A Tricyclic Hexaphosphane

Article abstract of DOI:10.1002/chem.201503808

The reaction of the functionalized cyclo-tetraphosphane [ClP(μ-PMes)]₂ (Mes=2,4,6-tri-tert-butylphenyl) with different Lewis bases led to the formation of an unprecedented tricyclic hexaphosphane, MesP₆Mes. The formation of this comp

Full text of DOI:10.1002/chem.201503808

DOI: 10.1002/chem.201503808
Communication
&
Phosphorus Chemistry
A Tricyclic Hexaphosphane
Jonas Bresien,[a] Axel Schulz,*[a, b] and Alexander Villinger[a]
Dedicated to Prof. Dr. F. Ekkehardt Hahn on the occasion of his 60th birthday
Abstract: The reaction of the functionalized cyclo-tetra-
phosphane [ClP(m-PMes*)]2 (Mes*=2,4,6-tri-tert-butylphen-
yl) with different Lewis bases led to the formation of an
unprecedented tricyclic hexaphosphane, Mes*P6Mes*. The
formation of this compound was investigated by spectro-
scopic and theoretical methods, revealing an unusual ring
expansion reaction. The title compound was fully charac-
terized by experimental and computational methods.
Phosphorus has fascinated chemists since its discovery by Hen-
ning Brand in 1669.[1] Especially since the development of
modern day chemistry, various modifications of phosphorus
have been identified,[2,3] and countless polyphosphorus sys-
tems with (organic) substituents have been reported that in-
corporate a plethora of different structural motifs.[2–5] The most
prominent ones are possibly monocyclic ring systems with
three to six phosphorus atoms, bicyclic tetra- or pentaphos-
phanes, derivatives of the heptaphosphatriide anion, as well as
higher aggregates which resemble strands of violet phospho-
rus. Moreover, with the development of computational meth-
ods, a broad range of hypothetical P clusters was studied to
identify possible minima on the potential energy surface.[6–13]
However, examples of polycyclic hexaphosphanes are still
rare. Baudler and co-workers reported on (tBu)4P6 (1,
Scheme 1) and the related oxide (tBu)4P6O (2), which were de-
rived from tBuPCl2 and PCl3 by reduction with Mg and subse-
quent oxidation by exposure to air.[14,15] As reported by the
group of Jutzi, thermolysis of (Cp*P)3 yielded the structurally
related bicyclic hexaphosphane Cp*4P6 (3, Cp*=pentamethyl-
cyclopentadienyl) as well as the tricyclic hexaphosphane
Cp*2P6 (4), which incorporates a benzvalene type P6 scaffold
(A).[16,17] Weigand and co-workers reported on the cationic
compound [Ph4P6]2+ (5), which can formally be derived from
the prismane type structure (B) of P6.[18] Additionally, a salt con-
taining a bicyclic hexastibine scaffold similar to 1 and 3 was re-
ported recently.[19]
Scheme 1. Top: Experimental examples of polycyclic hexaphosphanes.
Bottom: Computed minimum structures on the potential energy surface of
neutral P6 (relative energies given in kJmolÀ1).[13]
Theoretical investigations revealed that the benzvalene type
structure (A) of neutral P6 is the most favorable isomer in the
gas phase.[6–9,13] Nonetheless, other structures (B–F) were iden-
tified as local minima (Scheme 1),[13] which present desirable
new structural motives in phosphorus chemistry. To date, no
tricyclic hexaphosphanes that incorporate a P6 scaffold of type
C have been reported.[20] Therefore, we wish to present herein
our results concerning the synthesis of the tricyclic hexaphos-
phane Mes*P6Mes* (6, Mes*=2,4,6-tri-tert-butylphenyl), which
was derived from a P4 precursor by an unusual ring expansion
reaction (Scheme 2).
Scheme 2. Synthesis of the tricyclic hexaphosphane 6 (R=Mes*).
