Synthesis, spectroscopic characterisation and transmetalation of lithium and potassium diaminophosphanide-boranes

Article abstract of DOI:10.1039/c7dt04110a

A secondary diaminophosphane-borane (Et₂N)₂PH(BH₃) was prepared from a chlorophosphane precursor and LiBH₄ and metalated by reaction with anion bases (n-BuLi, KN(SiMe₃)₂) to yield the corresponding metal diaminophosphanide-boranes [(Et₂N)₂P(BH₃)]M (M = Li, K). Multinuclear NMR studies permitted the first spectroscopic characterisation of the metalation products and revealed the presence of monomeric (for M = Li) contact ion pairs in solution. NMR spectroscopic evidence that the ions in each pair interact via Li?P- rather than Li?H₃B-interactions as had been inferred for a Ph-substituted analogue was confirmed by DFT studies, which revealed also that the borane coordination plays a decisive role in boosting the PH-acidity of the original secondary diaminophosphane precursor. Transmetalation of the potassium and lithium diaminophosphanide-boranes with Cu(i) and Zn(ii) chlorides afforded the first functional transition metal complexes of a P-heteroatom-functionalised phosphanide-borane ligand. Both products were fully characterised. Thermolysis of the Cu-complex induced a reaction which involved transfer of an NHC ligand from the metal to the phosphorus atom and yielded a phosphaalkene NHCPH (NHC = N-heterocyclic carbene) as the major phosphorus-containing product.

Full text of DOI:10.1039/c7dt04110a

Dalton
Transactions
View Article Online
View Journal
PAPER
Synthesis, spectroscopic characterisation and
transmetalation of lithium and potassium
Cite this: DOI: 10.1039/c7dt04110a
diaminophosphanide-boranes†‡
a
M. Blum,a J. Kappler,a S. H. Schlindwein,a M. Nieger b and D. Gudat
*
A secondary diaminophosphane-borane (Et2N)2PH(BH3) was prepared from a chlorophosphane precursor
and LiBH4 and metalated by reaction with anion bases (n-BuLi, KN(SiMe3)2) to yield the corresponding
metal diaminophosphanide-boranes [(Et2N)2P(BH3)]M (M = Li, K). Multinuclear NMR studies permitted the
first spectroscopic characterisation of the metalation products and revealed the presence of monomeric
(for M = Li) contact ion pairs in solution. NMR spectroscopic evidence that the ions in each pair interact
via Li⋯P- rather than Li⋯H3B-interactions as had been inferred for a Ph-substituted analogue was
confirmed by DFT studies, which revealed also that the borane coordination plays a decisive role in boost-
ing the PH-acidity of the original secondary diaminophosphane precursor. Transmetalation of the potass-
ium and lithium diaminophosphanide-boranes with Cu(I) and Zn(II) chlorides afforded the first functional
transition metal complexes of a P-heteroatom-functionalised phosphanide-borane ligand. Both products
were fully characterised. Thermolysis of the Cu-complex induced a reaction which involved transfer of an
NHC ligand from the metal to the phosphorus atom and yielded a phosphaalkene NHCvPH (NHC =
N-heterocyclic carbene) as the major phosphorus-containing product.
Received 31st October 2017,
Accepted 27th November 2017
DOI: 10.1039/c7dt04110a
rsc.li/dalton
ents permit further post-functionalization after the initial
metathesis step.
Introduction
Alkaline metal phosphanides of general composition R2PM (M
In contrast to these well-established species, phosphanides
being mostly one of Li, Na, K) have since their discovery in the with electronegative substituents, in particular amino groups,
late 1950s1–3 become indispensable as synthetic tool for the are very scarce. The only reference to the preparation of such
formation of new phosphorus-element bonds via coupling of compounds is a patent which claims the synthesis of mixtures
the nucleophilic R2P-fragment with suitable electrophiles in of R2NPLi2 and (R2N)2PLi (R = Me, Et) from the reaction of the
metathesis reactions. The substituents R in the phosphanide appropriate amides with P4. However, no characterisation data
unit are generally alkyls or aryls, but the phosphorus atom are reported, and pure products do not seem to have been iso-
may also carry hydride4 or trialkylsilyl-moieties (mostly lated.7 Transition metal complexes [{(R2N)2P}Fe(CO)4]K 8 (I, R =
Me3Si),5,6 respectively. The resulting primary or secondary Et, Ph) and [{Me2N(Ph3C)P}W(CO)5]K 9 (II), which can formally
phosphanides (H2PM or R(H)PM) as well as silyl phosphanides be described as species containing anionic mono- or diamino-
(especially (Me3Si)2PM) are useful synthons for the assembly of phosphanide ligands, were produced by deprotonation of
complex molecular architectures since the ancillary substitu- appropriate PH-substituted precursor complexes and spectro-
scopically characterised. Even if it cannot be ruled out that the
metal atom in I, II participates to some extent in the stabilis-
ation of the negative charge, both species were shown to react
aInstitut für Anorganische Chemie, University of Stuttgart, Pfaffenwaldring 55,
as P-centred nucleophiles.8,9 Last, but not least, the group of
70550 Stuttgart, Germany. E-mail: gudat@iac.uni-stuttgart.de
bDepartment of Chemistry, University of Helsinki, P.O Box 55, 00014 Helsinki,
Knochel reported10 that reduction of a chlorophosphane-
borane 1 with lithium at −78 °C and subsequent addition of
Finland
†Dedicated to Prof. Dr P. P. Power on the occasion of his 65th birthday.
