)]
(2b), Figure 5 bottom) exhibits an ABCDE spin system due
to five different phosphorus atoms. The signal at 90.94 ppm
belongs to the central phosphorus atom of the dpepp ligand,
Pe. The COSY-45 spectrum shows an interaction with the
signal at 67.00 ppm (Pa) and a large coupling constant (106.7
Hz), indicative of trans-coupling. The cis-interaction occurs
with a coupling constant of 2J ) -16.6 Hz to the phosphorus
atom Pb which corresponds to the signal at 69.10 ppm.
Moreover, there is observable the small cis-coupling to Pc.
The corresponding coupling constant of 2/3J ) 7.2 Hz is the
result of two coupling pathways, one transmitted by the metal
center and one by the ethylene bridge. The low value is
caused by the opposite signs of these contributions, leading
to partial cancelation (vide supra). Because of the same
reason, the cis-coupling to Pd is very small (1.5 Hz) and can
only be detected in the COSY-45 experiment (see below).
The dddd-signal at 72.00 ppm is assigned to the phos-
phorus atom Pc. The trans-coupling to Pb is large (2J ) 101.6
Hz). Furthermore, there exist two metal-mediated cis-
couplings to Pa (2J ) -21.2 Hz) and Pd (2J ) -13.1 Hz).
The signal at 69.10 ppm with a dddd-structure is assigned
to the phosphorus atom Pb. Other than the already mentioned
couplings, there exist two cis-interactions to Pd and Pa with
N ) Jac/bd + Jad/bc
;
L ) Jac/bd - Jad/bc
;
K ) Jcd +
Jab; M ) Jcd - Jab (1)
The values of N, K, L, and M can in turn be obtained from
the relations
(
)(
)
(
)(
)
√
√
N ) a - b; L ) c - f d - e ) g - j h - i ; K ) g -
h ) i - j; M ) c - d ) e - f (2)
With +98.3 Hz, the trans-interactions 2Jad/bc are positive (see
below) and cause the largest splitting, as usual. The cis-
interactions between Pe and the phosphorus atoms of the depe
ligand, Pa and Pb, are only transmitted by the metal center;
the corresponding coupling constants are negative and have
values of 2Jea/eb) -15.5 Hz. This also applies to 2Jac/bd which
are -18.6 Hz. The relative signs of the above coupling
constants can also be inferred from a comparison with those
evaluated for compound 3 by COSY-45 measurements (cf.
Scheme 1). The coupling constant between Pa and Pb is due
to two interactions, one transmitted by the metal center and
2/3
one by the ethylene bridge, and has a value of
J ) -4.9
ab
Hz. The same applies to the coupling constants between Pc/
Pd and Pe. In this case, however, the sign cannot be inferred
from a comparison with 3 as the corresponding coupling
constants are very small and thus|
J
| ) 4.7 Hz. The
ed/ec
coupling constants of J ) -9.7 Hz and 2/3J ) 5.1 Hz,
2/3
2
2
remaining cis-interaction, Jcd, is small and negative (-9.5
Hz). Note that this coupling constant results from a metal-
mediated interaction and another interaction via Pe which
obviously acts to decrease the absolute value of the coupling
constant (this is even more pronounced in 3a and 3b; see
below).
respectively. The coupling constants of the signal at 67.0
ppm could not be determined unambiguously. The phos-
phorus atom Pa which gives rise to this signal couples with
four different phosphorus atoms, so theoretically a dddd-
structure should result. The fact that a ddt-structure is
observed can only be explained if the coupling constants to
Pd and Pc have exactly the same value. The corresponding
[Mo(N2)(dpepp)(dppe)] (2). This complex was obtained
in two isomeric forms (cf. Figure 1), as evident from the IR
spectrum which exhibits two clearly distinct bands for the
N-N stretch ν(NN) at 2007 and 1953 cm-1 (Figure 2). In
the Raman spectrum (Figure 3) the corresponding vibrations
appear at 2020 and 1960 cm-1. In the region of the
metal-ligand vibrations of 2, two IR-active vibrations can
be identified below 524 cm-1, at 507 cm-1 and at 497 cm-1.
These bands are attributed to the metal-N-N bending
2
simulation gives a value of J ) -22.0 Hz. The signal at
61.10 ppm, finally, does not show a large splitting attributable
to trans-coupling to another phosphine. It is therefore
assigned to the phosphorus atom Pd in trans-position to the
N2 ligand. For this signal the simulation gives a superposed
ddd-structure.
