Lithium and aluminium complexes supported by chelating phosphaguanidinates

Article abstract of DOI:10.1039/b506332a

Diisopropylcarbodiimide, ⁱPrN=C=NⁱPr, inserts into the lithium-phosphorus bond of in situ prepared Ph₂PLi(THF) _n to afford the lithium salt, [Li(Ph₂PC{N ⁱPr}₂)(THF)_n]_x (2a); alternatively, this compound can be made by deprotonation of the neutral phosphaguanidine, Ph₂PC{NⁱPr}{NHⁱPr} (1a) with ⁿBuLi. Displacement of the THF solvate in 2a is readily achieved with TMEDA to afford Li(Ph₂PC{NⁱPr}₂)(TMEDA) (3a). X-Ray crystallographic analyses show that 2a exists as a dimer in the solid state with a folded ladder structure and an N,N′ chelating phosphaguanidinate, while 3a is monomeric with N,P-coordination of the ligand to lithium. Compound 2a reacts via a transmetallation pathway with AlMe₂Cl to afford the dimethylaluminium complex, Al(Ph₂PC{NⁱPr} ₂)Me₂ (4a), which can also be prepared by protonation of a methyl group of AlMe₃ using 1a. The formation of a series of dialkylaluminium compounds has been investigated employing this latter pathway using both 1a and the N,N′-dicyclohexyl analogue, Ph₂PC{NCy} {NHCy} (1b), affording Al(Ph₂PC{NR}₂)Et₂ (5a, b), Al(Ph₂PC{NR}₂)ⁱBu₂ (6a, b) and the diphenylaluminium compound Al(Ph₂PC{NⁱPr} ₂)Ph₂ (7a). The oily nature of most of the dialkyl compounds and high sensitivity to oxygen and moisture lead to difficulty in manipulation and characterization; however, NMR spectroscopy indicated highly pure products (>95%) upon removal of the solvent. The molecular structures of the crystalline examples 4a and 7a are reported, showing monomeric aluminium species with symmetrically chelating phosphaguanidinate ligands. The series of aluminium compounds AlLCl₂ {L = [EC{NiPr}₂]^-: A, E = Me; B, E = Me₂N; C, E = (Me₃Si)₂N and D, E = Ph₂P} were investigated using density functional theory. In the more simple cases A and B, the delocalized electron density of the metallacycle was represented by a combination of the HOMO and an orbital of lower energy (A, HOMO-5; B, HOMO-6). The HOMO-1 in B was π-bonded across the Me₂N-C bond suggesting delocalization of electron density into the metallacycle. In the more complex systems C and D, delocalization within the metallacycle was less extensive due to the (Me₃Si)₂N-and Ph ₂P-moieties. A number of occupied orbitals in D, however, display phosphorus 'lone-pair' characteristics, indicating that these species have the potential to behave as Lewis bases in the formation of poly(metallic) systems. The Royal Society of Chemistry 2005.

Full text of DOI:10.1039/b506332a

View Article Online / Journal Homepage / Table of Contents for this issue
F U L L P A P E R
Lithium and aluminium complexes supported by chelating
phosphaguanidinates†
Natalie E. Mansfield, Martyn P. Coles* and Peter B. Hitchcock
The Department of Chemistry, University of Sussex, Falmer, Brighton, UK BN1 9QJ.
E-mail: m.p.coles@sussex.ac.uk; Fax: 01273 677196; Tel: 01273 877339
Received 5th May 2005, Accepted 7th July 2005
First published as an Advance Article on the web 21st July 2005
i
i
= =
Diisopropylcarbodiimide, PrN C N Pr, inserts into the lithium–phosphorus bond of in situ prepared
“Ph2PLi(THF)n” to afford the lithium salt, [Li(Ph2PC{NiPr}2)(THF)n]x (2a); alternatively, this compound
can be made by deprotonation of the neutral phosphaguanidine, Ph2PC{NiPr}{NHiPr} (1a) with nBuLi.
Displacement of the THF solvate in 2a is readily achieved with TMEDA to afford Li(Ph2PC{NiPr}2)(TMEDA)
(3a). X-Ray crystallographic analyses show that 2a exists as a dimer in the solid state with a folded ladder
structure and an N,Nꢀ chelating phosphaguanidinate, while 3a is monomeric with N,P-coordination of the ligand
to lithium. Compound 2a reacts via a transmetallation pathway with AlMe2Cl to afford the dimethylaluminium
complex, Al(Ph2PC{NiPr}2)Me2 (4a), which can also be prepared by protonation of a methyl group of AlMe3 using
1a. The formation of a series of dialkylaluminium compounds has been investigated employing this latter pathway
using both 1a and the N,Nꢀ-dicyclohexyl analogue, Ph2PC{NCy}{NHCy} (1b), affording Al(Ph2PC{NR}2)Et2 (5a,
b), Al(Ph2PC{NR}2)iBu2 (6a, b) and the diphenylaluminium compound Al(Ph2PC{NiPr}2)Ph2 (7a). The oily nature
of most of the dialkyl compounds and high sensitivity to oxygen and moisture lead to difficulty in manipulation
and characterization; however, NMR spectroscopy indicated highly pure products (>95%) upon removal
of the solvent. The molecular structures of the crystalline examples 4a and 7a are reported, showing monomeric
aluminium species with symmetrically chelating phosphaguanidinate ligands. The series of aluminium compounds
AlLCl2 {L = [EC{NiPr}2]−: A, E = Me; B, E = Me2N; C, E = (Me3Si)2N and D, E = Ph2P} were investigated using
density functional theory. In the more simple cases A and B, the delocalized electron density of the metallacycle was
represented by a combination of the HOMO and an orbital of lower energy (A, HOMO-5; B, HOMO-6). The HOMO-1
in B was p-bonded across the Me2N–C bond suggesting delocalization of electron density into the metallacycle.
In the more complex systems C and D, delocalization within the metallacycle was less extensive due to the (Me3Si)2N-
and Ph2P-moieties. A number of occupied orbitals in D, however, display phosphorus ‘lone-pair’ characteristics,
indicating that these species have the potential to behave as Lewis bases in the formation of poly(metallic) systems.
Introduction
Amidine and guanidine compounds EC{NRꢀ}{NHRꢀ}, in
which E = alkyl/aryl and amide respectively, have emerged
as a versatile class of ligand precursor, building on spo-
radic reports in the literature dating from the 1980’s.1 Work
has primarily focussed on their coordination as the anionic
amidinate/guanidinate, [EC{NRꢀ}2]−, which typically chelates
symmetrically through the nitrogen atoms to generate a strained,
four-membered metallacycle. Aluminium has featured promi-
nently in this area of chemistry as a metal to which these ligands
readily coordinate, being synthetically accessible via insertion
of carbodiimide into existing aluminium–alkyl/–amide bonds,
transmetallation of an aluminium–halide bond using the appro-
priate group 1 reagent, or, perhaps more successfully, via the
protonolysis reaction with trialkyl aluminium compounds.2–10
Examination of structural data pertaining to both amidinate
Scheme 1 Predominant resonance structure for the guanidinate and
and guanidinate compounds of aluminium has led to a better
understanding of the primary features which dictate the bonding
regime within these systems. For example, incorporation of
small substituents at either the nitrogen or carbon position in
amidinate compounds favours a bridged coordination mode,
typified in the dimer, [Al(MeC{NMe}2)Me2]2,2 while larger
substituents promote monomeric compounds in which the
amidinate chelates to the metal.5 Variation in the distribution
of p-electron density throughout the ligand is also possible with
guanidinate anions, where the incorporation of the R2N-amide
functionality presents a third possible resonance structure, in
phosphaguanidinate anions.
which the nitrogen lone-pair delocalises into the ‘CN3’ core of
the ligand (Scheme 1). The extent to which this ‘zwitterionic’
resonance form contributes to the overall bonding offers limited
control over the degree of electron donation to a metal centre.
