Formation of a new class of 7π radicals via sterically induced P-P bond cleavage of the dimers [(CH)2(NR)2P]2

Article abstract of DOI:10.1039/b821241d

The 2c-2e^- P-P bonded dimers [(CH)₂(NR) ₂P]₂ dissociate in solution to give the persistent new 7π radicals [(CH)₂(NR)₂P]^?, which are isoelectronic with the well known S/N thiazol

Full text of DOI:10.1039/b821241d

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Formation of a new class of 7p radicals via sterically induced P–P bond
cleavage of the dimers [(CH)2(NR)2P]2w
Ruth Edge,a Robert J. Less,*b Eric J. L. McInnes,a Kristine Mu
¨
ther,b Vesal Naseri,b
Jeremy M. Rawsonb and Dominic S. Wright*b
Received (in Cambridge, UK) 27th November 2008, Accepted 21st January 2009
First published as an Advance Article on the web 13th February 2009
DOI: 10.1039/b821241d
The 2c–2eꢀ P–P bonded dimers [(CH)2(NR)2P]2 dissociate in
solution to give the persistent new 7p radicals [(CH)2(NR)2P]ꢁ,
which are isoelectronic with the well known S/N thiazolyl
radicals.
structure of the free radical in the vapour phase has been
determined by electron diffraction methods.9,10 Structural
studies have suggested that dissociation of the group 15 dimers
R2EER2 (E = P, As) is largely driven by release of steric strain
between the bulky silyl substituents, even in cases where a
short (presumably strong) E–E bond is present.
The pursuit of stable molecules containing unpaired electrons
has been of interest to researchers for many years, and is at the
forefront of the development of new magnetic materials.1
Although the existence of main group radicals has been known
since the 1800s, it has only been since the 1960s that persistent
compounds of the heavier group 14 and 15 elements contain-
ing unpaired electrons have been isolated and characterised.2
Two main strategies have been employed to stabilise main
group radicals. One approach has been by the use of
p-delocalisation of the electron over several centres, combined
with the incorporation of electronegative elements, to form
unreactive, low-energy singly-occupied molecular orbitals
(SOMO) as seen in the extensive classes of 7p S/N and Se/N
radicals.3 In addition, kinetic stabilisation has been used in the
form of highly bulky steric groups to shield the reactive radical
centre against attack by other molecules or from forming
dimers.2 In fact, it is this tendency for formation of covalently
bonded spin-paired dimers that often represents the biggest
obstacle to the isolation of radicals.1
Our recent report of the synthesis of the s-bonded Sb–Sb
bonded anion [1,2-C6H4P2Sb]2
4ꢀ
(A),11 which is valence-
isoelectronic with the p-bonded benzo-1,3-dithia-2-azolyl
radical dimer (B),12 suggested that a broad class of new
radicals may be obtainable involving valence-isoelectronic
combinations of various main group elements. We show in
this paper that a new family of P-based 7p radicals is readily
accessible via low-energy P–P bond cleavage, driven by a
combination of steric stabilisation and electronic delocalisa-
tion within their C2N2P frameworks. The key precursors to the
new 7p radicals are the P–P bonded compounds 1 obtained in
crystalline form by reduction of the known p-chloro-diazaphos-
pholenes (CH)2(NR)2PCl13 with Mg powder in thf (Fig. 1).z
Importantly, the stability and reactivity of compounds of this
type have not been investigated previously, although the high
reactivity of P–P bonds in the unsymmetric species
(CH)2(NR)2P–PR02 has been noted.14 Significantly, however,
the potential intermediacy of [(CH2)2(NR)2P]ꢁ radicals
appears to have been overlooked in this case.
The chemistry of persistent radicals of the heavier elements
of group 15 (notably phosphorus), although less well devel-
oped than those based on group 16, is characterised by a high
degree of structural and chemical diversity, including radical
centres based on both P(III) and P(V).2,4 These include exam-
ples of pure s-radicals, where the unpaired electron is localised
in an orbital on a phosphorus centre, as well as p-based
systems where spin density is extended over several nuclei.
Several classes of phosphorus-based radicals have been
studied, including phosphinyls [R2P]ꢁ,5 diphosphanyls
[R2PPR]ꢁ,6 1,3-diphosphaallyls [(RP)2CR]ꢁ,7 and phos-
phoranyls [R4P]ꢁ.8 By far the most studied of these are the
phosphinyl and related arsinyl systems, and in a few cases the
All three compounds 1a–c exist as trans-oid dimers in the
solid state (Fig. 2) in order to minimise steric repulsion
between the bulky organic groups.y A large increase in the
P–P bond length is observed on moving from 2.2439(8) A in 1a
t
(R = Bu) to 2.324(2) A in 1b [R = Mes (mesityl)], to mean
2.33 A in 1c [R = Dipp (2,6-diisopropylphenyl)]; cf. 2.21 A
a School of Chemistry, University of Manchester, Oxford Road,
Manchester, UK M13 9PL
b Chemistry Department, University of Cambridge, Lensfield Road,
Cambridge, UK CB2 1EW. E-mail: rjl1003@cam.ac.uk.
E-mail: dsw1000@cam.ac.uk
w Electronic supplementary information (ESI) available: EPR spectra
for 1b and 1c; calculated bond lengths, angles and SOMO of the model
heterocyclic C2H2(NMe)2Pꢁ system by DFT at increasing levels.
CCDC reference numbers 707574 (1a), 707575 (1b) and 707576 (1c).
For ESI and crystallographic data in CIF or other electronic format
see DOI: 10.1039/b821241d
Fig. 1 Dissociation of diazaphospholene dimers into monomeric 7p
radicals. 1a (R = Bu), 1b (R = Mes), 1c (R = Dipp).
t
ꢂc
This journal is The Royal Society of Chemistry 2009
Chem. Commun., 2009, 1691–1693 | 1691

View Article Online
Fig. 2 Solid-state structures of 1a, 1b and 1c. Thermal ellipsoids are drawn at the 50% level. H-atoms have been omitted for clarity. Selected bond
lengths (A) and angles (1): 1a P(1)–P(1A) 2.2439(8), P(1)–N(1) 1.7388(11), P(1)–N(2) 1.7394(8), N(1)–C(1) 1.4217(10), N(2)–C(2) 1.4144(11), C(1)–C(2)
1.3383(13); N(1)–P(1)–N(2) 94.28(4), N(1)–P(1)–P(1A) 94.33(4), N(2)–P(1)–P(1A) 94.24(4), C(1)–N(1)–P(1) 106.31(6), C(2)–N(2)–P(1) 106.59(6),
C(2)–C(1)–N(1) 115.22(7), C(1)–C(2)–N(2) 115.34(7); 1b P(1)–P(1A) 2.3240(16), P(1)–N(1) 1.728(2), P(1)–N(2) 1.727(2), N(1)–C(1) 1.412(3), N(2)–C(2)
1.411(4), C(1)–C(2) 1.324(4); N(2)–P(1)–N(1) 89.76(11), N(2)–P(1)–P(1A) 97.44(9), N(1)–P(1)–P(1A) 97.95(9), C(1)–N(1)–P(1) 109.59(18),
C(2)–N(2)–P(1) 110.02(19), C(2)–C(1)–N(1) 113.6(3), C(1)–C(2)–N(2) 112.9(3); 1c (mean distances and angles) P–P 2.33, P–N 1.73, N–C 1.43, C–C
1.33; N–P–N 90.91, N–P–P 96.27, 114.70, C–N–P 109.88, C–C–N 113.79. Symmetry transformations used to generate equivalent atoms: 1a: ꢀx + 1,
ꢀy + 1, ꢀz + 1, 1b: ꢀx, ꢀy, ꢀz, for 1c only one of the molecules in the ASU is shown and disordered lattice solvent was omitted for clarity.
mean value observed for P–P single bonds in the literature.15
These P–P bond lengths represent relatively long P–P bonds,
although are exceeded by those present in the highly polarised,
unsymmetrical systems reported by Gudat et al. [up to
2.484(1) A] which utilised the same heterocyclic C2N2P ring
systems.14 In addition to the lengthening of the P–P bond,
other structural distortions arise in order to minimise the steric
clash between the (CH)2(NR)2 ring units. Most obvious are
the puckering of the (CH)2(NR)2 ring units, and the bending
of their organic substituents out of the heterocyclic ring plane
and the twisting of the P–P bond in the most sterically
congested dimer 1c (ca. 1111) away from the centrosymmetric
arrangements observed in 1a and 1b in order to accommodate
the very large Dipp groups.