To determine the relative signs of the above coupling
constants, a 45° COSY spectrum has been recorded (Sup-
porting Information, Figure S2). Figure 6 shows the cross-
peak resulting from the coupling between Pa and Pb. The
crosspeak consists of signals arranged in eight squares. Every
(12) Tuczek, F.; Horn, K. H.; Lehnert, N. Coord. Chem. ReV. 2003, 245,
107–120.
(13) Gu¨nther, H. Angew. Chem. 1972, 19, 907.
Inorganic Chemistry, Vol. 47, No. 14, 2008 6547

Klatt et al.
Figure 7. 31P{1H}-NMR (161.98 MHz, in THF-d8) of [Mo(N2)(dpepp)(1,2-dppp)] (3), simulated isomer A (top), measured spectrum (center), and simulated
isomer B (bottom).
square has the same side length given by the coupling constant
Jab. The shifts of the squares against each other result from the
couplings of Pa/Pb with the other phosphorus atoms Pc, Pd, and
Pe. The couplings of Pa can be determined from the shifts along
the x-axis, and the couplings of Pb from those along the y-axis.
The relative signs can be obtained by comparison of the shifts
due to the coupling with the same phosphorus atom. If both
shifts occur toward higher or lower frequency, they have the
same sign; if they point into different directions then the signs
are opposite (often the vector sum of these shifts is considered,
and the “tilt” of the resultant with respect to the x-axis is taken
to indicate the relative sign)14 In the Pa-Pb crosspeak, for
example, the coupling constants Jac and Jbc with phosphorus
atom Pc have different signs and the couplings Jad and Jbd with
Pd have the same sign. All crosspeaks of compound 2b have
been analyzed this way (Supporting Information, Figures
S3-S10). If, according to the literature, the value of the trans-
coupling is assumed to be positive, then the other absolute signs
can be derived in relation to this assumption.
and ∼500 cm-1) which can be assigned to the Mo-N-N
bending vibrations. In the lower-energy region, no identifica-
tion of ν(Mo-N) is possible withouth 15N substitution.
31P NMR spectroscopy was employed to obtain informa-
tion about the structure of compound 3 in solution. To
analyze the spectra and, in particular, to determine the signs
of the coupling constants, a COSY-45 spectrum was recorded
and analyzed as described for compound 2b (Supporting
Information, Figure S11). On the basis of this information,
the signals in the 31P NMR-spectrum of 3 (Figure 7 center)
can be assigned to isomers A and B, both of which are
ABCDE spin systems. First, the spectrum of isomer A is
analyzed (Figure 7, top; cf. Supporting Information, Figures
S12-S19). The signal at 71.90 ppm exhibiting a dd-structure
is attributed to the phosphorus atom Pe which is coordinated
in trans-position to the dinitrogen ligand. Two cis-couplings
are visible, one to Pb with a coupling constant of 2J ) -15.1
2
Hz and one to Pa with J ) -10.6 Hz. The coupling with
the phosphorus atom Pc can only be evaluated from an
analysis of the COSY-45 spectra, leading to a coupling
constant of +1.5 Hz. The coupling with Pd corresponds to
-1.5 Hz. This indicates that the interactions transmitted by
the metal center and the alkyl bridge have the same
magnitude and almost exactly cancel each other. The signal
at 69.81 ppm with a ddd-structure is caused by the
phosphorus atom Pd. The trans-interaction to Pa has a
coupling constant of 2J ) 99.2 Hz, and the cis-couplings to
Pb and Pc have J-values of -25.1 and -4.1 Hz, respectively.
[Mo(N2)(dpepp)(1,2-dppp)] (3). The product of the
synthesis described above contains two isomers A and B (see
Figure 1). This is also reflected by the IR spectrum of
[Mo(N2)(dpepp)(1,2-dppp)] (3) which shows a broad band
at 1957 cm-1 for ν(NN); the same vibration gives a broad
peak in the Raman spectrum at 1953 cm-1. Again several
shoulders appear in the IR spectrum below 520 cm-1 (at 510
(14) Fontaine, X. L. R.; Kennedy, J. D.; Shaw, B. L.; Vila Jose´, M. J. Chem.
Soc., Dalton Trans. 1987, 2401.