Attempts to optimise this effect using a bicyclic guanidinate
framework have been reported, where the enforced coplanar
arrangement of the two ligand components maximises the
orbital overlap to produce ‘electron-rich’ guanidinates.11
Delocalisation of the lone-pair into the ligand framework is
predicted to be much less prevalent upon replacement of the
amide moiety with a phosphide group, due to unfavourable
orbital overlap between the phosphorus and the sp2-carbon
† Electronic supplementary information (ESI) available: 1H, 13C and 31
P
NMR spectra of 2–7. See http://dx.doi.org/10.1039/b506332a
T h i s j o u r n a l i s
T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 5
D a l t o n T r a n s . , 2 0 0 5 , 2 8 3 3 – 2 8 4 1
2 8 3 3
©

View Article Online
(Scheme 1). The resulting phosphaguanidine compounds,
R2PC{NRꢀ}{NHRꢀ}, may therefore be considered as bifunc-
tional compounds with independently reactive amidine and
phosphine moieties.12 Our interest in such compounds predom-
inantly lies in the fact that phosphines remain one of the most
extensively used supporting ligands in coordination chemistry,
finding application in a host of different catalytic reactions.13 The
continued success of the PRꢀ3 ligands derives chiefly from the var-
ied steric and electronic profiles that are readily achieved through
the incorporation of different substituents, Rꢀ. The introduction
of an amidine unit at phosphorus provides a general method for
the incorporation of different metal fragments into a phosphine
compound, further extending the chemistry of this versatile
ligand class. While phosphaguanidine compounds were origi-
nally presented in the 1980’s,14,15 the synthesis of coordination
compounds were only briefly discussed,16 and no reference to bi-
functionality was reported. We have previously communicated a
general synthesis of phospha(III)-17 and phospha(V)guanidines,18
and discussed briefly their coordination potential as both
neutral19 and anionic17 species. In this submission we present
a series of lithium and aluminium complexes in which the
phosphaguanidinate anion predominantly bonds through the
nitrogen atoms of the amidine unit, leaving the phosphorus lone-
pair available for coordination to additional metal fragments.
1H NMR (2H8-toluene) 293 K: d 7.64 (mult, PPh2), 6.96 (mult,
PPh2), 3.96 (br sept, CHMe2), 3.70 (br sept, CHMe2), 3.51 (mult,
THF), 1.48 (br d, CHMe2), 1.40 (mult, THF), 0.74 (br mult,
CHMe2). 31P NMR (2H8-toluene, 202 MHz) 223 K: d 3.6 (s),
7
−20.4 (s), −23.4 (s). Li NMR (2H8–THF, 194 MHz) 298 K:
d 1.26.
Method
2
(in situ. generation):
A
solution of
Ph2PC{NiPr}{NHiPr} (0.500 g, 1.60 mmol) was dissolved
◦
n
in THF (∼30 mL) and cooled to 0 C. BuLi (0.64 mL of a
2.5 M solution in hexanes, 1.60 mmol) was added dropwise via
syringe and the resultant mixture allowed to warm to room
temperature. The reaction was stirred for 30 min and used
as a solution with no further purification, assuming 100%
conversion to the lithium species.
Li(Ph2PC{NiPr}2)(TMEDA) (3a). 2a (0.500 g, 1.34 mmol)
was dissolved in toluene to afford a pale yellow solution. Excess
TMEDA (0.40 mL, 2.82 mmol, 2.1 equiv.) was added dropwise
via syringe resulting in an immediate darkening of the solution
to deep yellow. The reaction was stirred for 3 h, concentrated to
incipient crystallisation and stored at 4 ◦C for 12 h, during which
time 3a formed as colourless crystals. Yield 0.566 g (2 crops),
96% [calculated for Li(Ph2PC{NiPr}2)(THF)].
Anal. Calc. for C25H40N4LiP: C, 73.86; H, 9.92; N,6.89. Found
C, 73.79; H, 9.79; N, 6.81. 1H NMR (2H8-toluene, 500.13 MHz)
368 K: d 7.64 (s, br, PPh2), 7.06 (t, PPh2), 6.97 (mult, PPh2),
3.87 (s, br, CHMe2), 2.09 (s, NCH2), 2.01 (s, NMe2), 0.96 (s, br,
CHMe2). 31P NMR (2H8-toluene, 202.5 MHz) 368 K: d 12.7 (s).
183 K: d 2.9 (d, br, JPLi = 16 Hz, N,P-bound isomer), −21.6 (s,
br, N,Nꢀ-bound isomer). 7Li NMR (2H8-toluene, 195 MHz) 368
K: d − 0.27 (s). 183 K: d −0.15 (d, JPLi = 5 Hz, N,Nꢀ-bound
isomer), −0.58 (d, JPLi = 16 Hz, N,P-bound isomer).
Experimental
General considerations
All manipulations were carried out under dry nitrogen using
standard Schlenk-line and cannula techniques, or in a con-
ventional nitrogen-filled glovebox. Solvents were dried over
appropriate drying agents and degassed prior to use. NMR
spectra were recorded using a Bruker Avance DPX 300 MHz
Al(Ph2PC{NiPr}2)Me2 (4a). Method
1
(metathesis):
1
1
spectrometer at 300.1 (1H), 75.4 (13C{ H}), 121.4 (31P{ H})
AlMe2Cl (1.34 mL of a 1.0 M solution in hexanes, 1.34 mmol)
was added dropwise at room temperature to a slurry of 2a
(0.500 g, 1.34 mmol) in hexanes. The resultant mixture was
stirred at room temperature for 4 h, during which time a
colourless solution and a white solid formed. The volatile
components were removed in vacuo to afford 4a as a colourless
oil, which gradually crystallised over a 7 day period. Analytically
pure samples were obtained by sublimation of the crude product
at 70 ◦C at 10−2 mmHg onto a water cooled probe. Overall yield
0.330 g, 67% [calculated for Li(Ph2PC{NiPr}2)(THF)].
and 116.6 (7Li{ H}) MHz and a Bruker AMX 500 MHz
1
spectrometer at 500.1 (1H), 125.7 (13C{ H}), 202.4 (31P{ H}) and
1
1
◦
194.4 (7Li{ H}) MHz, from samples at 25 C in [2H6]-benzene,
unless otherwise stated. Coupling constants are quoted in Hz.
Proton and carbon chemical shifts were referenced internally
to residual solvent resonances, phosphorus and lithium were
referenced externally to H3PO4 (aq) and LiCl (aq), respectively.
Elemental analyses were performed by S. Boyer at London
Metropolitan University.
1
i
i
= =
Method 2 (protonolysis): AlMe3 (0.16 mL of a 2.0 M solution
in hexanes, 0.32 mmol) was added dropwise at room temperature
to a rapidly stirred solution of 1a (0.100 g, 0.32 mmol) in toluene.
The resultant mixture was stirred at room temperature for 16 h,
after which time the volatile components were removed in vacuo
to afford 4a as a colourless oil. Overall yield 0.26 g, 82%.
Anal. Calc. for C21H30N2AlP: C, 68.46; H, 8.21; N, 7.60.
The reagents Ph2PH (Fluka), PrN C N Pr (Aldrich),
CyN C NCy (Aldrich) and BuLi (2.5 M in hexanes, Acros)
n
= =
were purchased from commercial sources and used as re-
ceived. Hexane solutions of the aluminium reagents AlMe3
(2.0 M) AlMe2Cl (1.0 M), AlEt3 (1.0 M) and AliBu2H
(1.0 M) were purchased from Aldrich chemical company
and used as received. The compound AlPh3 was made ac-
cording to literature procedures,20 and kindly donated by
Dr J. D. Smith; phosphaguanidines Ph2PC{NiPr}{NHiPr} (1a)
and Ph2PC{NCy}{NHCy} (1b) were synthesised according to
literature procedures.19
Found: C, 68.53; H, 8.27; N, 7.52. 1H NMR: d 7.46 (d t, 3JHH
=
7.7, 4JPH = 1.9, 4H, m-C6H5), 7.02–6.98 (m, 6H, o,p-C6H5), 3.81
3
4
3
(d sept, JHH = 6.3, JPH = 3.7, 2H, CHMe2), 0.89 (d, JHH
=
6.3 Hz, 12H, CHMe2), −0.16 (s, 6H, AlMe2). 13C NMR: d 172.7
1
(d, JPC = 51, PCN2), 133.0 (d, JPC = 19 Hz, C6H5), 132.7 (d,
Li(Ph2PC{NiPr}2)(THF)n (2a). Method 1 (isolation): nBuLi
(2.40 mL of a 2.5 M solution in hexanes, 5.90 mmol) was added
dropwise to a stirred solution of Ph2PH (1.00 g, 5.37 mmol)
in THF (∼50 mL) at −78 ◦C, resulting in the immediate
formation of a clear, orange solution. The mixture was allowed
to warm to room temperature and stirred for 30 min. A freshly
JPC = 7 Hz, C6H5), 129.2 (d, JPC = 18 Hz, C6H5), 129.0 (s, C6H5),
3
46.7 (d, JPC = 13, CHMe2), 25.4 (CHMe2), −9.1 (br, AlMe2).
