Coordination chemistry of cyclopropenylidene-stabilized phosphenium cations: Synthesis and reactivity of Pd and Pt complexes

Article abstract of DOI:10.1002/chem.201303686

A straightforward synthesis of cyclopropenylidene-stabilized phosphenium cations 1 a-g through the reaction of [(iPr₂N)₂C ₃⁺Cl]BF₄ with secondary phosphines is described. Their donor ability was evalu

Full text of DOI:10.1002/chem.201303686

DOI: 10.1002/chem.201303686
Full Paper
&
Cationic Ligands
Coordination Chemistry of Cyclopropenylidene-Stabilized
Phosphenium Cations: Synthesis and Reactivity of Pd and Pt
Complexes
ꢀ
gnes Kozma, Tobias Deden, Javier Carreras, Christian Wille, Jekaterina Petu sˇ kova,
Jçrg Rust, and Manuel Alcarazo*
[a]
0
Abstract: A straightforward synthesis of cyclopropenylidene-
stabilized phosphenium cations 1a–g through the reaction
ence of Pd species, an oxidative insertion of the Pd atom
into the Ccarbene–phosphorus bond takes place, providing di-
meric structures in which each Pd atom is bonded to a cyclo-
propenyl carbene while two dialkyl/diaryl phosphide ligands
serve as bridges between the two Pd centers. The catalytic
+
of [(iPr N) C Cl]BF with secondary phosphines is described.
2
2
3
4
Their donor ability was evaluated by analysis of the CO
stretching frequency in Rh complexes [RhCl(CO)L ](BF ) and
2
4 2
II
electrochemical methods. The cyclopropenium ring induces
a phosphite-type behavior that can be tuned by the other
two substituents attached to the phosphorus atom. Despite
of the positive charge that they bear, phosphenium cations
performance of the resulting library of Pt complexes was
tested; all of the cationic phosphines accelerated the proto-
type 6-endo-dig cyclization of 2-ethynyl-1,1’-biphenyl to
afford pentahelicene. The best ligand 1g was used in the
synthesis of two natural products, chrysotoxene and epime-
doicarisoside A.
1
a–g still act as two-electron donor ligands, forming ad-
II
II
ducts with Pd and Pt precursors. Conversely, in the pres-
Introduction
ties than their neutral counterparts. In this context, our group
has actively studied the synthesis and reactivity of cycloprope-
[3]
The ability to adjust finely the donor properties and steric
demand of the ligands attached to a metal center is crucial in
metal catalysis because it allows a dramatic control on the cat-
alyst reactivity and product selectivity. For this reason, phos-
phines, which can be readily modulated by just changing the
nature and size of the substituents on phosphorus, are proba-
bly the most employed and versatile ancillary ligands in the
nium-substituted phosphines. These compounds can be for-
mally described as intermediates between two mesomeric
forms: a phosphine containing a positively charged cyclopro-
[4,5]
+
penium substituent A,
or a phosphenium cation [PR2] sta-
bilized by donation from a cyclopropenylidene internal ligand
B (Scheme 1).
[
1]
catalysis arena. Specifically, when high Lewis acidity is re-
quired at a metal center to promote a transformation, poor do-
nating phosphites or polyfluorinated phosphines are the tradi-
tional ligands of choice.
Recently, an alternative approach for the design of weak
donor phosphines has emerged that makes use of cationic
[2]
substituents directly attached to the phosphorus atom. The
positive charge thus introduced lowers the energy of all of the
orbitals of the phosphine, including those responsible for the
s-donation and the p-back donation. This results in cationic
phosphines depicting weaker neat electron-releasing proper-
Scheme 1. Conceivable resonance extremes for cyclopropenylidene-stabi-
lized phosphenium cations.
Herein, we describe a general procedure for the synthesis of
these ligands, the evaluation of their donor ability by spectro-
scopic and electrochemical methods, and our studies on their
reactivity towards Pd and Pt centers. These experiments indi-
cate that both mesomeric forms A and B are in fact necessary
to rationalize their reactivity. Moreover, the catalytic activity of
a library of Pt complexes was tested on the model cycloisome-
rization of ortho-biaryl substituted alkynes into phenanthrenes.
[
a] ꢀ. Kozma, T. Deden, Dr. J. Carreras, C. Wille, Dr. J. Petu sˇ kova, J. Rust,
Dr. M. Alcarazo
Max-Planck-Institut fꢁr Kohlenforschung
Kaiser Wilhelm Platz 1
4
5470-Mꢁlheim an der Ruhr (Germany)
Fax: (+49)208-306-2994
E-mail: alcarazo@mpi-muelheim.mpg.de
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201303686.
These results defined 6g/AgBF as the most efficient catalytic
4
Chem. Eur. J. 2014, 20, 2208 – 2214
2208
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
system for this reaction. Finally, the impact of this catalytic mix-
ture was also demonstrated through the synthesis of phenan-
threne-derived natural occurring products, namely chrysotox-
ene and epimedoicarisoside A.
pound 1g, containing two very electron-withdrawing 3,5-bis-
(trifluoromethyl)phenyl substituents. This phosphine could not
be synthesized following the general procedure because the
low nucleophilicity of bis(trifluoromethyl)phenyl phosphine
prevents its direct condensation with 2. However, after basic
treatment with nBuLi, the phosphide generated in situ does in
fact react with salt 2, affording the desired 1g in moderate
yield.
Results and Discussion
We recently targeted the synthesis of cyclopropenium-substi-
tuted phosphines 1a–f by condensation of the readily avail-
able 1-chloro-2,3-bis(diisopropylamino)cyclopropenium tetra-
fluoroborate 2 with an array of secondary phosphines. Thus,
the desired salts 1a–f could be obtained in good to excellent
yields as air-stable white solids (Scheme 2). Furthermore, to
further increase the range of electronic properties of cyclopro-
penium-substituted phosphines, we have now prepared com-
Scheme 2 also shows the X-ray structures of compounds 1c
and 1 f. Interestingly, both C1ꢀP1 bond distances (C1ꢀP1
1
.820 ꢂ in 1c and 1.804 ꢂ in 1 f) are slightly shorter than the
other CꢀP single bonds present in the same molecules (C4ꢀP1,
.878 ꢂ in 1c and 1.827 ꢂ in 1 f), indicating a weak but still
1
[
3a]
measurable interaction between the phosphorus lone pair and
the p-system of the cyclopropenium cation. This delocalization
of the non-shared pair at phosphorus along the positively
charged substituent, which can be understood as a back-dona-
tion from phosphorus, is common in onium-substituted phos-
phines, and in extreme cases, may preclude the efficient over-
lap between the lone pair at P and the orbitals of a metal
avoiding the formation of the corresponding complexes.
Donor-releasing properties
The donor abilities of phosphines 1a–g were then evaluated
by classical analysis of the CO stretching frequencies in Rh car-
bonyl complexes 3a–g. These preliminary data indicated that
ligands 1a–f share with aryl phosphines the same area of the
Tolman stereoelectronic map, whereas 1g more closely ap-
proaches the values measured for phosphites (Scheme 3 and
Table 1). It is of note that these results are counterintuitive
2 4 2
Table 1. Carbonyl stretching frequencies in [RhCl(CO)L ](BF ) complexes
in the solid state and electrochemical redox potential of the ligands. The
values of commonly used phosphorus ligands are also included for com-
parison.
[
a]
Ligand
n˜ CO
E
p
[
b]
[
RhCl(CO)L
2
](BF
4
)
2
ox
1
1
1
1
1
1
a
c
d
e
f
1971
–
1.207
1.185
–
1.040
1.246
1.548
0.697
1.287
[
c]
[
c]
1969
1969
1976
1991
1979
2011
g
3
P
Ph
(
MeO) P
3
ꢀ
1
[
a] Values in cm . [b] Oxidation peak potentials reported in V. Calibrated
versus ferrocene/ferrocenium (E1/2 =0.00 V), Bu NPF (0.1m) in CH Cl
c] Not determined.
4
6
2
2
.
[
given the cationic nature of our ligands and they certainly
should be taken with caution. It is known that in Rh complexes
the IR frequencies of the cis-located CO ligand may not be
only sensitive to electronic effects but also reflect geometrical
distortions around the metal center owing to steric hindrance
or through-space interactions between the CO and the other
ligands. For these reasons, electrochemical methods were
[
6]
Scheme 2. Synthesis of 1a–f and X-ray structure of 1c and 1 f. Hydrogen
atoms are omitted for clarity; ellipsoids are set at 50% probability. Reagents
and conditions (product yields in parentheses): a) phosphine, 2 equiv, THF,
reflux, 24 h; b) NaBF
4
excess; 1a (90%); 1b (86%), 1c (79%); 1d (96%); 1e
76%); 1 f (80%); c) [3,5-bis(CF )C
PH, (1 equiv), ꢀ788C, THF, nBuLi
1 equiv) and then 2 (1 equiv), 608C, 48 h, 1g (38%).
(
(
3
6 3 2
H ]
Chem. Eur. J. 2014, 20, 2208 – 2214
www.chemeurj.org
2209
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
obtain the corresponding Pt allyl analogue 5a. In this case, the
31
diagnostic appearance of satellite signals in the P NMR spec-
1
trum ( J =2188.0 Hz) confirmed the formation of a PtꢀP
P–Pt
bond.
Furthermore, compounds 1a, 1d, 1 f, and 1g were reacted
with K [PtCl ], affording the trichloroplatinate complexes 6a,
2
4
6
d, 6 f, and 6g, respectively, as intensely yellow solids. In
these cases again the presence of satellite signals in the
31
1
P NMR spectrum with J =2011.0 Hz (6a), 2000.8 Hz (6d),
P–Pt
2016.8 Hz (6 f), and 2037.3 Hz (6g) indicated the formation of
the desired adducts, an extreme that could be later confirmed
by crystallographic analysis in the case of 6d. As expected, co-
ordination of the P atom to a metal center cancels any possi-
ble back-donation from the phosphorous to the cyclopropeni-
um ring. Thus, a slight elongation of the C1ꢀP1 bond is ob-
served in the coordinated phosphine moieties when compared
to the free ligands.
In an attempt to further study the reactivity of cyclopropeni-
um-substituted phosphines, we allowed a solution of 1a in di-
chloromethane to react with [Pd (dba) ] at room temperature.
2
3
Under these conditions a characteristic yellow color gradually
appeared together with the concomitant release of dba. After
work up, the newly formed compound exhibited a very indica-
31
tive P NMR upfield signal at (d=ꢀ130.6 ppm), which cannot
[
6]
Scheme 3. Synthesis of 3a–g and X-ray structure of 3e. Hydrogen atoms
and tetrafluoroborate counteranions are omitted for clarity; ellipsoids are set
at 50% probablility. Reagents and conditions (product yields in parentheses):
be attributed to the phosphorus atom of any of our phos-
phines; in fact, that is the region where anionic phosphide li-
ꢀ
a) phosphine (2 equiv), [{RhCl(CO)}
2
], (1 equiv) THF, ꢀ208C!RT, 3a (93%);
gands [R P] appear. Furthermore, the ESI mass spectrometry
2
3
b (98%), 3d (98%); 3e (98%); 3 f (97%); 3g (93%).
indicated that the newly formed complex is a salt whose cat-
ionic part contains two Pd atoms. Fortunately, after several at-
tempts, crystallization from acetonitrile afforded single crystals
of 7a,b that where suitable for a crystallographic analysis and
allowed the determination of their connectivity (Scheme 5). In-
chosen as a more reliable procedure to rank the donor proper-
[
7,8]
ties of cyclopropenium-substituted phosphines.
0
In sharp contrast to IR spectroscopy, the oxidation potentials
Ep (ox) of phosphines 1a–g, determined by cyclic voltammetry
are indicative of weak neat donations that are similar or even
poorer than those shown by phosphites (Table 1). This ranking
is now in complete agreement with our experiments on the
catalytic activity of Pt complexes bearing ligands 1a–g de-
scribed below. To conclude the stereoelectronic characteriza-
tion, the Tolman cone angle for ligands 1a,c–g has been deter-
mined from X-ray structures to be around 1748, which indi-
terestingly, an oxidative insertion of Pd into the CꢀP bond of
II
1a had taken place, providing a Pd -cyclopropenylidene com-
plex. Furthermore, two phosphide moieties act as bridging li-
gands between the two Pd centers and the remaining coordi-
II
nation position on each Pd is occupied by a molecule of ace-
tonitrile used as solvent for crystallization. From the distance
between the two Pd atoms (3.540 ꢂ) any significant interaction
between them can be ruled out. Their coordination is planar
with a sum of the angles around each Pd of exactly 360.08.
[
9]
0
cates a steric demand similar to PCy3.
If [Pd(PPh3)4] is employed as Pd source instead of
[
Pd (dba) ], more forcing conditions are necessary (toluene,
2 3
reflux) to observe the same kind of products. Note however
Coordination chemistry
that now PPh instead of acetonitrile occupies the coordination
3
We then decided to study the coordination chemistry of these
vacancy created at each Pd atom of 8a. It should be men-
tioned that the oxidative insertion of low-valent metals into
CꢀCl bonds of appropriate halo-imidinium salts is a well-
known process; however, to the best of our knowledge, the in-
sertion into the CꢀP bond of cyclopropenium salts is unprece-
II
II
cationic phosphines towards Pd and Pt species. Thus, reac-
tion of 1a and 1 f with a half equivalent of [{PdCl(allyl)} ] in
2
THF led to quantitative yields of the corresponding adducts 4a
and 4 f. The coordination of phosphorus to the palladium
31
[10]
atom was initially reflected by a significant shift of the P NMR
dented.
resonance signals from d=ꢀ23.6 and ꢀ24.6 ppm in 1a and
1
f to d=26.9 and 25.5 ppm in 4a and 4 f, respectively. More-
Catalysis
over, crystals of 4 f were obtained and its structure determined
unambiguously by X-ray diffraction, confirming the expected
To compare the catalytic performance of our cationic ligands
against standard phosphines, the platinum-catalyzed 6-endo-
dig cyclization of 2-ethynyl-1,1’-binaphtalene 9 into pentaheli-
connectivity (Scheme 4).
A similar procedure, but using
0
.25 equiv of [{PtCl(allyl)} ] instead, was also a viable way to
4
Chem. Eur. J. 2014, 20, 2208 – 2214
www.chemeurj.org
2210
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
this hydroarylation. This remark-
able behavior can be probably
ascribed to the synergic effect of
the cyclopropenium substituent
and the four CF groups that si-
3
multaneously withdraw electron
density from the Pt atom, and
the higher solubility of 6g in di-
chloroethane when compared
[3c]
with polycationic precatalysts.
Synthetic applications
In view of the excellent catalytic
activity depicted by 6g, its rele-
vance on the synthesis of natu-
rally occurring phenantrene de-
rivatives isolated from plants be-
longing to the Orchidaceae
family was explored. In particu-
lar, we focused our attention
on the synthesis of two com-
pounds, 7-hydroxy-2,3,4,8-tetra-
methoxy phenanthrene 11, also
known as chrisotoxene and 2-hy-
droxy-3,4,6,7-tetramethoxy-9,10-
dihydrophenanthrene-2-O-b-d-
glucopyranoside 12, named epi-
medoicarisoside A after being
isolated from the aerial parts of
[13]
Epimedium koreanum.
Chrisotoxene has been isolat-
ed from Dendrobium aurantia-
cum var. denneanum and Tamus
communis and displayed moder-
[
6]
ate cytotoxicities (IC =8.52 mm)
50
against the HeLa cell line.
Scheme 4. Synthesis of 4a, 4 f, 5a, 6a, 6d, 6 f, and 6g and X-ray structure of 4 f, 6d, and 6g. Hydrogen atoms,
[14,15]
tetrafluoroborate anions, and solvent molecules are omitted for clarity; ellipsoids are set at 50% probability. Re-
agents and conditions (product yields in parentheses): a) phosphine, (1 equiv), [{PdCl(allyl)}
2
] (0.5 equiv) or [{PtCl-
] (0.25 equiv), THF, ꢀ208C!RT, 30 min, 4a (98%); 4e (98%), 5a (94%); b) phosphine, (1 equiv), K [PtCl
1.1 equiv), CH CN, RT, 24 h, 6a (78%); 6d (88%); 6 f (68%), 6g (95%).
Our synthesis of this product
started with the Suzuki coupling
between the previously de-
(
(
allyl)}
4
2
4
]
3
scribed
6-bromo-3-hydroxy-2-
[16]
cene 10 was chosen as model reaction because the nucleophil-
methoxybenzaldehyde 13 and 2,3,4-trimethoxyphenylboron-
ic acid 14 to afford the corresponding biphenyl 15. Subse-
quently, the Ohira–Bestmann reaction produced the expected
alkyne 16 although contaminated with substantial amounts of
a second product that after NMR and X-ray analyses was
probed to be benzofurane 17. In line with our expectations,
when 16 was treated with precatalyst 6g, the key platinum-
catalyzed hydroarylation took place cleanly, affording crysotox-
ene in excellent yield after only ten minutes. This compound
crystallized quite readily, and thus its expected connectivity
could be confirmed by X-ray diffraction analysis (Scheme 6).
The detailed 1D and 2D NMR spectra and crystallographic anal-
ysis leaves no doubt that our synthetic material corresponds to
ic attack of the aryl ring to the activated alkyne is known to be
[
11]
the rate-determining step. This makes this transformation
[12]
ideal to evaluate strong p-acceptor ancillary ligands.
Figure 1 depicts the conversion versus time plot for Ph3P
(
(
blue), (PhO) P (yellow), (C F ) P (green) and precatalysts 6a
3 6 5 3
red), 6 f (violet), and 6g (light blue) under otherwise identical
conditions. As expected, both (PhO) P and (C F ) P performed
3
6 5 3
better than Ph P in terms of reactivity, while combinations of
3
6
a or 6 f with AgBF induced much faster reactions than any
4
of the neutral p-acceptor ligands tested.
The outstanding catalytic performance of precatalysts 6g
deserves to be discussed separately. In combination with
13
AgBF , it was able to promote complete conversion of 9 into
structure 11. However, comparison of the C NMR spectrum of
the synthetic and natural samples shows a significant mis-
match for the resonance assigned to C2, which appears at d=
4
pentahelicene 10 after only two minutes under the standard
reaction conditions, exceeding any other catalyst system for
Chem. Eur. J. 2014, 20, 2208 – 2214
www.chemeurj.org
2211
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Figure 1. Ligand effect on the Pt-catalyzed cyclization of 2-ethynyl-1,1’-bi-
naphthalene 9 into pentahelicene 10. a) Conditions for L=Ph
3
P, (PhO)
(5 mol%), and ligand (5 mol%), dichloroethane,
08C; b) Conditions for precatalysts 6a, 6 f, and 6g: 9 (0.05m), 6a, 6 f, or 6g
(5 mol%), dichloroethane, 808C.
3
P, and
(
6 5 3 2
C F ) P: 9 (0.05m), PtCl
8
(
5 mol%), AgBF
4
[19]
ratraldehyde following a reported bromination procedure.
Not unexpectedly, the Suzuki coupling between 19 and 20
provided the biaryl derivative 21; however, excellent yields
were only obtained when the reaction was carried out in a mi-
crowave oven at 1208C. Subsequent transformation of the al-
dehyde moiety into the corresponding terminal alkyne 22 by
the Ohira–Bestmann reaction set the stage for the key plati-
num-catalyzed 6-endo-dig cycloisomerization to the desired
phenanthrene 23 by employing 6g as precatalyst. This trans-
formation took place with excellent yield, and furthermore, the
structure of 23 could be unambiguously confirmed by single-
crystal X-ray analysis. At this stage, simultaneous hydrogena-
tion of the double bond at positions 9,10 of the phenanthrene
skeleton and deprotection of the benzyl group could be effect-
ed, giving access to the dihydrophenanthrene skeleton of epi-
medoicarisoside A.
[
6]
Scheme 5. Synthesis of 7a, 7b, and 8a and X-ray structure of 7b and 8a.
Hydrogen atoms and tetrafluoroborate anions were omitted for clarity; ellip-
soids are set at 50% probability. Reagents and conditions (product yields in
parentheses): a) phosphine, 1 equiv, [Pd
2
(dba)
3
] 0.5 equiv, CH
2 2
Cl , RT; 7a
(
44%); 7b (38%), b) phosphine, 1 equiv, [Pd(PPh )
3 4
], 1 equiv; toluene, reflux
8
a (73%).
1
52.0 and 148.0 ppm in the synthetic and natural chrysotox-
ene, respectively (for details, see the Supporting Information).
Epimedoicarisoside A is also an interesting synthetic target
that serves as testament of the utility of our catalysts towards
the preparation of phenanthrenes. Moreover, several studies
have revealed its cytotoxic effect on primary osteoblasts as
well as its potential application in the treatment of cardiovas-
cular and cerebrovascular diseases, such as myocardial infec-
With 24 in hand, only the installation of the b-d-glucopyra-
noside moiety was necessary to achieve our objective. To this
end the required monosacaride 25 was prepared following
[20]
known procedures and condensed with 24. After some opti-
mization process, we found that the use of two equivalents of
25 and a stoichiometric amount of BF ·OEt were the optimal
3
2
[
17]
tion or cerebral thrombosis. Our synthesis (Scheme 7) com-
menced with the preparation of 4-bromo-2,3-dimethoxy-1-ben-
zyloxybenzene 18 from pyrogallol following a reported proce-
dure. This compound could be efficiently transformed into
the necessary boronic acid 19 after treatment with nBuLi,
quenching with trimethylborate, and subsequent acidic hydrol-
ysis. The second coupling partner, 6-bromoveratraldehyde 20,
was obtained in only one step from commercially available ve-
conditions for this glycosidation step affording 26 in excellent
yield. Finally, deprotection of all of the benzyl groups from the
sugar fragment by hydrogenation over Pd on activated char-
coal allowed us to elaborate 26 into epimedoicarisoside A (12).
[
18]
1
13
Comparison of the H and C NMR shifts of the synthetic
sample with the reported data for the natural product closely
matched. Furthermore, the observed coupling constant
3
JH1,H2 =7.4 Hz in the sugar moiety of 12 is characteristic for an
Chem. Eur. J. 2014, 20, 2208 – 2214
www.chemeurj.org
2212
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Scheme 7. Synthesis of epimedoicarisoside A and X-ray structures of 23. Hy-
drogen atoms are omitted for clarity; ellipsoids are set at 50% probability.
Reagents and conditions (product yields in parentheses): a) nBuLi, ꢀ788C,
+
B(OMe)
3
, H , (59%); b) 21 (1 equiv) [Pd(PPh
3
)
4
] (2 mol%), microwave oven,
[
6]
Scheme 6. Synthesis of chrysotoxene 11 and X-ray structures of 17 and 11.
Hydrogen atoms are omitted for clarity; ellipsoids are set at 50% probability.
Reagents and conditions (product yields in parentheses): a) Pd(OAc)
.05 mol%, Ph P 0.05 mol%, DMF/H O, 608C., 12 h, (68%); b) Ohira–Best-
mann reagent (3 equiv), K CO (4 equiv), MeOH, 48 h, RT, 16 (52%), 17
22%); c) 6g (5 mol%), AgSbF (5 mol%), CH Cl 808C., 10 min.
1
208C, 25 min (98%); c) Ohira–Bestmann reagent, (3 equiv), K
RT, 48 h (83%); d) 6g (5 mol%), Cl(CH Cl, 808C, 3 h, (92%); e) H
Pd/C (20%), MeOH, RT, 48 h, (82%); f) 25 (2 equiv), BF
208C, 2 h, (93%); g) H (1 bar), Pd/C (10%), MeOH: EtOAc (2:1), 24 h, RT,
75%).
2
CO
3
, MeOH,
(30 bar),
2
)
2
2
2
3
2 2
·OEt (1 equiv), CH Cl ,
0
3
2
ꢀ
2
2
3
(
(
6
2
2
axial relationship between these two protons, confirming that
no epimerization at this site has taken place during the glyco-
sidation.
Experimental Section
All of the experimental details can be found in the Supporting In-
formation, which includes the characterization of all new com-
pounds, their NMR spectra, and X-ray analysis. The synthesis of 1g
is shown here as a representative example.
Conclusions
Compound 1g: A solution of bis[3,5-bis(trifluoromethyl)phenyl]-
phosphine (1 g, 2.18 mmol) in dry THF (15 mL) was cooled to
This study provides a new approach for the design of very
strong p-acceptor ligands through the incorporation of a cat-
ionic cyclopropenium substituent into an already polyfluorinat-
ed phosphine. The outstanding ability of these ligands to en-
ꢀ
788C, and then nBuLi (1.6m in hexane, 1.6 mL, 2.18 mmol) was
added and the resulting mixture was stirred for 2 h at this temper-
ature. After this, chlorocyclopropenium salt 2 (782 mg, 2.18 mmol)
was added and the mixture allowed to warm up slowly to RT and
further stirred at 608C for 2 days. After cooling to RT, the solvents
II
hance the natural p-acid properties of Pt has been first dem-
onstrated on the 6-endo-dig cyclization of 2-ethynyl-1,1’-bi-
naphtalene 9 into pentahelicene 10 and subsequently applied
to the synthesis of two natural products, chrysotoxene and ep-
imedoicarisoside A, which both contain a phenanthrene deriva-
tive in their structures. Moreover, considering the very general
mode of action of ligands 1a–g, which relies on a simple rein-
forcement of the metal-to-ligand back-donation, and the multi-
tude of transformations that may benefit from the use of
strong accepting ligands, we believe that our approach has
a vast potential in the development of metal catalysis.
were evaporated, the residue suspended in CH Cl2 (20 mL), and
2
washed with saturated aq. NaBF
solution (3ꢃ15 mL). Once dried
4
over Na SO , the organic phase was concentrated and the residue
2
4
purified by column chromatography (CH Cl /acetone: 9/1), afford-
2
2
ing the title compound as a pale yellow solid (653 mg, 38%).
1
H NMR (400 MHz, CDCl ); d=1.11 (d, J=6.8 Hz, 12H), 1.41 (d, J=
3
6
7
.8 Hz, 12H), 3.48 (sept, J=6.8 Hz, 2H), 4.16 (sept, J=6.8 Hz, 2H),
31
.93 (s, 2H), 7.95 (s, 2H), 8.02 ppm (2H). P NMR (162 MHz, CDCl3)
19
d=ꢀ26.9 ppm. F NMR (282 MHz, CD Cl ) d=ꢀ151.5, ꢀ151.5,
2
2
13
ꢀ63.1 ppm. C NMR (101 MHz, CDCl ) d=20.8, 20.8, 21.2, 52.8,
3
Chem. Eur. J. 2014, 20, 2208 – 2214
www.chemeurj.org
2213
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
5
1
1
6
1
4.6, 99.4 (d, J=61.4 Hz), 122.7 (q, 273.3 Hz), 124.8–125.1 (m),
[6] CCDC 960677 (1c), 960678 (1 f), 960679 (3e), 960680 (4 f), 960681 (6d),
960688 (7b), 960682 (8a), 966990 (11), 960689 (17), 960691 (23), and
33.3 (dq, J=34.3, 7.1 Hz), 133.6–134.1 (m), 134.3 (d, J=15.4 Hz),
+
973072 (6g) contain the supplementary crystallographic data for this
paper. These data can be obtained free of charge from The Cambridge
Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_request/cif.
7] O. Schuster, L. Yang, H. G. Raubenheimer, M. Albrecht, Chem. Rev. 2009,
40.0 ppm. HRMS calcd. for C H N F P : 693.226254; found
31
34
2 12
93.226826. IR n˜ =681, 704, 845, 899, 1057, 1095, 1122, 1180, 1278,
ꢀ
1
353, 1459, 1567, 1867, 2981 cm
.
[
[
1
09, 3445 and references cited therein.
8] a) S. Wolf, H. Plenio, J. Organomet. Chem. 2009, 694, 1487; b) M. Alcara-
zo, T. Stork, A. Anoop, W. Thiel, A. Fꢄrstner, Angew. Chem. 2010, 122,
Acknowledgements
2596; Angew. Chem. Int. Ed. 2010, 49, 2542.
[
9] J. M. Smith, N. J. Coville, Organometallics 2001, 20, 1210.
Generous financial support from the Deutsche Forschungsge-
meinschaft, the Fonds der Chemischen Industrie (Dozentensti-
pendium for M.A), and the European Research Council (ERC
Starting Grant to M.A.) is gratefully acknowledged. We also
thank Prof. Alois Fꢄrstner for constant support, our NMR spec-
troscopy department for assistance, and the DAAD (doctoral
fellowship to A.K.).
[
10] a) D. Kremzow, G. Seidel, C. W. Kehmann, A. Fꢄrstner, Chem. Eur. J. 2005,
1, 1833; b) Z. Yoshida, Pure Appl. Chem. 1982, 54, 1059; c) N. J. Hard-
1
man, M. B. Abrams, M. A. Pribisco, T. M. Gilbert, R. L. Martin, G. J. Kubas,
R. T. Baker, Angew. Chem. 2004, 116, 1989; Angew. Chem. Int. Ed. 2004,
4
3, 1955.
11] See reference [3c] and a) E. Soriano, J. Marco-Contelles, Organometallics
006, 25, 4542; b) V. Mamane, P. Hannen, A. Fꢄrstner, Chem. Eur. J.
[
[
2
2004, 10, 4556.
12] For reviews on Au and Pt p-catalysis, see: a) A. S. K. Hashmi, Chem. Rev.
2
007, 107, 3180; b) A. Fꢄrstner, P. W. Davies, Angew. Chem. 2007, 119,
Keywords: cations
·
cycloisomerization
·
homogeneous
3478; Angew. Chem. Int. Ed. 2007, 46, 3410; c) E. Jimꢈnez-Nfflnez, A. M.
catalysis · phosphines · platinum
Echevarren, Chem. Rev. 2008, 108, 3326; d) A. S. K. Hashmi, M. Rudolph,
Chem. Soc. Rev. 2008, 37, 1766; e) A. Fꢄrstner, Chem. Rev. 2009, 38,
3
208; f) A. S. K. Hashmi, Angew. Chem. 2010, 122, 5360; Angew. Chem.
[
[
1] C. A. Tolman, Chem. Rev. 1977, 77, 313.
2] a) N. Kuhn, J. Fahl, D. Blꢅser, R. Boese, Z. Anorg. Allg. Chem. 1999, 625,
29; b) N. Burford, P. J. Ragogna, R. MacDonald, M. J. Ferguson, J. Am.
Int. Ed. 2010, 49, 5232.
[
[
[
13] W. K. Li, J. Q. Pan, M. J. Lꢄ, R. Y. Zhang, P. G. Xiao, Phytochemistry 1995,
7
3
9, 231.
Chem. Soc. 2003, 125, 14404; c) J. Andrieu, M. Azuori, P. Richard, Inorg.
Chem. Commun. 2008, 11, 1401; d) E. Digard, J. Andrieu, H. Cattey, Inorg.
Chem. Commun. 2012, 25, 39; e) M. Azouri, J. Andrieu, M. Picquet, H.
Cattey, Inorg. Chem. 2009, 48, 1236; f) Y. Canac, N. Debono, L. Vendier,
R. Chauvin, Inorg. Chem. 2009, 48, 5562; g) Y. Canac, C. Maaliki, I. Abdel-
lah, R. Chauvin, New J. Chem. 2012, 36, 17; h) C. Barthes, C. Lepetit, Y.
Canac, C. Duhayon, D. Zargarian, R. Chauvin, Inorg. Chem. 2013, 52, 48.
3] a) J. Petu sˇ kova, H. Bruns, M. Alcarazo, Angew. Chem. 2011, 123, 3883;
Angew. Chem. Int. Ed. 2011, 50, 3799; b) J. Petu sˇ kova, M. Patil, S. Holle,
C. W. Lehmann, W. Thiel, M. Alcarazo, J. Am. Chem. Soc. 2011, 133,
14] L. Yang, Y. Wang, G. Zhang, F. Zhang, Z. Wang, L. Xu, Biomed. Chroma-
togr. 2007, 21, 687.
15] A. Kovꢆcs, P. Forgo, I. Zupkó, B. Rꢈthy, G. Falkay, P. Szabó, J. Hohmann,
Phytochemistry 2007, 68, 687.
16] T. Nakanishi, M. Suzuki, J. Nat. Prod. 1998, 61, 1263.
17] D. Zhang, Y. Cheng, J. Zhang, X. Wang, N. Wang, X. Yao, Pharm. Biol.
[
[
2
008, 46, 185.
[
[
18] a) H. Shigehisa, J. Takayama, T. Honda, Tetrahedron Lett. 2006, 47, 7301;
b) F. Dietl, G. Gierer, A. Merz, Synthesis 1985, 626.
[
[
19] S. Chandrasekhar, N. R. Reddy, Y. S. Rao, Tetrahedron 2006, 62, 12098.
20] a) B. Dasari, S. Jogula, R. Borhade, S. Balasubramanian, G. Chandrase-
khar, S. S. Kitambi, P. Arya, Org. Lett. 2013, 15, 432; b) R. R. Schmidt, M.
Stumpp, Liebigs Ann. Chem. 1983, 1249; c) R. R. Schmidt, J. Michel, M.
Roos, Liebigs Ann. Chem. 1984, 1343.
20758; c) J. Carreras, M. Patil, W. Thiel, M. Alcarazo, J. Am. Chem. Soc.
2012, 134, 16753.
[4] For a recent review on the chemistry of cyclopropenium ions see: J. S.
Bandar, T. H. Lambert, Synthesis 2013, 45, 2485.
[5] The amine analogues have been also synthesized: a) H. Bruns, M. Patil,
J. Carreras, A. Vꢆzquez, W. Thiel, R. Goddard, M. Alcarazo, Angew. Chem.
2010, 122, 3762; Angew. Chem. Int. Ed. 2010, 49, 3680; b) ꢀ. Kozma, G.
Received: September 19, 2013
Gopakumar, C. Farꢇs, W. Thiel, M. Alcarazo, Chem. Eur. J. 2013, 19, 3542.
Published online on January 23, 2014
Chem. Eur. J. 2014, 20, 2208 – 2214
www.chemeurj.org
2214
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim