A novel N,P,C cage complex formed by rearrangement of a tricyclic phosphirane complex: On the importance of non-covalent interactions

Article abstract of DOI:10.1002/chem.201305061

The reaction of Li/Cl P-CPh₃ phosphinidenoid tungsten(0) complex 2 with dimethylcyanamide afforded tricyclic phosphirane complex 4, an unprecedented rearrangement of which led to the novel N,P,C cage complex 6. On the basis of DFT calculations, formation and intramolecular [3+2] cycloaddition of the transient nitrilium phosphane ylide complex 3 to a phenyl ring of the triphenylmethyl substituent to give 4 is proposed. Furthermore, theoretical evidence for terminal N-amidinophosphinidene complex 7, formed by [2+1] cycloelimination from 4, is provided, and the role of the electronic structure and non-covalent interactions of intermediate 7 discussed.

Full text of DOI:10.1002/chem.201305061

DOI: 10.1002/chem.201305061
Full Paper
&
Strained Polycycles
A Novel N,P,C Cage Complex Formed by Rearrangement of
a Tricyclic Phosphirane Complex: On the Importance of
Non-covalent Interactions
Vitaly Nesterov,[a] Arturo Espinosa,*[b] Gregor Schnakenburg,[a] and Rainer Streubel*[a]
Dedicated to Professor Werner Uhl on the occasion of his 60th birthday
Abstract: The reaction of Li/Cl P-CPh3 phosphinidenoid
tungsten(0) complex 2 with dimethylcyanamide afforded tri-
cyclic phosphirane complex 4, an unprecedented rearrange-
ment of which led to the novel N,P,C cage complex 6. On
the basis of DFT calculations, formation and intramolecular
[3+2] cycloaddition of the transient nitrilium phosphane
ylide complex 3 to a phenyl ring of the triphenylmethyl sub-
stituent to give 4 is proposed. Furthermore, theoretical evi-
dence for terminal N-amidinophosphinidene complex 7,
formed by [2+1] cycloelimination from 4, is provided, and
the role of the electronic structure and non-covalent interac-
tions of intermediate 7 discussed.
Introduction
The chemistry of strained hydrocarbons[1] and their main-
group-element counterparts[2] continuously attracts attention,
as it often deals with fundamental questions of chemical struc-
ture, reactivity and bonding of a variety of unusual structural
motifs. Early milestones in P,C cage chemistry are represented
by tetraphosphacubane I[3] and hexaphosphapentaprismane
II[4] (Scheme 1), both obtained by phosphaalkyne cyclooligo-
merization.[5] Other synthetic strategies for P,C cage structures
include cycloaddition reactions of phosphaalkynes with cyclo-
butadienes[6] (III), thermolysis of 1,2,3-tris(pentamethylcyclo-
pentadienyl)cyclotriphosphene[7] (IV) and reactions of triphos-
phabenzene with phosphenium salts[8] (V). P,C cage structures
containing other heteroatoms directly bound to phosphorus,[9]
such as oxygen[10] and nitrogen,[11] are also known. In contrast
to P,C cages, for which ring expansions, rearrangements and
coordination chemistry have been reported, related N,P,C sys-
tems have not been studied intensively due to a lack of facile
synthetic protocols.
Scheme 1. Examples of P,C cage compounds.
bearing the P-pentamethylcyclopentadienyl group (Cp*) of
bridging[13] or terminal[14] transition metal phosphinidene com-
plexes, as well as on chemistry of P-Cp* oxaphosphirane com-
plexes.[15] In contrast, N,P,C cage formation involving transient
cycloaddition reactions of 1,3-dipoles such as nitrilium phos-
phane ylide complexes[16] is rare,[17] because the only estab-
lished route to these transient 1,3-dipole species relies on reac-
tions of thermally formed electrophilic terminal phosphinidene
complexes[18] with nitriles at elevated temperatures.[19,20] How-
ever, some evidence for a transient 1,3-dipole was obtained re-
cently, through the formation of P-iminoyl phosphane com-
plexes in the reaction of a Li/Cl phosphinidenoid complex and
dimethylcyanamide;[21] the formation of a C-amino-substituted
2H-azaphosphirene complex pointed in the same direction.
Herein, a reaction of a Li/Cl P-trityl phosphinidenoid complex
(trityl=CPh3) with dimethylcyanamide is presented. It leads to
an unprecedented sequence of rearrangements, yielding a tricy-
clic phosphirane complex and a novel N,P,C cage complex in
A versatile approach to a broad range of cage structures
(e.g., VI)[12] is based on the chemistry of highly reactive ligands
[a] Dr. V. Nesterov, Dr. G. Schnakenburg, Prof. Dr. R. Streubel
Institut fꢀr Anorganische Chemie
der Rheinischen Friedrich-Wilhelms-Universitꢁt Bonn
Gerhard-Domagk-Strasse 1, 53121 Bonn (Germany)
Fax: (+49)228-739616
E-mail: r.streubel@uni-bonn.de
[b] Dr. A. Espinosa
Departamento de Quꢂmica Orgꢃnica
Facultad de Quꢂmica, Universidad de Murcia
Campus de Espinardo, 30100 Murcia (Spain)
E-mail: artuesp@um.es
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201305061.
Chem. Eur. J. 2014, 20, 1 – 8
1
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
a diastereoselective fashion. DFT studies support the proposed
reaction mechanisms and describe the nature of reactive inter-
mediates involved.
troscopy showed selective formation of complex 4 from phos-
phinidenoid complex 2, which started at about 08C and went
to completion within the following 45 min at ambient temper-
ature. Neither the (expected) formation of a 2H-azaphosphir-
ene complex (cf. ref. [21]) nor a nitrilium phosphane ylide com-
plex nor any other intermediate was observed.
Results and Discussion
Density functional calculations on the reaction of phosphini-
denoid complex 2 with dimethylcyanamide were performed
(for computational details, see the Supporting Information).