‡Electronic supplementary information (ESI) available: Representation of 1H,
31P, 11B NMR spectra of 2, 5, 6, 8, 9, 1H, 7Li HOESY and 1H DOSY NMR spectra
for 2, 1H NOESY NMR spectrum for 6; computed energies and atomic coordi-
nates for (Et2N)2P(BH3)Li(OEt2)n (n = 0–3). CCDC 1582551 and 1582554. For ESI
and crystallographic data in CIF or other electronic format see DOI: 10.1039/
c7dt04110a
electrophiles afforded phosphane-boranes 3 (Scheme 1). The
reaction was claimed to proceed via a phosphanide-borane
intermediate 2 which was, however, neither positively identi-
fied nor further characterised. The process provides nonethe-
less an attractive entry to a new synthesis of substituted
diaminophosphanes (R2N)2PR′ which augments the known
This journal is © The Royal Society of Chemistry 2017
Dalton Trans.

View Article Online
Paper
Dalton Transactions
and some solvent was purified via flash chromatography over
silica (deactivated with 1% NEt3) using PE/EtOAc (3 : 2) as
eluent. After removing the volatiles, the product was obtained
as light yellow, highly viscous oil (1.17 g, 6.20 mmol, yield
82%).
Scheme 1 Reported synthesis10 of diaminophosphane-boranes 3 via
an intermediate lithium phosphanide-borane 2 (R = alkyl, allyl, benzyl,
aryl; X = halide or mesylate).
1
3
1H NMR (C6D6): δ = 6.26 (dq, JPH = 433 Hz, JHH = 4.3 Hz,
1
1 H, PH), 3.0–2.6 (m, NCH2, 8 H), 1.50 (broad q, JBH = 93 Hz,
3
3 H, BH3), 0.80 (t, JHH = 7.1 Hz, 12 H, CH3). 13C{1H} NMR
2
3
(C6D6): δ = 41.4 (d, JPC = 3 Hz, NCH2), 13.9 (d, JPC = 2 Hz,
route involving substitution of an electrophilic phosphorus-
containing precursor by suitable organometallic nucleophiles.
The target diaminophosphanes are amenable to further post-
functionalisation by nucleophiles rather than electrophiles.
This makes them valuable synthetic building blocks whose
reactivity complements that of secondary phosphanes or silyl
phosphanes, respectively.
Our interest in the chemistry and applications of neutral
metal complexes containing (R2N)2P-moieties11,12 prompted us
to explore the use of diaminophosphanide-boranes like 2 as
reagents for the introduction of nucleophilic aminophospha-
nyl fragments into transition metal complexes in greater
detail. Here, we report on a new access route to 2, its first spec-
troscopic characterisation, and on transmetalation reactions
which give rise to isolable transition metal diaminophospha-
nide-borane complexes.
CH3). 31P NMR (C6D6): δ = 80.3 (q, JPB = 72 Hz). 11B NMR
(C6D6): δ = −39.1 (dq, JPB = 73 Hz, JBH = 93 Hz). C8H24BN2P
(190.08): calcd C 50.55 H 12.73 N 14.74, found C 50.75 H 12.95
N 14.51. MS: m/z = 190.1771 (C8H24BN2P+).
1
1
1
Metalation of bisdiethylaminophosphane-borane
(a) With n-butyl lithium. Phosphane-borane 5 (40 mg,
0.21 mmol) was dissolved in Et2O (10 ml) and a solution of
n-BuLi (0.10 ml of a 2.5 M solution in hexane, 0.25 mmol) was
added dropwise. The mixture was stirred for 1 h. For NMR
spectroscopic characterisation, volatiles were removed under
reduced pressure and the residue dissolved in 0.5 mL THF-d8.
2
3
1H NMR (THF-d8): δ = 3.20 (ddq, JHH = 13.4 Hz, JPH = 5.7
Hz, 3JHH = 7.0 Hz, 4 H, NCH2), 3.06 (ddq, 2JHH = 13.4 Hz, 3JPH
=
3
3
6.8 Hz, JHH = 7.1 Hz, 4 H, NCH2), 0.97 (t, JHH = 7 Hz, 12 H,
CH3), 0.34 (qd, JBH = 85 Hz, JPH = 10 Hz, 3 H, BH3). 13C{1H}
1
2
3
NMR (C6D6): δ = 45.3 (broad, NCH2), 14.0 (d, JPC = 4.0 Hz,
CH3). 31P NMR (THF-d8): δ = 128.0 (q, JPB = 56 Hz); 11B NMR
1
1
1
7
(THF-d8): δ = −34.6 (dq, JPB = 56 Hz, JBH = 85 Hz), Li NMR
(THF-d8): δ = −0.27 (s).
Experimental
Materials and methods
(b) With KN(SiMe3)2. Bis(diethylamino)phosphane-borane
(5, 23 mg, 0.12 mmol) was added to a solution of KN(SiMe3)2
(960 mg, 0.48 mmol) in Et2O (5 ml). The mixture was stirred
for 1 h. Volatiles were evaporated under reduced pressure and
the residue taken up in THF-d8. NMR studies revealed the
presence of 6 beside unreacted KN(SiMe3)2. Alternatively,
dibenzo-18-crown-6 (51 mg, 0.14 mmol) was added to the orig-
inal reaction mixture and the mixture subjected to the same
work-up procedure as described above.
