A fluoroaryl substituent with spectator function: Reactivity and structures of cyclic and acyclic HF4C6-substituted phosphanes

Article abstract of DOI:10.1016/j.jorganchem.2009.10.025

The reactivity of a series of phosphanes with a fluoroaryl group (HF₄C₆-) carrying a spectator function in para position has been explored with respect to the formation of low coordinated and phosphorus rich phosphanes. An asymmetric diphosphene has been indentified as an intermediate in the synthesis of a linear 1,3-dihydrophosphane, while the symmetric diphosphene undergoes 2 + 2 cycloaddition under formation of the corresponding cyclotetraphosphetane for which a crystal structure could be obtained. Attempts to synthesize HF₄C₆-substituted iminophosphanes generally failed, which is attributed to the electronic nature of the corresponding precursors as suggested by quantum chemical calculations.

Full text of DOI:10.1016/j.jorganchem.2009.10.025

Journal of Organometallic Chemistry 695 (2010) 974–980
Contents lists available at ScienceDirect
Journal of Organometallic Chemistry
journal homepage: www.elsevier.com/locate/jorganchem
A fluoroaryl substituent with spectator function: Reactivity and structures
of cyclic and acyclic HF4C6-substituted phosphanes
*
Andreas Orthaber, Ferdinand Belaj, Rudolf Pietschnig
Karl-Franzens-Universität, Institut für Chemie, Schubertstraße 1, A-8010 Graz, Austria
a r t i c l e i n f o
a b s t r a c t
Article history:
The reactivity of a series of phosphanes with a fluoroaryl group (HF4C6-) carrying a spectator function in
para position has been explored with respect to the formation of low coordinated and phosphorus rich
phosphanes. An asymmetric diphosphene has been indentified as an intermediate in the synthesis of a
linear 1,3-dihydrophosphane, while the symmetric diphosphene undergoes 2 + 2 cycloaddition under
formation of the corresponding cyclotetraphosphetane for which a crystal structure could be obtained.
Attempts to synthesize HF4C6-substituted iminophosphanes generally failed, which is attributed to the
electronic nature of the corresponding precursors as suggested by quantum chemical calculations.
Ó 2009 Elsevier B.V. All rights reserved.
Received 25 September 2009
Received in revised form 21 October 2009
Accepted 22 October 2009
Available online 27 October 2009
Keywords:
Organo fluorine compounds
Phosphorus
Diphosphene
Inorganic rings
Ab initio calculations
1. Introduction
study we explore the influence of a fluorinated aryl group with
an additional spectator function on the electronic activation or
Phosphorus containing compounds with electron withdrawing
moieties attracted considerable attention in the last decade [1,2].
Electron deficient groups are known to influence both electronic
and optical properties of phosphorus containing compounds [3–
7]. Pentafluorophenyl and trifluoro methyl substituted phenyl
groups are the most frequently used substituents in the synthesis
of fluorinated phosphorus containing materials and complexes
[8–10]. Very recently perfluorinated phosphanoboranes have been
in the focus of interest for the synthesis of ‘‘frustrated” Lewis acid/
base pairs looking at applications as e.g. hydrogen storage or acti-
vation [11,12]. Tuning of different phosphorus based ligands has
been studied with respect to hydroformylation reactions [13]. Elec-
tron deficient units are essential for the synthesis of molecular n-
type conducting materials and different attempts have been made
to incorporate C6F5 groups into the backbone of phosphorus de-
rived oligomers and polymers [1]. Diphosphenes bearing electron
withdrawing groups are known to have reduced reactivity [14].
The steric situation alone cannot explain this behavior, because
the size of CF3 is comparable to that of the isoelectronic t-butyl
group. Consequently, electronic deactivation of the P@P group is
likely to be the crucial factor. Recently, we and others observed
the formation of P-rich compounds from diphosphenes [15–17].
Phosphorus rich compounds are of growing interest as stable acti-
vated products of white phosphorus [18,19]. In the underlying
deactivation of low coordinated phosphanes and their conversion
to linear and cyclic phosphanes.
2. Results and discussion
Pentafluorophenyl groups are well established as electron with-
drawing groups in chemistry [10,20]. We have chosen the tetra
fluorophenyl group (2,3,5,6-F4C6H-), which carries a hydrogen
atom in para position providing the possibility to follow reactions
easily by 1H spectroscopy. In addition, 2D 1H–13C coupled spectra
give the opportunity to detect the aromatic carbon atoms more
conveniently. We were able to synthesize the diethylaminophos-
phane (1) in good overall yields starting from the corresponding
lithiated fluoroaryl compound. Lithiation is carried out strictly be-
low ꢀ80 °C and the lithium compound is used without further
purification. During the addition of the chlorophosphane the tem-
perature is not allowed to rise above ꢀ60 °C. This procedure is ben-
eficial over literature procedures, where RF–MgBr is reacted with
Cl–P(NEt2)2, owing to the strikingly lower costs of the starting
materials RF–H compared with RF–Br. The synthesis of the dichlo-
rophosphane 3 (RF–PCl2) must by accomplished by the use of
diethylamino groups as protecting groups to prevent multiple sub-
stitution at the phosphorus atom. This is in contrast to the synthe-
sis of MesPCl2 where multiple substitution hardly occurs. The
diethylaminophosphane derivative 1 is obtained as viscous oil in
spectroscopically pure form after distillation. It shows a broad sin-
glet in the phosphorus NMR spectra at 84.9 ppm and neither cou-
* Corresponding author. Tel.: +43 316 380 5285; fax: +43 316 380 9835.