Firstly, potential ambiphilic behaviour of 2 was examined by
means of a stepwise approach of the nitrile carbon (path A) or
nitrogen atoms (path B) to the phosphorus centre of model
phosphinidenoid complex 2’ with a covalently P-bound lithium
atom instead of a fully solvated lithium cation (cf. ref. [21]; see
Supporting Information, Figure S1), for the sake of computa-
tional efficiency. This study pointed to a kinetic preference for
nucleophilic behaviour of phosphinidenoid complex 2’ (path A)
and subsequent ring closure to give 2H-azaphosphirene com-
plex 5·LiCl, which additionally results in the most exergonic
process.[23] However, an alternative electrophilic reactivity
(path B) affording a lithium chloride complex of the nitrilium
phosphane ylide (3·LiCl; Supporting Information, Figure S1)
cannot be entirely ruled out.
When P-CPh3 phosphinidenoid complex 2, prepared by chloro/
lithium exchange of dichlorophosphane complex 1 with tBuLi
in the presence of [12]crown-4 at ꢀ788C,[22] reacted in situ
with dimethylcyanamide, the novel tricyclic phosphane com-
plex 4 was selectively obtained in 45 min at ambient tempera-
ture (Scheme 2). Here, the intermediacy and intramolecular cy-
Since it is well known that 2H-azaphosphirene complexes do
not undergo ring opening at or below ambient temperature,[19]
the possibility of 5 serving as precursor for nitrilium phos-
phane ylide complex 3 was also checked. The reaction coordi-
nate for the elongation of the more compliant endocyclic PꢀN
bond[24] showed that this process results in dissociation
(Figure 2), that is, formation of dimethylcyanamide and the P-
Scheme 2. Reaction of phosphinidenoid complex 2 with Me2NCN.
cloaddition of the 1,3-dipole complex 3 can be inferred. Com-
plex 4 is unstable in solution at ambient temperature, but
could be isolated in good yields at ꢀ208C as a yellow solid
after filtration through Al2O3 and fully characterized. The mo-
lecular structure of 4 was confirmed by single crystal X-ray dif-
fraction studies (Figure 1).
In solution, 4 displayed a 31P chemical shift of ꢀ4.2 ppm
(1J(P,W)=292.0 Hz) and a set of four olefinic proton resonances
in the 1H NMR spectrum, which provided clear proof that trans-
formation of one phenyl group into a cyclohexadienyl motif
had occurred. Monitoring of the reaction by 31P{1H} NMR spec-
Figure 2. Calculated (COSMOTHF/B3LYP-D3/def2-SVPecp) reaction coordinates
for the stepwise elongation of the endocyclic PꢀC (black squares) or PꢀN
bond (grey circles) in azaphosphirene complex 5.
CPh3-substituted phosphinidene complex, after reaching an
energy barrier of about 34 kcalmolꢀ1 (at dPꢀN ꢁ2.73 ꢁ), whereas
elongation of the PꢀC bond affords the comparatively more
stable nitrilium phosphane ylide complex 3 in a a kinetically
more favoured process.[25] Possible ring-bond activation
through N complexation by naked lithium chloride (5·LiCl) was
ruled out because of the required higher TS energy.[26]
Figure 1. Molecular structure of complex 4 (50% probability level; hydrogen
atoms are omitted for clarity). Selected bond lengths [ꢁ] and angles [8]: PꢀW
2.4693(5), PꢀC8 1.8569(19), PꢀC7 1.8697(19), PꢀN1 1.6691(16), C1ꢀC2
1.534(3), C2ꢀC7 1.532(3), C5ꢀC6 1.341(3), C3ꢀC4 1.332(3); C7-P-C8 49.47(8),
N1-P-C7 97.40(8), C8-P-N1 103.31(8).
&
&
Chem. Eur. J. 2014, 20, 1 – 8
www.chemeurj.org
2
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
Remarkably, the calculated structure of 3 has the nitrilium
phosphane ylide unit almost parallel to one of the phenyl
groups (Supporting Information, Figure S2) with both N atoms
roughly facing the ipso (dCꢀN =2.921 ꢁ) and meta (dCꢀN
=
3.648 ꢁ) positions of the ring, which seems to entail activation
of the phenyl substituent.[27] This enables an intramolecular
(formal) [3+2] dipolar cycloaddition to one phenyl group[28] in
an exothermic and low-barrier process (Figure 3); early forma-
Scheme 3. Thermal rearrangement of 4 to 6.
Figure 3. Calculated (COSMOTHF/B3LYP-D3/def2-TZVPPecp//COSMOTHF/B3LYP-
D3/def2-SVPecp) reaction profile for the conversion 5!4.
tion of the CꢀC bond can be deduced from the TS(3!4)
structure and associated imaginary frequency (see Supporting
Information). The alternative “direct” pathway 5!4 without
the intermediacy of the nitrilium phosphane ylide complex 3
appeared to be significantly disfavoured due to its rather high
Figure 4. Molecular structure of complex 6[30] (50% probability level; hydro-
gen atoms are omitted for clarity). Selected bond lengths [ꢁ] and angles [8]:
PꢀW 2.4603(14), PꢀC6 1.830(5), PꢀC5 1.861(5), PꢀN1 1.648(5), C6ꢀC5
1.556(7), C5ꢀC4 1.461(8), C4ꢀC3 1.328(8), C3ꢀC2 1.516(7), C2ꢀC7 1.520(7),
C7ꢀC8 1.354(7), C6-P-C5 49.9(2), N1-P-C6 107.3(2), N1-P-C5 110.4(2).
z
barrier (DEZPE =45.63 kcalmolꢀ1, DG° =45.25 kcalmolꢀ1).
Rearrangement of complex 4
give the heteropolycyclic ligand structure in 6 by stereoselec-
tive [2+1] cycloaddition.