6548 Inorganic Chemistry, Vol. 47, No. 14, 2008

Molybdenum Dinitrogen Complexes
The dddd-signal at 62.2 ppm is associated with Pa. Other
than the mentioned interactions, there exist cis-interactions
with Pb (2/3J ) -6.6 Hz) and Pc (2J ) -21.0 Hz). The signal
at 58.40 ppm belongs to the phosphorus atom Pb and exhibits
a dddd-structure. The trans-coupling with Pc has a coupling
4. Discussion
In the preceding sections 31P NMR and vibrational
spectroscopic investigations of Mo dinitrogen complexes with
five phosphorus ligands have been presented. This coordina-
tion environment has been generated by a combination of a
triphosphine (dpepp) and different diphosphine ligands with
C2 bridges (depe, dppe, and 1,2-dppp). The structure and
bonding of this type of complexes has been described in
detail in reference 9. The emphasis of the present paper lies
on the NMR-spectroscopic characterization of these com-
plexes, in particular, on the spectroscopic identification of
different isomers in solution. This issue is of key relevance
whenever the synthesis of a catalytically active complex with
predefined properties on the basis of multidentate coligands
is pursued.
The 31P spectra of the title compounds were fully analyzed
and reproduced by simulation, leading to the identification
of one isomer for the complex [Mo(N2)(dpepp)(depe)] (1).
In contrast, two isomeric forms were evidenced for com-
pound 2, [Mo(N2)(dpepp)(dppe)] 2a and iso-[Mo(N2)(dpepp)-
(dppe)] (2b), as well as compound 3, [Mo(N2)(dpepp)(1,2-
dppp)] (isomers A and B). For the latter complex, the methyl
group of the 1,2-dppp ligand either points into the direction
of the N2 ligand (3a) or into the opposite direction (3b).
These results were found to be compatible with the corre-
sponding vibrational (IR and Raman) data, showing an
unsplit N-N stretch for 1, strongly split N-N stretching
vibrations for compound 2, and weakly split N-N stretches
for compound 3.
2
constant of J ) 98.0 Hz. The ddd-signal of Pc, finally, is
located at 55.90 ppm.
The remaining signals are assigned to the other isomer
(B, Figure 7 bottom; cf. Supporting Information, Figures
S20-S28). In particular, the signal at 77.76 ppm with dddd-
structure is associated with the phosphorus atom Pb. There
exist a large trans-coupling (2J ) 100.3 Hz) with Pc, two
2
cis-interactions with coupling constants of J ) -24.4 Hz
(Pd) and 2J ) -15.8 Hz (Pe) and a cis-coupling with Pa (2/3J
) -8.2 Hz). The signal at 72.50 ppm (Pd) is connected with
the signal at 43.17 ppm by a trans-coupling constant of 98.7
Hz. In addition to the mentioned cis-interaction, there exists
another cis-coupling with 2J ) -3.2 Hz to Pc. The signal at
71.02 ppm is associated with the phosphorus atom Pe. This
atom which is located in trans-position to the dinitrogen
ligand exhibits four cis-couplings, the already mentioned one
to Pb and a second one to Pa (2J ) -10.8 Hz). The two
interactions with Pc and Pd can only be recovered in the
COSY-45 spectra, having values of +1.5 and -1.5 Hz,
respectively. The signal at 56.60 ppm exhibits a ddd-structure
and belongs to Pc. In addition to the couplings to Pb and Pd,
there exists an interaction to Pa with a coupling constant of
2J ) -22.8 Hz. The signal at 43.17 ppm, finally, shows a
dddd-structure and is associated with Pa.
The frequencies of the N-N stretching vibrations (1952
cm-1 for 1, 1953 cm-1 for 2a, 2007 cm-1 for 2b, and 1957
cm-1 for 3a/b) reflect the different degrees of activation
imparted to the dinitrogen ligand by the phosphine coligands.
Compounds 1, 3a, and 3b have dinitrogen ligands in trans-
position to the central phosphorus atom of the dpepp ligand.
All of these complexes exhibit the N-N stretch at 1950-1960
cm-1. The same applies to compound 2a. Compound 2b, in
contrast, has a terminal phosphine group of the dpepp ligand
coordinated in trans-position to the N2 ligand. Correspond-
ingly, the N2 stretch is observed at higher frequency than in
1, 3a, 3b, and 2a, that is, at 2010-2020 cm-1. This
demonstrates the lower activation of N2 in a complex with
a trans-diphenylalkyl- compared to a complex with a trans-
dialkylphenylphosphine group. As expected, a much smaller
difference in ν(NN) is observed between the two isomers of
[Mo(N2)(dpepp)(dppe)] (3a and 3b) in which the methyl
group of the equatorial 1,2-dppp ligand points into different
directions (Figure 1).