E-mail address: rudolf.pietschnig@uni-graz.at (R. Pietschnig).
0022-328X/$ - see front matter Ó 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.jorganchem.2009.10.025

A. Orthaber et al. / Journal of Organometallic Chemistry 695 (2010) 974–980
975
pling to the fluorine nor to the hydrogen atoms could be resolved.
Also the 19F NMR spectra show only
single multiplet at
tial symmetric diphosphene is expected to be unstable because the
steric bulk of the fluorinated aryl moiety is comparable to that of
MesPCl2 and thus 2 + 2 cycloaddition of the intermediate diph-
osphene may take place. Ring sizes other than four have not been
observed in this reaction. The crystal structure analysis of 5 proves
the identity of tetrakis-(2,3,5,6-tetrafluorophenyl)tetraphosphe-
tane, whose molecular structure is depicted in Fig. 2. The mole-
cules are lying around two-fold rotation axes and have almost
mm2 symmetry (C2v). The P–P and P–C distances are not signifi-
cantly different for each ring position despite the different orienta-
tions of the phenyl rings: whereas the two phenyl rings at P2 and
P20 are almost co-planar with a torsion angle P20ꢁ ꢁ ꢁP2–C21–C22 of
2.9(4)°, the two phenyl rings at P1 and P10 have a torsion angle
P10ꢁ ꢁ ꢁP1–C11–C12 of 96.0(3)°. These structural parameters are in
good agreement with the published structure of the pentafluor-
ophenyl substituted cyclotetraphosphetane [21]. The packing of
the molecules shows remarkable tubular voids of ca. 3.8 Å in diam-
eter parallel to the a-axis confined by the F atoms F13, F23 and F26.
a
ꢀ141.6 ppm. Furthermore 13C NMR data and MS data are in agree-
ment with the proposed structure of 1. In order to explore the reac-
tivity of this compound we oxidized 1 with an excess of gray
selenium to the corresponding selenophosphorane 2 which pro-
ceeds in almost quantitative yields after 12 h. The constitution of
compound 2 was determined by X-ray crystallography revealing
an almost planar central phenyl ring which is nearly coplanar with
one of the P–N bonds (angle between ring plane and P1–N2
0.49(5)°) (Fig. 1). The P@Se bond is 2.1074(3) Å, which is in the nor-
mal range for this class of compounds and encloses an angle with
the least-squares (l.s.) plane of the phenyl ring of 60.77(4)°. The
phosphorus atom is slightly shifted out of the l.s. plane of the phe-
nyl ring by an angle of 7.14(5)° resulting in a slight elongation of
the P–C bond. Owing to this unsymmetrical orientation of the –
P(@Se)N2-fragment (with respect to the phenyl ring) quite differ-
ent bond angles [N1–P1–C1 103.37(5)°, N2–P1–C1 111.25(5)°,
N1–P1–Se1 117.84(4)°, N2–P1–Se1 110.57(4)°] are observed. On
the other hand the observed P–N distances are all almost equal
[P1–N1 1.647(1) Å, P1–N2 1.648(1) Å]. In the crystal structure the
tetrafluorophenyl rings of two molecules related by an inversion
center are packed with a distance of 3.310(8) Å between their ring
In order to stabilize a potential diphosphene moiety we coupled
t
*
*
Mes PH2 (Mes = 2,4,6-tris- butyl phenyl) with 4 which resulted in
the formation of the desired diphosphene (6) based on the 31P NMR
spectra (Fig. 3). Unfortunately the steric demand and the electronic
situation do not allow to stabilize the P@P moiety sufficiently and 6
undergoes subsequent diphosphene metathesis under formation of
l.s. planes showing weak
pꢀp interaction.
* *
the symmetric Mes P@PMes and tetraphosphetane 5. Besides
The amino groups can be replaced by halogen atoms with either
PX3 or anhydrous HX. For the scrambling reaction with PCl3 the
boiling point of the resulting product is only marginally different
from that of the second product (Et2N)PCl2. Therefore, it is prefer-
able to cleave the protecting groups with HCl. Additionally we ob-
served the formation of the mixed species RFPClNEt2 (3) under
distillation conditions if (Et2N)PCl2 was present. This might be rea-
soned by reverse scrambling reactions at higher temperatures.
Compound 3 is obtained in pure form from the distillation residue
as high viscous oil with a 31P NMR resonance at 112.7 ppm. Fur-
thermore the cleavage of the diethylamino groups by anhydrous
HCl proceeds rather slowly compared to other analogous reactions.
The formed ammonium salt can be easily removed by filtration and
the product HC6F4PCl2 (4) is collected by distillation under reduced
pressure.
*
*
*
unreacted Mes PH2 and the symmetric diphosphene Mes P@PMes
further products are also formed during this reaction. Attempts to
elucidate their structure by means of 31P–31P-correlated spectros-
copy showed the presence of R0PH–P(R0)–PHR for R = C6F4H and
R0 = Mes (7) (Fig. 4). Triphosphane 7 shows three 31P NMR signals
*
at PA ꢀ49.5 ppm, PB ꢀ55.1 ppm and PC ꢀ75.8 ppm with a character-
istic coupling pattern. Formation of 7 can be reasonably explained
*
by addition of unreacted Mes -PH2 to the diphosphene 6, which is
formed as an intermediate. Such a P–H addition to a diphosphene
was recently also observed for ferrocenyl mono- and bisdiphosph-
enes [16,17].