Selectively formed 4 further transformed spontaneously in the
reaction solution at ambient temperature (ca. 5 d) into three
new products (ratio ca. 2:1.8:0.6) having the following 31P NMR
According to DFT calculations, on changing the solvation
model from THF to toluene (as in the experiment), the phos-
phirane ring in 4 is cleaved in a [1+2] cycloelimination-type
data: d=ꢀ121.2 (sSat,1J(P,W)=304.2 Hz), 39.6 (dSat
,
1J(P,W)=
z
process after reaching a moderate energetic barrier (DEZPE =
268.3 Hz, 1J(P,H)=315.8 Hz) and 198.9 (sSat
,
J(P,W)=176.5,
J(P,W)=130.5 Hz). Heating a solution of isolated pure 4 in tolu-
ene (758C, 1 h) allowed selective formation of the product
with a 31P NMR shift of ꢀ121.2 ppm (1J(P,W)=304.2 Hz), which
was assigned as 6 (Scheme 3). Complex 6 was isolated in good
yield as a light yellow solid after removing the solvent and
washing with n-pentane.
26.68 kcalmolꢀ1), affording relatively stable intermediate termi-
nal phosphinidene complex 7 (Figure 5). It seems likely that
the higher energy barrier of TS(4!7)[31] compared to TS(5!3)
(Figure 3) allows the reaction to be stopped when performed
in THF at low temperature. Finally, the [2+1] cycloaddition of
the phosphinidene P atom to the double bond (central) of
triene moiety of 7 leads to N,P,C cage complex 6.
The surprisingly low energy of complex 7 raised interest, es-
pecially because electrophilic, neutral, terminal phosphinidene
complexes with p-acidic co-ligands[32] are unstable in general.
In this case, 7 receives strong stabilization from 1) through-
bond donation of electron density into the (formally) vacant p
orbital of the phosphorus atom and 2) back-donation from the
electron-enriched phosphorus atom into the adjacent amido
The X-ray structure of complex 6 confirmed diastereoselec-
tive formation of a novel N,P,C cage ligand (Figure 4), which
displayed shorter PꢀC bonds of the phosphirane unit than
complex 4 (PꢀC6 1.830(5) and PꢀC5 1.861(5) ꢁ). As it is known
that phosphirane complexes can thermally form terminal phos-
phinidene complexes (by alkene elimination),[29] we assumed
that the rearrangement of 4 to 6 might proceed through for-
mation of transient phosphinidene complex 7, which could
Chem. Eur. J. 2014, 20, 1 – 8
www.chemeurj.org
3
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
from the phosphorus atom of the P!p*(C=C) (in the NBO for-
malism, the transfer takes place to Rydberg-type orbitals at
both carbon atoms) and P!HꢀCPh types (in the NBO formal-
ism, transfer from P to Rydberg-type orbitals at ortho H and C
atoms via the central amidine C atom).[38] NBO deletion calcula-
tions[39] provided further evidence for the above-mentioned
main stabilizing interactions, the associated binding energies
!
of which were: CN!P donation, 82.2 kcalmolꢀ1; CN P back-
donation, 65.1 kcalmolꢀ1; through-space P!p*(C=C) transfer,
68.1 kcalmolꢀ1
;
through-space P!HꢀCPh transfer, 64.9 kcal
molꢀ1
.
Bonding and reactivity trends in phosphinidene complexes
Although electrophilic terminal phosphinidene complexes
have been the subject of various detailed studies,[18] the unex-
pected nucleophilic interaction in complex 7 suggested a com-
parative study of different model phosphinidene complexes
8a–i (Scheme 4) bearing a representative set of typical elec-
Figure 5. Calculated (COSMOToluene/B3LYP-D3/def2-TZVPPecp//COSMOToluene
B3LYP-D3/def2-SVPecp) reaction profile for the conversion 4!6.
/
group (through-bond back-donation). The net through-bond
transfer to the P atom is represented in the HOMO showing
extended conjugation along the NꢀC=NꢀP fragment (Support-
ing Information, Figure S4). In addition, two different non-cova-
lent (through-space) interactions were identified: 1) a side-on-
type interaction of the p* orbital of the central C=C bond (in
the triene unit) with the phosphorus atom, understood in
terms of a HOMO!LUMO+1 electron transfer (Supporting In-
formation, Figure S4) and 2) an interaction between an adja-
cent phenyl ortho H atom and the amidine-type carbon atom
(formally a C···HꢀCPh hydrogen bond).[33] These two non-cova-
lent interactions (NCIs) are fully supported by the location of
the corresponding bond critical points (BCPs),[34] and are con-
veniently visualized by colour-coded reduced density gradient
(RDG) isosurfaces by using the NCIplot[35] technique
(Figure 6).[36]
Scheme 4. Representative model phosphinidene complexes.
tron-withdrawing (8a–c) or electron-releasing (8e–i) groups at
phosphorus. From a previous study on uncomplexed phosphi-
nidenes,[40] it was concluded that the best stabilizing groups
are those having a N, P or S atom attached to phosphorus:
they have an electron lone pair which forms a dative p bond
with the empty orbital of the singlet phosphinidene. By com-
parison of relative stabilities derived from isodesmic reactions,
it was also shown that imino (H2C=Nꢀ) units exhibit even
higher stabilization than the parent amino group. In agree-
ment with these predictions on unligated phosphinidenes,
complex 8 f displays a planar geometry at N, and imino sub-
stituents in 8g–i show a linear geometry of the N atom to
According to natural bond orbital (NBO)[37] calculations on
complex 7, the (additional) through-space electron transfer
!
maximize dative p(N!P) bonding and N P back-bonding.
From the point of view of frontier molecular orbital (FMO)
theory,[41] typical C-substituted phosphinidene complexes (8a–
e) have rather low LUMOs (Supporting Information, Figure S5)
with small HOMO–LUMO gaps (Figure 7a) that account for
their characteristic electrophilic behaviour. On the contrary, N-
substituted complexes (8 f–i) feature large HOMO–LUMO gaps
as a consequence of the prevalence of their high-energy
LUMOs, which account for their reduced electrophilicity, pre-
sumably due to their p*(PꢀN) character (see Supporting Infor-
mation, Figure S6) arising from the antibonding combination
of a destabilized vacant 3p orbital at P with the nitrogen 2p
orbital.