The 31P NMR spectra of all compounds could be analyzed
and perfectly reproduced by simulation, giving a complete
set of chemical shifts and coupling constants. A consistent
set of coupling constants and corresponding signs was
derived for compounds 1-3 and the related complex
[Mo(N2)(dpepp)(dppm)] which has been studied previously
(Scheme 1). The 31P NMR trans-coupling constants were
found to be in the range of ∼100 Hz. For cis-coupling, two
cases have to be distinguished: If the interaction exclusively
In summary, the 31P NMR spectra of isomers A and B
could be fully analyzed and reproduced by simulation. The
coupling constants are very similar in both isomers. More-
over, they closely resemble those of the complex
[Mo(N2)(dpepp)(dppm)] (Scheme 1) which has been studied
previously (signs of coupling constants have been added for
this complex where they are evident from a comparison with
compound 3). However, there are important differences in
the chemical shifts: Whereas the chemical shifts of the dpepp
ligand are very similar for 3a and 3b, those of the 1,2-dppp
ligand greatly differ. In particular, the splitting between the
resonances of Pa and Pb, ∆δab, is comparable for isomer A
(∆δab ) 3.75 ppm) to that observed for compounds 1 and
2a (∆δab ) 0) whereas these resonances are one order of
magnitude more split for isomer B (∆δab ) 34.59 ppm). We
attribute this to a geometric distortion of isomer B which is
caused by an interaction between the phenyl substituent of
the phosphorus atom Pe in trans-position of dinitrogen and
the methyl group of the 1,2-dppp ligand. Note that this
methyl group is directed toward the Pe phenyl group in
isomer B whereas it points in the opposite direction for
isomer A and in this case does not interact with the phenyl
group of Pe. These considerations suggest that the 31P NMR
spectrum with the small splitting between Pa and Pb is
associated with structure 3a, and the isomer with the large
splitting between Pa and Pb (isomer B) belongs to structure
3b in Figure 1.
Inorganic Chemistry, Vol. 47, No. 14, 2008 6549

Klatt et al.
occurs via the metal, the coupling constant is ∼20 Hz and
negative. If the interaction occurs simultaneously over the
metal center and an ethylene bridge, the coupling constant
is small (sometimes in the order of the line width), indicating
that the corresponding coupling pathways correspond to
coupling constants with almost identical absolute values but
opposite signs. This is in agreement with literature data
the relative intensities in the 31P NMR spectrum 2a and 2b
occur in about equal amounts. Astonishingly, however, no
corresponding iso-forms could be evidenced for compounds
1 and 3. For the depe complex 1 this may have electronic
reasons; that is, a configuration where the diethylphosphine
groups are in trans-position to diphenylphosphine groups
may be particularly favorable as strongly electron-donating
groups are arranged in trans-position to strongly electron-
accepting groups. For the 1,2-dppp complex 3, on the other
hand, steric reasons may account for the lack of the iso-
form. Specifically, for 3b (with the methyl group pointing
downward) an iso-form may be unfavorable as in trans-
position of the dinitrogen ligand there would be two phenyl
groups instead of one coming in close contact with the methyl
group. For 3a, on the other hand, this argument would not
apply, and no principal reasons are visible which would
prevent the formation of an iso-form in this geometry.
To conclude, the vibrational and 31P NMR spectroscopic
properties of three complexes with five coordinating phos-
phine groups have been determined, and correlations between
their geometries and their spectroscopic properties have been
derived. In a previous paper it has been shown that the
activation of N2 in these complexes is higher than in
corresponding bis(diphosphine) complexes with two dini-
trogen ligands but lower than in trans-acetonitrile dinitrogen
complexes with two diphosphine coligands.9 The latter
systems, on the other hand, have the disadvantage that the
nitrile coligands are labile. As dinitrogen complexes with a
pentaphosphine ligation have also been shown to be proto-
natable at the N2 ligand,8 these systems appear to represent
a good compromise between sufficient activation of the
dinitrogen ligand on the one hand and thermodynamic/kinetic
stability of the ligand environment on the other. This agrees
with our theoretical finding that a catalytic cycle of ammonia
synthesis on the basis of a Mo pentaphosphine complex
should be feasible.10 Practically this concept is limited by
the fact that the phosphine ligand in trans-position of N2
gets labile at higher oxidation states of the metal (e.g., in
the nitrido complex) and sooner or later will be replaced by
another Lewis-basic ligand or a solvent molecule. Therefore,
the necessity persists to covalently attach the trans-ligand
to the complex in a way that it cannot dissociate at any stage
of the catalytic cycle. Concepts to achieve such an archi-
tecture on the basis of phosphine ligands for an efficient
homogeneous catalyst of ammonia synthesis have been
developed and are continued to be actively pursued in our
group.19
2
indicating that the J coupling constant via a metal has a
negative sign,15 and the 3J coupling constant via an ethylene
bridge a positive sign.16 It is also known that for second
2
and third-row metals J(P,P) trans-coupling constants are
large and positive.17,18 This in turn has been used to
determine the absolute signs of coupling constants for
compounds 2 and 3.