31P NMR: d −19.8.
Al(Ph2PC{NCy}2)Me2 (4b). Compound 4b was made ac-
cording to the procedure described for 4a (Method 2), using
0.100 g of 1b (0.25 mmol) and 0.14 mL AlMe3 (0.28 mmol, 1.1
equiv.). Removal of volatiles under reduced pressure afforded
i
i
= =
degassed solution of PrN C N Pr (0.680 g, 5.37 mmol) in
THF (∼20 mL) was added dropwise at 0 ◦C resulting in
decolourisation of the orange solution. The reaction was allowed
to warm to room temperature and stirred for 12 h. Removal of
the volatile components and washing of the solid with pentane
(3 × 20 mL) afforded a pale yellow powder. Yield 1.51 g, 72%
[calculated for Li(Ph2PC{NiPr}2)(THF)].
1
the product as a colourless oil. Yield 0.142 g, 74%. H NMR:
d 7.50 (d t, 3JHH = 7.7, 4JPH = 1.5, 4H, m-C6H5), 7.04–6.95 (m,
6H, o,p-C6H5), 3.39 (m, 2H, NCH), 1.66–1.62 (m, Cy*), 1.47–
1.45 (m, Cy*), 1.30–0.86 (m, Cy*), −0.11 (s, 6H, CH3) [* total
integral for cyclohexyl methylene groups = 22H]. 13C NMR: d
172.0 (d, 1JPC = 48, PCN2), 133.1 (d, JPC = 19, C6H5), 132.4 (d,
JPC = 8, C6H5), 129.3 (d, JPC = 20, C6H5), 129.1 (C6H5), 54.2
Anal. Calc. for C27H40N2LiO2P [≡Li(Ph2PC{NiPr}2)(THF)2]:
C, 70.11; H, 8.72; N, 6.06. Found C, 70.12; H, 8.69; N, 6.13.
2 8 3 4
D a l t o n T r a n s . , 2 0 0 5 , 2 8 3 3 – 2 8 4 1

View Article Online
(d, 3JPC = 11, NCH), 36.4 (Cy), 25.6 (Cy), 25.4 (Cy), −8.9 (br,
CH3). 31P NMR: d − 19.0. Low resolution MS (EI+, m/z): 433
[M]+ − Me. Accurate MS 433.234816 PPM 1.2.
Al(Ph2PC{NiPr}2)Et2 (5a). Compound 5a was made ac-
cording to the procedure described for 4a (Method 2), using
0.100 g of 1a (0.32 mmol) and 0.35 mL AlEt3 (0.35 mmol,
1.1 equiv.). Removal of volatiles under reduced pressure afforded
the product as a colourless oil. Yield 0.195 g, 79%.
Al(Ph2PC{NiPr}2)Ph2 (7a).
A
colourless solution of
Ph2PC{NiPr}{NHiPr} (0.10 g, 0.32 mmol in toluene (∼10 mL)
was added dropwise to a stirred suspension of AlPh3 (0.08 g,
0.32 mmol) in toluene (∼10 mL) at room temperature. The
resulting clear, colourless solution was stirred at ambient tem-
perature for 36 h. Removal of volatiles under reduced pressure
afforded the crude product as a white powder. X-Ray quality
crystals were obtained by slow cooling of a hot heptane to
ambient temperature. Yield 0.162 g, 59%.
3
4
1H NMR: d 7.49 (d t, JHH = 7.8, JPH = 2.0, 4H, m-C6H5),
3
1H NMR: d 8.14 (d, JHH = 7.7, 4H, i-C6H5{Al}), 7.52
7.05–6.98 (m, 6H, o,p-C6H5), 3.82 (d sept, 3JHH = 6.3, 4JPH = 3.6,
3
4
3
(dt, JHH = 7.8, JPH = 1.5, 4H, m-C6H5{P}), 7.44 (t, JHH
=
2H, CHMe2), 1.47 (t, 3JHH = 8.1, 6H, CH2CH3), 0.89 (d, 3JHH
=
3
7.2, 4H, o-C6H5{Al}), 7.38 (d, JHH = 7.5, 2H, p-C6H5{Al}),
6.3, 12H, CHMe2), 0.42 (q, 3JHH = 8.1, 4H, CH2CH3). 13C NMR:
3
7.06–7.01 (m, 6H, o,p-C6H5{P}), 3.85 (dsept, JHH = 6.3,
1
4JPH = 3.6, 2H, CHMe2), 0.82 (d, JHH = 6.3, 12H, CHMe2).
3
d 173.0 (d, JPC = 52, PCN2), 132.9 (d, JPC = 19, C6H5), 132.7
13C NMR (125.5 MHz): d 177.2 (d, JPC = 54, PCN2),
1
(d, JPC = 13, C6H5), 129.2 (d, JPC = 6, C6H5), 129.0 (C6H5), 46.6
3
(d, JPC = 13, CHMe2), 25.4 (CHMe2), 9.7 (CH2CH3), 0.7 (br,
145.5 (br, i-C6H5{Al}), 138.4 (C6H5{Al}), 133.2 (d, JPC = 26,
C6H5{P}), 129.5 (C6H5{Al}), 129.1 (d, JPC = 7, C6H5{P}), 128.6
CH2CH3). 31P NMR: d −19.6. Low resolution MS (EI+, m/z):
367 [M]+ − Et. Accurate MS 367.189339 PPM −2.6.
(C6H5{P}), 128.3 (C6H5{P}), 127.8 (C6H5{Al}), 47.1 (d, 3JPC
=
12, CHMe2), 25.6 (CHMe2). 31P NMR: d −18.0. Low resolution
MS (EI+, m/z): 492 [M]+. Accurate MS 492.229181 PPM −3.4.
Al(Ph2PC{NCy}2)Et2 (5b). Compound 5b was made ac-
cording to the procedure described for 4a (Method 2), using
0.100 g of 1b (0.25 mmol) and 0.28 mL AlEt3 (0.28 mmol,
1.1 equiv.). Removal of volatiles under reduced pressure afforded
the product as a colourless oil. Yield 0.194 g, 81%.
Crystallography
Details of the crystal data, intensity collection and refinement
are listed in Table 1. Crystals were covered in oil and suitable
single crystals were selected under a microscope and mounted
on a Kappa CCD diffractometer. The structures were refined
with SHELXL-97.21 Additional features are described below.
CCDC numbers 182913 (2a), 182914 (3a), 193950 (4a), 262953
(7a).
3
4
1H NMR: d 7.55 (d t, JHH = 7.7, JPH = 1.4, 4H, m-C6H5),
7.05–6.95 (m, 6H, o,p-C6H5), 3.42 (m, 2H, NCH), 1.69–1.65 (m,
3
Cy*), 1.54 (t, JHH = 8.1, 6H, CH2CH3), 1.48–1.44 (m, Cy*),
1.31 (m, Cy*), 1.19–1.12 (m, Cy*), 0.93–0.79 (m, Cy*), 0.48
3
(q, JHH = 8.1, 4H, CH2CH3) [* total integral for cyclohexyl
methylene groups = 22H]. 13C NMR: d 173.1 (d, JPC = 52,
1
See http://dx.doi.org/10.1039/b506332a for crystallographic
data in CIF or other electronic format.
PCN2), 133.0 (d, JPC = 19, C6H5), 133.0 (d, JPC = 12, C6H5),
129.3 (d, JPC = 7, C6H5), 128.8 (C6H5), 54.2 (d, 3JPC = 12, NCH),
36.4 (Cy), 25.6 (Cy), 25.5 (Cy), 9.9 (CH2CH3), 0.8 (br, CH2CH3).
31P NMR: d −19.0. Low resolution MS (EI+, m/z): 477 [M]+ −
Et. Accurate MS 447.249321 PPM 3.7.