Figure 6. Calculated (COSMOTHF/B3LYP-D3/def2-SVPecp) most stable struc-
ture for complex 7 with NCIplot highlighting key stabilizing NCIs, the two
most significant of which are explicitly indicated. The RDG s=0.2 a.u. isosur-
face is coloured over the range ꢀ0.2 <sign(l2)1<0.2 a.u.: blue denotes
strong attraction, green stands for moderate interaction and red indicates
strong repulsion.
C-Amino-substituted imino groups (N-amidino groups) in
8h,i have an additional effective electron transfer (mesomeric
+M effect) to phosphorus that rises the energy of the HOMO
&
&
Chem. Eur. J. 2014, 20, 1 – 8
www.chemeurj.org
4
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
on a MAT 95 XL Thermo Finnigan
spectrometer (selected data given).
Infrared spectra were recorded on
a Thermo Nicolet 380 FT-IR spec-
trometer (selected data given).
Synthesis of complex 4
A solution of tBuLi (1.6m, 0.7 mL,
1.12 mmol) was slowly added to
a stirred solution of phosphane
complex 1 (669.0 mg, 1 mmol) and
[12]crown-4 (0.175 mL, 1.12 mmol)
in THF (12.5 mL) at ꢀ788C. After
stirring for 20 min at ꢀ788C dime-
thylcyanamide (90 mL, 1.1 mmol)
was added. The mixture was
slowly warmed up to ꢀ208C (ca.
1 h), the cooling bath was re-
moved and the reaction mixture
was stirred further for about 1 h at
248C until Li/Cl phosphinidenoid
complex
2
was
consumed
(31P NMR monitoring). Volatile sub-
stances were removed in vacuo
(0.02 mbar). The residue was dis-
solved in cold dichloromethane
Figure 7. Trends of selected reactivity descriptors of terminal phosphinidene complexes. a) HOMO–LUMO gap [eV]
and Lçwdin charge at P [e]. b) Electron affinity [eV] and local philicity at P [kcalmolꢀ1].
(ꢀ208C) and the resulting solution was filtered through Al2O3 at
ꢀ208C and washed with about 100 mL of dichloromethane. Evapo-
ration of the solvent (0.02 mbar) gave 4 as yellow solid, m.p.
1608C. Yield: 543 mg (0.81 mmol, 81%). 1H NMR (CDCl3, ꢀ408C):
d=7.6 (m, 2H; Ph), 7.13–7.36 (m, 8H; Ph), 6.2 (m, 1H; H’C=CH’),
6.1 (m, 1H; HC=CH), 5.9 (m, 1H; H’C=CH’), 5.47 (m, 1H; HC=CH),
4.4 (brd, 4J(H,P)=27.1 Hz, 1H; CHP), 2.7 (s, 3H; CH3), 2.6 (s, 3H;
CH3); 13C{1H} NMR (CDCl3, 258C): d=199.1 (d, 2J(C,P)=35.9 Hz;
and therefore enables some nucleophilic reactivity to emerge.
This phenomenon is also reflected in a decrease of the
(Lçwdin) electric charge (less positive) at the P atom (Fig-
ure 7a) and lowering of their electron affinity (Figure 7b),
which reaches a minimum for complex 8i. The local electrophi-
licity index[42] at P wPþ also points to the weakly electrophilic
character and thus higher nucleophilicity of the N-substituted
model phosphinidene complexes 8 f–i.
trans-CO), 194.9 (dSat
,
2J(C,P)=126.4 Hz, 1J (C,W)=8.5 Hz; cis-CO),
2
2
173.9 (d, J(C,P)=5.5 Hz; C=N), 138.3 (d, J(C,P)=5.1 Hz; Ph), 135.9
(s; Ph), 131.9 (s; Ph), 131.8 (s; Ph), 128.6 (s; Ph), 128.0 (s; Ph), 127.2
4
Conclusion
(s; Ph), 127.0 (d, J(C,P)=3.0 Hz; HC’=C’H), 126.6 (s; Ph), 125.0 (d,
4J(C,P)=2.8 Hz; HC’=C’H), 120.9 (d, 3J(C,P)=4.7 Hz; HC=CH), 117.5
(s; HC=CH), 59.4 (d, J(C,P)=15.1 Hz; Ph2CP), 50.5 (s; CH), 41.7 (d,
1J(C,P)=15.1 Hz; CP), 39.0 (s; NCH3), 38.9 (s; NCH3); 31P{1H} NMR
A facile transformation of a P-trityl group enabled highly dia-
stereoselective and high-yield formation of new polycyclic P-
ligand scaffolds. DFT studies provided strong evidence for an
unprecedented rearrangement sequence, which involves con-
secutively 1) 1,3-dipolar [3+2] cycloaddition of the nitrilium
phosphane ylide and a phenyl substituent, 2) [2+1] cycloelimi-
nation on the resulting phosphirane ring and 3) [2+1] cycload-
dition of a transient N-amidino-substituted phosphinidene
complex leading to the novel N,P,C cage ligand. The study also
1
1
~
(CDCl3, 258C): d=ꢀ4.2 (sSat, J(P,W)=292.0 Hz); IR (neat): n=850.0
(s), 1572.7 (s), 1902.5 (vs), 1926.6 (s), 2073 cmꢀ1 (m); MS (70 eV):
m/z (%): 668 (10) [M+]; elemental analysis calcd (%) for
C27H21N2O5PW: C 48.53, H 3.17, N 4.19; found: C 49.11, H 3.63, N
4.20.
Synthesis of complex 6
revealed that
a novel nucleophilic reactivity emerges if
Toluene (4 mL) was added to complex 4 (120 mg, 0.18 mmol) and
the resulting suspension was stirred for 2 h at 708C to give
a yellow solution. The solvent was evaporated (0.02 mbar) and the
resulting solid was washed with n-pentane to give 6 as a light
yellow solid, m.p. 1618C. Yield: 96 mg (0.14 mmol, 80%). 1H NMR
(CDCl3, 258C): d=7.16–7.40 (m, 10H; 2C6H5), 6.37 (m, 1H; HC=CH),
5.6 (m, 1H; HC=CH), 4.41 (m, 1H), 3.40 (m, 1H; CHP), 3.04 (s, 3H;
CH3), 2.66 (s, 3H; CH3), 2.51 (m, 1H; CHP); 13C{1H} NMR (CDCl3,
a strong electron donor such as an N-amidino group is bound
to the phosphinidene P centre, and hence new reactivity ave-
nues are opened for exploration.