Interestingly, cyclotetraphosphetane 5 is also obtained by
reduction with other phosphanes. The ‘‘Phospha-Wittig” route to
phosphaalkenes as established by Protasiewicz and coworkers in-
volves the reaction of dichlorophosphanes with PMe3 in the pres-
ence of Zn dust and aldehydes [22]. Reaction of 4 with PMe3 in
the presence of ferrocenyl aldehyde gave only the cyclotetrapho-
The reductive coupling of 4 with elemental magnesium resulted
in the formation of cyclotetraphosphetane 5 (Scheme 1). A poten-
Fig. 1. Molecular structure of 2. Ortep plot at 50% probability level. Hydrogen atoms
are omitted for clarity. Selected bond lengths, distances [Å] and angles [°]: Se1–P1
2.1073(3), P1–N1 1.647(1), P1–N2 1.648(1), P1–C1 1.845(1), N1–P1–N2 106.0(1),
N1–P1–C1 103.4(1), N2–P1–C1 111.3(1), N1–P1–Se1 117.84(4), N2–P1–Se1
110.57(4), C1–P1–Se1 107.66(4), N1–P1–C1–C6 –121.0(1), N2–P1–C1–C6–7.7(1),
Se1–P1–C1–C6 113.58(1).
Fig. 2. Molecular structure of 5. Ortep plot at 50% probability level. Selected bond
lengths, distances [Å] and angles [°]: P1–P2 2.237(1), P1–P20 2.238(1), 1.833(3), P2–
C2 1 1.829(3), P2–P1–P20 79.87(5), P1–P2–P10 83.34(5), P20–P1–P2–P10 41.82(5).

976
A. Orthaber et al. / Journal of Organometallic Chemistry 695 (2010) 974–980
Scheme 1. Synthetic routes to phosphaalkenes and diphosphenes including follow up reactions. (a) PMe3; (b) Zn, R0-CHO (c) (1) LAH; (2) n-BuLi; (3) TMS-Cl; (d) R0-COCl
(Becker route); (e) (1) R-PHLi; (2) DBU; (f) +R0PH2; (g) Mg; (h) [2 + 2]-cycloaddition; (i) + R-PCl2 (4).
*
**
*
*
Fig. 3. 31P NMR of the low field region of the diphosphene 6 ( ArF-P; 2,4,6-tBu–C6H2–P). Formation of the symmetric diphosphene Mes -P@P-Mes (s) is already observed.
Fig. 4. 31P NMR of the high field region of the PP-COSY spectrum showing the triphosphane 7, besides other unidentified side products. Dotted circles indicate the
corresponding cross peaks.

A. Orthaber et al. / Journal of Organometallic Chemistry 695 (2010) 974–980
977
sphetane and unreacted aldehyde but not the intended phos-
phaalkene. The electron withdrawing fluoroaryl substituent may
lead to an increased electrophilicity at phosphorus which could fa-
vor the competing reaction of 4 with the corresponding ‘‘Phospha-
Wittig” reagent (12) once it is formed. The resulting products are
then the kinetically unstable symmetric diphosphene (HF4C6)2P2
which may undergo 2 + 2 cycloaddition to 5 and PMe3Cl2. In our
studies we were able to form the ‘‘Phospha-Wittig” reagent already
in the absence of Zn dust. Compound 12 could be detected by 31P
did not trigger any conversion. Some conversion could be achieved
using SbF3 in a toluene/diethyl ether mixture. The reaction pro-
ceeds slowly and is accompanied by by-products. For the reaction
of 8 with SbF3 two new resonances are detected in the 31P NMR
1
1
spectra at 160.9 ppm (d, JPF 1044 Hz) (9) and 135.0 ppm (d, JPF
1007 Hz) besides those of the starting material. Also in the MS
spectra the formation of the desired product 9 could be confirmed.
The second product which was formed in about 20% could not be
identified unambiguously, but an addition/oxidation product with
antimony can be expected. The reaction of the 8 with KF in chloro-
form-d1 turned out to be the most successful approach. Complete
conversion into the fluorine derivative is achieved after 4 h at
slightly elevated temperatures (i.e. 50 °C). This reaction seems to
be quite solvent dependent since no conversion was obtained if
dichloromethane was used as solvent instead of chloroform. Com-
pound 9 shows a phosphorus NMR resonance at 160.9 ppm. In the
19F NMR spectra a typical doublet at ꢀ123.9 ppm (1JPF 1043.6 Hz)
was observed. To our surprise also in this case the envisaged elim-
ination of TMS-F could not be achieved. The series of unsuccessful
attempts to induce halosilane eliminations (AlCl3, TMS-N3, Ag-tri-
flate) for these fluoroaryl substituted halophosphanes suggests a
very strong affinity of the phosphorus atom towards the adjacent
halogen. This behavior may be reasonably explained by the elec-
tron withdrawing character of the aryl group, which increases
the electrophilicity of the phosphorus atoms thus strengthening
the P–X bond.