[A(14N) = 5.8 G] nuclei. Couplings to phosphorus are lower
and to nitrogen higher than observed in non-cyclic phos-
phinyls [typically A(31P) = 63–108 G, A(14N) = 3.7–5.1 G],1
corresponding to a delocalisation of the unpaired electron
about the p-system of the heterocyclic framework rather than
being overwhelmingly P-based. No 1H coupling was observed,
either to the organic substituents (tBu) or to heterocyclic C–H
atoms. The longer, weaker P–P bond in 1b permitted a similar
spectrum to be observed at the lower temperature of 315 K
[g = 2.01775, A(31P) = 40 G, A(14N) = 5.2 G] whilst 1c,
having the bulkiest steric groups and longest P–P bond,
showed significant dissociation into monomeric radicals
at room temperature, as shown by EPR spectroscopy [g =
2.0248, A(31P) = 42 G, A(14N) = 5.4 G, see ESIw]. It is
worthwhile to compare EPR spectra we have obtained with
those reported by West et al. for a series of structurally similar
heterocyclic radicals (CH)2(NtBu)2M–R, where the P atom in
1a has been substituted by M–R groups (M = Si or Ge).16
Seven-line EPR spectra with strong coupling to heterocyclic
C–1H nuclei were reported. However, anomalously no
coupling to 29Si was observed for the (CH)2(NtBu)2Si–R
radical, even though calculations predicted a large hyperfine
coupling to this nucleus.
As a result of the presence of P–P bonded arrangements,
1a–c are all EPR-silent in the solid state. No solution EPR
signal could be detected for 1a in toluene at room temperature.
However on increasing the temperature to 353 K the colour of
the solution changed from colourless to faint yellow, signalling
the shifting of the monomer–dimer equilibrium in favour
of the radical monomer form. The EPR spectrum of 1a recorded
at 353 K (Fig. 3) shows a doublet further split into 1 : 2 : 3 : 2 : 1
quintets (g = 2.00088), consistent with hyperfine coupling to
one phosphorus [A(31P) = 41 G] and two equivalent nitrogen
Density functional theory (DFT) calculations were under-
taken to gain insight into the nature of the frontier orbitals
and the dimerisation energy of these species.17 The model
heterocyclic (CH)2(NMe)2Pꢁ system was geometry optimised
at the double-zeta UB3LYP/6-31G, UB3LYP/6-31G* and
triple zeta UB3LYP/6-311G and UB3LYP/6-311G* levels in
order to assess the basis set dependence of the calculations.
Reassuringly, the nature of the frontier orbitals was not
affected on increasing the size of the basis set. The frontier
orbitals obtained at the 6-31G level for (CH)2(NAr)2Pꢁ (Ar =
2,6-Me2C6H3), models for 1b and 1c, are also similar to those
determined for (CH)2(NMe)2Pꢁ (Fig. 4). The SOMO show
that the majority of the unpaired electron density lies on the
heterocyclic C2N2P ring, broadly consistent with EPR studies
Fig. 3 EPR spectrum of 1a in toluene at 353 K.
1692 | Chem. Commun., 2009, 1691–1693
ꢂc
This journal is The Royal Society of Chemistry 2009

View Article Online
0.0135. 1c; Empirical formula C53H74N4O0.25P2, FW = 833.10, crystal
ꢀ
system triclinic, space group P1, a = 12.1056(1), b = 20.2299(2),
c = 21.6313(2) A, a = 72.598(1), b = 77.804(1), g = 81.880 (1)1, V =
4923.51(8) A3, Z = 4, rcalcd = 1.124 Mg mꢀ3, m(Mo-Ka) =
0.127 mmꢀ1, reflections collected 57 801, independent reflections
17 212 [Rint = 0.0337]. R1 = 0.0378, wR2 = 0.0956. Data were
Fig. 4 SOMO (left) and LUMO (right) of 1c.
collected on
(1a, 1b) and 150(2) K (1c), solved by direct methods and refined by
a Nonius KappaCCD diffractometer at 180(2) K
full-matrix least squares on F2.19
z Calculations by Oakley on the radical HCN2E2 (E = S, Se) have
ꢁ
shown that accurate estimates of dimerisation energy are not easily
obtained even using large basis sets or additional correlation func-
tions.18 In addition, these are gas-phase calculations which neglect
solvation effects.
1 J. M. Rawson, A. Alberola and A. Whalley, J. Mater. Chem., 2006,
16, 2560.