Experimental Section
All manipulations involving air- and moisture-sensitive compounds
were carried out under an atmosphere of purified argon by using
standard Schlenk-line techniques or in a glovebox. Solvents were
dried with appropriate drying agents and degassed before use.
258C): d=199.74 (d, 2J(C,P)=36.2 Hz; trans-CO), 196.1 (dSat
,
1 2
2J(C,P)=125.9 Hz, J (C,W)=9.6 Hz; cis-CO), 163.32 (d, J=13.4 Hz;
C=N), 143.0 (s; C=CPh2), 142.2 (d, 4J(C,P)=1.0 Hz; Ph), 142.1 (d,
3
4J(C,P)=1.1 Hz; Ph), 135.67 (d, J(C,P)=3.3 Hz; C=CPh2), 129.98 (s;
1
The H, 13C{1H} and 31P{1H} NMR spectroscopic data were recorded
Ph), 129.76 (d, 5J(C,P)=1.3 Hz; Ph), 129.43 (s; Ph), 128.6 (s; Ph),
2
on a Bruker DMX 300 spectrometer. Mass spectra were recorded
128.5 (s; Ph), 127.2 (s; Ph), 127.2 (s; Ph), 126.8 (d, J(C,P)=5.5 Hz;
Chem. Eur. J. 2014, 20, 1 – 8
www.chemeurj.org
5
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
CH=CH), 124.22 (d, 2J(C,P)=4.3 Hz; C=CPh2), 118.70 (d, 3J(C,P)=
3.6 Hz; CH=CH), 40.37 (d, J(C,P)=11.3 Hz), 37.9 (s; NCH3), 37.4 (s;
NCH3), 35.27 (d, 1J(C,P)=31.1 Hz; CHP), 30.91 (d, 1J(C,P)=23.2 Hz;
opment in the chemistry of hexaphosphapentaprismanes, see: C. Hçhn,
I. Keller, L. Rohwer, F. W. Heinemann, U. Zenneck, Eur. J. Inorg. Chem.
2013, 5769–5780, and references therein.
3
CHP); 31P{1H} NMR (CDCl3, 258C): d=ꢀ121.2 (sSat
,
1J(P,W)ꢀ=1
[5] a) R. Streubel, Angew. Chem. 1995, 107, 478–480; Angew. Chem. Int. Ed.
Engl. 1995, 34, 436–438; b) J. M. Lynam, Organomet. Chem. 2007, 36,
170–178; c) M. Regitz, A. Hoffmann, U. Bergstrꢂbter, Phosphaalkynes -
Starting Point for the Synthesis of Phosphorus-Carbon Cage Compounds,
in Modern Acetylene Chemistry (Eds. P. J. Stang, F. Diederich), Wiley-VCH
Verlag GmbH, Weinheim, 2007, pp. 173–202.
~
304.2 Hz); IR (neat): n=699.1 (s), 1545.2 (m), 1913 (vs), 2069 cm
(m); MS (70 eV): m/z (%): 668 (9) [M+]; elemental analysis calcd (%)
for C27H21N2O5PW: C 48.53, H 3.17, N 4.19; found: C 49.09, H 3.42, N
3.93.
[6] J. Fink, W. Rçsch, U.-J. Vogelbacher, M. Regitz, Angew. Chem. 1986, 98,
265–266; Angew. Chem. Int. Ed. Engl. 1986, 25, 280–282.
[7] P. Jutzi, N. Brusdeilins, Z. Anorg. Allg. Chem. 1994, 620, 1375–1380.
[8] N. S. Townsend, S. R. Shadbolt, M. Green, C. A. Russell, Angew. Chem.
2013, 125, 3565–3568.
[9] For N,P,C cages such as 1,3,5-triaza-7-phosphaadamantane derivatives
having CH2 linkages between P and N, see: L. Gonsalvi, A. Guerriro, F.
Hapiot, D. A. Krogstad, E. Monflier, G. Reginato, M. Peruzzini, Pure Appl.
Chem. 2013, 85, 385–396.
[10] L. Abdrakhmanova, V. F. Mironov, T. P. Gryaznova, S. A. Katsuba, M. N. Di-
mikhametov, Phosphorus Sulfur Silicon Relat. Elem. 2011, 186, 652–656.
[11] J. Liedtke, S. Loss, G. Alcaraz, V. Gramlich, H. Grꢃtzmacher, Angew. Chem.
1999, 111, 1724–1727; Angew. Chem. Int. Ed. 1999, 38, 1623–1626.
[12] U. Rohde, F. Ruthe, P. G. Jones, R. Streubel, Angew. Chem. 1999, 111,
158–160; Angew. Chem. Int. Ed. 1999, 38, 215–217.
[13] a) M. Stubenhofer, C. Kuntz, M. Bodensteiner, A. Y. Timoshkin, M. Scheer,
Organometallics 2013, 32, 3521–3528; b) M. Stubenhofer, G. Lassandro,
G. Balꢄzs, A. Y. Timoshkin, M. Scheer, Chem. Commun. 2012, 48, 7262–
7264; c) M. Scheer, C. Kuntz, M. Stubenhofer, M. Zabel, A. Y. Timoshkin,
Angew. Chem. 2010, 122, 192–196; Angew. Chem. Int. Ed. 2010, 49,
188–192; d) M. Scheer, D. Himmel, B. P. Johnson, C. Kuntz, M. Schiffer,
Angew. Chem. 2007, 119, 4045–4049; Angew. Chem. Int. Ed. 2007, 46,
3971–3975.
[14] a) J. M. Villalba Franco, A. Espinosa, G. Schnakenburg, R. Streubel, Chem.
Commun. 2013, 49, 9648–9650; b) M. Bode, G. Schnakenburg, J. Dan-
iels, A. Marinetti, R. Streubel, Organometallics 2010, 29, 656–661; c) M.
Bode, G. Schnakenburg, J. Daniels, R. Streubel, Organometallics 2009,
28, 4636–4638.
Computational details
DFT calculations were performed with the ORCA program.[43] All
geometry optimizations were run in redundant internal coordi-
nates by using the B3LYP functional[44] together with the def2-SVP
basis set[45] (level A). For W atoms the [SD(60,MWB)] effective core
potential[46] (ECP) was used. The latest Grimme semiempirical
atom-pair-wise London dispersion correction (DFT-D3) was includ-
ed in all calculations.[47] Solvent effects (THF or toluene) were taken
into account through the COSMO solvation method.[48] Harmonic
frequency calculations verified the nature of ground states or tran-
sition states (TS) having all positive frequencies or only one imagi-
nary frequency, respectively. From these optimized geometries all
reported data were obtained by means of single-point (SP) calcula-
tions with the more flexible and polarized def2-TZVPP[49] basis set
(level B), which is the working level of theory throughout the text,
unless otherwise stated. Reported energies were corrected for the
zero-point energy (ZPE) term at the optimization level, except for
model solvated phosphinidenoid complex 2’·[12]crown-4·OEt2 and
model phosphinidene complexes 8. Wiberg bond indices (WBI)
were obtained from the natural bond orbital (NBO) population
analysis.[50] Charges from the Lçwdin population analysis[51] were
used for computing the Fukui indices, although Mulliken[52] charges
resulted in very similar variation along the series. Bader’s AIM-de-
rived topological analysis of the electron density[53] was conducted
with AIM2000[54] and by using the so-called wavefunctions (elec-
tron densities) generated with Gaussian 09,[55] whereas that com-
puted at the B3LYP/def2-SVP(ecp) level for 6 was used as input for
the NCIplot program. Figure 4 and Figure S3 (Supporting Informa-
tion) were obtained with VMD.[56]
[15] M. Bode, G. Schnakenburg, P. G. Jones, R. Streubel, Organometallics
2008, 27, 2664–2667.
[16] R. Streubel, Top. Curr. Chem. 2003, 223, 91–109.
[17] R. Streubel, U. Schiemann, N. Hoffmann, Y. Schiemann, P. G. Jones, D.
Gudat, Organometallics 2000, 19, 475–481.
[18] a) F. Mathey, Dalton Trans. 2007, 1861–1868; b) K. Lammertsma, Top.
Curr. Chem. 2003, 229, 95–119; c) F. Mathey, N. H. Tran Huy, A. Marinetti,
Helv. Chim. Acta 2001, 84, 2938–2957.
Acknowledgements
[19] R. Streubel, Coord. Chem. Rev. 2002, 227, 175–192.
[20] R. Streubel, U. Schiemann, P. G. Jones, N. H. Tran Huy, F. Mathey, Angew.
Chem. 2000, 112, 3845–3847; Angew. Chem. Int. Ed. 2000, 39, 3686–
3688.
[21] A. ꢅzbolat, G. von Frantzius, J. M. Pꢆrez, M. Nieger, R. Streubel, Angew.
Chem. 2007, 119, 9488–9491; Angew. Chem. Int. Ed. 2007, 46, 9327–
9330.
We are grateful to the Deutsche Forschungsgemeinschaft for
financial support (STR 411/26-1), SFB 813 “Chemistry at spin
centers” and European COST actions CM0802 (PhosSciNet) and
CM1302; G.S. thanks Prof. A. C. Filippou for support.
[22] V. Nesterov, G. Schnakenburg, A. Espinosa, R. Streubel, Inorg. Chem.
2012, 51, 12343–12349.
Keywords: cage compounds · noncovalent interactions · P
ligands · rearrangement · tungsten
[23] At the working level of theory, the conversion of 2 and Me2NꢀCꢂN to 5
and LiCl is markedly exergonic by 17.00 kcalmolꢀ1. At the non-ZPE-cor-
rected COSMOTHF/B3LYP-D3/def2-TZVPPecp//COSMOTHF/B3LYP-D3/def2-
SVPecp level, the more realistic model reaction between 2’·[12]crown-
4·OEt2 and Me2NꢀCꢂN leading to 5 and LiCl·[12]crown-4·OEt2 was
[1] Strained Hydrocarbons: Beyond the van’t Hoff and Le Bel Hypothesis.
(Ed.: H. Dodziuk), WILEY-VCH, Weinheim, 2009.
found to be endergonic by only 0.64 kcalmolꢀ1
.
[2] Carbocyclic and Heterocyclic Cage Compounds and Their Building
Blocks (Ed. K. K. Laali), JAI Press, Stamford, 1999. Supplement 1 in Ad-
vances in Strained and Interesting Organic Molecules (Ed.: B. Halton).
[3] T. Wettling, J. Schneider, O. Wagner, C. G. Kreiter, M. Regitz, Angew.
Chem. 1989, 101, 1035–1037; Angew. Chem. Int. Ed. Engl. 1989, 28,
1013–1014.
[4] a) B. Breit, U. Bergstrꢂsser, G. Maas, M. Regitz, Angew. Chem. 1992, 104,
1043–1046; Angew. Chem. Int. Ed. Engl. 1992, 31, 1055–1058; b) M. M.
Al-Ktaifani, W. Bauer, U. Bergstrꢂßer, B. Breit, M. D. Francis, F. W. Heine-
mann, P. B. Hitchcock, A. Mack, J. F. Nixon, H. Pritzkow, M. Regitz, M.
Zeller, U. Zenneck, Chem. Eur. J. 2002, 8, 2622–2633; c) for recent devel-
[24] The compliance-constants method showed that the endocyclic PꢀC
bond in complex 5 is much stiffer (higher relaxed force constant of k0 =
4.066 mdynꢁꢀ1) and therefore less prone to deformation than the en-
docyclic PꢀN bond (k0 =2.551 mdynꢁꢀ1).
[25] At the working level of theory, the PꢀC bond-cleavage transformation
(5!3) was found to be slightly endothermic with a moderately low
z
energy barrier of DEZPE =24.70 kcalmolꢀ1, DG° =24.52 kcalmolꢀ1 for
a coordinate dPꢀC =2.288 ꢁ.
z
[26] The energy barrier for PꢀC bond cleavage in 5·LiCl is DEZPE =29.80 kcal
molꢀ1, DG° =30.61 kcalmolꢀ1
.
&
&
Chem. Eur. J. 2014, 20, 1 – 8
www.chemeurj.org
6
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
[27] According to the weaker aromatic character in comparison to the other
two phenyl groups, as roughly estimated from structural descriptors
such as the Jugl index AJ (A. Jugl, P. FranÅoise, Theor. Chim. Acta 1967,
7, 249) and the GEO component of the HOMA parameter (T. M. Krygow-
ski, M. K. Cyranski, Tetrahedron 1996, 52, 10255–10264). AJ =0.997
versus 0.998 and 0.999; GEOꢇ103 =2.39 versus 1.93 and 0.80.
[28] For an example of an 1,3-dipolar intramolecular cycloaddition of a tran-
sient nitrilium phosphane ylide complex to an activated P-phenyl
group of a Wittig ylide, see: a) N. Hoffmann, P. G. Jones, R. Streubel,
Angew. Chem. 2002, 114, 1226–1228; Angew. Chem. Int. Ed. 2002, 41,
1186–1188; b) Intermolecular 1,4-cycoaddition of terminal phosphini-
dene complex to a benzene ring: c) M. J. van Eis, C. M. D. Komen, F. J. J.
de Kanter, W. H. de Wolf, K. Lammertsma, F. Bickelhaupt, M. Lutz, A. L.
Spek, Angew. Chem. 1998, 110, 1656–1658; Angew. Chem. Int. Ed. 1998,
37, 1547–1550.
[40] Z. Benkç, R. Streubel, L. Nyulaꢄszi, Dalton Trans. 2006, 4321–4327.
[41] a) T. A. Nguyen, Frontier Orbitals: a practical manual, Wiley, Hoboken,
2007; b) I. Fleming, Molecular Orbitals and organic chemical reactions,
Reference ed., Wiley, Hoboken, 2010.
[42] Global philicity (electrophilicity) w measures the stabilization in energy
when the system acquires an additional electric charge and was calcu-
lated from the electron affinity (EA) and ionization potential (IP): w=
[(IP+EA)2]/[4(IPꢀEA)]. It can be transformed into the local electrophilici-
ty index w+ =w·f+, where f+ is the Fukui function (R. G. Parr, W. Yang,
J. Am. Chem. Soc. 1984, 106, 4049–4050) for reaction with a nucleophile
at a particular atom.
[43] ORCA - An ab initio, DFT and semiempirical SCF-MO package. Written
by F. Neese, Max Planck Institute for Bioinorganic Chemistry, d-45470
Mꢃlheim/Ruhr, 2012. Version 2.9.1. http://www.mpibac.mpg.de/bac/
logins/neese/description.php. F. Neese, WIREs Comput. Mol. Sci. 2012, 2,
73–78.
[29] F. Mathey, M. Regitz in Phosphorus-Carbon Heterocyclic Chemistry: The
Rise of
pp. 17–55.
a
New Domain (Ed.: F. Mathey), Pergamon, Oxford, 2001,
[44] A. D. Becke, J. Chem. Phys. 1993, 98, 5648–5652. C. T. Lee, W. T. Yang,
R. G. Parr, Phys. Rev. B 1988, 37, 785–789.
[30] CCDC 965979 (4) and 965980 (6) contain the supplementary crystallo-
graphic data for this paper. These data can be obtained free of charge
from The Cambridge Crystallographic Data Centre via www.ccdc.cam.a-
c.uk/data_request/cif.
[45] A. Schꢂfer, H. Horn, R. Ahlrichs, J. Chem. Phys. 1992, 97, 2571–2577.
[46] D. Andrae, U. Haeussermann, M. Dolg, H. Stoll, H. Preuss, Theor. Chim.
Acta 1990, 77, 123–141. ECP basis sets for W [SD(60,MWB)] have been
obtained from Turbomole basis set library at ftp://ftp.chemie.uni-karls-
ruhe.de/pub/basen/.
[31] The energy of TS(4!7) in THF solution was found to be slightly higher
(26.55 kcalmolꢀ1) than in toluene solution (see Figure 5).
[47] S. Grimme, J. Antony, S. Ehrlich, H. Krieg, J. Chem. Phys. 2010, 132,
154104.
[32] For examples of stabilized yet reactive terminal amino phosphinidene
complexes, see: a) J. Wit, G. van Eijkel, M. Schakel, K. Lammertsma, Tet-
rahedron 2000, 56, 137–141; b) M. L. G. Borst, N. van der Riet, R. H. Lem-
mens, F. J. J. de Kanter, M. Schakel, A. W. Ehlers, A. M. Mills, M. Lutz, A. L.
Spek, K. Lammertsma, Chem. Eur. J. 2005, 11, 3631–3642; c) cf.[15a]
[33] A rough estimate of the global effect of NCIs, amounting to 1.1 kcal
molꢀ1, can be obtained from comparison of ZPE-corrected energies
with a conformer 7conf featuring the P atom far away from the triene
moiety, but having lower Gibbs free energy due to its most favourable
entropic term (see Figure S4 of the Supporting Information for a DG re-
action profile).
[48] a) A. Klamt, G. Schꢃꢃrmann, J. Chem. Soc. Perkin Trans. 2 1993, 220,
799–805; b) A. Klamt, J. Phys. Chem. 1995, 99, 2224–2235.
[49] A. Bergner, M. Dolg, W. Kuchle, H. Stoll, H. Preuss, Mol. Phys. 1993, 80,
1431–1441. Obtained from the EMSL Basis Set Library at https://
bse.pnl.gov/bse/portal. D. Feller, J. Comput. Chem. 1996, 17, 1571–
1586.
