An N,N′-chelated phosphenium cation supported by a β-diketiminate ligand

Article abstract of DOI:10.1039/b607549e

The first example of a phosphenium cation supported by N,N′-chelation of a β-diketiminate ligand has been prepared and structurally characterized. The Royal Society of Chemistry 2006.

Full text of DOI:10.1039/b607549e

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
An N,N9-chelated phosphenium cation supported by a b-diketiminate
ligand
Dragoslav Vidovic, Zheng Lu, Gregor Reeske, Jennifer A. Moore and Alan H. Cowley
Received (in CCAUS) 27th May 2006, Accepted 7th June 2006
First published as an Advance Article on the web 22nd June 2006
DOI: 10.1039/b607549e
The first example of a phosphenium cation supported by
N,N9-chelation of a b-diketiminate ligand has been prepared
and structurally characterized.
substituent from H to alkyl in order to disfavor phosphorus
substitution at this site. Accordingly, the b-diketiminate ligands
2 6
[HC(CMe)2(NC6F5)2]2 (L1
2 7
)
and [MeC(CMe)2(NDipp)2]2
(L2
)
were employed for the syntheses described herein.
N,N9-chelated b-diketiminate complexes are known for an
increasingly large number of transition metal, lanthanide and
main group elements.1 Several of these complexes have found
useful applications, particularly in the field of catalysis. In terms of
main group chemistry, b-diketiminate ligands have been used
advantageously for the stabilization of unusual oxidation states
and novel bonding arrangements.2 However, there is a gap in our
knowledge in this area of chemistry in the sense that there are no
examples of uncoordinated N,N9-chelated b-diketiminate com-
plexes3 of the electron-rich elements from groups 15, 16 and 17.
Previous attempts to prepare b-diketiminates that incorporate
chelated phosphorus-containing moieties resulted in the c-carbon-
substituted acyclic derivatives, 14 and 2.5
Treatment of L1H with NaNH2 in Et2O/hexane solution resulted
in the in situ formation of L1Na which, in turn, was allowed to
react with a hexane solution of PCl3. The initially formed product
of the latter reaction is the bis-imine derivative 3a. However, when
crystals of 3a are dissolved in e.g. C6D6 or CD2Cl2, the conversion
of 3a to the iminoamine isomer, 3b, is detectable within minutes by
NMR spectroscopy.8
At equilibrium, which is achieved in y1 day, the molar
composition of the 3a/3b mixture is approximately 2 : 3. The
structure of 3a was determined by single-crystal X-ray diffraction
(Fig. 1).9 That 3a exists in the diimine form is clear from the
presence of two short carbon–nitrogen bonds, C(2)–N(1) and
C(4)–N(2), two single carbon–carbon bonds, C(2)–C(3) and C(3)–
C(4), the tetrahedral geometry at C(3), and the trigonal planar
geometries at C(4) and C(5).10 X-Ray crystallographic data for
compounds 14 and 2,5 which are closely related to 3b, have been
published already.
It has been suggested that this outcome is due to steric
encumbrance by the bulky 2,6-diisopropylphenyl (Dipp) ligands
and the unfavorable bite angle of the six-membered b-diketiminate
ring with respect to phosphorus chelation.4a A further factor
disfavoring the chelation of electron-rich phosphorus-containing
fragments is the donor character of the flanking nitrogen atoms of
the b-diketiminate ligand. Taking the foregoing issues into
consideration, it occurred to us that the N,N9-chelated phos-
phorus-containing fragment needed to be as small and electro-
negative as possible. Cationic halophosphorus moieties, [PX]+,
seemed like excellent potential candidates. Two further choices
were made in an effort to optimize the conditions required for
N,N9-chelation, namely (i) reduction of the electron density on the
nitrogen atoms of the b-diketiminate ligand by the use of
electronegative substituents, and (ii) changing the c-carbon
Treatment of freshly crystallized 3a with trimethylsilyl triflate
(TMSOTf) in CH2Cl2 solution results in Me3SiCl elimination and
formation of the chlorophosphenium salt 4. The elemental com-
position of the cation was established as C17H7ClF10N2P on the
basis of HRMS (CH4, CI, positive mode) and the presence of the
H, CH3 and C6F5 ring substituents and the triflate anion was clear
from 1H and 19F NMR spectroscopic data.8 The 31P chemical shift
of d106.81isclosetothatofstructurallyauthenticated6(d97.23).8
Department of Chemistry and Biochemistry, The University of Texas at
Austin, 1 University Station A5300, Austin, Texas, 78712, USA.