Li(Ph2PC{NiPr}2)(TMEDA) 3a. Two independent molecules
with the same geometry were present in the unit cell.
Al(Ph2PC{NiPr}2)iBu2 (6a). Compound 6a was made ac-
cording to the procedure described for 4a (Method 2), using
0.100 g of 1a (0.32 mmol) and 0.35 mL AliBu2H (0.32 mmol, 1.1
equiv.). Removal of volatiles under reduced pressure afforded
the product as a colourless oil. Yield 0.222 g, 73%.
Results and discussion
Lithium phosphaguanidinates
The reaction between lithium amide derivatives, LiNR2, and
ꢀ
ꢀ
= =
carbodiimides, R N C NR , with subsequent quenching of
the intermediate guanidinate salt with a suitable electrophile
represents a highly versatile route to conventional guanidine
compounds, R2NC{NRꢀ}{NHRꢀ}.22,23 We have adopted an
analogous approach for the synthesis of phosphaguanidines,
using diphenylphosphine and diisopropyl- and dicyclohexylcar-
bodiimide starting reagents (Scheme 2). To access the neutral
compound, we have found it most convenient to quench the
reaction with [HNEt3]Cl, and have shown that the resultant
Ph2PC{NRꢀ}{NHRꢀ} compounds will chelate to group 6 metal
carbonyl fragments.19 Although under many circumstances it is
more convenient to react the lithium salt in situ, several inter-
esting features concerning the different modes with which these
ligands coordinate to a metal are clearly illustrated by the lithium
salts. We therefore report here our findings on the solution-
and solid-state properties of the lithium phosphaguanidinate
compound, “Li(Ph2PC{NiPr}2)”, which can be isolated as either
the THF (2a) or TMEDA (3a) adduct.
3
4
1H NMR: d 7.47 (d t, JHH = 8.5, JPH = 1.9, 4H, m-C6H5),
3
4
7.06–6.99 (m, 6H, o,p-C6H5), 3.88 (d sept, JHH = 6.3, JPH
=
3.9, 2H, iPr–CHMe2), 2.21 (sept, 3JHH = 6.7, 2H, iBu–CHMe2),
1.26 (d, 3JHH = 6.5, 12H, iBu–CHMe2), 0.92 (d, 3JHH = 6.3, 12H,
iPr–CHMe2), 0.50 (d, 3JHH = 7.2, 4H, CH2CHMe2). 13C NMR:
1
d 173.2 (d, JPC = 52, PCN2), 132.8 (d, JPC = 19, C6H5), 132.7
(d, JPC = 13, C6H5), 129.1 (d, JPC = 5, C6H5), 129.0 (C6H5),
46.9 (d, 3JPC = 13, iPr–CHMe2), 28.9 (iBu–CHMe2), 27.1 (iBu–
CHMe2), 25.5 (iPr–CHMe2), 23.4 (CH2CHMe2). 31P NMR: d −
19.9. Low resolution MS (EI+, m/z): 395 [M]+ − Bu. Accurate
i
MS 395.219757 PPM −0.2.
Al(Ph2PC{NCy}2)iBu2 (6b). Compound 6b was made ac-
cording to the procedure described for 4a (Method 2), using
0.100 g of 1b (0.25 mmol) and 0.25 mL AliBu2H (0.25 mmol, 1.0
equiv.). Removal of volatiles under reduced pressure afforded
the product as a colourless oil. Yield 0.156 g, 76%.
3
4
i
i
1H NMR: d 7.52 (d t, JHH = 7.9, JPH = 1.3, 4H, m-C6H5),
7.06–6.98 (m, 6H, o,p-C6H5), 3.47 (m, 2H, NCH), 2.27 (sept,
3JHH = 6.7, 2H, iBu–CHMe2), 1.72–1.68 (m, Cy*), 1.50–1.46 (m,
Cy*), 1.32–1.18(m, Cy*), 1.31 (d, 3JHH = 6.6, 12H, iBu–CHMe2),
0.94–0.82 (m, Cy*), 0.56 (d, 3JHH = 7.2, 4H, CH2CHMe2) [* total
integral for cyclohexyl methylene groups = 22H]. 13C NMR
(125.5 MHz): d 173.3 (d, 1JPC = 53, PCN2), 135.8 (d, JPC = 15,
C6H5), 134.3 (d, JPC = 19, C6H5), 133.0 (d, JPC = 33, C6H5),
Addition of a THF solution of “Ph2PLi” to PrN C N Pr
generates the lithium compound, [Li(Ph2PC{NiPr}2)(THF)n]x
(2a) which can be isolated as a white powder in excellent
yield (Scheme 2). Analytical data suggest the formation of a
mixture of different species, both in solution and the solid-state,
differing by the extent of oligomerization and solvation. For
example, combustion analysis of the bulk sample is consistent
with formation of the bis-THF adduct (n = 2 and x = 1,
or multiples thereof) whilst X-ray structural data of isolated
crystals showed formation of the monosolvated dimer (n = 1
and x = 2). 31P NMR experiments at 223 K [2H8-toluene] reveal
the presence of one major species d 3.6, with minor signals at
= =
3
129.1 (C6H5), 54.8 (d, JPC = 12, NCH), 36.6 (Cy), 29.0 (iBu–
CHMe2), 27.2 (iBu–CHMe2), 25.7 (Cy), 23.5 (CH2CHMe2). 31
P
i
NMR: d − 19.3. Low resolution MS (EI+, m/z): 475 [M]+ − Bu.
Accurate MS 475.280098 PPM 4.6.
D a l t o n T r a n s . , 2 0 0 5 , 2 8 3 3 – 2 8 4 1
2 8 3 5

View Article Online
Table 1 Crystal structure and refinement data for [Li(Ph2PC{NiPr}2)(THF)]2 (2a), Li(Ph2PC{NiPr}2)(TMEDA) (3a), Al(Ph2PC{NiPr}2)Me2 (4a)
and Al(Ph2PC{NiPr}2)Ph2 (7a)
2a
3a
4a
7a
Formula
Formula weight
Temperature/K
C46H64Li2N4O2P2
780.83
173(2)
C25H40LiN4P
434.52
173(2)
C21H30AlN2P
368.42
173(2)
C31H34AlN2P
492.55
173(2)
˚
Wavelength/A
0.71073
0.71073
0.71073
0.71073
Crystal/mm
Crystal system
Space group
0.40 × 0.30 × 0.30
Monoclinic
P21/n (No. 14)
9.9739(7)
19.6236(13)
11.6860(8)
90
94.531(3)
90
2280.1(3)
2
0.30 × 0.20 × 0.20
0.30 × 0.30 × 0.10
Monoclinic
P21/n (No. 14)
16.0218(9)
8.3303(6)
17.6419(9)
90
107.443(4)
90
2246.3(2)
4
0.20 × 0.20 × 0.20
Monoclinic
Cc
15.9891(5)
11.2621(3)
15.4755(3)
90
104.277(2)
90
2700.62(12)
4
Triclinic
¯
P1 (No. 2)
˚
a/A
9.5099(2)
˚
b/A
11.9348(2)
28.6122(6)
98.125(1)
˚
c/A
a/◦
b/◦
c /◦
75.908(1)
122.665(3)
2651.37(9)
4
1.09
0.12
3.73 to 23.06
22110
3
˚
V/A
Z
Dc/Mg m−3
1.14
0.14
3.73 to 25.05
8491
1.09
0.17
4.23 to 23.02
9341
1.21
0.16
3.76 to 25.02
19322
Absorption coefficient/mm−1
h range for data collection/◦
Reflections collected
Independent reflections
Reflections with I > 2r(I)
Data/restraints/parameters
Goodness-of-fit on F2
Final R indices [I > 2r(I)]
R indices (all data)
3937 [Rint = 0.088]
7271 [Rint = 0.065]
3095 [Rint = 0.041]
4664 [Rint = 0.055]
3047
5713
2623
4382
3937/0/253
1.035
R1 = 0.063, wR2 = 0.162
R1 = 0.085, wR2 = 0.177
0.60 and −0.28
7271/0/559
1.040
R1 = 0.058, wR2 = 0.141
R1 = 0.081, wR2 = 0.156
0.39 and −0.29
3095/0/226
1.042
R1 = 0.041, wR2 = 0.096
R1 = 0.052, wR2 = 0.101
0.26 and −0.26
4664/2/316
1.014
R1 = 0.030, wR2 = 0.072
R1 = 0.034, wR2 = 0.074
0.16 and −0.14
−3
˚
Largest diff. peak and hole/e A
Scheme 2 Synthetic procedures for the synthesis of lithium and aluminium phosphaguanidinate compounds.
d −20.4 and −23.4 (relative intensities 0.03 and 0.07). We
species are assigned as the N,Nꢀ- and N,P-bonded complexes,
respectively, (Scheme 3) where the magnitude of the one-bond
Li–P coupling in 3a is considerably less than the closely related
[Ph2PLi(TMEDA)]2 compound (JLiP = 45 Hz),24 consistent with
a weak interaction. In the former bonding situation inequivalent
nitrogen substituents are observed at 183 K, indicated by two
interpret these data as indicating the presence of different
bonding modes for the ligand at low temperature, more clearly
illustrated for the TMEDA adduct (vide infra).