[50] NBO 5.9 code. E. D. Glendening, J. K. Badenhoop, A. E. Reed, J. E. Car-
penter, J. A. Bohmann, C. M. Morales and F. Weinhold, Theoretical
Chemistry Institute, University of Wisconsin, Madison, 2001.
[51] a) P.-O. Lçwdin, J. Chem. Phys. 1950, 18, 365–375; b) P.-O. Lçwdin, Adv.
Quantum Chem. 1970, 5, 185–199; c) A. Szabo and N. S. Ostlund,
Modern Quantum Chemistry. Introduction to advanced electronic structure
theory, Dover Publications, Mineola, NewYork, 1989.
[34] P···C(p): d=3.494 and 3.494 ꢁ, WBI=0.023, 1(r)=0.76ꢇ10ꢀ2 eaoꢀ3
PNC···HCPh: d=2.790 ꢁ, WBI=0.006, 1(r)=0.73ꢇ10ꢀ2 eaoꢀ3
;
.
[35] NCIplot, version 1.1. Department of Chemistry, Duke University (USA),
2011. http://www.chem.duke.edu/~yang/Software/software NCI.html;
a) E. R. Johnson, S. Keinan, P. Mori-Sꢄnchez, J. Contreras Garcꢈa, A. J.
Cohen, W. Yang, J. Am. Chem. Soc. 2010, 132, 6498–6506; b) J. Contrer-
as Garcꢈa, E. R. Johnson, S. Keinan, R. Chaudret, J.-P. Piquemal, D. N. Be-
ratan, W. Yang, J. Chem. Theory Comput. 2011, 7, 625–632. Here, l2
stands for the second highest eigenvalue of the electron density Hessi-
an matrix.
[52] R. S. Mulliken, J. Chem. Phys. 1955, 23, 1833–1840.
[53] R. F. W. Bader, in Atoms in Molecules: A Quantum Theory, Oxford Univer-
sity Press, Oxford, 1990.
[54] a) AIM2000 v. 2.0, designed by F. Biegler-Kçnig and J. Schçnbohm,
2002. Home page: http://www.aim2000.de/. F. Biegler-Kçnig, J. Schçn-
bohm, D. Bayles, J. Comput. Chem. 2001, 22, 545–559; b) F. Biegler-
Kçnig, J. Schçnbohm, J. Comput. Chem. 2002, 23, 1489–1494.
[55] Gaussian 09, Revision A.02, M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E.
Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Men-
nucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian,
A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara,
K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O.
Kitao, H. Nakai, T. Vreven, J. A. Montgomery, Jr., J. E. Peralta, F. Ogliaro,
M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Ko-
bayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyen-
gar, J. Tomasi, M. Cossi, N. Rega, N. J. Millam, M. Klene, J. E. Knox, J. B.
Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann,
O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin,
K. Morokuma, V. G. Zakrzewski, G. A.Voth, P. Salvador, J. J. Dannenberg,
S. Dapprich, A. D. Daniels, ꢅ. Farkas, J. B. Foresman, J. V. Ortiz, J. Cio-
slowski and D. J. Fox, Gaussian Inc., Wallingford CT, 2009.
[36] Based on electron density and its derivatives, in real space it represents
regions with electron-density regime below that for typical covalent
bonds and ranging from strong or moderate NCIs (blue; e.g., hydrogen
bonds), weak interactions (green; e.g., van der Waals interactions) and
steric (repulsive) clashes (red). Some of the interactions belonging to
the important van der Waals category have very low bond strength; for
instance, for the methane dimer computed at the working level of
theory (gas-phase): dCꢀC =3.715 ꢁ, 1(r)=0.24ꢇ10ꢀ2 eaoꢀ3
.
[37] a) A. E. Reed, F. Weinhold, J. Chem. Phys. 1983, 78, 4066–4073; b) A. E.
Reed, R. B. Weinstock, F. Weinhold, J. Chem. Phys. 1985, 83, 735–746.
NBO 5.0 code: E. D. Glendening, J. K. Badenhoop, A. E. Reed, J. E. Car-
penter, J. A. Bohmann, C. M. Morales, F. Weinhold, Theoretical Chemistry
Institute, University of Wisconsin, Madison (2001).
[38] The interactions corresponding to second-order perturbation analysis
of the Fock matrix in NBO basis amount to: CN!P donation,
[56] VMD - Visual Molecular Dynamics. W. Humphrey, A. Dalke, K. Schulten,
J. Molec. Graphics 1996, 14, 33–38. Home page: http://www.ks.uiuc.e-
du/Research/vmd/.
!
P ; through-
back-donation, 105.04 kcalmolꢀ1
219.52 kcalmolꢀ1
;
CN
space P!p*(C=C) transfer, 19.16 kcalmolꢀ1; through-space P!HꢀCPh
transfer, 11.19 kcalmolꢀ1
.
[39] a) E. M. Sproviero and G. Burton, J. Phys. Chem. A 2002, 106, 7834–
7843; b) E. M. Sproviero, G. Burton, J. Phys. Chem. A 2003, 107, 5544–
5554.
Received: December 28, 2013
Published online on && &&, 0000
Chem. Eur. J. 2014, 20, 1 – 8
www.chemeurj.org
7
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
FULL PAPER
&
Strained Polycycles
V. Nesterov, A. Espinosa,*
G. Schnakenburg, R. Streubel*
&& – &&
A Novel N,P,C Cage Complex Formed
by Rearrangement of a Tricyclic
Phosphirane Complex: On the
Importance of Non-covalent
Interactions
Trityl group in action: Unusual reactivi-
ty of the triphenylmethyl group bound
to phosphorus leads to intramolecular
heterocyclization and formation of
bly via transient N-amidinophosphini-
dene complex II, to give diastereoselec-
tively novel N,P,C cage complex III. The
role of non-covalent interactions in II
was analyzed theoretically.
strained polycycle I via [3+2] cycloaddi-
tion. Complex I rearranges, most proba-
&
&
Chem. Eur. J. 2014, 20, 1 – 8
www.chemeurj.org
8
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!