The 31P spectra also sensitively reflect the symmetry of
the complexes. Quasi-tetragonal symmetry is exhibited by
complex 2a which has one phosphine group in axial and
four diphenylphosphine endgroups in equatorial position. The
signals deriving from the equatorial phosphines a,b,c,d appear
almost unsplit. A similar geometry is present in compound
1. However, in this case the equatorial diethylphosphine and
diphenylphosphine groups Pa/Pb and Pc/Pd, respectively,
appear at different chemical shifts (48.80 and 62.12 ppm,
respectively). Diethylphosphine is a stronger σ-donor than
diphenylphosphine and the latter is a stronger π-acceptor
than diethylphosphine; Pa and Pb are therefore more shielded
than Pc and Pd. In compound 3, finally, the symmetry of
complex 2a is broken by an additional methyl group attached
to the ethylene bridge of the equatorial diphos ligand. The
methyl group can point “upward” (i.e., in the direction of
the N2 ligand; isomer 3a) or “downward” (i.e., in the opposite
direction; isomer 3b). Importantly, isomer B exhibits a
significantly larger splitting between the 31P resonances of
the bidentate ligand than found for the corresponding depe
and dppe complexes. This is attributed to a distortion of the
complex because of a steric interaction between the additional
methyl group of the diphosphine and the phenyl group of
the phosphorus atom Pe in trans-position to the N2 ligand
and correlates isomer B with structure 3b. Correspondingly,
isomer A is correlated with structure 3a of Figure 1.
For complex 2 an iso-form (2b) has been identified in
which the central phosphorus atom of the dpepp ligand is in
equatorial position; that is, cis to the dinitrogen group. From
(15) Pregosin, P. S.; Kunz, R. W. NMR Basic Principles and Progress.
31P and 13C NMR of Transition Metal Phosphine Complexes; Springer-
Verlag: New York, 1979.
(16) Friebolin, H. Ein- und zweidimensionale NMR-Spektroskopie; Wiley
VCH: Weinheim, 2006.
(17) (a) Bertrand, R. D.; Ogilie, F. B.; Verkade, J. G. J. Am. Chem. Soc.
1970, 92, 1916. (b) Ogilie, F. B.; Jenkins, J. M.; Verkade, J. G. J. Am.
Chem. Soc., 1970, 92, 1908. (c) Crumbliss, A. L.; Topping, R. J.
Phosphorus-31 NMR: Spectral Properties in Compound Characteriza-
tion and Structural Analysis; VCH: New York, 1987; Vol. 15. (d)
Hughes, A. N. Phosphorus-31 NMR: Spectral Properties in Compound
Characterization and Structural Analysis; VCH: New York, 1987; Vol.
19. (e) Clark, H. C.; Kapoor, P. N.; McMahon, I. J. J. Organomet.
Chem. 1984, 190, C101. (f) Airey, A. A.; Swiegers, G. F.; Willis,
A. C.; Wild, S. B. Inorg. Chem. 1997, 36, 1588.
Acknowledgment. Support of this research by DFG Tu58-
12-2 (SPP1118 “Sekunda¨re Wechselwirkungen”) is gratefully
acknowledged.
Supporting Information Available: IR, Raman, and COSY-
45 spectra of the compounds (PDF). This material is available free
of charge via the Internet at http://pubs.acs.org.
(18) This does not apply to first-row metals where 2J(P,P) for trans-coupling
may be negative, cf. ref. 17a. Likewise, 2J(P,P)cis and 2J(P,P)trans values
for metal complexes in the first transition-metal series can be of similar
magnitude whereas in the second and third series the trans-coupling
is usually much larger.
IC800389G
(19) Stephan, G.; Na¨ther, C. N.; Sivasankar, C.; Tuczek, F. Inorg. Chim.
Acta 2008, 361, 1008–1019.
6550 Inorganic Chemistry, Vol. 47, No. 14, 2008