Addition of excess TMEDA to a toluene solution of 2a affords
Li(Ph2PC{NiPr}2)(TMEDA), 3a, which is isolated as colourless
crystals. As with 2a, spectroscopic data suggest compound 3a
is present in a number of isomeric forms in solution. NMR
analysis of a 2H8-toluene solution at 183 K shows two doublets
1
different methine resonances in the H NMR spectrum at d
5.18 and 3.97, consistent with restricted rotation about the P–C
bond.25 As the solution is warmed to ∼190 K, free rotation
causes these signals to collapse to a single peak at d 4.57.
in the 7Li NMR spectra at d −0.2 (JPLi = 5 Hz) and −0.6 (JPLi
=
16 Hz), with corresponding 31P resonances at d −21.6 (coupling
to lithium not resolved) and 2.9 (q, br, JPLi = 16 Hz). These
Two isomers are also observed at low temperature for the
i
=
N,P-bound species, depending on the position of the C N Pr
2 8 3 6
D a l t o n T r a n s . , 2 0 0 5 , 2 8 3 3 – 2 8 4 1

View Article Online
Scheme 3 Fluxional processes occurring in solution for Li(Ph2PC-
{NiPr}2)(TMEDA) (3a).
substituent [i.e. cis or trans with respect to the double bond].
Fig.
2
ORTEP representation of the molecular structure of
Li(Ph2PC{NiPr}2(TMEDA) (3a) with ellipsoids drawn at the 30% level
Initial interconversion of these two isomers is observed at 243 K,
1
7
and at high temperature (368 K), all H, Li and 31P NMR
signals collapse into one set of resonances, corresponding to
rapid exchange between all of the different isomeric forms.
The molecular structure of 2a is illustrated in Fig. 1; crystallo-
graphic and refinement details are given in Table 1 and selected
bond distances and angles in Table 2. Compound 2a crystallised
as the centrosymmetric dimer, [Li(Ph2PC{NiPr}2)(THF)]2, in
and hydrogen atoms omitted.
˚
C(1)–N(2) 1.339(3) A], where the longer distance reflects a
charge localisation at the nitrogen atom, as a consequence of
incorporation in the Li2N2 ring.26 The C(1)–P distance [1.915(2)
A] and pyramidal geometry at the phosphorus [Rangles = 315.61◦]
˚
are consistent with a single P–C bond and retention of the lone
pair at phosphorus.
1
2
which the ligand is present as a l–g ,g -bridge between two
lithium centres. Whilst this bonding pattern has not been
previously reported in phosphaguanidinate chemistry, related
dimeric structures have been observed in a number of metal
amidinate26,27 and guanidinate23,28 compounds. The core of the
molecule consists of three fused four-membered rings in a folded
ladder-type arrangement (angle between CN2Li and N2Li2
planes = 124.4◦), with an anti-arrangement about the central
Li2N2 metallacycle. Bond lengths are consistent with unequal
The molecular structure of 3a (Fig. 2; crystallographic and
refinement details Table 1; selected bond distances and angles
in Table 3) consists of two independent molecules in the unit
cell, each of which differ only slightly in geometry. For brevity
the distances and angles associated with one molecule will be
discussed. The most notable structural difference upon displace-
ment of the lithium bound THF by TMEDA is the different
bonding mode for the phosphaguanidinate ligand. The resultant
complex is monomeric with an N,P-bound phosphaguanidinate
and chelating TMEDA molecule defining the four-coordinate
˚
delocalisation within the ‘CN2’ moiety [C(1)–N(1) 1.316(3) A,
˚
lithium atom. Although the P–Li bond distance of 2.625(5) A in
3a is comparable to that reported in [Ph2PLi(TMEDA)]2 (2.523–
24
˚
2.629 A), severe distortion from ideal tetrahedral geometry
results at both the P- and Li-atoms [angles at Li: 66.54–140.3◦;
angles at P: 74.49–139.88◦], suggesting that such an interaction
is weak (consistent with the low 1JLiP value observed in solution).
The C–N bond distances indicate a considerably more localised
bonding situation than for 2a, with the C(1)–N(2) distance
˚
=
[1.299(4) A] close to the value found for C N double bonds.
˚
As a consequence, the N(1)–Li bond [1.927(5) A] is appreciably
shorter than the corresponding distance in 2a.
Aluminium phosphaguanidinates
We have previously reported that the lithium salts 2a and 3a
react with AlMe2Cl via transmetallation to afford the dimethy-
laluminium compound, Al(Ph2PC{NiPr}2)Me2 4a. We generally
find it more convenient, however, to adopt a protonolysis
Fig.
1
ORTEP representation of the molecular structure of
[Li(Ph2PC{NiPr}2(THF)]2 (2a) with ellipsoids drawn at the 30% level
and hydrogen atoms omitted. ꢀ: −x, −y + 1, −z.
◦
˚
Table 3 Selected bond lengths (A) and angles ( ) for 3a
◦
˚
Table 2 Selected bond lengths (A) and angles ( ) for 2a
Li–N(1)
Li–N(3)
P–C(1)
P–C(14)
C(1)–N(2)
1.927(5)
2.107(6)
1.892(3)
1.825(3)
1.299(4)
Li–P
2.625(5)
2.081(6)
1.834(3)
1.354(4)
Li(1)–N(1)
Li(1)–N(2ꢀ)
P–C(1)
P–C(14)
C(1)–N(2)
2.024(5)
2.078(5)
1.915(2)
1.831(3)
1.339(3)
Li(1)–N(2)
Li(1)–O
P–C(8)
2.175(5)
1.923(5)
1.827(3)
1.316(3)
Li–N(4)
P–C(8)
C(1)–N(1)
C(1)–N(1)
P–Li–N(1)
P–Li–N(4)
N(1)–Li–N(4)
C(1)–P–Li
C(1)–P–C(14)
C(8)–P–Li
66.54(16)
122.1(2)
140.3(3)
74.49(14)
105.06(13)
139.88(15)
P–Li–N(3)
114.4(2)
125.3(3)
88.4(2)
107.82(13)
105.43(13)
112.64(14)
N(1)–Li(1)–N(2)
N(1)–Li–O
N(2)–Li–O
C(1)–P–C(8)
C(8)–P–C(14)
65.28(15)
113.3(2)
115.5(2)
107.13(11)
104.03(12)
N(1)–Li–N(2ꢀ)
N(2)–Li–N(2ꢀ)
N(2ꢀ)–Li–O
126.0(2)
109.70(2)
115.9(2)
N(1)–Li–N(3)
N(3)–Li–N(4)
C(1)–P–C(8)
C(8)–P–C(14)
C(14)–P–Li
C(1)–P–C(14)
104.45(11)
D a l t o n T r a n s . , 2 0 0 5 , 2 8 3 3 – 2 8 4 1
2 8 3 7

View Article Online
◦
˚
Table 4 Selected bond lengths (A) and angles ( ) for 4a
procedure, and report here the synthesis of a range of dialkylalu-
minium species incorporating the phosphaguanidinate anions,
i
[Ph2PC{NRꢀ}2]− (Rꢀ = Pr, Cy).
Al–N(1)
Al–C(8)
P–C(1)
1.928(2)
1.948(3)
1.869(2)
1.829(2)
1.336(3)
Al–N(2)
Al–C(9)
P–C(10)
C(1)–N(1)
1.927(2)
1.972(3)
1.824(2)
1.337(3)
The reaction between the neutral phosphaguanidines 1a and
1b and a solution of appropriate aluminium reagent, in 1 : 1
stoichiometry, proceeds to afford the desired mono-guanidinate
compound directly upon removal of the volatile components
P–C(16)
C(1)–N(2)
N(1)–Al–N(2)
N(1)–Al–C(9)
N(2)–Al–C(9)
C(1)–P–C(10)
C(10)–P–C(16)
68.95(8)
N(1)–Al–C(8)
N(2)–Al–C(8)
C(8)–Al–C(9)
C(1)–P–C(16)
115.96(13)
116.11(11)
120.14(13)
102.44(10)
of the reaction (Scheme 2). Using this protocol we have
i
synthesized the series Al(Ph2PC{NRꢀ}2)Rꢀꢀ (a, Rꢀ = Pr; b,
112.30(10)
112.75(10)
101.07(10)
106.12(10)
2
Rꢀ = Cy) for the dimethyl (4a and 4b, Rꢀꢀ = Me), diethyl (5a
i
and 5b, Rꢀꢀ = Et), diisobutyl (6a and 6b, Rꢀꢀ = Bu) complexes
and the aryl derivative Al(Ph2PC{NiPr}2)Ph2 (7a). There
was no evidence for formation of stable P-donor adducts of
the type (ligand)·AlMe3 during these syntheses, despite the
ability of triarylphosphine to form the adducts Ar3P·AlMe3
(Ar = Ph, o-tolyl),29 indicating preferable reaction at the
amidine nitrogen in the case of phosphaguanidines. The 2 : 1
reaction of ligand with AlR3 reagents did not proceed beyond
formation of the mono-ligand product, despite prolonged
reaction times at elevated temperature (100 ◦C, 18 h). This is
somewhat surprising given the ready formation of the related
bis(amidinate)/bis(guanidinate) complexes Al(EC{NRꢀ}2)2X
◦
˚
Table 5 Selected bond lengths (A) and angles ( ) for 7a
Al–N(1)
Al–C(20)
P–C(1)
P–C(14)
C(1)–N(2)
1.927(2)
1.970(2)
1.886(2)
1.836(2)
1.339(2)
Al–N(2)
Al–C(26)
P–C(8)
1.923(2)
1.970(2)
1.828(2)
1.336(2)
C(1)–N(1)
N(1)–Al–N(2)
N(1)–Al–C(26)
N(2)–Al–C(26)
C(1)–P–C(8)
69.30(7)
115.98(8)
113.97(8)
104.04(8)
106.84(9)
N(1)–Al–C(20)
N(2)–Al–C(20)
C(20)–Al–C(26)
C(1)–P–C(14)
120.10(8)
113.63(8)
115.43(8)
98.57(8)
i
i
(E = Me, Rꢀ = Pr, X = Cl;5 E = Me, Rꢀ = Pr, X = Me;9
E = Bu, Rꢀ = Pr, Cy, X = Cl;5 E = Ph, Rꢀ = SiMe3, X =
t
i
C(8)–P–C(14)
H;4 E = Me2N, Rꢀ = Pr, X = Me;9 E = Me2N, Rꢀ = Pr,
i
i
X = Me2N, Cl10) and the tris(amidinate)/tris(guanidinate)
compounds, Al(PhC{NPh}2)3,8 Al(MeC{NiPr}2)3
and
9
Al(Me2NC{NiPr}2)3.10
Attempted purification by crystallisation has, for the most
part, been prevented by the high solubility and oily nature of
these products. As a consequence, despite numerous attempts,
accurate data from combustion analyses have not been forth-
coming during this study which we attribute to both the inherent
difficulties in handling oils and the high moisture/oxygen
sensitivity of aluminium alkyl species. Multinuclear (1H, 13C, 31P)
NMR studies, however, indicate that a single species is formed
in each case, consistent with the predicted product of a 1 : 1
reaction. To illustrate the high level of purity of these compounds
attained upon removal of the volatile components of the reaction
mixture copies of the 1H, 13C and 31P NMR spectra of 4–7 have
been included as ESI. †The spectra indicate a C2v-symmetric
structure in which the phosphaguanidinate is symmetrically
bound to the metal centre through both of the nitrogen atoms. In
order to maintain the observed symmetry in solution, rotation
about the Ph2P–C bond must be rapid on the NMR timescale
and, in contrast to the lithium salt 2a, this motion could not be
frozen out at low temperature (e.g. 198 K, 4a). The 31P chemical
shift in each compound (range −19.9 to −18.0 ppm) is similar to
that found for the neutral ligands (−17.4 and −16.9 ppm for 1a
and 1b, respectively), giving no indication of any multiple bond
character, in which a low field shift between +116 and +170 ppm
has previously been noted.14 The high-resolution mass spectra
have also been recorded, predominantly indicating loss of one
alkyl group during the experiment, highlighting the reactivity of
this linkage.
Fig.
3
ORTEP representation of the molecular structure of
Al(Ph2PC{NiPr}2Me2 (4a) with ellipsoids drawn at the 30% level.
Structural analyses of the favourably crystalline compounds
Al(Ph2PC{NiPr}2)Me2 (4a) and Al(Ph2PC{NiPr}2)Ph2 (7a) have
been performed using X-ray diffraction. The molecular struc-
tures are illustrated in Figs. 3 and 4; crystallographic and
refinement details are given in Table 1 and selected bond
distances and angles in Tables 4 and 5.
Each compound crystallised as the monomeric species with a
distorted tetrahedral aluminium centre and an N,Nꢀ-chelating
phosphaguanidinate ligand. The acute bite angle of the lig-
and [4a, 68.95(8)◦; 7a, 69.30(7)◦] is comparable to those in
related dimethylaluminium complexes containing amidinate and
guanidinate anions3,5 suggesting that the nature of the carbon
substituent does not efficiently transmit differences in steric de-
mand to the metal coordination sphere. The ligands are bonded
symmetrically with no discernable difference in the metal–
Fig.
4
ORTEP representation of the molecular structure of
Al(Ph2PC{NiPr}2Ph2 (7a) with ellipsoids drawn at the 30% level.
nitrogen distances, and equal distribution of p-electron density
across the ‘CN2’ amidine unit is implied from equivalent C–N
˚
distances. The Ph2P–C distances [4a, 1.869(2) A; 7a, 1.886(2)
˚
A] are shorter than in the lithium salt 2a, attributed to the
greater electronegativity associated with Al vs. Li. The retention
2 8 3 8
D a l t o n T r a n s . , 2 0 0 5 , 2 8 3 3 – 2 8 4 1

View Article Online
Table 6 Calculated and observed bond lengths for complexes
Al(EC{NiPr}2)Cl2
of a pyramidal geometry at the phosphorus atom [Rangles: 4a =
309.63◦; 7a = 309.45◦] encourages us in the potential application
of these Al(Ph2PC{NiPr}2)R”2 compounds as ‘metal-containing
phosphine ligands’ for formation of poly(metallic) compounds.
E = Me (A)
E = Me2N (B)
Calculated Observed
Observed
Calculated
Calculations
C(1)–N(1)
C(1)–N(2)
C(1)–E
Al–N(1)
Al–N(2)
1.339(3)
1.336(3)
1.495(3)
1.879(2)
1.879(2)
1.340
1.340
1.505
1.919
1.922
1.360(5)
1.356(6)
1.343(5)
1.873(3)
1.870(4)
1.356
1.356
1.366
1.915
1.915
To further our understanding of the differing extents to which
the p-electron density may become delocalized throughout the
core of ligands of this type, the electronic structures of a
series of aluminium compounds have been examined using the
B3LYP density functional theory and the 6–31+G(d) basis set,
implemented by Gaussian 03.30 The aluminium compounds
AlLCl2 {L = [EC{NiPr}2]−; A, E = Me; B, E = Me2N;6 C,
E = (Me3Si)2N;6 D, E = Ph2P17} were investigated, where the
variation in predicted p-donor/acceptor ability from an alkyl-
group (A) to two substantially different amide-substituents (B,
C) to a phosphide group (D) has generated a series of results
with which to compare experimentally determined data. The
input coordinates for each calculation were generated using
crystallographic data,‡ and full optimization of the structures
led to bond lengths and angles which were, in general, found to
be within 3% of the experimentally determined values (Table 6).
For compounds A and B (Fig. 5), the LUMOs are ligand-
based and p* in character, predominantly extending across
the NCN moiety of the metallacycle (Fig. 6). The delocalised
electron density within this moiety is incorporated within two
E = (Me3Si)2N (C)
E = Ph2P (D)a
Observedb
Calculated Observed
Calculated
C(1)–N(1)
C(1)–N(2)
C(1)–E
Al–N(1)
Al–N(2)
1.341(4)
—
1.400(6)
1.878(3)
—
1.349
1.349
1.403
1.913
1.913
1.337(2)
1.336(2)
1.869(2)
1.928(2)
1.927(2)
1.343
1.342
1.898
1.921
1.922
a Observed bond lengths for dimethyl complex, Al(Ph2PC{NiPr}2)Me2;
Calculated bond lengths for the dichloride complex, Al(Ph2PC-
{NiPr}2)Cl2. b Molecule contains a crystallographic centre of symmetry.
orbitals for each compound. The HOMOs in A and B contribute
to the p-bonding within one of the C–N bonds, with an orbital
of lower energy (A, HOMO-5; B, HOMO-6) p-bonding across
the remaining C–N bond. The key difference between the two
bonding schemes is the presence of a p-bonding orbital (HOMO-
1) in B, demonstrating the delocalization of the lone-pair from
the exocyclic nitrogen into the core of the ligand.
As may be expected, the molecular orbital schemes are
considerably more complex for the silyl substituted guanidinate,
C, and the model‡ phosphaguanidinate compound, D (Fig. 5).
Substituting the Me2N-group in B for (Me3Si)2N- in C imposes
an approximately perpendicular arrangement of this amide
substituent relative to the metallacycle, and as such overlap
between the nitrogen lone-pair and the sp2-carbon centre is
predicted to be much less favourable. The energy level diagram
(Fig. 7) indicates a destabilization of the LUMO relative to
‡ In the case of the phosphaguanidinate compound D, ligand coordi-
nates for the dimethyl compound Al[Ph2PC{NiPr}2]Me2 (4a) were used,
and the methyl ligands substituted for chlorides. The Al–Cl and Al–
N distances used in the calculation were modified to those values for
complex (B) and a full optimization performed.
Fig. 5 Atom labeling scheme used during calculations.
Fig. 6 Molecular orbital diagram of the amidinate compound, Al(MeC{NiPr}2)Cl2 (A)5 and the guanidinate compound Al(Me2NC{NiPr}2)Cl2
(B).6
D a l t o n T r a n s . , 2 0 0 5 , 2 8 3 3 – 2 8 4 1
2 8 3 9

View Article Online
Fig. 7 Molecular orbital diagram of the guanidinate compound, Al[(Me3Si)2NC{NiPr}2]Cl2 (C)6 and the phosphaguanidinate compound
Al(Ph2PC{NiPr}2)Cl2 (D).‡
the other systems under investigation which, rather than being
entirely ligand-based, indicates an element of Al–N p-bonding.
The contribution of the HOMO to the bonding within the
metallacycle essentially consists of nitrogen-based lone-pairs
with the electron density associated with the HOMO-1 being
focussed in N–Si bonding of the silyl substituents. Indeed,
the orbitals containing significant contributions to the delo-
calization of electron density within the metallacycle are much
lower in energy (e.g. HOMO-2, HOMO-9). The silazide nitrogen
lone-pair is delocalised within the Si–C bond regime (e.g.
HOMO-6), resulting in the observed trigonal planar geometry.
The distribution of electron density within the metallacycle in
phosphaguanidinate, D, is also more complex than in A and
B. The LUMO is delocalised across the Al–N bonds, with
both the HOMO and the HOMO-1 containing elements of the
orbital corresponding to electron density in a phosphorus-based
‘lone-pair’. This is augmented by additional contributions from
lower energy orbitals (e.g. HOMO-6, HOMO-9). These data
suggest that electronically these species should retain the ability
to behave as Lewis-bases.
oils which prevented characterization by elemental analysis.
1
However, the H, 13C and 31P NMR spectra showed, unequiv-
ocably, that the product had been generated in >98% purity.
Two examples demonstrated favourable crystalline properties
allowing study by single-crystal X-ray analysis. The com-
pounds Al(Ph2PC{NiPr}2)Rꢀꢀ (Rꢀꢀ = Me, Ph) were shown to
2
be monomeric with symmetrical coordination of the phosph-
aguanidinate anion.
Analysis of the electronic structure of a series of aluminium
dichloride compounds incorporating amidinate, guanidinate
and phosphaguanidinate ligands was performed using density
functional theory. The electron density corresponding to the
delocalized bonding within the NCN component of the met-
allacycle was located to two occupied molecular orbitals for
the [MeC{NiPr}2]− and [Me2NC{NiPr}2]− anions, where in the
latter case, additional electron density was contributed from
the exocyclic nitrogen lone pair, in the HOMO-1. The silyl-
substituted guanidinate and phosphaguanidinate complexes
showed considerably more complicated bonding schemes. In the
former compound, the lone-pair on the silyl substituted nitrogen
in [(Me3Si)2NC{NiPr}2]− is delocalised within the Si–C bond
regime, likely due to the increased electronegativity of the silicon
atom and unfavourable orbital alignment due to the relatively
large SiMe3 groups. In the case of the model phosphaguanidine
compound, Al(Ph2PC{NiPr}2)Cl2, several occupied molecular
orbitals contained elements consistent with location of electron
density in a phosphorus-based ‘lone-pair’. It is predicted that
this will enable these (and related) phosphaguanidinate com-
pounds to coordinate to additional metal fragments, enabling
poly(metallic) systems to be developed. Our results in this area
will be presented in a forthcoming publication.
Conclusions
Carbodiimide insertion into the phosphorus–lithium bond of
solvated ‘Ph2PLi’ serves as a general route to lithium phosph-
aguanidinate compounds. The mono-solvated THF adduct is
dimeric in the solid-state with the N,Nꢀ-chelated ligand serving as
1
2
a l-g ,g -bridge between the two lithium centres. In contrast, the
molecular structure of the TMEDA adduct is monomeric, with
an N,P-bonding motif for the phosphaguanidinate ligand. Both
compounds show complicated fluxional processes in solution
involving different extents of oligomerization and variable
coordination modes of the ligands.
Acknowledgements
The neutral phosphaguanidines, Ph2PC{NRꢀ}{NHRꢀ}, react
cleanly with aluminium alkyls via alkyl elimination, or in the
case of AliBu2H evolution of hydrogen gas, to afford the
We would like to thank Dr M. Suter and Dr M. S. Hill for
helpful discussion regarding the DFT calculations. We also wish
to acknowledge the use of the EPSRC’s Chemical Database
Service at Daresbury.
monoligand compounds, Al(Ph2PC{NRꢀ}2)Rꢀꢀ . The majority
2
of these compounds were isolated as air- and moisture-sensitive
2 8 4 0
D a l t o n T r a n s . , 2 0 0 5 , 2 8 3 3 – 2 8 4 1

View Article Online
16 H. P. M. M. Ambrosius, A. H. I. M. van der Linden and J. J.
Steggerda, J. Organomet. Chem., 1980, 204, 211; D. H. M. W.
Thewissen, H. P. M. M. Ambrosius, H. L. M. V. Gaal and J. J.
Steggerda, J. Organomet. Chem., 1980, 192, 101; F. Lindenberg, J.
Sieler and E. Hey-Hawkins, Polyhedron, 1996, 15, 1459.
17 M. P. Coles and P. B. Hitchcock, Chem. Commun., 2002, 2794.
18 J. Grundy, M. P. Coles, A. G. Avent and P. B. Hitchcock, Chem.
Commun., 2004, 2410.
References
1 J. Barker and M. Kilner, Coord. Chem. Rev., 1994, 133, 219; F. T.
Edelmann, Coord. Chem. Rev., 1994, 137, 403; P. J. Bailey and S.
Pace, Coord. Chem. Rev., 2001, 214, 91.
2 H.-D. Hausen, F. Gerstner and W. Schwarz, J. Organomet. Chem.,
1978, 145, 277.
3 R. Lechler, H.-D. Hausen and U. Weidlein, J. Organomet. Chem.,
1989, 359, 1; M. P. Coles, D. C. Swenson and R. F. Jordan,
Organometallics, 1998, 17, 4042; S. Dagorne, I. A. Guzei, M. P. Coles
and R. F. Jordan, J. Am. Chem. Soc., 2000, 122, 274; J. Grundy, M. P.
Coles and P. B. Hitchcock, J. Organomet. Chem., 2002, 662, 178;
J. A. R. Schmidt and J. Arnold, Organometallics, 2002, 21, 2306; B.
Clare, N. Sarker, R. Showmaker and J. R. Hagadorn, Inorg. Chem.,
2004, 43, 1159.
4 R. Duchateau, A. Meetsma and J. H. Teuben, Chem. Commun., 1996,
223.
5 M. P. Coles, D. C. Swenson, R. F. Jordan and V. G. Young, Jr.,
Organometallics, 1997, 16, 5183.
6 S. L. Aeilts, M. P. Coles, D. C. Swenson, R. F. Jordan and V. G.
Young, Jr., Organometallics, 1998, 17, 3265.
7 C.-C. Chang, C.-S. Hsiung, H.-L. Su, B. Srinivas, M. Y. Chiang, G.-H.
Lee and Y. Wang, Organometallics, 1998, 17, 1595; K. Kincaid, C. P.
Gerlach, G. R. Giesbrecht, J. R. Hagadorn, G. D. Whitener, A. Shafir
and J. Arnold, Organometallics, 1999, 18, 5360; J. Barker, D. R. Aris,
N. C. Blacker, W. Errington, P. R. Phillips and M. G. H. Wallbridge,
J. Organomet. Chem., 1999, 586, 138; D. Doyle, Y. K. Gun’ko, P. B.
Hitchcock and M. F. Lappert, J. Chem. Soc., Dalton Trans., 2000,
4093; V. Amo, R. Andre´s, E. de Jesu´s, F. J. de la Mata, J. C. Flores,
R. Go´mez, M. P. Go´mez-Sal and J. F. C. Turner, Organometallics,
2005, 24, 2331.
19 J. Grundy, M. P. Coles and P. B. Hitchcock, Dalton Trans., 2003,
2573.
20 T. A. Neely, W. W. Schwarz, H. W. Vaughan, Jr., Organic Syntheses,
coll. vol. 5, p 1116.
21 G. M. Sheldrick, SHELXL-97, Program for the Refinement of Crystal
Structures, 1997.
22 S. H. Oakley, D. B. Soria, M. P. Coles and P. B. Hitchcock, Dalton
Trans., 2004, 537.
23 M. P. Coles and P. B. Hitchcock, Eur. J. Inorg. Chem., 2004, 2662.
24 R. E. Mulvey, K. Wade, D. R. Armstrong, G. T. Walker, R. Snaith,
W. Clegg and D. Reed, Polyhedron, 1987, 6, 987.
25 The observed inequivalence of the nitrogen substituents may be
=
explained by other situations, including, for example, localised C
N
double and C–N single bonds or monodentate N-coordination to
lithium. From examination of the space filling models generated from
crystal structure data we feel that restricted rotation about the Ph2P–
C bond is a reasonable explanation.
26 J. Barker, D. Barr, N. D. R. Barnett, W. Clegg, I. Cragg-Hine, M. G.
Davidson, R. P. Davies, S. M. Hodgson, J. A. K. Howard, M. Kilner,
C. W. Lehmann, I. Lopez-Solera, R. E. Mulvey, P. R. Raithby and
R. Snaith, J. Chem. Soc., Dalton Trans., 1997, 951.
27 J. R. Hagadorn and J. Arnold, Inorg. Chem., 1997, 36, 132; F. A.
Cotton, S. C. Haefner, J. H. Matonic, X. Wang and C. A. Murillo,
Polyhedron, 1997, 16, 541; F. A. Cotton, L. M. Daniels, J. H. Matonic
and C. A. Murillo, Inorg. Chim. Acta, 1997, 256, 277; A. R. Sadique,
M. J. Heeg and C. H. Winter, Inorg. Chem., 2001, 40, 6349; C. L.
Boyd, B. R. Tyrrell and P. Mountford, Acta Crystallogr., Sect. E:
Struct. Rep. Online, 2002, 58, m597.
8 R. P. Davies, D. J. Linton, P. Schooler, R. Snaith and A. E. H.
Wheatley, Eur. J. Inorg. Chem., 2001, 619.
9 C. N. Rowley, G. A. DiLabio and S. T. Barry, Inorg. Chem., 2005, 44,
1983.
10 A. P. Kenney, G. P. A. Yap, D. S. Richeson and S. T. Barry, Inorg.
28 S. R. Foley, G. P. A. Yap and D. S. Richeson, Inorg. Chem., 2002, 41,
4149.
Chem., 2005, 44, 2926.
11 M. P. Coles and P. B. Hitchcock, J. Chem. Soc., Dalton Trans., 2001,
1169; S. R. Foley, G. P. A. Yapp and D. S. Richeson, Polyhedron,
2002, 21, 619; M. P. Coles and P. B. Hitchcock, Organometallics, 2003,
22, 5201; D. B. Soria, J. Grundy, M. P. Coles and P. B. Hitchcock,
Polyhedron, 2003, 22, 2731; M. P. Coles and P. B. Hitchcock, Inorg.
Chim. Acta, 2004, 357, 4330; S. J. Birch, S. R. Boss, S. C. Cole, M.P.
Coles, R. Haigh, P. B. Hitchcock and A. E. H. Wheatley, Dalton
Trans., 2004, 3568; D. B. Soria, J. Grundy, M. P. Coles and P. B.
Hitchcock, J. Organomet. Chem., 2005, 690, 2278.
29 D. A. Wierda and A. R. Barron, Polyhedron, 1989, 8, 831.
30 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A.
Robb, J. R. Cheeseman, J. A. Montgomery, Jr.., T. Vreven, K. N.
Kudin, J. C. Burant, J. M. Millam, S. S. Iyengar, J. Tomasi, V. Barone,
B. Mennucci, M. Cossi, G. Scalmani, N. Rega, G. A. Petersson, H.
Nakatsuji, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa,
M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, M. Klene,
X. Li, J. E. Knox, H. P. Hratchian, J. B. Cross, C. Adamo, J.
Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin,
R. Cammi, C. Pomelli, J. W. Ochterski, P. Y. Ayala, K. Morokuma,
G. A. Voth, P. Salvador, J. J. Dannenberg, V. G. Zakrzewski, S.
Dapprich, A. D. Daniels, M. C. Strain, O. Farkas, D. K. Malick,
A. D. Rabuck, K. Raghavachari, J. B. Foresman, J. V. Ortiz, Q. Cui,
A. G. Baboul, S. Clifford, J. Cioslowski, B. B. Stefanov, G. Liu,
A. Liashenko, P. Piskorz, I. Komaromi, R. L. Martin, D. J. Fox,
K. T., M. A. Al-Laham, C. Y. Peng, A. Nanayakkara, M. Challa-
combe, P. M. W. Gill, B. Johnson, W. Chen, M. W. Wong, C. Gon-
zalez, and J. A. Pople, in ‘Gaussian 03, Revision C. 02’, Wallingford,
2004.
12 An example of silyl migration from phosphorus to nitrogen, resulting
=
in formation of a P C double bond has been reported in solution for
ꢀ
R P C{N(SiMe3)R}2 (from 31P NMR data, d ∼ 120). See reference
=
14).
13 ‘Applied Homogeneous Catalysis with Organometallic Compounds:
A Comprehensive Handbook in Two Volumes’, ed. B. Cornnils and
W. A. Hermann, VCH Publishers, 1996.
14 K. Issleib, H. Schmidt and H. Meyer, J. Organomet. Chem., 1980,
192, 33.
15 D. H. M. W. Thewissen and H. P. M. M. Ambrosius, Rec. Trav. Chim.
Pays-Bas, 1980, 99, 344.
D a l t o n T r a n s . , 2 0 0 5 , 2 8 3 3 – 2 8 4 1
2 8 4 1