E-mail: cowley@mail.utexas.edu; Fax: (+1) 512-471-6822;
Tel: (+1) 512-471-7484
This journal is ß The Royal Society of Chemistry 2006
Chem. Commun., 2006, 3501–3503 | 3501

View Article Online
Fig. 2 View of the cation of 6 showing the atom numbering scheme and
thermal ellipsoids at 40% probability and hydrogen atoms omitted for
˚
clarity. Selected bond distances (A) and angles (u): P(1)–Cl(1) 2.0923(11),
P(1)–N(1) 1.731(2), P(1)–N(2) 1.701(2), N(1)–C(1) 1.359(4), N(2)–C(3)
1.381(4), C(1)–C(2) 1.405(4), C(2)–C(3) 1.377(4); N(1)–P(1)–N(2)
97.51(12), N(1)–P(1)–Cl(1) 97.41(9), N(2)–P(1)–Cl(1) 102.65(9), P(1)–
N(1)–C(1) 123.96(19), P(1)–N(2)–C(3) 122.80(18), N(1)–C(1)–C(2)
121.4(2), N(2)–C(3)–C(2) 121.2(3), C(1)–C(2)–C(3) 121.7(3), C(1)–C(2)–
C(5) 119.0(2), C(3)–C(2)–C(5) 118.6(3).
Fig. 1 View of 3a showing the atom numbering scheme and thermal
ellipsoids at 40% probability with hydrogen atoms omitted for clarity.
˚
Selected bond distances (A) and angles (u): P(1)–C(3) 1.864(5), P(1)–Cl(1)
2.0747(15), P(1)–Cl(2) 2.0675(15), C(2)–C(3) 1.514(5), C(3)–C(4) 1.532(5),
C(2)–N(1) 1.284(5), C(4)–N(2) 1.281(5); P(1)–C(3)–C(2) 107.3(2), P(1)–
C(3)–C(4) 107.0(3), C(2)–C(3)–C(4) 111.5(3), C(1)–C(2)–C(3) 117.4(4),
C(1)–C(2)–N(1) 128.5(4), C(3)–C(2)–N(1) 114.0(4), C(3)–C(4)–C(5)
117.2(4), C(3)–C(4)–N(2) 115.0(4), C(5)–C(4)–N(2) 127.8(4).
As illustrated in Fig. 3, one of the salient aspects of the structure
…
of 3a relates to the short P(1) N(1) and P(1) N(2) contacts of
…
˚
2.690(4) and 2.894(4) A, respectively. Both of these distances are
appreciably shorter than the sum of van der Waals radii for P and
12
N (3.40 A). We believe that this structural feature may provide
Compound 5, which is formed via the reaction of L2H with
n-BuLi in hexanes, followed by treatment with PCl3, has not been
obtained in a pure state due to the concomitant formation of as yet
unidentified products, the quantities of which increase with time
(as monitored by 31P{1H} NMR spectroscopy). In order to
maximize the yield of the desired chlorophosphenium salt 6, it is
best to treat 5 with TMSOTf in CH2Cl2 solution as soon as the
reaction of L2Li with PCl3 is complete. After work-up of the
reaction mixture and recrystallization from pentane/CH2Cl2,
colorless, crystalline 6 was isolated in an overall yield of 54%
based on the consumption of L2H.
˚
an important clue regarding the mechanism by which 3a and 5 are
converted into 4 and 6, respectively. Phosphenium ion formation
can be envisaged as occurring by a concerted process in which
P(1)–N(1) and P(1)–N(2) bond formation is accompanied by
concomitant rupture of the P(1)–C(3) bond as a chloride anion is
abstracted from P(1). Alternatively, P(1)–N(1) formation and
P(1)–C(3) cleavage could occur in a first step, followed by rotation
of the C(4)–N(2) imine group around the C(3)–C(4) bond to effect
ring closure.