2 P. P. Power, Chem. Rev., 2003, 103, 78.
3 A. W. Cordes, R. C. Haddon and R. T. Oakley, in Chemistry of
Inorganic Ring Systems, ed. R. Steudel, Elsevier, Amsterdam, 1992,
ch. 16, pp. 295–232.
Fig.
5
(Left) NBO charges (bold) and bond orders in the
[(CH)2(NMe)2P]2 dimer, and (right) bonding interaction between
radicals in the HOMO.
4 A. Armstrong, T. Chivers and R. T. Boere, The Diversity of Stable
´
of 1a–c discussed previously. Calculations on the dimer
[(CH)2(NMe)2P]2 at the B3LYP/6-31G level revealed it to be
a minimum on the potential energy surface and showed a small
dimerisation energy (DE) of B3 kJ molꢀ1, perhaps lower than
expected considering the elevated temperatures required to
detect EPR signals. Accurate estimates of dimerisation ener-
gies of radicals are, however, difficult to obtainz (being highly
dependent on the basis set used) though the observation of a
stable dimer on the potential energy surface is significant.
Solution is viable at accessible temperatures. The calculated
parameters are consistent with a weakly s-bonded dimer
formed through a trans-cofacial p*–p* interaction equivalent
to a 2c–2eꢀ P–P bond involving p-orbitals (Fig. 5).
and Persistent Phosphorus-Containing Radicals, Modern Aspects
of Main Group Chemistry, ACS Symp. Ser., 2006, 917, 66.
5 M. J. S. Gynane, A. Hudson, M. F. Lappert and P. P. Power,
J. Chem. Soc., Chem. Commun., 1976, 623; M. J. S. Gynane,
A. Hudson, M. F. Lappert, P. P. Power and H. Goldwhite,
J. Chem. Soc., Dalton Trans., 1980, 2428.
6 S. Loss, A. Magistrato, L. Cataldo, S. Hoffmann, M. Geoffroy,
U. Rothlisberger and H. Grutzmacher, Angew. Chem., Int. Ed.,
2001, 40, 723.
¨
7 Y. Canac, A. Baceiredo, W. W. Schoeller, D. Gigmes and
G. Bertrand, J. Am. Chem. Soc., 1997, 119, 7579.
8 S. Marque, Y. Berchadsky, K. Lang, M. Moussavi, A. Fournel,
P. Bertrand, E. Belorizky and P. Tordo, J. Phys. Chem. A, 1997,
101, 5640.
9 S. L. Hinchley, C. A. Morrison, D. W. H. Rankin, C. L. B.
MacDonald, R. J. Wiacek, A. H. Voigt, A. H. Cowley,
M. F. Lappert, G. Gunderson, J. A. C. Clyburne and
P. P. Power, J. Am. Chem. Soc., 2001, 123, 9045.
10 J-P. Bezombes, K. B. Borisenko, P. B. Hitchcock, M. F. Lappert,
J. E. Nycz, D. W. H. Rankin and H. E. Robertson, Dalton Trans.,
2004, 1980.
In conclusion, applying simple isoelectronic principles has
allowed us to predict homolytic cleavage of the P–P bonds in
dimers of the type [(CH)2(NR)2P]2 and to establish a new class
of 7p radicals.
11 F. Garcıa, R. J. Less, V. Naseri, M. McPartlin, J. M. Rawson and
´
D. S. Wright, Angew. Chem., Int. Ed., 2007, 46, 7827.
12 E. G. Awere, N. Burford, R. C. Haddon, S. Parsons, J. Passmore,
J. V. Waszczak and P. S. White, Inorg. Chem., 1990, 29, 4821.
13 (a) D. Gudat, A. Haghverdi, H. Hupfer and M. Nieger,
Chem.–Eur. J., 2000, 6, 3414; (b) S. Burck, D. Gudat, M. Neiger
and W.-W. du Mont, J. Am. Chem. Soc., 2006, 128, 3946.
14 S. Burck, D. Gudat and M. Neiger, Angew. Chem., Int. Ed., 2004,
43, 4801.
15 Cambridge Crystallographic Database (ConQuest) 21 October
2008. I. J. Bruno, J. C. Cole, P. R. Edgington, M. Kessler,
C. F. Macrae, P. McCabe, J. Pearson and R. Taylor, Acta Crystallogr.,
B, 2002, 58, 389.
16 (a) B. Tumanskii, P. Pine, Y. Apeloig, N. J. Hill and R. West, J. Am.
Chem. Soc., 2004, 126, 7786; (b) B. Tumanskii, P. Pine, Y. Apeloig,
N. J. Hill and R. West, J. Am. Chem. Soc., 2005, 127, 8248.
17 The geometry optimisations were undertaken using DFT with a
B3LYP functional (A. D. Becke, Phys. Rev., 1988, A38, 3098;
A. D. Becke, J. Chem. Phys., 1993, 98, 5648) using GAMESS-UK
(M. F. Guest, I. J. Bush, H. J. J. van Dam, P. Sherwood, J. M. H.
Thomas, J. H. van Lenthe, R. W. A Havenith and J. Kendrick,
Mol. Phys., 2005, 103, 719). NBO analysis utilised a routine within
GAMESS-UK based upon Version 3.0 of the NBO program from
the Quantum Chemistry Program Exchange (No. 408, 1980)
(A. E. Reed, L. A. Curtiss and F. Weinhold, Chem. Rev., 1988,
88, 899).
Notes and references
z All manipulations were performed using Schlenk-line techniques
under an atmosphere of dry argon. Chloro-diazaphospholenes were
synthesised according to published procedures.13b The synthesis of 1a
is described below. Compounds 1b and 1c were obtained in crystalline
form using a similar procedure, but have only currently been obtained
in low yields. Optimised syntheses of both will be published in a full
paper on this work in due course. Synthesis of 1a, [1,10,3,30-tetra-tert-
butyl-1,10,3,30-tetrahydro-2-20-bi(1,3,2-diazaphosphole)]: Mg powder
(21 mg, 0.85 mmol) was added to a solution of 1,3-bis(tert-butyl)-2-
chloro-1,3,2-diazaphospholene (200 mg, 0.85 mmol) in 5 ml thf, and
the mixture stirred overnight at room temperature. The volume of
solvent was then reduced to approx. 1 ml and the mixture left to
crystallise at ꢀ20 1C for 1 d. The pale yellow crystals were filtered and
dried in vacuo. Yield 66 mg, 39%. 1H NMR (500 MHz, C6D6): d =
6.02 (t, 4H, CHQCH, JHꢀP = 1.9 Hz), 1.33 (s, 36H, CH3); 13C NMR
(500 MHz, C6D6): d = 121.19 (s, CQC), 54.15 (t, N–C–Me3, JCꢀP
=
7.7 Hz), 30.22 (t, CH3, JCꢀP = 3.9 Hz); 31P{1H} NMR (400 MHz,
C6D6) d = 80.8.
y Crystal data for 1a: Empirical formula C20H40N4P2, FW = 398.50,
crystal system triclinic, space group P1, a = 6.3138(1), b = 9.8033(2),
c = 10.3563(2) A, a = 73.7030(8), b = 72.8375(8), g = 87.2143(9)1,
V = 587.45(2) A3, Z = 1, rcalcd = 1.126 Mg mꢀ3, m(Mo-Ka) =
0.196 mmꢀ1, reflections collected 12 404, independent reflections 4070
[Rint = 0.022]. R1 = 0.0285, wR2 = 0.0802. 1b; Empirical formula
ꢀ
18 A. W. Cordes, C. D. Bryan, W. M. Davis, R. H. de Laat,
S. H. Glarum, J. D. Goddard, R. C. Haddon, R. G. Hicks,
D. K. Kennepohl, R. T. Oakley, S. R. Scott and N. P. C.
Westwood, J. Am. Chem. Soc., 1993, 115, 7232.
19 G. M. Sheldrick, SHELXL-97, Program for refinement of crystal
structures, University of Gottingen, Germany, 1997.
¨
ꢀ
C40H48N4P2, FW = 646.76, crystal system triclinic, space group P1,
a = 8.6059(3), b = 10.2485(4), c = 10.8287(4) A, a = 85.8725(16),
=
b = 74.5382(17), g = 73.0208(14)1, V = 880.36(6) A3, Z = 1, rcalcd
1.220 Mg mꢀ3, m(Mo-Ka) = 0.158 mmꢀ1, reflections collected 9748,
independent reflections 3565 [Rint = 0.067]. R1 = 0.063, wR2
=
ꢂc
This journal is The Royal Society of Chemistry 2009
Chem. Commun., 2009, 1691–1693 | 1693