All reactions were carried out under an atmosphere of inert
Argon and in flame-dried glassware. Solvents were dried by
standard procedures. NMR spectra were recorded on Bruker
Avance 250 (1H/13C/31P/11B 250.0/62.9/101.2/80.2 MHz) or
Avance 400 (1H/13C/31P/11B 400.1/100.5/161.9/128.3 MHz)
instruments. Chemical shifts were referenced to ext. TMS (1H,
13C), 85% H3PO4 (Ξ = 40.480747 MHz, 31P), BF3·Et2O (Ξ =
32.083974 MHz, 11B). Elemental analyses were done with an
Elementar Micro Cube elemental analyzer. Mass spectra were
obtained with a Bruker Daltonics Mikrotof-Q-Mass spectro-
meter. (Et2N)2PCl13 and (IDipp)CuCl14 (IDipp = 1,3-bis-(2,6-
1
2
2
6: H NMR (THF-d8): δ = 3.21 (ddt, JHH = −13.4 Hz, JPH
=
3
2
5.7 Hz, JHH = 7.1 Hz, 4 H, NCH2), 3.06 (ddt, JHH = −13.4 Hz,
2JPH = 6.4 Hz, 3JHH = 7.0 Hz, 4 H, NCH2), 0.98 (t, 3JHH = 7.1 Hz,
12 H, CH3), 0.48 (dq, 2JPH = 10 Hz, 1JBH = 86 Hz, 3 H, BH3). 31
P
diisopropylphenyl)-imidazolylidene)
were
prepared
as
NMR (THF-d8): δ = 130.7 (broad q, JPB = 52 Hz); 11B NMR
1
reported. ZnCl2(tmeda) was obtained from Sigma Aldrich and
used as received.
(THF-d8): δ = −32.3 (dq, 1JPB = 53 Hz, 1JBH = 85 Hz).
1
6(Dibenzo-18-crown-6): H NMR (THF-d8): δ = 6.95 (m, 4 H,
Synthetic procedures
C6H4), 6.88 (m, 4 H, C6H4), 4.24–4.14 (m, 16 H, OCH2), 3.26
Bisdiethylaminophosphane-borane (5). (Et2N)2PCl (1.60 g, (ddt, 2JHH = −13.3 Hz, 2JPH = 5.7 Hz, 3JHH = 7.0 Hz, 4 H, NCH2),
2
2
3
7.60 mmol) was dissolved in anhydrous Et2O (85 ml). LiBH4 3.11 (ddt, JHH = −13.3 Hz, JPH = 6.6 Hz, JHH = 7.0 Hz, 4 H,
3
2
(2.3 ml of a 4 M solution in THF, 9.2 mmol) was added drop- NCH2), 0.98 (t, JHH = 7.0 Hz, 12 H, CH3), 0.76 (dq, JPH = 10
wise and the reaction mixture stirred for 2 h at room tempera- Hz, JBH = 86 Hz, 3 H, BH3). 31P NMR (THF-d8): δ = 135.1 (q,
1
ture. To complete the reduction, LiALH4 (0.25 ml of a 1 M solu- 1JPB = 52 Hz).
tion in THF, 0.25 mmol) was added and the reaction mixture
{Bis(diethylamino)phosphane-borane}-{1,3-bis(2,6-diisopropyl-
was stirred at room temperature for 24 h. The further work-up phenyl)-imidazol-2-ylidene}-copper(I) (8). n-BuLi (0.31 ml of a
was carried out in air. The reaction mixture was washed care- 2.5 M solution in hexane, 0.79 mmol) was added dropwise to
fully with demineralized water (4 × 100 mL). The organic layer a solution of 5 (0.15 g, 0.79 mmol) in Et2O (10 ml). The solu-
was dried over MgSO4, filtered, and concentrated under tion was stirred overnight. (IDipp)CuCl (384 mg, 0.79 mmol)
reduced pressure. The residue containing the crude product was added and stirring continued for 4 h. Volatiles were
Dalton Trans.
This journal is © The Royal Society of Chemistry 2017

View Article Online
Dalton Transactions
Paper
then removed in vacuum. The residue was treated with isotropically, and all other hydrogen atoms with a riding
hexane. The resulting suspension was filtered over dried silica model. Complex 8 co-crystallizes with a solvent molecule
and the remaining colourless solid extracted with Et2O (Et2O) part of which is disordered. The disordered atoms were
(10 ml). Storing the solution at 8 °C afforded 8 as colourless refined isotropically and with restraints. CCDC 1582554 (8)
crystals (yield 100 mg, 20%).
1H NMR (C6D6): δ = 7.18 (m, 2 H, C6H3), 7.05 (m, 4 H, data for this paper.
C6H3), 6.25 (s, 2 H, NCH), 3.23–3.00 (m, N(CH2)2, 8H), 2.54 8: colourless prisms, C35H59BCuN4P·C4H10O, M = 715.30
(sept, JHH = 6.9 Hz, 4 H, CH), 1.43 (d, JHH = 6.9 Hz, 12 H, g mol−1, crystal size 0.63 × 0.40 × 0.38 mm, monoclinic, space
and 1582551 (9)‡ contain the supplementary crystallographic
3
3
1
3
CCH3), 1.4 (br q, JBH ≈ 95 Hz, 3 H, BH3), 1.05 (d, JHH = 6.9 group P21/c (no. 14), a = 12.4854(9) Å, b = 16.7096(12) Å, c =
Hz, 12 H, CCH3), 0.95 (t, JHH = 7.0 Hz, 12 H, CH3). 13C{1H} 20.7813(17) Å, β = 90.058(7)°, V = 4335.5(6) Å3, Z = 4, ρ(calcd) =
3
NMR (C6D6): δ = 130.7 (p-C6H3), 123.7 (m-C6H3), 122.4 (NCH), 1.096 Mg m3, F(000) = 1552, θmax = 30.5°, μ = 0.572 mm−1
,
44.8 (NCH2), 28.7 (CHMe2), 24.8 (CH3), 23.7 (CH3), 15.1 (CH3). max./min. transmission 0.715/0.811, 53 570 reflections
31P NMR (C6D6): δ = 123.8 (broad). 11B{1H} NMR (C6D6): δ = measured, 13 161 unique reflections [Rint = 0.048] for structure
1
−33.6 (broad d, JPB = 68 Hz). 11B NMR (C6D6): δ = −33.6 solution and refinement with 430 parameters and 30
(broad dq, 1JPB = 68 Hz, 1JBH ≈ 95 Hz). C35H59BCuN4P (641.21): restraints, R1 (I > 2σ(I)) = 0.042, wR2 = 0.101, GoF = 1.008,
calcd C 65.56 H 9.27 N 8.74, found C 65.57 H 9.35 N 8.56.