The X-ray crystal structure of 69 reveals an ensemble of
Density Functional Theory (DFT) calculations have been
performed on phosphenium ion 6 at the B3LYP level of theory
using the 6-31+G(d) basis set. The input parameters for the
geometry optimization were generated from the X-ray crystal-
lographic data set for 6. In general, the computed metrical
parameters lie within 2% of the experimental values. The HOMO
and LUMO for 6 are aryl ring bonding and b-diketiminate p*
in character, respectively, and the HOMO–LUMO gap is
chlorophosphenium cations and triflate anions and the closest
˚
cation–anion contact is 3.312(7) A between Cl(1) and a triflate
fluorine atom. The cyclic cation (Fig. 2) consists of a PCl+ moiety
which is N,N9-chelated by the b-diketiminate ligand, L2. The
metrical parameters for the six-membered PN2C3 ring indicate that
the geometry tends towards an alternating, somewhat localized
˚
distribution of electron density as reflected by the 0.2–0.3 A
differences in the C–C, C–N and N–P distances. The geometry of
the PN2C3 ring is non-planar (sum of bond angles = 708.6(2)u).
However, the atoms N(1), C(1), C(2), C(3) and N(2) are
approximately coplanar and P(1) deviates from the said plane by
˚
˚
0.549(3) A. The average N–P bond distance is 1.716(2) A. Due to
the presence of a stereochemically active lone pair at P(1), the
chloride ligand is approximately orthogonal to the PN2C3
ring. (The Cl(1)–P(1)–N(1) and Cl(1)–P(1)–N(2) angles are
97.41(9) and 102.65(9)u, respectively). Overall, the structure of
the cation of 6 bears a close resemblance to that of the neutral
germanium(II) b-diketiminate [HC(CMe)2(NDipp)2]GeCl with
which it is isovalent.11
…
Fig. 3 View of the skeleton of 3a showing the short N P contacts.
3502 | Chem. Commun., 2006, 3501–3503
This journal is ß The Royal Society of Chemistry 2006

View Article Online
5 P. B. Hitchcock, M. F. Lappert and J. E. Nycz, Chem. Commun., 2003,
1142.
6 A. Panda, M. Stender, R. J. Wright, M. M. Olmstead, P. Klavins and
P. P. Power, Inorg. Chem., 2002, 41, 3909.
7 The new ligand L2H was prepared via the reaction of 3-methyl-2,4-
pentanedione, DippNH2 and p-toluenesulfonic acid following the
protocol described in ref. 6.
8 Spectroscopic data: for 3a: 1H NMR (299.89 MHz, C6H6): d 1.44 (d, 6H,
2
4JPH = 0.9 Hz, Me), 4.04 (d, 1H, JPH = 4.5 Hz, CH); 19F NMR
3
(282.41 MHz, C6D6): d 2151.94 (m, 4F, o-F), 2161.62 (t, 2F, JFF
=
20.9 Hz, p-F), 2163.25 (m, 4F, m-F); 31P{1H} NMR (121.50 MHz,
C6D6): d 160.36 (s). HRMS (CI, CH4): calc. for C17H7Cl2F10N2P [(M +
H)+] 530.9642; found 530.9642. For 3b: 1H NMR (299.89 MHz, C6D6):
d 2.11 (d, 6H, JPH = 3.9 Hz, Me), 14.17 (s, 1H, NH); 19F NMR
4
(282.41 MHz, C6D6): d 2149.16 (m, 4F, o-F), 2158.04 (t, 2F, 3J =
20.9 Hz, p-F), 2162.26 (m, 4F, m-F); 31P{1H} NMR (121.50 MHz,
C6D6): d 160.93 (s). HRMS (CI, CH4): calc. for C17H7Cl2F10N2P [(M +
H)+] 530.9642; found 530.9642. For 4: 1H NMR (300.14 MHz, CH2Cl2):
d 2.64 (s, 6H, Me), 7.36 (s, 1H, CH); 19F NMR (282.41 MHz, CH2Cl2):
d 279.92 (s, 3F, OTf2), 2146.43 (m, 4F, o-F), 2158.08 (m, 2F, p-F),
2161.51 (m, 4F, m-F); 31P{1H} NMR (121.50 MHz, CH2Cl2): d 106.81
(s) HRMS (CI, CH4): calc. for C17H7ClF10P [(M + H)+] 494.9876; found
Fig. 4 The HOMO25 orbital of the cation of 6 showing the phosphorus
lone pair.
1
494.9845. For 6: H NMR (300.14 MHz, CD2Cl2): d 1.10 (d, 6H, J =
86.55 kcal mol21. As shown in Fig. 4, the orbital that features the
most phosphorus lone pair character is the HOMO25.
In summary, we have prepared, structurally characterized and
theoretically modeled the first example of a phosphenium ion
supported by N,N9-chelation of a b-diketiminate ligand. Diimine
isomers of acyclic, phosphorus–carbon bonded b-diketiminates
have also been isolated and structurally authenticated. These
diimine isomers play an important role as precursors to the cyclic
N,N9-chelated phosphenium ions.
6.6 Hz, CHMe2), 1.29 (d, 6H, J = 6.6 Hz, CHMe2), 1.32 (d, 6H, J =
6.6 Hz, CHMe2), 1.34 (d, 6H, J = 6.6 Hz, CHMe2), 2.43 (s, 6H, Me),
2.44 (s, 3H, Me), 2.71 (sept, 2H, J = 6.6 Hz, CHMe2), 2.90 (sept, 2H, J =
6.6 Hz, CHMe2), 7.41–7.54 (m, 6H, Ar); 19F NMR (282.41 MHz,
CD2Cl2): d 279.26 (s, 3F, OTf2); 31P{1H} NMR, 121.50 MHz):
d 97.23 (s).
9 Crystal data: for 3a; C17H7Cl2F10N2P, Mr = 531.12, orthorhombic,
˚
space group Pna21, a = 23.649(5), b = 6.050(5), c = 13.385(5) A, V =
3
1.914.8(18) A , Z = 4, T = 153(2) K, m = 0.526 mm21, reflections
˚
collected/independent = 6534/3808 (Rint = 0.0423), R1 (I > 2s(I)) =
0.0457 and wR2 (I > 2s(I)) = 0.0865. For 6: C31H43ClF3N2O3PS,
Mr = 647.16, monoclinic, space group P21/c, a = 14.742(3), b =
We are grateful to the Robert A. Welch Foundation and the
American Chemical Society Petroleum Research Fund (38970-
AC1) for financial support of this work.
3
˚
˚
12.911(3), c = 18.370(4) A, b = 110.87(3)u, V = 3267.0(14) A , Z = 4, T =
153(2) K, m = 0.281 mm21, reflections collected/independent = 13840/
7433 (Rint = 0.0474), R1 (I > 2s(I)) = 0.0633 and wR2 (I > 2s(I)) =
0.1554. Crystals of 3a and 6 were covered with a mineral oil prior to
mounting on the goniometer of a Nonius Kappa CCD diffractometer
equipped with an Oxford Cryostream liquid nitrogen cooling system.
Both data sets were corrected for adsorption. CCDC 603059 (3a) and
603060 (6). For crystallographic data in CIF or other electronic format
see DOI: 10.1039/b607549e.
Notes and references
1 For an excellent review, see: L. Bourget-Merle, M. F. Lappert and
J. R. Severn, Chem. Rev., 2002, 102, 3031.
2 See, for example: (a) H. W. Roesky and S. S. Kumar, Chem. Commun.,
2005, 4027; (b) N. J. Hardman and P. P. Power, Chem. Commun., 2001,
1184.
10 We have also prepared the PhPCl analogue of 3a by substituting
PhPCl2 for PCl3 in the synthetic procedure. The X-ray crystal
…
structure is very similar to that of 3a and also features short P
N
3 The sole example of
a coordinated phosphorus-containing
˚
contacts (2.633(9) and 3.054(10) A). Akin to 3a, the PhPCl
analogue also undergoes conversion to the corresponding iminoamine
isomer.
N,N9-chelated b-diketiminate is [P{N(H)C(C5Me5(C(H)C(H)NH}
{W(CO)5}2]. See: M. Schiffer and M. Scheer, Angew. Chem., Int. Ed.,
2001, 40, 3413.
11 Y. Ding, H. W. Roesky, M. Noltemeyer and H.-G. Schmidt,
Organometallics, 2001, 20, 1190.
12 Values taken from: A. Bondi, J. Phys. Chem., 1964, 68, 441.
4 (a) P. J. Ragogna, N. Burford, M. D’eon and R. McDonald, Chem.
Commun., 2003, 1052; (b) N. Burford, M. D’eon, P. J. Ragogna,
R. McDonald and M. J. Ferguson, Inorg. Chem., 2004, 43, 734.
This journal is ß The Royal Society of Chemistry 2006
Chem. Commun., 2006, 3501–3503 | 3503