Vanadium and niobium diamidophosphine complexes and their reactivity

Article abstract of DOI:10.1139/v03-153

The tridentate ligand precursors R′(CH₂SiMe ₂NR″)₂ (^R′R″[NPN]: R′ = Cy, Ph; R″ = Ph, Mes, Me) were prepared from metathesis reactions of a lithiated amine, chloro(chloromethyl)dimethylsilane, the appropriate 1°

Full text of DOI:10.1139/v03-153

1431
Vanadium and niobium diamidophosphine
complexes and their reactivity
Michael P. Shaver, Robert K. Thomson, Brian O. Patrick, and Michael D. Fryzuk
Abstract: The tridentate ligand precursors R′P(CH2SiMe2NR′′)2 (R′R ′′[NPN]: R′ = Cy, Ph; R′′ = Ph, Mes, Me) were
prepared from metathesis reactions of a lithiated amine, chloro(chloromethyl)dimethylsilane, the appropriate 1° phos-
phine, and n-butyl lithium and were isolated as solvent adducts. Metathesis between CyPh[NPN]Li2(OEt2), 2, and
VCl3(THF)3 afforded (CyPh[NPN]VCl)2, 7, whose solid-state structure was established by X-ray crystallography. Reduc-
tion attempts of the (R ′R ′′[NPN]VCl)2 species with KC8 incorporated molecular nitrogen but were complicated by imide
formation and ligand decomposition. Metathesis of 2 with NbCl2Me3 afforded the highly unstable complex
CyPh[NPN]NbMe3, 15. Attempts to hydrogenate this species were unsuccessful.
Key words: vanadium, niobium, metathesis, coordination chemistry, reduction, hydrogenation.
Résumé : Les précurseurs de ligand R′P(CH2SiMe2NR′′)2 (R ′ R ′′[NPN] : R′ = Cy, Ph; R′′ = Ph, Mes, Me) ont été pré-
paré des réactions de métathèse du amine lithié correspondant, chloro(chloromethyl)dimethylsilane, le phosphine corres-
pondant et n-butyl lithium. Les ligands ont été isolés comme adducts de solvant de selde dilithium. Les réactions de
métathèse du CyPh[NPN]Li2(OEt2), 2, avec VCl3(THF)3 ont donné (CyPh[NPN]VCl)2, 7, dont de la structure d’état solide
a été établie par crystallographie de rayons-X. Les tentatives de réduction de l’espèce (R′R ′′[NPN]VCl)2 avec KC8 a in-
corporé l’azote moléculaire mais a été compliquée par la formation d’imide et la décomposition de ligand. Les réac-
tions de métathèse de 2 avec NbCl2Me3 ont donné le complexe extrêmement instable CyPh[NPN]NbMe3, 15. Les
tentatives de hydrogéner était infructueux.
Mots clés : vanadium, niobium, la métathèse, la chimie de coordination, la réduction, l’hydrogénation. Shaver et al.
1437
Introduction
Experimental
Recent work from our laboratory has focused on the
chemistry of tantalum dinitrogen complexes stabilized by the
[NPN] ligand, where [NPN] = PhP(CH2SiMe2NPh)2 (1–4).
One of the most intriguing species is ([NPN]Ta)2(µ-H)2(µ-
η1-η2-N2), A, with a side-on, end-on bound dinitrogen unit.
This complex has displayed remarkable reactivity with sim-
ple organometallic hydride reagents. The coordination chem-
istry of [NPN] with Ta — the heaviest member of group
5 — piqued our interest in the lighter elements of this group,
namely V and Nb. In this manuscript we explore some of
our successes and failures in trying to extend dinitrogen
chemistry to low valent vanadium and niobium [NPN] com-
plexes.
General information
Unless otherwise stated, all manipulations were performed
under an atmosphere of dry, oxygen-free dinitrogen by
means of standard Schlenk or glovebox techniques (Vacuum
Atmospheres HE-553-2 glovebox equipped with an MO-40-
2H purification system and a –40 °C freezer). Thick-walled
Pyrex reaction vessels with 5 or 10 mm Teflon needle valves
and ground glass joints were used to maintain an inert atmo-
sphere at pressures between 10–6 and 4 atm (1 atm =
101.325 kPa). Hexanes and toluene were purchased anhy-
drous from Aldrich and further dried by passage through a
tower of alumina and degassed by passage through a tower
of Q-5 catalyst under a positive pressure of nitrogen (5). An-
hydrous diethyl ether, tetrahydrofuran (THF), benzene, and
hexamethyldisiloxane (HMDS) were stored over sieves and
distilled from sodium benzophenone ketyl under argon. Ni-
trogen was dried and deoxygenated by passage through a
column containing activated molecular sieves and MnO.
Deuterated benzene was dried by refluxing with molten po-
tassium metal in a sealed vessel under partial pressure, then
trap-to-trap distilled, and degassed by freeze-pump-thaw
three times. Unless otherwise stated, H, 31P{1H}, and Li
1
7
Received 3 June 2003. Published on the NRC Research Press Web site at http://canjchem.nrc.ca on 13 October 2003.
M.P. Shaver, R.K. Thomson, B.O. Patrick, and M.D. Fryzuk.1 University of British Columbia, Department of Chemistry,
2036 Main Mall, Vancouver, BC V6T 1Z1, Canada.
1Corresponding author (e-mail: fryzuk@chem.ubc.ca).
Can. J. Chem. 81: 1431–1437 (2003)
doi: 10.1139/V03-153
© 2003 NRC Canada

1432
Can. J. Chem. Vol. 81, 2003
Table 1. Crystallographic data and details of refinement for
CyMes[NPN]Li2(THF), 4.
NMR spectra were recorded on a Bruker AMX-500 instru-
1
1
ment operating at 500.1 MHz for H spectra. H NMR spec-
tra were referenced to internal C6D5H (7.15 ppm), 31P{1H}
NMR spectra to external P(OMe)3 (141.0 ppm with respect
Empirical formula
C34H57Li2N2OPSi2
7
to 85% H3PO4 at 0.0 ppm), and Li{1H} NMR spectra to a
Formula weight
Crystal system
Space group
610.86
Monoclinic
P21/a
0.3 mol L–1 LiCl solution in MeOH (0.00 ppm). Elemental
analyses were performed in the departmental facility; mass
spectra were recorded on a Kratos MS 50. Clusters assigned
to specific ions show appropriate isotopic distribution pat-
terns, as calculated for the atoms present. NbCl5 was pur-
chased from Strem Chemicals and used without further
purification. Anhydrous ZnCl2 was purchased from Aldrich
and dried by refluxing in excess SOCl2 under N2.
a, b, c (Å)
β (°)
19.597(1), 8.9414(4), 21.492(2)
105.846(3)
3623.0(4)
4
V (Å3)
Z
Dcalcd (g cm–3)
1.120
µ (Mo Kα) (cm–1)
T (K)
2θ range (°)
1.69
173 1
55.7
+
MeNH3 Cl– was purchased from Fisher Chemicals and
recrystallized out of hot ethanol and dried in vacuo prior to
use. The reagents PhNH2, ClSiMe2CH2Cl, CyPH2, MesNH2
(MesNH2 = 2,4,6-Me3C6H2NH2), and PhPCl2 were pur-
chased from Aldrich and purified by standard procedures
(6). Solutions of MeLi (1.4 mol L–1 in diethylether) and
nBuLi (1.6 mol L–1 in hexanes) were obtained from Acros
Organics and used as received. The compounds PhPH2 (7),
Total reflections
Unique reflections
Parameters
34 076
8455
508
a
R1
0.086
a
Rw
0.109
Goodness-of-fit
0.87
PhPh[NPN]Li2(THF)2
(2),
CyPh[NPN]Li2(OEt2)
(8),
aR1 = Σ||Fo| – |Fc||/Σ|Fo|; Rw = Σw(|Fo | – |Fc2|)2/Σw|Fo | )1/2
.
2
2 2
VCl3(THF)3 (9), KC8 (10), and NbCl2Me3 (11) were pre-
pared by literature methods.
Crystallographic data2 for compounds 4 and 7 appear in
Tables 1 and 2, respectively. All measurements were made
on a Rigaku–ADSC CCD area detector with graphite mono-
chromated Mo Kα radiation. The data were processed (12)
and corrected for Lorentz and polarization effects. The struc-
ture was solved by direct methods (13) and expanded by us-
ing Fourier (14) techniques. All nonhydrogen atoms were
refined with anisotropic thermal parameters. Neutral atom
scattering factors and anomalous dispersion corrections were
taken from the International tables for X-ray crystallography
(15, 16).
Table 2. Crystallographic data and details of refinement for
CyPh[NPN]VCl)2, 7.
(
Empirical formula
C24H37N2PClSi2V
Formula weight
Crystal system
Space group
a, b, c (Å)
α, β, γ (°)
V (Å3)
1054.22
Triclinic
P1
10.2811(5), 11.4776(4), 13.3016(6)
103.948(3), 111.556(1), 92.914(3)
1399.8(1)
1
1.250
Z
Dcalcd (g cm–3)
CyMes[NPN]Li2(THF) (4)
µ (Mo Kα) (cm–1)
T (K)
6.06
173
A
300 mL ethereal solution containing 12.1 mL
1
(11.640 g, 86.1 mmol) of MesNH2 was cooled to –78 °C in
a dry ice – isopropanol slush bath. A solution of 1.6 mol L–1
nBuLi in hexanes was added dropwise via syringe. After
30 min at –78 °C, the solution was warmed to room temper-
ature (r.t.) and stirred for 60 min. Upon recooling the solu-
tion to –78 °C, 11.4 mL (12.320 g, 86.1 mmol) of
ClSiMe2CH2Cl was added and stirred for 30 min. After
warming to r.t. and stirring for an additional 30 min, the so-
lution was cooled to 0 °C and 5.8 mL (43.1 mmol) of CyPH2
2θ range (°)
55.8
Total reflections
Unique reflections
Parameters
12 014
5342
280
a
R1
0.047
a
Rw
0.095
Goodness-of-fit
aR1 = Σ||Fo| – |Fc||/Σ|Fo|; Rw = Σw(|Fo | – |Fc2|)2/Σw|Fo | )1/2
1.48
2
2 2
.
n
and 107.6 mL (172.2 mmol) of BuLi was added consecu-
tively via syringe. This solution was stirred for 30 min and
allowed to warm to r.t. The resulting white suspension was
concentrated in vacuo, and the white residue was dissolved
in toluene (32 g) and filtered through Celite. Removal of to-
luene in vacuo yielded a foamy white solid, which was dis-
solved in hexanes (150 mL). Addition of 5.25 mL
(0.0647 mol) of THF caused CyMes[NPN]Li2(THF) (4) to
precipitate as a white microcrystalline solid, which was col-
lected on a frit and dried under vacuum (79% yield). The
other ligand precursors, PhMes[NPN]Li2(diox)2 (3),
PhMe[NPN]Li2(diox)2 (5), and CyMe[NPN]Li2(diox) (6)
(diox = dioxane) were synthesized using the appropriate
phosphines and amines in an analogous manner. Precipita-
tion of 3, 5, and 6 was accomplished by addition of 3 equiv
of dioxane to a hexanes solution of each product. Recovered
yields were 81%, 74%, and 66%, respectively. For the N-
2 Supplementary data (crystallographic data (excluding structure factors) for CyMes[NPN]Li2(THF), 4, and (CyPh[NPN]VCl)2, 7, may be pur-
chased from the Depository of Unpublished Data, Document Delivery, CISTI, National Research Council Canada, Ottawa, ON K1A 0S2,
Canada (http://www.nrc.ca/cisti/irm/unpub_e.shtml for information on ordering electronically). CCDC 211498 and 211499, respectively,
contain the supplementary data for this paper. These data can be obtained, free of charge, via www.ccdc.cam.ac.uk/conts/retrieving.html (or
from the Cambridge Crystallographic Data Centre, 12 Union Road, Cambridge, U.K.; fax +44 1223 336033; or deposit@ccdc.cam.ac.uk).