Recently, we reported on the synthesis of the cyclophos-
phane [ClP(m-PMes*)]2 (7),[21] which allows for selective func-
tionalization of the P4 ring system due to the chlorine substitu-
ents. Cyclophosphane 7 was synthesized from Mes*PH2 and
PCl3 in the presence of NEt3; thus, it can be regarded as a P4
building block that was derived from P1 units, as opposed to
the commonly applied strategy to functionalize white phos-
phorus (P4) directly using Lewis acids, (transition) metals, or
singlet carbenes.[22–25] After investigating the reactivity of 7 to-
[a] J. Bresien, Prof. Dr. A. Schulz, Dr. A. Villinger
Institut für Chemie, Universität Rostock
Albert-Einstein-Straße 3a, 18059 Rostock (Germany)
E-mail: axel.schulz@uni-rostock.de
[b] Prof. Dr. A. Schulz
Leibniz-Institut für Katalyse an der Universität Rostock e.V.
Albert-Einstein-Straße 29a, 18059 Rostock (Germany)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201503808.
Chem. Eur. J. 2015, 21, 18543 – 18546
18543
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
wards Lewis acids,[26] we aimed at nucleophilic substitution of
the Cl atoms at the P4 scaffold. Since 4-(N,N-dimethylamino)-
pyridine (dmap) had previously been proven to be a good nu-
cleophile and can effectively stabilize positive charges,[27–29] the
reaction of 7 with dmap was investigated. In analogy to the re-
action of [TfOP(m-NDmp)]2 (Dmp=2,6-dimethylphenyl) with
dmap (Scheme 3),[28] the formation of a base-stabilized cyclo-
the formation of 6 can be rationalized in terms of formal
Mes*PCl2 extrusion and subsequent dimerization of the three-
membered ring system. Comparable three-membered inter-
mediates have been discussed previously in P4 activation
routes.[2,3,31]
The molecular structure of the tricyclic hexaphosphane 6
was determined by single-crystal X-ray diffraction (Figure 2).[32]
Scheme 3. Reaction of a cyclodiphosphadiazane with dmap (R=Dmp).[28]
phosphenium cation could be expected. In fact, in situ
31P NMR spectroscopy revealed the formation of a P4 species as
indicated by an ABMX spin system (À99.3, À91.2, +9.7, and
+79.3 ppm); however, the chemical shifts were indicative of
a species containing a three- rather than a four-membered
ring system (Figure 1).[21,30] Comprehensive DFT studies helped
Figure 2. Molecular structure of 6. Ellipsoids are set at 50% probability
(123 K). Selected bond lengths [Š] and angles [8]: P1ÀC1 1.876(1), P1ÀP2
2.2189(5), P1ÀP3 2.2151(5), P2ÀP3 2.2373(5), P2ÀP3’ 2.2368(5); P3-P1-P2
60.61(2), P1-P2-P3 59.61(2), P1-P3-P2 59.78(2), P1-P2-P3’ 92.87(2), P2’-P3-P2
90.01(2); P1-P2-P3-P2’ 93.33(2).
The molecule is located on a crystallographic inversion center,
hence the central P4 ring system is perfectly planar. All bond
angles within the four-membered ring system are close to 908,
the bond angles within the three-membered rings are close to
608. The six-membered ring molecule adopts a chair conforma-
tion with two transannular PÀP bonds. The P1ÀP2 and P1ÀP3
bonds are slightly shorter (average 2.217 Š) than the P2ÀP3
and P2ÀP3’ bonds (average 2.237 Š). Therefore, the central P4
scaffold adopts an almost square geometry, while the three-
membered rings correspond to isosceles triangles. All PÀP
bond lengths correspond to typical single bonds (Srcov
=
2.22 Š).[33] The angle between the least-squares planes of the
three- and four-membered ring systems is 93.38, which com-
pares to to those of other polycyclic phosphanes, such as bicy-
clic tetraphosphanes (ꢀ 1008)[22,34–36] or the aforementioned bi-
cyclic hexaphosphanes 1 and 3 (89.88 and 99.28, respective-
ly).[14,17]
Figure 1. In situ 31P NMR spectrum of the reaction mixture after 5 h. The
main signals were attributed to the intermediate 8+ (* denotes other spe-
cies in solution).
us assign the NMR signals to the intermediate
[Mes*P3(dmap)P(Cl)Mes*]Cl (8Cl, Scheme 2). The calculated
NMR data agrees well with the experimental spectrum
(Table S3, Supporting Information). With respect to the starting
material, the formation of 8+ can be understood in terms of
a formal 1,2-Cl shift and rearrangement of one PÀP bond.
Other signals in the in situ NMR spectrum were likely caused
by different (configurational or rotational) isomers of 8+. How-
ever, due to low intensity of the resonances, these could not
be assigned unambiguously.
In the solid-state Raman spectrum, the P6 scaffold can be
identified by three characteristic bands at 414, 442, and
537 cmÀ1, which were assigned on the basis of computed vi-
brational data (Table 1, for more details see Figure S5 in the
Supporting Information)
In the 31P NMR spectrum, the tricyclic hexaphosphane 6 is
characterized by an AA’BB’B’’B’’’ spin system (À107.7 and
À96.1 ppm; Figure 3). Due to dynamic rotation of the tBu
groups in solution, all four P atoms of the central P4 scaffold
become equivalent, resulting in apparent C2h symmetry. The
chemical shifts are similar to the NMR shifts of the P3 unit of
the structurally related compounds 1 and 3,[14,17] but also com-
parable to either cyclotriphosphanes[21,30] or the bridging P
atoms in bicyclic tetraphosphanes.[22,34–36] The experimental
NMR data are in good agreement with calculated NMR shifts
and coupling constants (Table 2).
All attempts to crystallize the intermediate 8Cl failed. In-
stead, colorless crystals of the novel tricyclic hexaphosphane 6
were obtained (yield: 58%). According to 31P NMR spectrosco-
py, compound 6 and Mes*PCl2 were in fact the only products
after a total reaction time of about two days. Based on NMR in-
tegrals, the ratio of Mes*PCl2 and 6 was found to be 2:1; thus,
Chem. Eur. J. 2015, 21, 18543 – 18546
www.chemeurj.org
18544
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
Table 1. Characteristic Raman vibrations of 6.
Vibration
mode (*=P)
Main
contribution
Frequency
[cmÀ1
]
in-phase P3
ring mode
537
out-of-phase P4
ring mode
442
414
Figure 4. Depiction of the electron localization function, showing the PÀP
bonding system and the lone pairs at the P atoms; left: three-membered
ring plane; right: four-membered ring plane.
out-of-phase P3
ring mode
Figure 5. Relative stabilities of R2P6 isomers (DG298 in kJmolÀ1).
Figure 3. Experimental (top) and simulated (bottom) 31P NMR spectrum of 6.
tal evidence.[16] Thus, the unusual tricyclic structure of type C is
stabilized by the bulky Mes* moiety.
The tricyclic hexaphosphane 6 is poorly soluble in common
organic solvents. It decomposes at 2128C under inert condi-
tions. Intriguingly, it was found to be stable in air. Even upon
adding water to solution of 6 in CD2Cl2, 60% of the substance
(by mole fraction) remained intact over a period of ten days at
ambient temperature. This prompted us to develop an alter-
nate synthetic protocol: starting from Mes*PH2 and PCl3, the
precursor 7 was generated in situ by base-assisted HCl elimina-
tion. After removal of the hydrochloride, the crude product
was treated with dmap without further purification. This led to
precipitation of the tricyclic hexaphosphane, which was subse-
quently washed with water and acetone to remove all polar
and nonpolar impurities. In this vein, compound 6 could be
synthesized in 47% yield (based on Mes*PH2), which is a signifi-
cant improvement over the synthesis from pure 7 (combined
yield based on Mes*PH2: 26%).
Table 2. 31P NMR data of 6. Calculated values [PBE0/6-31G(d,p), GIAO
method] are given in brackets.
d [ppm]
J [Hz]
PA
PB
À107.7 (À126.2)
J
AA’ =À45 (À56)
À96.1 (À88.8)
J
J
AB =JAB’ =À191 (À158)
AB’’ =JAB’’’ = +54 (+35)
J
J
J
BB’ =À111 (À78)
BB’’ =À146 (À90
BB’’’ =À30 (À45)
To investigate the bonding situation within the P6 scaffold,
DFT calculations were performed.[37] As expected, natural bond
orbital (NBO)[38] analysis revealed eight PÀP single bonds, of
which six are considerably bent out of the line of nuclear cen-
ters due to the small bonding angles within the three-mem-
bered rings. This effect is also reflected in the symmetry and
shape of the Kohn–Sham orbitals and can be nicely visualized
by the electron localization function (ELF, Figure 4), where the
local maxima are located beside the bond axes. Hence, the
bonding within the P3 units is comparable to that of tetrahe-
dral P4.[39] To evaluate the stability of compound 6, the corre-
sponding cis-tricycle, which can formally be derived from the
prismane-type structure (B) of P6, as well as the benzvalene
type isomer (A) were calculated (Figure 5). These were found
to be 40.2 or 72.6 kJmolÀ1 higher in energy (DG298) than the
experimental structure, respectively. However, in case of the
smaller Cp* compound (4) or tBu substituents (9), the benzva-
lene type isomers (A) were found to be the most stable ones,
in accordance with the findings for neutral P6 and experimen-
When treating the cyclophosphane 7 with other Lewis
bases, such as PPh3 or tetramethylethylenediamine (TMEDA),
practically no reaction was observed. Slow consumption of the
starting material could only be detected after prolonged stir-
ring, and a mixture of different products was observed in the
31P NMR spectrum. In the case of PPh3, trace amounts of 6
were identified in the NMR spectrum after nine days. On the
other hand, the reaction of 7 with tetramethylimidazolylidene
led to degradation of the P4 ring system; the main products
were identified as the endo–exo isomer of the bicyclic phos-
phane Mes*P4Mes*, the diphosphene Mes*PPMes*, and the pri-
mary phosphane Mes*PCl2 (5:1:7 ratio).[40] Accordingly, the un-
usual reactivity of the cyclophosphane 7 towards dmap cannot
be generalized to other Lewis basic systems.
Chem. Eur. J. 2015, 21, 18543 – 18546
www.chemeurj.org
18545
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
In conclusion, the first [3.1.0.02,4]-tricyclic hexaphosphane,
Mes*P6Mes* (6), could be prepared starting from a P4 precursor
molecule by formal ring expansion. Due to its symmetry, com-
pound 6 displayed an unusual AA’BB’B’’B’’’ spin system in the
31P NMR spectrum. The P6 scaffold could be well discerned in
the solid-state Raman spectrum due to three characteristic
bands. Furthermore, DFT calculations showed that the structur-
al motif is stabilized by the bulky Mes* substituents; in case of
smaller moieties, such as Cp* or tBu, a benzvalene type struc-
ture of the P6 scaffold is preferred. This demonstrates the ad-
vantages of bulky substituents for the stabilization of unusual
bonding patterns.
[11] M. Häser, O. Treutler, J. Chem. Phys. 1995, 102, 3703–3711.
[12] L. Guo, H.-S. Wu, Z.-H. Jin, J. Mol. Struct. 2004, 677, 59–66.
[13] T. R. Galeev, A. I. Boldyrev, Phys. Chem. Chem. Phys. 2011, 13, 20549–
20556.
[14] M. Baudler, Y. Aktalay, K.-F. Tebbe, T. Heinlein, Angew. Chem. Int. Ed. Engl.
1981, 20, 967–969; Angew. Chem. 1981, 93, 1020–1022.
[15] M. Baudler, M. Michels, M. Pieroth, J. Hahn, Angew. Chem. Int. Ed. Engl.
1986, 25, 471–473; Angew. Chem. 1986, 98, 465–467.
[16] P. Jutzi, R. Kroos, A. Müller, M. Penk, Angew. Chem. Int. Ed. Engl. 1989,
28, 600–601; Angew. Chem. 1989, 101, 628–629.
[17] P. Jutzi, R. Kroos, A. Müller, H. Bçgge, M. Penk, Chem. Ber. 1991, 124,
75–81.
[18] J. J. Weigand, M. Holthausen, R. Frçhlich, Angew. Chem. Int. Ed. 2009,
48, 295–298; Angew. Chem. 2009, 121, 301–304.
[19] S. S. Chitnis, N. Burford, J. J. Weigand, R. McDonald, Angew. Chem. Int.
Ed. 2015, 54, 7828–7832; Angew. Chem. 2015, 127, 7939–7943.
[20] A tricyclic structure of type C was proposed for the salt [H3NCy][HS2P6],
but no analytical evidence could prove this connectivity. See: M. Becke-
Goehring, H. Hoffmann , Z. Anorg. Allg. Chem. 1969, 369, 73–82.
[21] J. Bresien, C. Hering, A. Schulz, A. Villinger, Chem. Eur. J. 2014, 20,
12607–12615.
Experimental Section
All manipulations were carried out under an inert atmosphere of
argon. A mixture of [{ClP(m-PMes*)}2] (350 mg, 0.51 mmol) and 4-
(N,N-dimethylamino)pyridine (dmap, 69 mg, 0.56 mmol) was dis-
solved in CH2Cl2 (3 mL) and stirred at ambient temperature for
2.5 days. Subsequently, the solution was slowly cooled to 58C,
whereupon crystallization of colorless crystals was observed. The
yellowish supernatant was removed by syringe and the crystals
were washed with cold CH2Cl2 (0.5 mL) to remove adhering impuri-
[22] R. Riedel, H.-D. Hausen, E. Fluck, Angew. Chem. Int. Ed. Engl. 1985, 24,
1056–1057; Angew. Chem. 1985, 97, 1050–1050.
[23] J. D. Masuda, W. W. Schoeller, B. Donnadieu, G. Bertrand, Angew. Chem.
Int. Ed. 2007, 46, 7052–7055; Angew. Chem. 2007, 119, 7182–7185.
[24] J. E. Borger, A. W. Ehlers, M. Lutz, J. C. Slootweg, K. Lammertsma, Angew.
Chem. Int. Ed. 2014, 53, 12836–12839; Angew. Chem. 2014, 126, 13050–
13053.
ties. Drying in vacuo yielded
a colorless product (100 mg,
[25] S. Heinl, S. Reisinger, C. Schwarzmaier, M. Bodensteiner, M. Scheer,
Angew. Chem. Int. Ed. 2014, 53, 7639–7642; Angew. Chem. 2014, 126,
7769–7773.
[26] J. Bresien, K. Faust, A. Schulz, A. Villinger, Angew. Chem. Int. Ed. 2015,
54, 6926–6930; Angew. Chem. 2015, 127, 7030–7034.
[27] E. Rivard, K. Huynh, A. J. Lough, I. Manners, J. Am. Chem. Soc. 2004, 126,
2286–2287.
[28] R. J. Davidson, J. J. Weigand, N. Burford, T. S. Cameron, A. Decken, U.
Werner-Zwanziger, Chem. Commun. 2007, 4671–4673.
[29] C. Hering-Junghans, M. Thomas, A. Villinger, A. Schulz, Chem. Eur. J.
2015, 21, 6713–6717.
0.15 mmol, 58%): m.p. 2128C (decomposition); elemental analysis
calcd (%) : C 63.90, H 8.64; found: C 63.75, H 8.82; 31P{1H} NMR
(CD2Cl2, 121.5 MHz): d=À107.7 (m, 1JAB =À191.3 Hz, 2JAB’ = +
54.3 Hz, 3JAA’ =À44.8 Hz,
2
P, (Mes*P)2P4), À96.1 (m, JAB
=
1
À191.3 Hz, 1JBB’ =À110.5 Hz, 1JBB’’ =À146.1 Hz, 2JBB’’’ =À29.7 Hz,
2JAB’ = +54.3 Hz, 4 P, (Mes*P)2P4); MS (CI positive, iso-butane): m/z
(%): 677 [M]+. Further experimental details can be found in the
Supporting Information.