Calculations of free energies for these reactions indicate that the
formation of the iminophosphane from 8 is slightly more favorable
than from 9, but the energy differences are very small and solva-
tion effects might alter these findings drastically. The energy bal-
ance of the calculations shows that in principle the reactions are
thermodynamically favored. Naturally, the energy barrier of the
transition state will be relevant as well, but is hard to estimate
and even harder to locate on the PES, owing to intermolecular pro-
cesses involved. In conclusion, the calculations suggest that intro-
duction of a fluorine atom at phosphorus does not affect the
elimination reaction for the HF4C6-substituted compounds signifi-
cantly. Elimination to the corresponding iminophosphanes is ther-
modynamically feasible but no appropriate reaction conditions
have been found in our hands. Table 1 summarizes the energy dif-
ferences for reactions: (i) 8 ? R–P@N-TMS + TMS-Cl and (ii)
9 ? R–P@N-TMS + TMS-F [25].
1
spectroscopy from the crude reaction mixture (28.2 ppm, d, JPP
1
403 Hz and ꢀ55.8 ppm, d, JPP 403 Hz). In this mixture tetrapho-
sphetane 5 and PMe3Cl2 are the only observed products after 1 day.
Formal replacement of one phosphorus atom in diphosphenes
by nitrogen leads to the class of iminophosphanes. The synthetic
strategy towards iminophosphanes usually involves the elimina-
tion of LiCl in the first step forming an aminophosphane [23]. If a
lithium amide of a primary amine is used (route a, Scheme 2),
the second step proceeds either via lithiation of RPCl–NHR0 fol-
lowed by further LiCl elimination or via dehydrohalogenation with
non-nucleophilic bases. Similarly, in the reaction of dichlorophos-
phane 4 with LiNTMS2 the first step also involves LiCl elimination,
the second step however is driven by the elimination of TMSCl. In
many cases the latter step proceeds only at elevated temperatures
which may require heating. To lower the required temperature for
the silyl elimination, chlorine may be replaced by fluorine mini-
mizing the possibility of thermally induced dimerization or poly-
merization. By contrast, the trend is just the opposite for lithium
halide elimination (TBr < TCl < TF) [23].
To achieve P–N bond formation, we reacted one equivalent of
Wannagat’s base [24] with 4 at a low temperature. LiCl is filtered
off and after removal of the solvent product 8 is obtained as a col-
orless oil. It shows phosphorus spectra similar to
3 (br. s,
119.8 ppm), where couplings to fluorine and hydrogen could not
be resolved either. Further attempts to convert aminophosphane
8 into an iminophosphane failed in our hands. Even at elevated
temperatures (up to 165 °C) the elimination of TMS-Cl did not oc-
cur indicating unusual thermal stability of this chlorophosphane
(vide infra). As already mentioned before, the elimination might
be facilitated by fluorine exchange owing to the strength of the
Si–F bond. Therefore, we tried to exchange Cl by F using different
fluorinating reagents like AgF, SbF3. Unfortunately, none of them
led to the envisaged fluorophosphane in a clean reaction. Different
activators like [18]-crown-[6] or silver triflate were employed but
Alternatively we tried to synthesize iminophosphanes via the
second route (Scheme 2), by the reaction of a primary lithium
amide with 3. At a low temperature t-Bu–NHLi gives a clean reac-
tion with 3 and the desired amino phosphane (10) is obtained
quantitatively as colorless oil. The phosphorus NMR signal is in
the expected region (89.0 ppm) showing the expected coupling
3
3
pattern of a doublet of triplets arising from JPF and JPH, respec-
tively. The constitution of this compound is further corroborated
by 2D-NMR experiments. Addition of a base to this compound
should lead to the abstraction of the NH proton, but likewise care
must be taken to avoid Ar-H activation. In an initial attempt n-BuLi
was used at low temperatures (ꢀ80 °C), but instead of acting as a
base it turned out to undergo substitution of the chlorine atom
resulting in the formation of HC6F4P(n-Bu)–NH(t-Bu) (11). The
31P signal is shifted to higher field (18.5 ppm) and constitution of
Scheme 2. Formation of 8, 9, 10 and possible ways to iminophosphanes, (a) R0 = t-
Bu, R0-NHLi, (b) LiN(TMS)2ꢁEt2O for 8 X = Cl and 9 X = F and R = TMS.
Table 1
Absolute energies of TMS-X (X = Cl, F), 8, 9, and R–P@N-TMS in [hartree]. Reaction energies for reactions (i) and (ii) in [kcal/mol].
TMS-Cl
TMS-F
8
9
R–P@N-TMS
(i)
9.6
7.5
ꢀ6.7
(ii)
7.5
10.6
ꢀ9.1
D
D
D
Eelec
Eelec + ZPE
G298
ꢀ869.6156
ꢀ869.5034
ꢀ869.5350
ꢀ509.2745
ꢀ509.1529
ꢀ509.1928
ꢀ2303.8559
ꢀ2303.5662
ꢀ2303.6229
ꢀ1943.5116
ꢀ1943.2207
ꢀ1943.2770
ꢀ1434.2251
ꢀ1434.0509
ꢀ1434.0987

978
A. Orthaber et al. / Journal of Organometallic Chemistry 695 (2010) 974–980
11 was confirmed by 2D-NMR spectroscopy. Further attempts to
induce HCl elimination of 10 with Wannagat’s base or LDA did
not give any satisfactory results. Reaction of 10 with DBU gave so
far unidentified products, but formation of the iminophosphane
or the corresponding 2 + 2 cycloaddition product can be excluded
by comparison of 31P NMR shifts with literature data [26].