© 2003 NRC Canada

Shaver et al.
1433
+
methyl derivatives, MeNH3 Cl– was used to prepare the lith-
C28H54Cl2N4P2Si4V2 (%): C 42.36, H 6.86, N 7.06; found: C
42.09, H 6.89, N 6.90.
n
ium amide, and an additional equivalent of BuLi was used.
1
3: H NMR (C6D6, 25 °C, 500 MHz) δ: 0.22, 0.33 (s, 12H
total, Si-CH3), 2.20 (s, 18H, NPh-CH3), 3.17 (s, 16H,
dioxane), 0.89, 1.28 (m, 4H, P-CH2), 7.15, 7.24, 7.61 (m,
5H, P-Ph-H), 6.80 (s, 4H, NPh-m-H). 31P{1H} NMR (C6D6,
Reduction of 7 with 2.2 KC8 under 4 atm N2
A thick-walled reaction vessel was charged with a mag-
netic stir bar, 0.218 g (0.413 mmol) of 7, and 0.056 g
(0.909 mmol) of KC8. Et2O (50 mL) was vacuum transferred
into the flask, frozen at –196 °C, and charged with 4 atm of
N2. The brick-red solution turned deep purple upon warming
to r.t. The reaction was stirred for 24 h. Filtration through
Celite and removal of solvent gave a waxy purple, pentane-
soluble solid. EI-MS m/z (%): 505 ([M+]: CyPh[NPN]VϵN,
100), 491 (CyPh[NPN]V, 22).
25 °C) δ: –34.1 (q, JPLi = 18.9 Hz). Li{1H} NMR (C6D6,
1
7
1
25 °C) δ: –0.7 (s, 1Li), –0.8 (d, JPLi = 18.9 Hz, 1Li). Anal.
calcd. for C38H59Li2N2O4PSi2 (%): C 64.38, H 8.39, N 3.95;
1
found: C 64.54, H 8.49, N 3.86. 4: H NMR (C6D6, 25 °C,
500 MHz) δ: 0.21, 0.35 (s, 12H total, Si-CH3), 2.20 (s, 18H,
NPh-CH3), 1.13, 2.96 (s, 8H, THF), 0.85, 0.97 (m, 4H, P-
CH2), 1.22, 1.38, 1.68, 1.77, 1.91 (m, 11H, P-C6H11), 6.82
(s, 4H, NPh-m-H). 31P{1H} NMR (C6D6, 25 °C) δ: –30.2 (s).
7Li{1H} NMR (C6D6, 25 °C) δ: –0.8 (s, 2 Li). Anal. calcd.
for C34H57Li2N2OPSi2 (%): C 66.85, H 9.41, N 4.59; found:
C 66.98, H 9.60, N 4.39. 5: 1H NMR (C6D6, 25 °C,
500 MHz) δ: –0.09, –0.10 (s, 12H total, Si-CH3), 2.83 (s,
6H, N-CH3), 3.57 (s, 16H, dioxane), 0.89, 0.91 (m, 4H, P-
CH2), 7.15, 7.22, 7.49 (m, 5H, P-Ph-H). 31P{1H} NMR
Reduction of 7 with 3.3 KC8 under 4 atm N2
Diethylether (50 mL) was vacuum transferred into a thick-
walled reaction vessel charged with a magnetic stir bar,
0.218 g (0.413 mmol) of 7, and 0.084 g (1.364 mmol) of
KC8 and frozen at –196 °C. The vessel was charged with 4
atm of N2 and warmed to r.t. The brick-red solution changed
to olive green and was stirred for an additional 24 h. Filtra-
tion through Celite and removal of solvent gave a dark
green, pentane-soluble powder. EI-MS m/z (%): 1049 ([M+]:
K(CyPh[NPN]VϵN)2, 12), 505 ([M+] – K(CyPh[NPN]VϵN),
100), 491 (CyPh[NPN]V, 20).
1
7
(C6D6, 25 °C) δ: –37.5 (q, JPLi = 26.6 Hz). Li{1H} NMR
1
(C6D6, 25 °C) δ: –1.4 (s, 1 Li), –1.5 (d, JPLi = 26.6 Hz,
1 Li). Anal. calcd. for C22H43Li2N2O4PSi2 (%): C 52.78, H
1
8.66, N 5.60; found: C 52.52, H 8.41, N 5.75. 6: H NMR
(C6D6, 25 °C, 500 MHz) δ: 0.35 (s, 12H total, Si-CH3), 3.42
(s, 8H, dioxane), 0.88, 1.22 (m, 4H, P-CH2), 1.68, 1.79, 1.94
(m, 11H, P-C6H11), 3.02 (s, 6H, N-CH3). 31P{1H} NMR
Reduction of 9 with 2.2 KC8 under 1 atm N2
7
(C6D6, 25 °C) δ: –36.1 (s). Li{1H} NMR (C6D6, 25 °C) δ: –
Diethylether (50 mL) was transferred via syringe into a
thick-walled reaction vessel charged with a magnetic stir
bar, 0.252 g (0.413 mmol) of 9, and 0.056 g (0.909 mmol) of
KC8 under an atmosphere of N2. The vessel was sealed and
the contents stirred for 24 h. Filtration through Celite and re-
moval of solvent gave a brown, hexanes-soluble powder.
Washing the powder with pentane produced an impure
brown solid (CyMes[NPN]V=NMes, 13) and a brown solution,
which, upon cooling to –37 °C, afforded brown needle crys-
tals of MesN(SiMe2CH2)2PCy, 14. Reductions of 8 and 10–
12 were conducted in the same manner. 13: EI-MS m/z (%):
708 ([M+] CyMes[NPN]V=NMes, 50), 575 ([M+] – NMes,
1.1 (s, 1 Li). Anal. calcd. for C18H41Li2N2O2PSi2 (%): C
51.65, H 9.87, N 6.69; found: C 51.93, H 9.85, N 6.60.
(
CyPh[NPN]VCl)2 (7)
An ethereal solution (100 mL) of CyPh[NPN]Li2(OEt2)
(3.00 g, 4.23 mmol) maintained at –78 °C was added via
cannula to a 3:1 mixture of Et2O and THF containing
VCl3(THF)3 (1.58 g, 4.23 mmol). After the addition, the
solution was warmed to r.t. and stirred for 90 min. Upon re-
moval of volatiles, the resultant brick-red powder was dis-
solved in toluene (100 mL); the solution was filtered through
Celite to remove LiCl, and the toluene was removed in
vacuo. The brick-red powder was washed with pentane and
dried under vacuum (75.8%, 1.69 g). Crystals of 7 suitable
for X-ray diffraction were grown from a concentrated solu-
tion of layered benzene–HMDS. The compounds
1
100). 14: H NMR (C6D6, 25 °C, 200 MHz) δ: 2.28, 2.32 (s,
6H total, o-Ph-CH3), 2.40 (s, 3H, p-Ph-CH3), 6.90, 7.25 (s,
2H total, m-Ph-H), 0.05, 0.58 (s, 12H total, Si-CH3), 0.85,
1.39, 1.97 (m, 15H total, CH2 and C6H11). 31P{1H} NMR
(C6D6, 25 °C) δ: –25.8 (s). Anal. calcd. for C21H38NPSi2
(%): C 64.40, H 9.78, N 3.58; found: C 64.45, H 9.70, N
3.51. Reduction of 8: EI-MS m/z (%): 576 ([M+]
PhPh[NPN]V=NPh, 20), 485 ([M+] – NPh, 100). Reduction
of 10: EI-MS m/z (%): 702 ([M+] PhMes[NPN]V=NMes, 35),
569 ([M+] – NMes, 100). Reduction of 11: EI-MS m/z (%):
396 ([M+] CyMe[NPN]V=NMe, 8), 367 ([M+] – NMe, 100).
Reduction of 12: EI-MS m/z (%): 390 ([M+]
PhMe[NPN]V=NMe, 12), 361 ([M+] – NMe, 100).
(
PhPh[NPN]VCl)2, 8, (CyMes[NPN]VCl)2, 9, (PhMes[NPN]VCl)2,
10, (CyMe[NPN]VCl)2, 11, and (PhMe[NPN]VCl)2, 12, were
prepared analogously. 7: EI-MS m/z (%): 526 ([M+]
CyPh[NPN]VCl, 100). Anal. calcd. for C48H74Cl2N4P2Si4V2
(%): C 54.69, H 7.08, N 5.31; found: C 54.97, H 7.26, N
5.05. 8: EI-MS m/z (%): 520 ([M+] PhPh[NPN]VCl, 100).
Anal. calcd. for C48H62Cl2N4P2Si4V2 (%): C 55.32, H 6.00,
N 5.38; found: C 55.50, H 5.88, N 5.10. 9: EI-MS m/z (%):
610 ([M+] CyMes[NPN]VCl, 100). Anal. calcd. for
C60H98Cl2N4P2Si4V2 (%): C 58.95, H 8.08, N 4.58; found: C
58.97, H 8.26, N 4.55. 10: EI-MS m/z (%): 604 ([M+]
PhMes[NPN]VCl, 100). Anal. calcd. for C60H86Cl2N4P2Si4V2
(%): C 59.54, H 7.16, N 4.63; found: C 59.37, H 7.00, N
4.39. 11: EI-MS m/z (%): 402 ([M+] CyMe[NPN]VCl, 100).
Anal. calcd. for C28H66Cl2N4P2Si4V2 (%): C 41.73, H 8.25,
N 6.95; found: C 41.90, H 8.02, N 6.85. 12: EI-MS m/z (%):
396 ([M+] PhMe[NPN]VCl, 100). Anal. calcd. for
CyPh[NPN]NbMe3 (15)
In the dark, a solution of 5.601 g (0.0106 mol) of 2 in
150 mL of Et2O maintained at –78 °C was added via can-
nula to a solution of 2.200 g (0.0106 mol) of NbCl2Me3 in
150 mL of Et2O. The solution was stirred at –78 °C for 1 h,
then warmed to –10 °C, producing an orange solution. Ex-
cess Et2O was removed in vacuo, and the resulting solid was
dissolved in toluene and filtered through Celite. Toluene was
© 2003 NRC Canada

1434
Can. J. Chem. Vol. 81, 2003
removed in vacuo and the solid washed with pentane to af-
ford 15 in 72% yield (4.42 g). The solid was highly light
and thermally sensitive and was stored in a darkened
vessel at –37 °C. Preparation of PhPh[NPN]NbMe3, 16, and
CyMes[NPN]NbMe3, 17, was accomplished in the same man-
ner. Reactions of 3, 5, and 6 did not afford the correspond-
ing R′R′′[NPN]NbMe3 species, as decomposition pathways
product. PhMe[NPN]Li2(diox)2, 5, and CyMe[NPN]Li2(diox), 6,
were prepared by an analogous procedure, except the requi-
+
site amine hydrochloride salt MeNH3 Cl– is used to prepare
MeNHLi prior to metathesis. The ligand precursors were
1
7
characterized by 31P{1H}, H, and Li{1H} NMR spectros-
copy. Colorless platelet crystals of 4 were recrystallized
from a saturated hexanes–toluene solution and analyzed by
X-ray crystallography.
1
prevented isolation of pure product. 15: H NMR (C6D6,
25 °C, 500 MHz) δ: 0.00, 0.34 (s, 12H, Si-(CH3)), 0.75–1.42
(ov, m, 15H total, CH2 and C6H11), 1.68 (s, 9H, Nb-(CH3)3),
6.96–7.12 (ov, m, 10H, N-Ph-H). 31P{1H} NMR (C6D6,
The molecular structure of 4 is shown as an ORTEP3 (17)
drawing in Fig. 1. Relevant bond lengths and angles are
listed in Table 3 and crystallographic data is located in Ta-
ble 1. The X-ray data unambiguously show that 4 has a sin-
gle coordinated THF molecule. The bond lengths from the
amido nitrogen donors N1 and N2 to Li2 are approximately
0.2 Å shorter than the corresponding bond lengths from N1
and N2 to Li1, to which the THF molecule is coordinated.
The P1—Li1 bond length is 2.589(4) Å, an average length
for phosphorus—lithium bonds. Some disorder was evident.
The crystalline lattice contained two different conformations
of the ligand within the same unit cell. This manifested as
overlapping opposite chair conformations of the cyclohexyl
ring, as well as overlapping silyl methyl fragments.
1
25 °C) δ: 10.8 (s). 16: H NMR (C6D6, 25 °C, 500 MHz) δ:
–0.08, 0.35 (s, 12H total, Si-(CH3)), 1.20 (m, 4H total, CH2),
1.63 (s, 9H, Nb-(CH3)3), 6.80–7.38, 7.60–7.85 (s, 12H total,
1
N-Ph, P-Ph). 31P{1H} NMR (C6D6, 25 oC) δ: 6.2 (s). 17: H
NMR (C6D6, 25 °C, 500 MHz) δ: 2.29 (s, 9H, Nb-(CH3)3),
2.10, 2.16 (s, 12H total, o-Ph-(CH3)), 2.39 (s, 6H, p-Ph-
CH3), 0.21, 0.29 (s, 12H total, Si-(CH3)), 0.91–2.08 (m, 15H
total, CH2 and C6H11), 6.82, 7.02 (s, 4H total, m-Ph-H).
31P{1H} NMR (C6D6, 25 °C) δ: 11.2 (s).
Attempted hydrogenations of 15–17
Hydrogenations of the three R′R′ ′[NPN]NbMe3 complexes
were attempted using the same methodology as that shown
for 15 below. A 75 mL ethereal solution of 0.500 g
(0.864 mmol) of 15 was prepared in a thick-walled reaction
vessel and immediately cooled to –78 °C. N2 was removed
by successive freeze-pump-thaw cycles. The solution was
then frozen at –196 °C, flushed with H2, sealed, warmed to
r.t., and stirred for 24 h. Solvent was removed in vacuo,
yielding a dark brown solid. Extensive decomposition of 15
complicated analysis. In situ NMR studies (1 atm) con-
firmed that no hydride resonances were observable during
the hydrogenation attempts.
Several interesting trends have been observed in the solu-
tion characterization of the six dilithio [NPN] precursors
used in this report. The three ligand precursors with cyclo-
hexyldonors (2, 4, and 6) are each monosolvent adducts,
compared with 1, 3, and 5, which have two coordinating sol-
vent molecules. In each case, the solvent adduct isolated is
the least soluble product. These same cyclohexylphosphine
derivatives (2, 4, and 6) show no 31P–7Li coupling in solu-
tion (r.t.) that could suggest their solution behaviour is dif-
ferent than that found for 4 in the solid state (Fig. 1). One
possibility is that the phosphine donor does not bind to the
lithium ions in solution. However, a low temperature 31P
7
spectrum (–25 °C) of 2 shows a doublet due to Li coupling
(1JPLi = 17.4 Hz). It is likely that the lithium ions are under-
going intermolecular exchange in solution, fast enough at
room temperature to lose coupling information. However, at
lower temperatures, this process is slow, and coupling be-
Results and discussion
One of the key advantages of the diamidophosphine lig-
ands, referred to as [NPN], is their inherent ability to tune
steric and electronic properties by varying the substituents
on the amido and phosphine donors. In this report,
R′R ′′[NPN] refers to the ligand R′P(CH2SiMe2NR′′)2. The
ligand precursors are isolated as solvent adducts of the
dilithium diamidophosphines both for ease of purification
and to facilitate reaction with metal chlorides. The synthesis
of PhPh[NPN]Li2(THF)2, 1 (2), and CyPh[NPN]Li2(OEt2), 2
(8), were reported previously. This report details the prepara-
tion of four new [NPN] precursors: PhMes[NPN]Li2(diox)2, 3,
CyMes[NPN]Li2(THF), 4, PhMe[NPN]Li2(diox)2, 5, and
CyMe[NPN]Li2(diox), 6. A general synthesis of these deriva-
tives is shown in Scheme 1.
7
tween Li and 31P is observed. In solution, the structure of
the dilithio derivative is therefore likely the same as in the
solid state with phosphine bound to lithium.
PhPh[NPN] and CyPh[NPN] ligand sets have already been
applied to other group five metal-based systems. The complex
PhPh[NPN]TaMe3 reacts with H2 to form (PhPh[NPN]Ta)2(µ-H)4
(1, 2). Under an atmosphere of dinitrogen, this complex
loses 1 equiv of H2 and binds dinitrogen to form
(
PhPh[NPN]Ta)2(µ-H)2(µ-η1:η2-N2). This complex has an
intriguing side-on, end-on coordination mode that promotes
reactivity with E-H species (E = B, Si, Al) (3, 4). Dinitrogen
activation has also been observed by a related niobium
system supported by a [P2N2] ligand set. Upon thermolysis,
the complex ([P2N2]Nb)2(µ-N2) (where [P2N2] = PhP(CH2Si-
Me2NSiMe2CH2)2PPh) decomposes into a bridging nitride
species where one N atom from the activated N2 inserts into
the macrocycle backbone, forming the complex [P2N2]Nb(µ-
N)Nb[PN3] (where [PN3] = PhPMe(CHSiMe2NSiMe2CH2P-
Synthesis of 4 and 5 proceeds by metathesis of MesNHLi
with ClCH2SiMe2Cl at –78 °C to form a silylamine that re-
acts in situ with the desired primary phosphine, RPH2, and
n
4 equiv (with respect to MesNHLi) of BuLi. When phenyl-
phosphine is used in this methodology, the ligand precursor
precipitates as a pale yellow solid, 3, after addition of
2.5 equiv of dioxane to a hexanes solution of the product.
Similarly, when cyclohexylphosphine is used in this method-
ology, white crystals (4) were isolated after addition of
1.5 equiv of tetrahydrofuran to a hexanes solution of the
(Ph)CH2SiMe2NSiMe2N)) (18). As
well, the complexes
RPh[NPN]NbCl(DME),
(
PhPh[NPN]NbCl)2(µ-N2),
and
RPh[NPN]NbCl2 (R = Cy, Ph; DME = dimethoxyethane)
have been synthesized. The complex CyPh[NPN]NbCl2
© 2003 NRC Canada

Shaver et al.
1435
Scheme 1.
Table 3. Selected bond distances (Å) and
angles (°) for CyMes[NPN]Li2(THF), 4.
Fig. 1. ORTEP 3 plot for 4 (50% probability, THF carbons omit-
ted for clarity).
Bond distances (Å)
P(1)—Li(1)
N(1)—Li(1)
N(1)—Li(2)
N(2)—Li(1)
N(2)—Li(2)
O(1)—Li(1)
2.589(4)
2.106(4)
1.943(4)
2.133(4)
1.964(4)
1.923(4)
Bond angles (°)
C(7)-N(1)-Li(1)
C(7)-N(1)-Li(2)
C(22)-N(2)-Li(2)
C(22)-N(2)-Li(1)
Li(2)-N(2)-Li(1)
Li(2)-N(1)-Li(1)
O(1)-Li(1)-P(1)
103.74(15)
123.97(17)
94.27(16)
119.11(17)
73.74(15)
74.77(15)
121.74(16)
try and elemental analysis confirmed the empirical formula.
Recrystallization of 7 from a layered benzene–HMDS solution
at –35 °C gave red block crystals of (CyPh[NPN]VCl)2. X-ray
analysis of the crystals, isolated in 78% yield, showed that
the compound exists as a chloride-bridged dimer. The mo-
lecular structure of 7 is shown as an ORTEP drawing in
Fig. 2. Relevant bond lengths and angles and crystallo-
graphic data are listed in Tables 2 and 4, respectively. The
V(1)—Cl(1) and V(1*)—Cl(1) bond lengths of 2.3632(4)
and 2.4913(4) Å, respectively, are average for a loosely
bound chloride dimer (22); the compound appears as the
monomer in mass spectrometry studies. The vanadium am-
ide bond lengths of 1.935(1) and 1.934(1) Å agree with
other amido vanadium chloride systems (23). While the
V(1)—P(1) bond length of 2.4323(5) Å is not unusual for
vanadium phosphine systems (23), it is short for related va-
nadium chloride dimers (~2.6–2.7 Å) (24–27). The overall
complex is distorted trigonal bipyramidal with chlorine and
phosphine ligands occupying the apical positions.
decomposes
in
solution
to
form
an
imide,
CyPh[NPN]NbCl(=NPh), a cyclized [NP] ligand, CyP(CH2-
SiMe2)2NPh, and a niobium(III) product by an inter-
molecular N—Si bond scission (8). The goal of this work is
to examine how ligand modifications would affect other
group 5 precursors.
Application of these ligand sets to vanadium(III) could af-
ford species of the form [NPN]VCl from which a number of
reactivities can be explored. Behavior of these systems
might mimic that seen for niobium and generate a dinitrogen
complex directly, analogous to
(
PhPh[NPN]NbCl)(µ-N2)
chemistry. Alternatively, if the [NPN]VCl system were
reduced under a dinitrogen atmosphere it would give a
complex isoelectronic with Mo(NtBuAr)3, which coordinates
and splits molecular nitrogen to form the nitride
(NtBuAr)3MoϵN (19, 20). The cleavage of molecular nitro-
gen was also observed when a bisamidoamine vanadium
chloride complex was reduced (21).
Treatment of a cooled solution of VCl3(THF)3 in THF
with an Et2O solution of CyPh[NPN]Li2(OEt2) gave a brick-
red paramagnetic solid, CyPh[NPN]VCl, 7. Mass spectrome-
Similar reactions of ligand precursors 1, 3, 4, 5, and 6
with VCl3(THF)3 produced the vanadium chlorides
R′R ′′[NPN]VCl (R′′ = Ph, Cy; R′′ = Ph, Mes, Me; 8–12).
Characterization by mass spectrometry and elemental analy-
sis confirmed the empirical formula in each case. A general
synthesis is shown in eq. [1].
© 2003 NRC Canada

1436
Can. J. Chem. Vol. 81, 2003
Fig. 2. ORTEP 3 plot for 7 (50% probability; silyl methyl and
R′ and R′′ groups, save the attached carbons, have been omitted
for clarity).
Table 4. Selected bond distances (Å)
and angles (°) for (CyPh[NPN]VCl)2, 7.
Bond distances (Å)
V(1)—Cl(1)
V(1)—Cl(1*)
V(1)—P(1)
V(1)—N(2)
V(1)—N(1)
2.3632(4)
2.4913(4)
2.4323(5)
1.934(1)
1.935(1)
Bond angles (°)
Cl(1)-V(1)-Cl(1*)
Cl(1)-V(1)-N(1)
Cl(1)-V(1)-P(1)
Cl(1)-V(1)-N(2)
N(1)-V(1)-N(2)
P(1)-V(1)-N(1)
P(1)-V(1)-N(2)
84.68(2)
115.77(4)
88.90(2)
127.32(4)
116.17(6)
86.46(4)
86.00(4)
reducing agent, or temperature were unsuccessful. Similarly,
reductions of other vanadium chlorides (8–12) were unpro-
ductive. In several cases, there was no evidence of nitride
formation.
The reduction of (CyMes[NPN]VCl)2, 9, illustrated the
problems associated with reduction of the vanadium chlo-
rides. Upon workup of the reaction of 9 with 2.2 equiv of
KC8, a peak was observed in the mass spectrum at m/z 709,
which corresponds to CyMes[NPN]V=NMes, as supported by
theoretical isotope pattern calculations. The vanadium imide
moiety CyMes[NPN]V=NMes is the likely product of ring-
closing of the [NPN] ligand on itself. Scissioning of one of
the Si—N bonds and one of the N—V bonds in the
([NPN]VCl)2 species during reduction would result in the
vanadium imide species 13 and the elimination of a six-
membered heteroatomic ring 14, as illustrated in eq. [2].
Crystals of 14 formed out of a toluene–hexanes solution, and
[1]
Reductions of 7–12 with potassium graphite (KC8) were
conducted under four conditions: 1 atm N2 and 2.2 equiv of
KC8, 4 atm N2 and 2.2 equiv of KC8, 1 atm N2 and 3.3 equiv
of KC8, 4 atm N2 and 3.3 equiv of KC8. Reduction of 7 at
4 atm of N2 with 2.2 equiv of KC8 at –78 °C gave a strongly
purple colored solution upon warming to r.t. Removal of
KCl followed by recrystallization from a saturated pentane
solution gave a deep purple paste, highly soluble in common
organic solvents. Mass spectral analysis of the product indi-
cated the formation of a CyPh[NPN]VϵN species with a peak
at m/z 505 (C24H37N3Si2PV), but repeated attempts to grow
diffractable crystals failed. A paramagnetic impurity pre-
vented purification and full characterization of the product.
The observation of a mononuclear species by mass spec-
trometry does not, of course, prove its formulation. The spe-
cies could have been generated in the experiment from a
dinuclear bisnitride complex; however, it is unlikely to be
generated from a compound with an intact N2 moiety (18).
Reduction of 7 at 4 atm of N2 with 3.3 equiv of KC8 at
–78 °C gave a deep olive green solution upon warming to r.t.
From the solution, a dark green solid was isolated and char-
acterized by mass spectrometry. A small peak at m/z 1050
(C48H74KN6P2Si4V2), corresponding to K(CyPh[NPN]VϵN)2,
as well as a substantial CyPh[NPN]VϵN peak (m/z 505,
100%), suggested dinitrogen incorporation during the reduc-
tion process. Unfortunately, the reduced products could not
be purified. Attempts to improve the synthesis of the vana-
dium nitride by changing solvent, reaction time, N2 pressure,
1
the structure was verified by H NMR. Peaks corresponding
to imide formation were observed in the mass spectra of
each of the reductions conducted, no matter which substrate
was reduced. As mentioned, decomposition of [NPN] metal
complexes to form imides has complicated a previous study
as well: the imide CyPh[NPN]NbCl(=NPh) was generated
from the decomposition of CyPh[NPN]NbCl2 (8).
[2]
The [NPN] ligand may provide access to attractive Nb(V)
complexes since the Ta(V) complex PhPh[NPN]TaMe3 reacts
with H2 to form (PhPh[NPN]Ta)2(µ-H)4, a precursor to the
dinitrogen complex (PhPh[NPN]Ta)2(µ-H)2(µ-η1:η2-N2) (2).
We examined the formation of R′R ′′[NPN]NbMe3 and its re-
activity with dihydrogen.
Production of CyPh[NPN]NbMe3 (15) was accomplished
via addition of a –78 °C solution of ligand to a cooled solu-
© 2003 NRC Canada

Shaver et al.
1437
tion of NbCl2Me3, freshly sublimed before use. After solvent
removal at –10 °C, the resultant brown solid was extracted
with pentane to generate an orange solid that was character-
(NSERC) in the form of a Research Grant to M.D.F. and a
postgraduate scholarship to M.P.S. and the Killam Founda-
tion for a postgraduate scholarship to M.P.S.
1
ized by H and 31P{1H} NMR and mass spectrometry. The
product was extremely photo- and thermo-sensitive and was
stored in an opaque vessel at –37 °C. Thermal instability
precluded the use of elemental analysis and 13C{1H} NMR
for characterization. A general reaction pathway is shown in
eq. [3].
References
1. M.D. Fryzuk, S.A. Johnson, and S.J. Rettig. J. Am. Chem.
Soc. 120, 11 024 (1998).
2. M.D. Fryzuk, S.A. Johnson, B.O. Patrick, A. Albinati, S.A.
Mason, and T.F. Koetzle. J. Am. Chem. Soc. 123, 3960 (2001).
3. M.D. Fryzuk, B.A. MacKay, S.A. Johnson, and B.O. Patrick.
Angew. Chem. Int. Ed. 41, 3709 (2002).
[3]
4. M.D. Fryzuk, B.A. MacKay, and B.O. Patrick. J. Am. Chem.
Soc. 125, 3234 (2003).
5. A.B. Pangborn, M.A. Giardello, R.H. Grubbs, R.K. Rosen, and
F.J. Timmers. Organometallics, 15, 1518 (1996).
6. D.D. Perrin and W.L.F. Armarego. Purification of laboratory
chemicals. 3rd ed. Butterworth-Heinemann Ltd., Oxford. 1994.
7. M. Baudler and A. Zarkdas. Chem. Ber. 104, 1034 (1965).
8. M.D. Fryzuk, M.P. Shaver, and B.O. Patrick. Inorg. Chim.
Acta. 350, 293 (2003).
Reactions of NbCl2Me3 with ligand precursors 1, 3, 4, 5,
and 6 were not as successful. The products from these
reactions were more thermally unstable and frequently
decomposed prior to spectroscopic analysis. Only
PhPh[NPN]NbMe3 (16) and CyMes[NPN]NbMe3 (17) were sta-
ble enough to characterize spectroscopically.
9. L.E. Manzer. Inorg. Synth. 21, 135 (1982).
Hydrogenation of complexes 15–17 was attempted at 4
atm pressure. Following the addition of H2, the reaction mix-
tures turned black and a precipitate formed. Analysis of the
reaction mixture by NMR showed extensive decomposition.
No hydride resonances were observed at the expected chem-
ical shift for a niobium hydride. Attempts to vary reaction
conditions to isolate the putative polyhydride species, by
changing the concentration of reagents, temperature, expo-
sure to light, pressure of H2, solvent, or time of reaction,
were unsuccessful. It is suspected that the trimethyl starting
materials are too unstable to withstand the hydrogenation
process.
10. D.E. Berbreiter and J.M. Killough. J. Am. Chem. Soc. 100,
2126 (1978).
11. G.W.A. Fowles, D.A. Rice, and J.D. Wilkins. J. Chem. Soc.
Dalton Trans. 9, 961 (1973).
12. Molecular Structure Corp. teXsan for Windows: Crystal struc-
ture analysis package [computer program]. Molecular Struc-
ture Corp., The Woodlands, TX. 1996.
13. A. Altomane, M.C. Burla, G. Cammali, M. Cascarano, C.
Giacovazzo, A. Gagliardi, A.G.G. Moliterni, G. Polidoi, and
A. Spagna. SIR 97: an integrated package of computer pro-
grams for the solution and refinement of crystal structures us-
ing single crystal data [computer program]. CNR-IRMEC,
Universitario, Bari, Italy. 1999.
14. P.T. Beurkens, G. Admiraal, G. Baurskens, W.P. Bosman, R.
de Gelder, R. Israel, and J.M.M. Smits. DIRDIF-94 program
system [computer program]. University of Nijmegen, The
Netherlands. 1994.
15. International Union of Crystallography. International tables for
X-ray crystallography. Kynoch Press, Birmingham, U.K. 1974.
16. International Union of Crystallography. International tables for
crystallography. Kluwer Academic, Boston, MA. 1992.
17. L.J. Farrugia. J. Appl. Cryst. 30, 565 (1997).
18. M.D. Fryzuk, C.M. Kozak, M.R. Bowdridge, B.O. Patrick, and
S.J. Rettig. J. Am. Chem. Soc. 124, 8389 (2002).
19. C.E. LaPlaza and C.C. Cummins. Science (Washington, D.C.),
268, 861 (1995).
Conclusions
While a number of reports have shown the applicability of
Ta(V), Nb(III), and V(III) systems to dinitrogen activation,
in this work we have shown that small changes to the elec-
tronics and sterics of these systems can prevent such reac-
tion pathways. Six variants of the [NPN] diamidophosphine
ligand were used to prepare (R′R ′′[NPN]VCl)2 complexes, to
investigate their ability to coordinate and cleave dinitrogen
upon reduction. Unfortunately, reduction of these complexes
under a variety of conditions was complicated by ligand deg-
radation to form a V(IV) imide. While some nitride-
containing material was formed, it was not isolable. Simi-
larly, the complexes R′R ′′[NPN]NbMe3 were prepared, with
the goal that these highly unstable materials could be hydro-
genated in an analogous manner to the formation of
([NPN]Ta)2(µ-H)2(µ-η1:η2-N2); however, their thermal and
photochemical decomposition occurred prior to hydrogena-
tion. What becomes evident from this work is just how
changes in ligand design or metal centres can lead to quite
different and unpredictable reaction behaviour.
20. C.C. Cummins. Chem. Commun. 1998, 1777 (1998).
21. G.K.B. Clentsmith, V.M.E. Bates, P.B. Hitchcock, and F.G.N.
Cloke. J. Am. Chem. Soc. 121, 10 444 (1999).
22. Cambridge Crystallography Data Centre. Cambridge crystal-
lography database. [cited Nov. 2002].
23. P. Berno, M. Moore, R. Minhas, and S. Gambarotta. Can. J.
Chem. 74, 1930 (1996).
24. J. Nieman and J.H. Teuben. Organometallics, 5, 1149 (1986).
25. F.A. Cotton, J. Lu, and T. Ren. Inorg. Chim. Acta, 215, 47
(1994).
26. F.A. Cotton, S.A. Duraj, L.R. Falvello, and W.J. Roth. Inorg.
Acknowledgments
Chem. 24, 4389 (1985).
27. F.A. Cotton, S.A. Duraj, and W.J. Roth. Inorg. Chem. 23, 4113
(1984).
Funding for this research was provided by the Natural
Sciences and Engineering Research Council of Canada
© 2003 NRC Canada