[30] M. Baudler, B. Makowka, Z. Anorg. Allg. Chem. 1985, 528, 7–21.
[31] A. Lorbach, A. Nadj, S. Tüllmann, F. Dornhaus, F. Schçdel, I. Sänger, G.
Margraf, J. W. Bats, M. Bolte, M. C. Holthausen, Inorg. Chem. 2009, 48,
1005–1017.
[32] CCDC 1417832 (6) contain the supplementary crystallographic data for
this paper. These data are provided free of charge by The Cambridge
Crystallographic Data Centre..
Acknowledgements
Financial support from the Fonds der Chemischen Industrie
(scholarship for J.B.) and the Deutsche Forschungsgemein-
schaft (SCHU 1170/11-1) is gratefully acknowledged.
[33] P. Pyykkç, M. Atsumi, Chem. Eur. J. 2009, 15, 12770–12779.
[34] E. Niecke, R. Rüger, B. Krebs, Angew. Chem. Int. Ed. Engl. 1982, 21, 544–
545; Angew. Chem. 1982, 94, 553–554.
Keywords: Lewis base · 4-(N,N-dimethylamino)pyridine · NMR
spectroscopy
·
phosphorus compounds
·
polycyclic
[35] P. Jutzi, U. Meyer, J. Organomet. Chem. 1987, 333, C18–C20.
[36] A. R. Fox, R. J. Wright, E. Rivard, P. P. Power, Angew. Chem. Int. Ed. 2005,
44, 7729–7733; Angew. Chem. 2005, 117, 7907–7911.
phosphanes · structure elucidation
[37] Detailed information on experimental and computational methods is
given in the Supporting Information.
[1] A. Oppenheim, in Allgemeine Deutsche Biographie, Dritter Band, Duncker
& Humblot, Leipzig (Germany), 1876, p. 236.
[38] E. D. Glendening, J. K. Badenhoop, A. E. Reed, J. E. Carpenter, J. A. Boh-
mann, C. M. Morales, C. R. Landis, F. Weinhold, NBO 6.0, Theoretical
Chemistry Institute, University of Wisconsin, Madison, 2013.
[39] M. F. Guest, I. H. Hillier, V. R. Saunders, J. Chem. Soc.Faraday Trans. 2
1972, 68, 2070–2074.
[2] M. Scheer, G. Balµzs, A. Seitz, Chem. Rev. 2010, 110, 4236–4256.
[3] N. A. Giffin, J. D. Masuda, Coord. Chem. Rev. 2011, 255, 1342–1359.
[4] M. Baudler, K. Glinka, Chem. Rev. 1993, 93, 1623–1667.
[5] G. He, O. Shynkaruk, M. W. Lui, E. Rivard, Chem. Rev. 2014, 114, 7815–
7880.
[6] R. O. Jones, D. Hohl, J. Chem. Phys. 1990, 92, 6710–6721.
[7] R. Janoschek, Chem. Ber. 1992, 125, 2687–2689.
[8] M. Häser, U. Schneider, R. Ahlrichs, J. Am. Chem. Soc. 1992, 114, 9551–
9559.
[40] J. Bresien, K. Faust, C. Hering-Junghans, J. Rothe, A. Schulz, A. Villinger,
Dalton Trans. 2016, DOI 10.1039/C5DT02757H.
Received: September 23, 2015
[9] D. S. Warren, B. M. Gimarc, J. Am. Chem. Soc. 1992, 114, 5378–5385.
[10] M. Häser, J. Am. Chem. Soc. 1994, 116, 6925–6926.
Published online on November 5, 2015
Chem. Eur. J. 2015, 21, 18543 – 18546
www.chemeurj.org
18546
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim