Communications
diazasilaphosphetidine give this particular diphosphineimine
its unique stability remains to be seen. The potential rewards
for finding stable diphosphineimines are clear, and we are
currently exploring the reaction chemistry of 1 to understand
the properties of these intriguing molecules.
Experimental Section
1: A solution of Me2Si(m-NtBu)2PN(H)Ph[11a] (11.2 mmol) in hexanes
(20 mL) was treated dropwise at 08C with nBuLi (11.3 mmol) in
hexanes (15 mL). The reaction mixture was heated at reflux for 1 h,
cooled to RT, and then treated dropwise with a solution of Ph2PCl
(11.2 mmol) in hexanes. This caused a white precipitate to form. The
mixture was stirred overnight and then filtered to afford a colorless,
clear solution from which colorless, rectangular crystals separated
upon cooling; yield: 3.69 g (64.9%). M.p. 170–1728C; elemental
analysis (%) calcd for C28H39N3P2Si: C 66.24, H 7.74, N 8.28; found: C
Figure 4. Solid-state structure of 4 (the THF molecule and the hydro-
gen atoms have been omitted). Selected bond lengths [ꢀ] and angles
[8]: Ni(1)-P(1) 2.1270(10), Ni(1)-P(2) 2.1151(10), Ni(1)-Br(1)
2.3481(6), Ni(1)-Br(2) 2.3302(6), P(1)-N(1) 1.714(3), P(2)-N(1)
1.714(3), P(1)-N(2) 1.658(3), P(1)-N(3) 1.670(3); P(1)-N(1)-P(2)
97.17(14), P(1)-Ni(1)-P(2) 74.62(4), Br(1)-Ni(1)-Br(2) 98.07(2), N(2)-
P(1)-N(3) 87.67(15).
1
66.01, H 7.84, N 8.11; H NMR (500.1 MHz, CD2Cl2, 218C): d = 7.94
(m, 4H, Ph), 7.37 (m, 6H, Ph), 7.15 (t, J = 7.1 Hz, 2H, Ph), 7.06 (d, J =
8.1 Hz, 2H, Ph), 6.72 (t, J = 7.1 Hz, 1H, Ph), 1.13 (s, 18H, tBu), 0.56 (s,
3H, Me), 0.20 ppm (s, 3H, Me); 31P[1H] NMR (202.5 MHz, CD2Cl2,
218C): d = 2.66 (d, J(P,P) = 182.4 Hz), À30.82 ppm (d, J(P,P) =
182.8 Hz).
À
À
bonds are almost equidistant (the Ni P1 and Ni P2 bond
lengths are 2.1270(10) and 2.1151(10) ꢀ) and are slightly
shorter than those in the related cis-[NiBr2{Ar2PN(Me)PAr2}]
2: Compound 1 (0.235 g, 0.486 mmol) and elemental sulfur
(0.016 g, 0.50 mmol) were dissolved in THF (20 mL), and the mixture
was heated at reflux for 3 h. Following removal of THF, the ensuing
white powder was dissolved in toluene to afford colorless crystals
upon cooling; yield: 0.213 g (81.2%). M.p. 202–2068C; elemental
analysis (%) calcd for C28H39N3P2SSi: C 62.31, H 7.28, N 7.79; found:
C 61.94, H 7.35, N 7.72; 1H NMR (500.1 MHz, C6D6, 218C): d = 9.178
(dd, J = 12.4, 8.0 Hz, 4H, Ph), 7.43 (d, J = 8.3 Hz, 2H, Ph), 7.34 (t, J =
7.5 Hz, 2H, Ph), 7.11 (td, J = 8.0, 3.2 Hz, 2H, Ph), 6.99 (t, J = 7.4 Hz,
2H, Ph), 6.95 (t, J = 7.3 Hz, 1H, Ph), 0.94 (s, 18H, tBu), 0.83 (s, 3H,
Me), 0.38 ppm (s, 3H, Me); 31P[1H] NMR (202.5 MHz, CD2Cl2,
218C): d = 10.94 (d, J(P,P) = 238.8 Hz), À8.40 ppm (d, J(P,P) =
238.8 Hz).
À
(Ar= o-tBuC6H4), which features symmetrical Ni P bonds
(2.1611(6) ꢀ).[4a] This latter nickel complex, with its bulky
bis(diphenylphosphine)amine ligand, was shown to be a very
good polyolefin catalyst, and the more electron-rich 4 may
exhibit a similar reactivity.
The 31P NMR spectrum gave no indication that the
isomerization of 1 had occurred as it showed two doublets
at d = 55.8 and 43.9 ppm for the P1 and P2 atoms, respectively,
with larger coupling constants (J(P,P) = 137 Hz) than those
seen in 3. Thus, 31P NMR spectroscopy, which is usually
considered a diagnostic tool, was neither reliable for predict-
ing the connectivity in these phosphorus–nitrogen compounds
nor did it show a correlation between the coupling constants
3: Compound 1 (0.250 g, 0.500 mmol) and cis-[Mo(CO)4(piper-
idine)2] (0.195 g, 0.500 mmol) were dissolved in THF (50 mL) and the
mixture was heated at reflux for 3 h. A yellow powder precipitated
upon cooling, which was recrystallized from toluene to afford light-
yellow crystals; yield: 0.263 g (65.1%). M.p. 1768C (decomp);
elemental analysis (%) calcd for C39H47MoN3O4P2Si: C 57.99, H
À
and the P P bond lengths. These results are in keeping with
earlier studies that showed large and variable two-bond P P
coupling constants in diphosphinoamines that ranged from 15
À
1
to 665 Hz.[12]
5.86, N 5.20; found: C 58.04, H 5.98, N 5.34; H NMR (500.1 MHz,
CD2Cl2, 218C): d = 8.09 (m, 4H, Ph), 7.52 (s, 6H, Ph), 7.39 (t, J =
7.6 Hz, 2H, Ph), 7.22 (m, 4H, Ph), 6.91 (t, J = 7.3 Hz, 1H, Ph), 2.34 (s,
3H, Me), 0.84 (s, 21H, tBu, Me), 0.66 ppm (s, 3H, Me); 31P[1H] NMR
(202.5 MHz, CD2Cl2, 218C): d = 49.93 (d, J(P,P) = 69.5 Hz),
44.97 ppm (d, J(P,P) = 69.0 Hz); 13C[1H] NMR (CD2Cl2),
121.5 MHz, 218C): d = 221.4 (s, COeq), 218.3 (d, J = 32.9 Hz, COeq),
210.2 (d, J = 8.5 Hz, COax), 132.8 (d, J = 26.4 Hz, Ph), 131.9 (s, Ph),
129.6 (s, Ph), 128.9 (s, Ph), 126.1 (s, Ph), 122.4 (s, Ph), 32.4 (s, tBu), 5.0
(s, Me), 4.4 ppm (s, Me); IR (nujol): n˜ = 2010.4 (s), 1889.9 (vs),
1850.4 cmÀ1 (vs).
4: A suspension of cis-[NiBr2(diglyme)] (0.321 g, 0.912 mmol) in
THF (10 mL) was treated dropwise with a solution of 1 (0.398 g,
0.816 mmol) in THF (25 mL) at RT. The ensuing orange mixture
slowly turned red and was stirred overnight. The dark red–brown
solution was filtered through a medium-porosity frit and stored in a
freezer. Several crops of red–brown, needle-shaped crystals formed
upon cooling; yield: 0.369 g (59.0%). M.p. 273–2828C; elemental
analysis (%) calcd for C32H47 Br2N3NiOP2Si: C 48.15, H 5.93, N 5.26;
found: C 47.91, H 5.98, N 5.23; 1H NMR (500.1 MHz, CD2Cl2, 218C):
d = 8.11 (dd, J = 12.4, 7.9 Hz, 4H, Ph), 7.77 (s, 2H, Ph), 7.49 (s, 4H,
Ph), 7.23 (s, 4H, Ph), 7.01 (s, 3H, Ph), 3.58 (s, 4H, THF), 1.85 (s, 4H,
THF) 1.50 (s, 18H, tBu) 0.78 (s, 3H, Me), 0.401 ppm (s, 3H, Me);
Given the catalytic activity of nickel and the inertness of
the [Mo(CO)4] moiety, it is not surprising that 1 isomerized in
the presence of NiBr2 even at room temperature, although it
was unaffected by [Mo(CO)4] when they were heated
together to reflux in THF. However, it is surprising that the
soft {Mo(CO)4} fragment is coordinated by the hard nitrogen
atom, although the harder NiBr2 moiety is coordinated by the
soft phosphorus atoms. This disparity further underlines the
kinetic control demonstrated in Equations (4) and (5). Our
results also suggest that attempts to stabilize diphosphine-
imines with certain Group 10 metal compounds, as had been
done previously, may cause isomerization instead.[2a] The ease
with which 1 isomerizes in the presence of NiBr2 clearly rules
out its use as a diphosphineimine ligand in Group 10 metal
catalysis.
In summary, the syntheses and solid-state structures of a
kinetically stable diphosphineimine, its molybdenum com-
plex, and its sulfide analogue have been reported. Whether
the electronic properties or the cyclic structure of the
3274
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2005, 44, 3271 –3275