ature (30 °C). The excess of metal is removed by filtration and
washed with dichloromethane. The filtrate is dried in vacuum
yielding almost quantitatively 2 as a white crystalline solid
(76 mg, 94%). Crystals suitable for X-ray diffraction were obtained
by recrystallization from chloroform-d1. M.p.: 183 °C; 1H NMR
(C6D6, 400 MHz): (m, 1H, aryl-H), (m, 8H), (t, 12H); 13C{1H} NMR
3
2
(CDCl3, 100 MHz), 14.10 (d, JPC 2.3 Hz), 41.43 (d, JPC 5.3 Hz),
3
1
1
108.32 (t, JCF 21.5 Hz), 118.68 (d, JPC 89.5 Hz), 145.30 (d, JPF
3. Conclusion
251.0 Hz), 146.1 (d, JPF 254.0 Hz). 31P{1H} NMR (CDCl3,
1
162 MHz): d = 55.3 (br.s. JPSe 798 Hz). EI-MS [m/z]: M+ 404.1
1
In summary we have explored the reactivity of a series of phos-
phanes with a new fluoroaryl group (HF4C6-), which carries a spec-
tator function in para position. An asymmetric diphosphene is
obtained as an intermediate in the synthesis of a linear 1,3-dihy-
drophosphane, while the symmetric diphosphene undergoes
2 + 2 cycloaddition under the formation of the corresponding
cyclotetraphosphetane for which a crystal structure could be ob-
tained. Attempts to synthesize HF4C6-substituted iminophos-
phanes generally failed, which could be attributed to the
electronic nature of the corresponding precursors as suggested
by quantum chemical calculations. Further studies are necessary
to understand the reactivity of these phosphanes in order to devel-
op strategies to incorporate this substituent into phosphorus rich
materials with interesting bonding situations and reactivity.
(7%), M+ꢀSe 323.1 (9%), M+ꢀSeꢀNEt2 252.1 (100%), P(@Se)NEt+
+
181.0 (10%), NEt2 72.1 (24%).
4.3. Synthesis of 3
Distillation of the crude mixture of 3 and Cl2PNEt2 (ca. 15 g)
gives at higher temperatures scrambling reactions yielding 2 as
the distillation residue. Extraction of the residue with Et2O gives
2 besides small amounts of 3. Yield: ca. 0.7 g, 2.4 mmol. 1H NMR
3
3
(C6D6, 400 MHz): 1.11 (6H, t, JHH 7.10 Hz, CH3), 3.21 (4H, dq, JPH
3.2 Hz, 3JHH 7.5 Hz, CH2), 7.13 (1H, m, aryl-H); 13C{1H} NMR (CDCl3,
100 MHz): 14.02 (d, 3JPC 6.3 Hz, CH3), 45.20 (d, 2JPC 17.2 Hz), 108.33
3
1
3
(t, JCF 22.7 Hz), 119.21 (dt, JPC 62.2 Hz, JCF 17.9 Hz), 145.95 (dm,
1JCF 247.8 Hz), 146.33 (dm, JCF 248.9 Hz); 31P{1H} NMR (CDCl3,
1
162 MHz): 112.7 (ttm, JPF 55.2 Hz, JPF 11.9 Hz). EI-MS [m/z]: M+
3
4
4. Experimental
287.0 (57%, Cl pattern), M+ꢀCH3 272.0 (100%), M+ꢀCl 252.1
+
(61%), M+ꢀNEt2 214.9 (63%), ClPNEt2 138.0 (23%), PNEt2Cl+–CH3
All reactions are carried out with a modified Schlenk technique.
Solvents are degassed and dried over alumina using the PureSolv
system. All reagents are obtained from ABCR (1,2,4,5-tetrafluoro-
benzene) or Sigma–Aldrich (n-BuLi, t-BuNH2, LDA) and used with-
out further purification. DBU was dried and degassed prior to use.
LiNTMS2ꢁEt2O is prepared according to a published procedure [24].
NMR measurements are carried out on a Varian Unity INOVA400
and a Bruker AvanceIII 300 operating at 400 MHz an 300 MHz pro-
ton frequencies, respectively. X-ray studies were performed on a
STOE four-circle diffractometer or a Bruker SMART-APEX II diffrac-
tometer with a CCD detector. Mass spectra were recorded with an
Agilent 5975C mass spectrometer equipped with a direct injection
unit (DI-EI) at current of 70 eV.
127.0 (23%).
4.4. Synthesis of 4
Compound 1 (18.2 g, 56.1 mmol) was dissolved in Et2O (300 ml)
and cooled with an ice bath, while anhydrous HCl is bubbled over
the solution. A white voluminous precipitate is formed immedi-
ately. Stirring is continued for 60 min under continued HCl stream.
The precipitate is removed by filtration and washed with diethyl-
ether. The solvent is removed under reduced pressure (100 mbar)
and the product is collected by fractional distillation (4 mbar,
54 °C, 8.1 g, 58%). 1H NMR (C6D6, 400 MHz): 6.31 (m, 1H, aryl-H);
3
4
13C{1H} NMR (CDCl3, 100 MHz): 110.60 (tt, JCF 22.7 Hz, JCF
1
2
1
1.5 Hz), 119.17 (dt, JPC 80.7 Hz, JCF 17.4 Hz), 145.54 (dm, JCF
4.1. Synthesis of 1
250.9 Hz), 146.43 (dm, JCF 253.9 Hz); 31P{1H} NMR (CDCl3,
1
162 MHz): 135.9 (t, JPF 63.7 Hz) 19F{1H} ꢀ131.2 (dm., JPF
63.0 Hz, o-aryl-F) ꢀ136.4 (m, m-aryl-F). EI-MS [m/z]: M+ 349.9
(3%), M+ꢀCl 214.0 (100%), M+ꢀPCl2 99.0 (55%), 150.0 (98%), PCl+
65.0 (61%).