(Bis(diethylamino)phosphanido-borane)-chlorido-(N,N′-tetra-
largest diff. peak and hole: 0.679 and −0.504 e Å−3
.
9: colourless blocks, C14H39BClN4PZn, M = 406.09 g mol−1
,
methylethylenediamine)-zinc(II) (9). A solution of 5 (389 mg, crystal size 0.18 × 0.08 × 0.06 mm, monoclinic, space group
2.00 mmol) in Et2O (10 ml) was added drop wise to a P21/c (no. 14), a = 7.7586(8) Å, b = 29.766(3) Å, c = 9.2778(8) Å,
stirred solution of KN(SiMe3)2 (398 mg, 2.00 mmol) in Et2O β = 96.393(4)°, V = 2129.3(4) Å3, Z = 4, ρ(calcd) = 1.267 Mg m3,
(10 ml). After 15 min, Zn(TMEDA)Cl2 (504 mg, 2.00 mmol) was F(000) = 872, θmax = 26.5°, μ = 1.356 mm−1, max./min. trans-
added, and the stirring was continued overnight. The volatiles mission 0.927/0.796, 16 133 reflections measured, 4361 unique
were then removed under reduced pressure. The residue was reflections [Rint = 0.054] for structure solution and refinement
treated with hexane (10 ml) and the resulting suspension fil- with 208 parameters and 9 restraints, R1 (I > 2σ(I)) = 0.042,
tered over dried silica. The remaining colourless solid was wR2 = 0.091, GoF = 1.002, largest diff. peak and hole: 0.924 and
then extracted with Et2O (10 ml). Storing the ether solution at −0.398 e Å−3
.
8 °C afforded 9 as colourless crystals (yield 39%, 320 mg).
2
1H NMR (C6D6): δ = 3.49 (m, JHH = −15.1 Hz, 2JPH = 1.5 Hz,
Computational studies
2
3JHH = 7.2 Hz, 4 H, NCH2), 3.44 (m, 2JHH = −15.1 Hz, JPH = 2.3
RI-DFT16 calculations were carried with the TURBOMOLE17
program suite. Energy optimisation of the molecular structures
of (Et2N)2P(BH3)Li(OEt2)n (Li⋯P-bound isomer, n = 0–3) and
(Et2N)2P–BH3⋯Li(OEt2)n (Li⋯H3B-bound isomer, n = 0, 2) was
carried out using the BP86 functional18 with a def2-tzvp basis
set,19 Grimme’s D3BJ formalism20 to include dispersion
effects, and the conductor-like screening model21 (COSMO)
with the solvent parameters for Et2O to simulate solvation
effects. The final molecular structures were identified as local
minima on the energy hypersurface by harmonic vibrational
frequency calculations. Natural population analysis22 was per-
formed with the NBO module implemented in the software.
3
Hz, JHH = 7.0 Hz, 4 H, NCH2), 2.34 (s, 6 H, tmeda-CH3), 2.04
(s, 6 H, tmeda-CH3), 1.84 (m, 2 H, tmeda-CH2), 1.63 (m, 2 H,
1
tmeda-CH2), 1.47 (broad q, JBH = 93 Hz, 3 H, BH3), 1.13 (t,
3JHH = 7.1 Hz, 12 H, CH3). 13C{1H} NMR (C6D6): δ = 56.4 (s,
tmeda-CH2), 47.9 (s, tmeda-CH3), 47.5 (s, tmeda-CH3), 43.8 (s,
3
NCH2), 14.8 (d, JPC = 3.6 Hz, CH3). 31P NMR (C6D6): δ = 94.2
(q, 1JPB = 74 Hz); 11B NMR (C6D6): δ = −33.7 (d, 1JPB = 70 Hz).
Thermolysis of 8. A solution of 8 (12 mg, 18 μmol) in
toluene-d8 (2 mL) was heated to 95 °C for 3 h. After cooling to
room temperature, 31P, 31P{1H}, and 1H,31P HMQC NMR
spectra were recorded. The identification of 10 as the main
product (80% by spectral integration) was derived from a com-
parison of the observed spectral data with those of an auth-
entic sample. No work-up was attempted.
Results and discussion
31P{1H} NMR (toluene-d8): δ = −5.7 (s), −133.6 (s); 31P NMR
1
1
(toluene-d8): δ = −5.7 (d, JPH = 643 Hz), −133.6 (d, JPH = 165 Metal alkyl and aryl phosphanides are usually accessed via one
Hz; 10).
of two generally applicable routes, i.e. reductive bond cleavage
in a suitable precursor (frequently a halogeno-phosphane), or
metalation of an appropriate secondary phosphane with a
Crystallography
X-ray diffraction studies were carried out using a Bruker Kappa strong base, respectively. Both strategies fail for diamino-phos-
Apex II diffractometer equipped with a Duo CCD-detector and phanides (R2N)2PM. Reduction of diamino-halogeno-phos-
a KRYO-FLEX cooling device with Mo-Kα radiation (λ = phane precursors affords diphosphanes which withstand
0.71073 Å) at T = 130(2) K. The structures were solved by direct attempts toward further reductive bond cleavage,23 and Cowley
methods (SHELXS-9715) and refined with a full-matrix least- et al. established that sterically protected secondary amino-
squares scheme on F2 (SHELXL-2014 and SHELXL-9715). Semi- phosphanes resist all attempts to deprotonation at phos-
empirical absorption corrections were applied. Non-hydrogen phorus.24 We found that the same is also true for the sterically
atoms were refined anisotropically, hydrogen atoms at boron less shielded phosphane (Et2N)2PH (4). Realising that the work
This journal is © The Royal Society of Chemistry 2017
Dalton Trans.

View Article Online
Paper
Dalton Transactions
borane 2 but impeded concise interpretation of the spectral
data. We found, however, that well resolved spectra without
interference from the effects of chemical exchange could be
obtained when the metalation of 5 was carried out with a large
excess of n-BuLi (the excess reagent acting in this case as
efficient scavenger for adventitious H2O). Careful analysis of
multinuclear (1H, 11B, 13C, 31P) NMR spectra recorded from
such solutions allowed to establish unambiguously that the
P-bound hydrogen was absent but the remaining (Et2N)2P
(BH3) fragment was still intact. A 1H,7Li-HOESY NMR spec-
trum§ showed correlations between a strong 7Li NMR signal
(δ = −0.27) and the resonances of NCH2CH3- and BH3-protons
of the (Et2N)2P(BH3)-unit as well as residual protons in the
OCD2-moiety of the deuterated solvent (THF-d8). Based on
Scheme 2 Preparation and metalation of diaminophosphane-borane 5.
by Knochel et al.10 implies that borane coordination to a these findings, we assign the metalation product 2 the struc-
diaminochlorophosphane boosts its liability to undergo reduc- ture of a tight contact ion pair of composition [{(Et2N)2P(BH3)}
tive PCl-bond cleavage under formation of phosphanide- {Li(THF)n}]. A 1H DOSY spectrum of a mixture of 2 and 5
borane 2 rather than a diphosphane, we wondered if borane recorded at −50 °C where the chemical exchange was found to
coordination can also enhance the acidity of a secondary be negligible disclosed a ratio of diffusion coefficients of D(2)/
aminophosphane towards strong bases.
D(5) = 0.91. This value implies that both species possess very
As a starting material for this study we prepared phos- similar hydrodynamic radii and exhibit thus the same degree
phane-borane 5, which had previously only been obtained in a of aggregation. Since we have no evidence for the formation of
mixture with other compounds,25 by borohydride reaction of any higher aggregates in case of 5, we therefore assign also to
(Et2N)2PCl (Scheme 2). Reaction monitoring by 31P NMR spec- 2 a monomeric structure.
troscopy revealed that the starting material is initially con-
A similar reaction as with n-BuLi was observed when 5 was
verted into a mixture of the target compound 5 and the chloro- treated with KN(SiMe3)2. A 31P NMR spectroscopic survey
phosphane-borane 1. The further reaction of 1 and LiBH4 revealed clean formation of a single product which showed a
under Cl/H-exchange is very slow and sluggish, but rapid and comparable sensitivity towards hydrolysis as 2 but was also
quantitative (by 31P NMR) conversion into 5 was accomplished readily spectroscopically characterised in solution in the pres-
after addition of an appropriate amount of LiAlH4 to the reac- ence of excess amide. The NMR data revealed as in 2 the pres-
tion mixture. The product was finally isolated after aqueous ence of an intact (Et2N)2P(BH3) fragment and the absence of a
work-up as analytically pure, colourless oil and its identity P-bound hydrogen atom, and we assign the product thus as
established by multinuclear NMR data (see Experimental). The potassium phosphanide-borane 6 (Scheme 2). The fact that the
presence of the P-bound hydrogen atom was readily estab- observed NMR chemical shifts are similar but not identical to
lished by the observation of appropriate 1H and 31P NMR those of 2 (e.g. δ31P 130.7 (6) vs. 128.0 (2) and δ11B −34.6 (6) vs.
signals which showed highly characteristic doublet splittings −32.2 (2) in THF at 30 °C) is attributable to the influence of
arising from 1H,31P spin coupling (1JPH = 433 Hz). The product the different cations and suggests that 6 should likewise be for-
can be handled for short times under air and is thermally mulated as a contact ion pair. In line with this hypothesis is
stable in refluxing benzene but decomposes to an intractable the observation of a further change in the 31P NMR chemical
yellow solid in refluxing toluene.
shift (Δδ31P +4.4 ppm) upon addition of an equimolar amount
Reaction of 5 with an equimolar amount of n-butyl lithium (with respect to K+) of crown ether (dibenzo-18-crown-6).
in ether solution afforded a new product which gave rise to a Interestingly, the 1H NOESY spectrum of this solution revealed
31P NMR signal that had been shifted to higher frequency and cross peaks which connect the signals of the OCH2-units in
1
showed no splitting by JPH. The newly formed species proved the crown ether with those of the NCH2- and BH3-moieties.
thermally stable but extremely sensitive to traces of moisture This finding indicates that binding of the cation to the crown
which induced hydrolysis under re-formation of 5. As this sen- ether did not totally disrupt the interaction with the phospha-
sitivity thwarted as yet all attempts to purify the product by nide fragment so that the resulting complex exhibits still some
crystallisation, we decided to establish its identity by spectro- ion pair character. A similar contact between a crown ether
scopic analysis of samples obtained after evaporation of vola- decorated lithium cation and a phosphanide anion had pre-
tiles and dissolution of the residue in a suitable solvent.
viously been observed in a solid phosphinidenoid metal-penta-
The observation that NMR spectra of samples containing a carbonyl complex.26 Apart from the sensitivity towards hydro-
mixture of the newly formed species and 5 (arising from lysis, solutions of 6 containing dibenzo-18-crown-6 were ther-
partial hydrolysis or incomplete conversion of the starting
material) displayed signal broadenings attributable to
dynamic exchange between both components was interpreted
§The spectrum was recorded at −90 °C. At room temperature, no correlation
as a first proof for the presence of litthiated phosphanide- signals at all were observable, presumably due to unfavourable correlation times.
Dalton Trans.
This journal is © The Royal Society of Chemistry 2017

View Article Online
Dalton Transactions
Paper
mally instable and decayed within hours under formation of (3.791 Å) is in the order of the sum of van der Waals radii.
appreciable amounts of 5 along with products arising from Similar changes had been noticed in an earlier DFT study on
degradation of the crown ether.
lithium phosphanide-borane solvates27 and reflect in the
The reported 1H NOESY results are best rationalised by present case presumably deficits of the continuum models in
assuming that the ion pairing creates both Li⋯P and Li⋯H3B- a realistic description of cation solvation rather than an obser-
contacts between the metal cation and the phosphorus atom vable structural trend. Even if in this context comparison of
of the anion fragment. A somewhat different situation had the energies of “naked” Li⋯P- and Li⋯H3B-bound contact ion
been inferred for lithium diphenylphosphanide-borane. Here, pairs cannot be assumed to give meaningful results, the per-
NMR data indicated a structure Ph2P–BH3⋯Li(THF)2 (7) in sisting energetic preference for the Li⋯P-bound isomer even
which the cation is exclusively connected with the hydrides on after the inclusion of explicit solvent molecules strongly
boron.27 Since the higher stability of the Li⋯H3B- over a Li⋯P- suggests that the Li⋯H3B-binding mode that had been
bound isomer had in that case been confirmed by DFT model inferred for 7 27 is in case of 2 disfavoured.28 Even if the DFT
studies, we initiated a computational study in order to analyse results allow no explicit explanation, this finding is in accord
also the relative stabilities of the two different isomers of 2. To with the commonly acknowledged higher nucleophilicity of
this end, calculations at the COSMO-RI-BP86-D3BJ/def2-tzvp aminophosphanes as compared to PPh- or PH-substituted
level were carried out on a “naked” ion pair [(Et2N)2P(BH3)]Li derivatives.
where solvation was only simulated by a continuum model,
It should be noted that computing the energetics of the iso-
and on a species [(Et2N)2P(BH3)][Li(OEt2)2] under explicit mod- desmic Li/H exchange reactions shown in Scheme 3, which
elling of the cation solvation by two ether molecules. The indicated that metalation of secondary phosphane-borane 5 is
results of harmonic frequency analyses on energy optimised by 69 kJ mol−1 more exergonic than the analogous reaction of
molecular structures suggest that both the Li⋯P- and Li⋯H3B- the appropriate free secondary phosphane, confirmed also our
bound isomers (see Fig. 1) constitute local minima on the initial hypothesis of borane coordination boosting the acidity
energy hypersurface. The former is favoured by 30.5 kJ mol−1 if of the PH-bond in secondary aminophosphanes.
explicit solvation is neglected, and still by 4.9 kJ mol−1 when
The synthetic potential of diaminophosphanide-boranes
lithium solvation by two ether molecules is explicitly con- had first been demonstrated by Knochel et al. in the prepa-
sidered. Comparison of the energy optimised structures for a ration of organic diethylaminophosphanes R–P(NEt2)2 and
series of Li⋯P-bound solvates [(Et2N)2P(BH3)][Li(OEt2)n] (n = (Et2N)2P–R–P(NEt2)2 via nucleophilic substitution of mono-
0–3, see ESI‡) reveals that an increasing number of solvent and difunctional organic halides or mesylates with 2.10 Having
molecules induces an overall increase in the P⋯Li distance up to now been unable to isolate crystalline samples of
from 2.400 Å (n = 0) to 2.509 Å (n = 3). At the same time, the lithium and potassium phosphanide-boranes 2 or 6, we antici-
lithium cation is shifted from a P,BH2-bridging position with pated that analogous reactions of these substrates with suit-
agnostic Li⋯H contacts (1.960 Å) into a “terminal” position able metal-based electrophiles might give way to isolable phos-
where it now completes a regular tetrahedral coordination phanide-borane derivatives by transmetalation. To this end, we
sphere at the phosphorus atom. The closest Li⋯H contact studied reactions of 2 and 6 with the carbene complex (IDipp)
CuCl and Zn(tmeda)Cl2 (IDipp
=
1,3-bis-(2,6-diisopropyl-
phenyl)-imidazolylidene; tmeda
=
tetramethyl-
ethylenediamine), respectively. Reaction monitoring by 31P
NMR spectroscopy disclosed in each case the formation of a
single main product beside 5 (possibly through partial hydro-
lysis) and minor amounts of as yet unidentified by-products.
The major products were isolated in moderate yields after
work-up and fully characterised.
Fig. 1 Molecular structures and relative Gibbs free energies ΔGrel for
Li⋯P- (left) and Li⋯H3B-bound isomers (right) of (Et2N)2P(BH3)Li(OEt2)n
(n = 0, 2) calculated at the COSMO-RI-BP86-D3BJ/def2-tzvp level of
theory.
Scheme 3 Electronic energies including zero-point vibrational correc-
tion (ΔE + zpe) and Standard Gibbs free energy (ΔG) for isodesmic
lithium transfer reactions for 4 (Y = l.p.) and 5 (Y = BH3) calculated at the
COSMO-RI-BP86-D3BJ/def2-tzvp level of theory.
This journal is © The Royal Society of Chemistry 2017
Dalton Trans.

View Article Online
Paper
Dalton Transactions
Scheme 4 Metathesis reactions of 2 with (IDipp)CuCl and Zn(tmeda)
Cl2 (R = 2,6-diisopropylphenyl, tmeda = tetramethylethylenediamine).
Identification of the products as the expected complexes 8
and 9 (Scheme 4) was most easily feasible from the single-
crystal X-ray diffraction data. Crystals of both compounds
contain mononuclear complexes that are assembled around
metal atoms with quasi-linear (8, Fig. 2) or tetrahedral (9,
Fig. 3) coordination geometry, respectively. The metal-phos-
phorus distances (8: Cu1–P1 2.2112(5) Å; 9: Zn1–P1 2.3589(8)
Å) are close to standard bond lengths in phosphane complexes
(Cu–P 2.207 Å, Zn–P 2.413 Å29) and, in case of 8, slightly
shorter than the Cu–P distances in diphenylphosphanide-
borane complexes containing four-coordinate Cu(I) centres30
(Cu–P 2.2609(6)–2.2640(5) Å). The various distances in the
phosphanide unit and the tetrahedral coordination geometry
of the phosphorus atom are normal and differ not significantly
between both complexes.
Fig. 3 Representation of the molecular structure of 9 in the crystal. For
clarity, hydrogen atoms of CH-bonds were omitted and the carbon
atoms in the tmeda-ligand represented using a wire model. Selected
distances [Å]: Zn1–Cl1 2.2445(8), Zn1–P1 2.3589(8), P1–N2 1.682(3), P1–
N1 1.701(3), P1–B1 1.949(4).
First reactivity studies revealed that the copper complex 8
resisted all attempts to cleave the coordinated borane unit by
action of tertiary amines such as Et3N or DABCO, respectively.
Quite interestingly, however, we noticed that heating solutions
of 8 in toluene to 95 °C, both in the presence or absence of a
tertiary amine, resulted in transformation into a major phos-
phorus-containing compound (>75%) which was identified by
comparison of the observed spectral data with an authentic
sample as the phosphaalkene 10 31 (Scheme 5). The identity of
the by-products formed is still unknown, and the elucidation
Scheme 5 Thermolysis of complex 8 (R = 2,6-diisopropylphenyl).
of the reaction mechanism and the origin of the P-bound
hydrogen atom in 10 are currently being investigated.
Conclusions
The metalation of a secondary diaminophosphane-borane
with strong anion bases provides an alternative to the syn-
thesis of thermally stable diaminophosphanide-boranes by
metal reduction of diamino-chlorophosphane-borane precur-
sors with appealing prospects to be developed into a widely
applicable synthetic protocol. NMR studies suggest that the
metalated products exist in solution as contact ion pairs the
constituents of which interact through M⋯P interactions and
undergo, at least for M = Li, no aggregation to form higher
oligomers. DFT studies indicate that ion pairing via close Li⋯P
contacts is energetically slightly preferable over formation of a
hydride-bridged structure R2P–BH3⋯Li(solv) which had pre-
viously been reported for lithium diphenylphosphanide-
borane. This finding confirms once more that the structures of
the ion pairs exhibit a subtle dependence on substituent
effects.28 The reactivity of the diaminophosphanide-boranes as
P-centred nucleophiles is emphasised by transmetalation reac-
tions with zinc and copper halides which afforded the first
fully characterised metal complexes of the functionalised
phosphanide-boranes. The thermal decomposition of the
copper complex 8 surprisingly yields phosphaalkene 10 under
Fig. 2 Representation of the molecular structure of 8 in the crystal. For
clarity, hydrogen atoms of CH-bonds were omitted and the Dipp-substi-
tuents represented using a wire model. Selected distances [Å] and
angles [°]: Cu1–C1 1.9123(16), Cu1–P1 2.2112(5), P1–N3 1.6956(14), P1–
N4 1.6975(15), P1–B1 1.944(2), C1–Cu1–P1 172.00(5).
Dalton Trans.
This journal is © The Royal Society of Chemistry 2017

View Article Online
Dalton Transactions
Paper
concomitant transfer of a metal-bound carbene to phos- 13 P. G. Chantrell, C. A. Pearce, C. R. Toyer and R. Twaits,
phorus. This finding draws attention to the fact that metalated J. Appl. Chem., 1964, 14, 563.
diaminophosphanide-boranes may show similar behaviour as 14 R. Marion, F. Sguerra, F. Di Meo, E. Sauvageot, J.-F. Lohier,
phosphinidenoid complexes9,26,32 which can undergo reac-
tions under cleavage of electrofugic (M+) and nucleofugic
R. Daniellou, J.-L. Renaud, M. Linares, M. Hamel and
S. Gaillard, Inorg. Chem., 2014, 53, 9181.
(Et2N−) substituents and formal transfer of a P(I) fragment 15 G. M. Sheldrick, Acta Crystallogr., Sect. A: Fundam.
(“phosphinidene”) to a different substrate. The investigation of
Crystallogr., 2008, 64, 112.
the scope of such reactions, as well as the elucidation of the 16 M. Sierka, A. Hogekamp and R. Ahlrichs, J. Chem. Phys.,
mechanism of the formation of 10, is the objective of ongoing
2003, 118, 9136.
research in our laboratory.
17 TURBOMOLE V7.0 2015, a development of University of
Karlsruhe and Forschungszentrum Karlsruhe GmbH,
1989–2007, TURBOMOLE GmbH, since 2007; available
from http://www.turbomole.com.
Conflicts of interest
18 A. D. Becke, Phys. Rev. A, 1988, 38, 3098.
19 F. Weigend and R. Ahlrichs, Phys. Chem. Chem. Phys., 2005,
7, 3297.
There are no conflicts to declare.
20 S. Grimme, S. Ehrlich and L. Goerigk, J. Comput. Chem.,
2011, 32, 1456.
Acknowledgements
21 A. Klamt and G. Schüürmann, J. Chem. Soc., Perkin Trans.
2, 1993, 799.
The computational studies were supported by the state of
Baden-Württemberg through bwHPC. We further thank
B. Förtsch for elemental analyses, and Dr W. Frey and Mr
J. Trinkner (both Institute of Organic Chemistry, University of
Stuttgart) for the collection of X-ray data sets and mass
spectra, respectively.
22 A. E. Reed, R. B. Weinstock and F. Weinhold, J. Chem.
Phys., 1985, 83, 735.
23 Selected references: (a) H. Nöth and J. Vetter, Chem. Ber.,
1961, 94, 1505; (b) N. Kuhn and K. Jendral, Z. Naturforsch.,
B: Chem. Sci., 1991, 46, 280; (c) H. R. G. Bender, E. Niecke,
M. Nieger and H. Westermann, Z. Anorg. Allg. Chem., 1994,
620, 1194; (d) J.-P. Bezombes, P. B. Hitchcock,
M. F. Lappert and J. E. Nycz, Dalton Trans., 2004, 499;
(e) R. Edge, R. J. Less, E. J. L. McInnes, K. Müther,
V. Naseri, J. M. Rawson and D. S. Wright, Chem. Commun.,
2009, 1691; (f) D. Förster, H. Dilger, F. Ehret, M. Nieger
and D. Gudat, Eur. J. Inorg. Chem., 2012, 3989;
(g) O. Puntigam, D. Förster, N. A. Giffin, S. Burck,
J. Bender, F. Ehret, A. D. Hendsbee, M. Nieger,
J. D. Masuda and D. Gudat, Eur. J. Inorg. Chem., 2013,
2041.
Notes and references
1 K. Issleib and A. Tzschach, Chem. Ber., 1959, 92, 1118.
2 K. Issleib and G. Grams, Z. Anorg. Allg. Chem., 1959, 299,
58.
3 K. Issleib and H. O. Fröhlich, Z. Naturforsch., B: Anorg.
Chem. Org. Chem. Biochem. Biophys. Biol., 1959, 14, 349.
4 N. Kreutzkamp, Chem. Ber., 1954, 87, 919.
5 G. Fritz and W. Hölderich, Z. Anorg. Allg. Chem., 1976, 422, 24 A. H. Cowley and R. A. Kemp, Inorg. Chem., 1983,
104. 22, 541.
6 F. Uhlig, S. Gremler, M. Dargatz, M. Scheer and 25 M. Stankevič, A. Włodarczyk and D. Nieckarz, Eur. J. Org.
E. Herrmann, Z. Anorg. Allg. Chem., 1991, 606, 105.
7 L. Maier, German patent, DE124418019670713, 1967.
8 (a) J.-J. Brunet, R. Chauvin, O. Diallo, B. Donnadieu,
Chem., 2013, 4351.
26 A. Özbolat, G. v. Frantzius, W. Hoffbauer and R. Streubel,
Dalton Trans., 2008, 2674.
J. Jaffart and D. Neibecker, J. Organomet. Chem., 1998, 570, 27 G. Barozzino Consiglio, P. Queval, A. Harrison-Marchand,
195; (b) Erratum: J.-J. Brunet, R. Chauvin, O. Diallo,
B. Donnadieu, J. Jaffart and D. Neibecker, J. Organomet.
Chem., 1999, 584, 389.
A. Mordini, J.-F. Lohier, O. Delacroix, A.-C. Gaumont,
H. Gérard, J. Maddaluno and H. Oulyadi, J. Am. Chem. Soc.,
2011, 133, 6472.
9 P. K. Majhi, A. W. Kyri, A. Schmer, G. Schnakenburg and 28 The coexistence between M⋯P and M⋯H3B bonding was
R. Streubel, Chem. – Eur. J., 2016, 22, 15413.
10 A. Longeau and P. Knochel, Tetrahedron Lett., 1996, 37,
6099.
11 M. Gediga, C. M. Feil, S. H. Schlindwein, J. Bender,
M. Nieger and D. Gudat, Chem. – Eur. J., 2017, 23,
11560.
also established in other crystalline alkaline metal phos-
phanide-boranes: (a) G. Müller and J. Brand,
Organometallics, 2003, 22, 1463; (b) K. Izod, J. M. Watson,
W. Clegg and R. W. Harrington, Eur. J. Inorg. Chem., 2012,
1696; (c) K. Izod, J. M. Watson, W. Clegg and
R. W. Harrington, Inorg. Chem., 2013, 52, 1466.
12 M.
Gediga,
S.
H.
Schlindwein,
J.
Bender, 29 Median values returned by queries in the CSD data base
M. Nieger and D. Gudat, Angew. Chem., Int. Ed., 2017, 56,
15718.
for phosphane complexes containing two-coordinate Cu(I)
and four-coordinate Zn(II) cations, respectively.
This journal is © The Royal Society of Chemistry 2017
Dalton Trans.

View Article Online
Paper
Dalton Transactions
30 I. Abdellah, E. Bernoud, J.-F. Lohier, C. Alayrac, L. Toupet, 32 See also: (a) A. Özbolat, G. v. Frantzius, J. Marinas Perez,
C. Lepetit and A.-C. Gaumont, Chem. Commun., 2012, 48, 4088.
31 M. Cicač-Hudi, J. Bender, S. H. Schlindwein,
M. Bispinghoff, M. Nieger, H. Grützmacher and D. Gudat,
Eur. J. Inorg. Chem., 2016, 649.
M. Nieger and R. Streubel, Angew. Chem., Int. Ed., 2007, 46,
9327; (b) R. Streubel, A. Özbolat-Schön, G. v. Frantzius,
H. Lee, G. Schnakenburg and D. Gudat, Inorg. Chem., 2013,
52, 3313.
Dalton Trans.
This journal is © The Royal Society of Chemistry 2017