3
3
To a solution of 1,2,4,5-tetrafluorobenzene (13.46 g, 89.7 mmol)
in THF one equivalent of n-BuLi (hexane solution 1.6 M, 56 ml,
89.6 mmol) is added in ca. 80 ml THF at ꢀ100 °C over a period of
1 h. Stirring is continued for 1/2 h before ClP(NEt2)2 (19.30 g,
91.6 mmol) is added very slowly. The reaction mixture is stirred
for 1 h at ꢀ100 °C then 1 h at ꢀ60 °C and subsequently allowed
to warm to room temperature over night with continued stirring.
The resulting precipitate is removed by filtration and from the
solution the solvent is stripped off in vacuum yielding 23.48 g
spectroscopically pure product (81%) which can be distilled b.p.:
54 °C (0.05 mbar). 1H NMR (C6D6, 400 MHz): 6.88 (m, 1H, aryl-H),
4.5. Synthesis of (tetrakis(2,3,5,6-tetrafluorophenyl)-
tetraphosphetane), 5
Compound 4 (154 mg, 0.6 mmol) is dissolved in 3 ml Et2O and
3 ml dichloromethane and magnesium powder (100 mg, 4.1 mmol)
is added. The resulting mixture is stirred for 3 days at 30 °C. All vol-
atiles are removed in vacuum and the remaining residue is ex-
tracted with diethylether. After recrystallization from chloroform
colorless crystals suitable for X-ray diffraction are obtained
(50 mg, 47%). 1H NMR (CDCl3, 400 MHz): 7.03 (m, 1H, aryl-H);
3
3.04 (m, 8H, CH2), 1.02 (t, JHH, 12H, CH3); 13C{1H} NMR (CDCl3,
3
2
100 MHz), 14.73 (d, JPC 3.5 Hz, CH3), 44.27 (d, JPC 18.6 Hz, CH2),
104.89 (t, JCF 23.1 Hz, Caryl-H), 123.14 (m, Caryl-P), 31P{1H} NMR
3
(CDCl3, 162 MHz): d = 84.2 (br.s). EI-MS [m/z]: M+ 324.1 (5%), M+-
+
NEt2 252.1 (23%), P(NEt2)+ 175.2 (44%), NEt2 73.0 (62%), NEt+
3
13C{1H} NMR (CDCl3, 100 MHz): 108.15 (t, JPF 23 Hz, aryl-H),
44.0 (77%) Et+ 28.1 (100%).
1
1
145.82 (dm, JPF 253 Hz), 146.91 (dm, JPF 244 Hz), Caryl-P n.d.
31P{1H} NMR (CDCl3, 162 MHz): ꢀ63.3 (m); 19F{1H} ꢀ126.2 (m, o-
aryl-F), ꢀ137.0 (m, m-aryl-F).
4.2. Synthesis of 2
Compound
5 is also obtained by reaction of 4 (237 mg,
To a solution of 1 (65 mg, 0.2 mmol) in dichloromethane (2 ml)
an excess of gray selenium (158 mg, 2.0 mmol) is added and the
resulting mixture is stirred over night at slightly elevated temper-
0.9 mmol) with PMe3 (1.5 ml, 1 M in hexane) in the presence of
Zn dust (ca. 100 mg, 1.5 mmol) in 3 ml CDCl3. The mixture is stir-
red for 24 h at ambient temperatures. After removal of all volatiles

A. Orthaber et al. / Journal of Organometallic Chemistry 695 (2010) 974–980
979
the residue is extracted with chloroform-d1 and benzene yielding
pure 5 in quantitative yield (160 mg, >99%).
ꢀ80 °C. After stirring over night, the solvent is removed and the
residue extracted with pentane yielding the crude product
4
(725 mg, 74%). 1H NMR (C6D6, 300 MHz): 1.35 (d, JPH 1.5 Hz, 9H,
2
4.6. Attempted synthesis of diphosphenes 6 and resulting side products
7
t-Bu), 3.38 (d, JPH 4.4 Hz, 1H, NH), 7.12 (m, 1H, Ar-H); 13C NMR
3
2
(C6D6, 100.6 MHz): 31.31 (d, JPC 11.5 Hz, CH3), 53.90 (d, JPC
2
1
16.3 Hz, CCH3), 108.37 (t, JCF 22.6 Hz, Caryl-H), 121.45 (dt, JPC
2
1
*
2,4,6-Tris-t-butyl-phenylphosphane
(Mes PH2,
163 mg,
48.2 Hz, JCF 68.0 Hz, Caryl-P), 145.97 (br d, JCF 250.1 Hz, o,m-Car-
3
0.6 mmol) is dissolved in ca. 10 ml Et2O and cooled to ꢀ10 °C
yl-F); 19F NMR (C6D6, 282.4 MHz): ꢀ134.4 (dm, JPF 59.2 Hz, o-
and one equivalent (370 l) n-BuLi is added very slowly. The slurry
l
aryl-F), ꢀ137.33 (m, m-aryl-F); 31P{1H} NMR (C6D6, 121.5 MHz):
3
3
is allowed to warm to room temperature and stirred over night.
89.0 (t, JPF 59.2 Hz), 31P NMR (C6D6, 121.5 MHz): 89.0 (td, JPF
Mes -PHLi is added to a cooled solution (ꢀ80 °C, 10 ml Et2O) of 4
59.2 Hz, JPH 8.9 Hz). EI-MS [m/z]: M+ 287.1 (22%, Cl pattern),
2
*
(125 mg, 0.5 mmol). The solution is allowed to warm to room tem-
perature within 2 h. This intermediate was not isolated, but di-
rectly reacted with one equivalent DBU at ꢀ80 °C. The precipitate
is removed and the filtrate is dried in vacuum yielding the crude
product as the above described mixture. After recrystallization
from pentane/C6D6 the product was obtained as a mixture of the
diphosphene and other side products. Owing to the crude mixture
of products only 31P resonances are reported. Compound 6: 31P
M+ꢀCH3 272.1 (100%), M+ꢀCl 252.0 (37%), M+ꢀBu 231.1 (55%),
M+ꢀNHBu 215.0 (65%), M+ꢀBuCl 196.0 (99%), 127.0 (25%), Bu+
57.1 (54%).
4.10. Reaction of 10 with n-BuLi (11)
Reaction of 10 (170 mg, 0.6 mmol) with one equivalent n-BuLi
(380 lL, 0.6 mmol) is carried out in Et2O at low temperatures
1
3
NMR (C6D6, 121.5 MHz): 565.7 (dt, JPP 555 Hz, JPF 21.3 Hz,
(ꢀ80 °C). The precipitate is filtered off and an oily residue is ob-
1
*
*
tained after removal of the solvent in 60% yield (111 mg). 1H
HF4C6-P), 391.3 (d, JPP 555 Hz, Mes -P). Compound 7 Mes PA(H)–
F
1
1
*
3
PB(Mes )–PC(H)R : ꢀ49.5 (ddt JPP 307 Hz, JPP 223 Hz, PB) ꢀ55.1
NMR (CDCl3, 400 MHz): 0.74 (3H, t, JHH 7.0 Hz, CH3), 1.01 (s, 9H,
1
2
1
3
(dd JPP 307 Hz, JPP 17 Hz, PA), ꢀ75.8 (dtd, JPP 223 Hz, JPF
t-Bu CH3), 1.22 (m, 2H, CH2–CH3), 1.60 (m, 2H, CH2–P), 1.80 (m,
2
2
63.1 Hz, JPP 17 Hz, PC).
2H, CH2CH2CH2) 2.25 (1H, d, JPH 12.4 Hz, NH), 6.23 (m, 1H, Caryl
-
H); 13C NMR (100.6 MHz): 13.6 (s, CH2CH3) 27.40 (d, JPC 17.2 Hz),
31.52 (d, JPC 9.4 Hz C(CH3)3) 31.82 (m, PCH2), 51.08 (2JPC 16.4 Hz
3
4.7. Synthesis of 8
C(CH3)3). EI-MS [m/z]: M+ 309.1 (7%, Cl pattern), M+ꢀCH3 293.2
(27%), M+ꢀBu 251.1 (14%), M+ꢀNHBu 237.1 (38%), 210.1(42%),
M+ꢀ2xBu 196.0 (100%), P(NH)Bu+ 127.1 (12%).
To a cooled solution of LiNTMS2ꢁEt2O (120 mg, 0.5 mmol) in
15 ml toluene one equivalent of 4 (124 mg, 0.5 mmol) is added
slowly. The mixture is warmed to room temperature and the initial
clear solution becomes a cloudy and slightly yellow suspension.
The reaction progress was monitored by 31P NMR spectroscopy.
Complete consumption of the 4 was observed after one day where
the formation of a single reaction product with a phosphorus res-
onance at 119.8 ppm could be observed. After removal of the sol-
vent, 8 is obtained as a colorless oil (b.p. > 300 °C). 1H NMR
Acknowledgements
Financial support by the Austrian Science Fund (FWF) (Grants
P18591-03 and P20575-N19) and the EU-COST Action CM0802
‘‘PhoSciNet” is gratefully acknowledged.
3
(C6D6, 300 MHz): 0.23 (d, JPH 1.5 Hz, 18H, CH3), 6.16 (m, 1H,
Appendix A. Supplementary material
aryl-H); 13C NMR (by HMBC/HSQC measurements): 3.13 (9C,
1
SiMe3), 107.00 (1C Caryl-H) 146.04 (dm, JCF 261.9 Hz, o,m-aryl);
CCDC 748985 and 748984 contain the supplementary crystallo-
graphic data for 2 and 5. These data can be obtained free of charge
from The Cambridge Crystallographic Data Centre via http://
www.ccdc.cam.ac.uk/data_request/cif. Supplementary data associ-
ated with this article can be found, in the online version, at
doi:10.1016/j.jorganchem.2009.10.025.
19F NMR (C6D6, 282.4 MHz): ꢀ133.2 (m, o-aryl-F), ꢀ138.4 (m, m-
aryl-F); 31P NMR (C6D6, 121.5 MHz): 119.8 (br s); EI-MS [m/z]:
M+ 375.0 (Cl pattern 5%), M+ꢀCH3 360.0 (22%), M+ꢀCl 340 (98%),
M+ꢀTMSCl 267.1 (25%), 248.0 (33%), 122.0 (37%), TMS+ 73.1
(100%).
4.8. Synthesis of 9
References
To a solution of 8 in CDCl3 (41 mg in 650 ll CDCl3) a large excess
[1] Y. Ren, Y. Dienes, S. Hettel, M. Parvez, B. Hoge, T. Baumgartner, Organometallics
28 (2009) 734–740.
of strictly anhydrous KF was added. The mixture was stirred for
1 day at slightly elevated temperatures (65 °C) and complete con-
version was monitored by 31P NMR spectroscopy. After removal
of the solvent the residue is extracted with Et2O giving pure prod-
uct after evaporation of the solvent (40 mg, 91%). 1H NMR (CDCl3,
300 MHz): 0.26 (br. s, 18H, CH3), 7.05 (m, 1H, Caryl-H); 31P NMR
(CDCl3, 121.5 MHz): 160.9 (1JPF 1045.6 Hz); 19F NMR (CDCl3,
[2] B. Hoge, W. Wiebe, Angew. Chem., Int. Ed. 47 (2008) 8116–8119.
[3] P. Le Floch, Coord. Chem. Rev. 250 (2006) 627–681.
[4] M. Hissler, C. Lescop, R. Réau, C.R. Chim. 11 (2008) 628–640.
[5] M.G. Hobbs, T. Baumgartner, Eur. J. Inorg. Chem. (2007) 3611–3628.
[6] T. Baumgartner, R. Reau, Chem. Rev. 106 (2006) 4681–4727.
[7] C. Lescop, M. Hissler, in: B. Pignataro (Ed.), Tomorrow’s Chemistry Today, 2009,
pp. 295–319.
[8] J. Grobe, D. Le Van, B. Broschk, M. Hegemann, B. Luth, G. Becker, M. Bohringer,
E.-U. Würthwein, J. Organomet. Chem. 529 (1997) 177–187.
[9] P. Espinet, A.C. Albéniz, J.A. Casares, J.M. Martínez-Ilarduya, Coord. Chem. Rev.
252 (2008) 2180–2208.
[10] B. Hoge, J. Bader, J. Fluorine Chem. 128 (2007) 857–861.
[11] P.A. Chase, D.W. Stephan, Angew. Chem., Int. Ed. 47 (2008) 7433–7437.
[12] P.A. Chase, G.C. Welch, T. Jurca, D.W. Stephan, Angew. Chem., Int. Ed. 119
(2007) 8196–8199.
[13] M.L. Clarke, D. Ellis, K.L. Mason, A.G. Orpen, P.G. Pringle, R.L. Wingad, D.A.
Zaher, R.T. Baker, Dalton Trans. (2005) 1294–1300.
4
282.4 MHz): ꢀ123.9 (1JPF 1043.4, JPF 26.3 Hz), ꢀ136.3 (m, m-
aryl-F), ꢀ138.4 (m, o-aryl-F). EI-MS [m/z]: M+ 359.1 (29%), M+ꢀF
340.0 (60%), M+ꢀTMS 286.1 (40%), M+ꢀNTMS2ꢀF 192.9 (100%),
+
PNTMS2 191.1 (76%), TMSF+ 92.0 (31%).
4.9. Synthesis of 10
[14] M. Scholz, H.W. Roesky, D. Stalke, K. Keller, F.T. Edelmann, J. Organomet. Chem.
366 (1989) 73–85.
[15] C. Moser, F. Belaj, R. Pietschnig, Chem. Eur. J. (2009), in press, doi:10.1002/
chem.200902255.
To t-butyl amine (250 mg, 3.42 mmol) in heptane/Et2O (25 ml,
1:1) one equivalent of n-BuLi (2.14 ml, 3.42 mmol) is added at
0 °C. The mixture is stirred for 1 h and warmed to room tempera-
ture. The addition of the dichlorophosphane 3 is then carried out at
[16] C. Moser, M. Nieger, R. Pietschnig, Organometallics 25 (2006) 2667–2672.

980
A. Orthaber et al. / Journal of Organometallic Chemistry 695 (2010) 974–980
[17] N. Nagahora, T. Sasamori, N. Takeda, N. Tokitoh, Organometallics 24 (2005)
3074–3080.
[18] P. Barbaro, M. Di Vaira, M. Peruzzini, S.S. Costantini, P. Stoppioni, Angew.
Chem., Int. Ed. 47 (2008) 4425–4427.
[19] J.M. Lynam, Angew. Chem., Int. Ed. 47 (2008) 831–833.
[20] A.G. Massey, A.J. Park, J. Organomet. Chem. 5 (1966) 218–225.
[21] F. Sanz, J.J. Daly, J. Chem. Soc., Chem. Commun. (1971) 1083–1086.
[22] R.C. Smith, X. Chen, J.D. Protasiewicz, Inorg. Chem. 42 (2003) 5468–5470.
[23] E. Niecke, D. Gudat, Angew. Chem., Int. Ed. 30 (1991) 217–237.
[24] U. Wannagat, H. Niederprüm, Chem. Ber. 94 (1961) 1540–1547.
[25] Calculations on the corresponding non-fluorinated congeners showed that the
reactions energies are of the same order of magnitude. Thus influence of the
the perfluorinated groups seems equal on both asymptotes of the eliminiation
reaction and the energy barrier due to the transistion state is crucial for this
reactivity (further information in the supplementary material).
[26] K. Miqueu, J.-M. Sotiropoulos, G. Pfister-Guillouzo, V.L. Rudzevich, H.
Gornitzka, V. Lavallo, V.D. Romanenko, Eur. J. Inorg. Chem. (2004) 2289–2300.