Bis(cyclopropenium)phosphines: Synthesis, Reactivity, and Applications

Article abstract of DOI:10.1002/chem.201601759

A straightforward route for the preparation of a set of bis(cyclopropenium)-substituted phosphines is reported. Due to their dicationic nature, these ligands depict an excellent π-acceptor character. The effect of the ligand substituent pattern on the catalytic activity of the metal complexes thereof derived is also studied. Whereas sterically demanding biaryl groups directly attached to the phosphorus atom seem to facilitate elementary steps such as the product release from the catalyst, long chain dialkylamino groups on the cyclopropenium units maximize the catalysts solubility and, thus, allow the use of typical apolar solvents such as toluene. Importantly, all new ligands prepared can be easily handled in air. Finally, the impact of the newly prepared dicationic phosphines in hydroarylation reactions is demonstrated. In particular, their use in the synthesis of several naphtho[1,2-b]furanes and naturally occurring naphthalene derivatives such as Calanquinone C is reported.

Full text of DOI:10.1002/chem.201601759

DOI: 10.1002/chem.201601759
Full Paper
&
Synthetic Methods
Bis(cyclopropenium)phosphines: Synthesis, Reactivity, and
Applications
Gerlinde Mehler,[a] Pawel Linowski,[a] Javier Carreras,[a] Alessandro Zanardi,[a]
Jonathan W. Dube,[b] and Manuel Alcarazo*[b]
Abstract: A straightforward route for the preparation of
a set of bis(cyclopropenium)-substituted phosphines is re-
ported. Due to their dicationic nature, these ligands depict
an excellent p-acceptor character. The effect of the ligand
substituent pattern on the catalytic activity of the metal
complexes thereof derived is also studied. Whereas sterically
demanding biaryl groups directly attached to the phos-
phorus atom seem to facilitate elementary steps such as the
product release from the catalyst, long chain dialkylamino
groups on the cyclopropenium units maximize the catalysts
solubility and, thus, allow the use of typical apolar solvents
such as toluene. Importantly, all new ligands prepared can
be easily handled in air. Finally, the impact of the newly pre-
pared dicationic phosphines in hydroarylation reactions is
demonstrated. In particular, their use in the synthesis of sev-
eral naphtho[1,2-b]furanes and naturally occurring naphtha-
lene derivatives such as Calanquinone C is reported.
Introduction
To increase the p-acceptor properties of phosphines, one
might be tempted to introduce a second (or even a third) posi-
tively charged substituent on their structure. However, this ra-
tionale underestimates the effect of the additional charges. In
addition to the desired increment of the p-acceptor properties,
new charges reduce, even more dramatically, the s-donor abili-
ty of the ligand and, consequently, most dicationic phosphines
are unable to coordinate metals.[4]
Two steps are crucial in the process of designing a new cata-
lyst: 1) The choice of the metal, which determines the inherent
reactivity and therefore the type of transformations that might
be promoted, and 2) the adjustment of this primary reactivity
by the use of ancillary ligands. This second step is of upmost
importance, since an appropriate ligand selection allows a fine
control of the kinetics of the whole process, the optimization
of regio- and/or enantioselectivity issues and the elimination
of undesired reaction pathways. In this context, it is not sur-
prising that phosphanes play a prominent role as ancillary li-
gands due to the easy adjustment of their steric and electronic
parameters at convenience by simple modification of the sub-
stituents bonded to phosphorus.
One of the very few exceptions to this rule are bis(cyclopro-
penium)phosphines 1 (Figure 1).[5] In these ligands, the forma-
During the last years our group has explored the effect of
attaching positively charged groups at the central phosphorus
atom of phosphines.[1] As expected, this modification signifi-
cantly decreases the nucleophilicity of the PIII center, rendering
the corresponding a-cationic phosphines weaker s donors, but
stronger p acceptors than the corresponding neutral counter-
parts.[2] In fact, monocationic phosphines retain their ability to
form coordination complexes with a variety of metals and
depict donor properties that resemble those of phosphites.[3]
Figure 1. General structure of bis(cyclopropenium)-substituted phosphines
1.
tion of coordination complexes is still possible, because the
positive charges are efficiently stabilized along the three mem-
bered rings. Nevertheless, the strong electron-withdrawing
effect of the two cyclopropenium substituents confers donor
properties to the resulting phosphine that are comparable to
those depicted by the very p-acidic P(CF3)3.
[a] G. Mehler, P. Linowski, Dr. J. Carreras, Dr. A. Zanardi
Max-Planck-Institut fꢀr Kohlenforschung
Kaiser Wilhelm Platz 1, 45470 Mꢀlheim an der Ruhr (Germany)
[b] Dr. J. W. Dube, Prof. Dr. M. Alcarazo
Georg-August-Universitꢁt Gçttingen
As part of our ongoing program aimed at the synthesis and
development of new electron-poor phosphines with applica-
tions in catalysis, we decided to deepen in the chemistry of
bis(cyclopropenium)-substituted phosphines. Hence, we de-
Tammanstraße 2, 37077 Gçttingen (Germany)
E-mail: Manuel.Alcarazo@chemie.uni-goettingen.de
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201601759.
Chem. Eur. J. 2016, 22, 1 – 9
1
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
scribe herein a general synthetic route suitable for the prepara-
tion of a complete set of these compounds, depicting signifi-
cantly differentiated steric requirements and solubilities in or-
ganic media (Figure 1). In addition, we also report the reactivi-
ty of these phosphorus-centered dications towards oxidants,
their coordination to AuI and PtII complexes, and the impact of
these ancillary ligands in selected model transformations
under different reaction conditions. Specifically, the potential
of the new catalysts is illustrated by the preparation of a series
of naphtho[1,2-b]furans and naphtho[2,1-b]furans, and the syn-
thesis of Calanquinone C, a p-phenanthroquinone isolated
from the roots of the terrestrial orchid Calanthe arisanensis.
Results and Discussion
Synthesis of bis[(dialkylamino)cyclopropenium]phosphines
We recently reported the condensation of the chlorocyclopro-
penium salt 2 with phenylphosphine to afford the cycloprope-
nium-substituted secondary phosphine 5a.[5a] Similarly, the ad-
dition of other primary alkyl- or arylphosphines to 2 also result-
ed in the isolation of compounds 5b–g as white solids in mod-
erate to excellent yields (Scheme 1). These condensations were
also satisfactorily performed in multigram scale.
The atom connectivity in intermediates 5a–g was first sug-
gested by their characteristic 31P NMR signals (d=ꢀ59 to
ꢀ101 ppm.) and subsequently confirmed by single-crystal X-
ray diffraction studies of compounds 5a and 5c. These struc-
tures reveal that the phosphorus centers are in a slightly dis-
torted pyramidal environment with an accessible lone pair
(sum of angles around P1, 307.78 and 301.98 in 5a and 5c, re-
spectively). The P1ꢀC4 distance is slightly shorter in 5a than in
5c, which is probably due to the different hybridization at C4
(sp2 in 5a and sp3 in 5c; Figure 2).
Scheme 1. Synthesis of bis[(dialkylamino)cyclopropenium]phosphines. Re-
agents and conditions: a) nBuLi, ꢀ788C, THF. b) R’PCl2, ꢀ788C–rt; 1b 63%,
1c 41%, 1d 73%, 1 f 49%. c) R’PH2, THF or diglyme, reflux; 5a 76%,
5b 65%, 5c 87%, 5d 71%, 5e 36%, 5 f 61%, 5g 62%. d) KHMDS (potassium
hexamethyldisilazide), THF, ꢀ308C. e) 2 (1 equiv), ꢀ308C–rt; 1a 69%,
1b 42%, 1c 50%, 1d 32%, 1e 23%, 1 f 49%. f) 3 (1 equiv), ꢀ408C–rt;
7a 56%. g) 4 (1 equiv), ꢀ408C–rt; 8a 65%. h) NaSbF6 (7 equiv), CH3CN;
1a(SbF6) 74%, 8a(SbF6) 49%.
Deprotonation of 5 in the presence of a strong base such as
potassium hexamethyldisilazide (KHMDS) and subsequent reac-
tion of the in situ generated phosphaalkene 6 with a second
equivalent of 2 afforded the desired dications 1a–f as air
stable colorless solids in moderate yields. Under these condi-
tions, the dication 1g could never be isolated in analytically
pure form. Most probably, the tremendous steric requirement
of the R’ group inhibits the condensation step, resulting in
complex and dirty reaction mixtures as indicated by 31P NMR
spectroscopy. Alternatively, the biscyclopropenium phosphines
1b–f could be prepared in even better yields by treatment of
the chlorocyclopropenium 2 with nBuLi followed by trapping
of the in situ prepared carbene–lithium adduct with the corre-
sponding dichlorophosphine.[7] This route has the advantage
of avoiding the employment of primary phosphines
(Scheme 1, steps a and b).
The phosphaalkene 6a could also be trapped by the chloro-
cyclopropenium salts 3 and 4 to afford the hybrid salts 7a and
8a, in which one of the cyclopropenium rings is decorated
with dibutylamino or dihexylamino units, respectively, while
the second cyclopropenium keeps the original diisopropylami-
no substituents. This modification, which has no apparent in-
fluence on the donor properties of the resulting ligands,
Figure 2. Molecular structures of compounds 5a (left) and 5c (right) in the
solid state. Anisotropic displacement parameters are shown at 50% proba-
bility level and hydrogen atoms, except those bonded to phosphorus, are
omitted for clarity.[6]
&
&
Chem. Eur. J. 2016, 22, 1 – 9
www.chemeurj.org
2
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
strongly influences their solubilities. Thus, while 1a can only
be dissolved in acetonitrile or chlorinated solvents (CH2Cl2,
Cl(CH2)2Cl), 7a and 8a are soluble in Et2O, and 8a(SbF6) is per-
fectly soluble in toluene. It is worth mentioning that once pre-
pared all bis(cyclopropenium)phosphines here described are
bench-stable compounds, and therefore their use is distin-
guished by operational simplicity.
Evaluation of the donor properties
At this point, the newly prepared ligands were investigated by
cyclic voltammetry. The depicted oxidation peaks of the phos-
phines 1a, 1b, and 1 f are at relatively high potentials (1.541,
1.526, and 1.425 V, respectively). Assuming that the electron
richness of a phosphine is, in a first approach, inversely propor-
tional to its oxidation potential Ep(ox), these values speak for
a very efficient stabilization of the phosphorus electron pair by
the two cyclopropenium groups (Table 1). Note, however, that
The dicationic nature of 1a–f, 7a, and 8a was first reflected
by the shielding of their 31P NMR signals (d=ꢀ43 to
ꢀ62 ppm.), which are comparable to that of the bis(imidazoli-
um)-substituted sister compound 9 (d=ꢀ50.8 ppm.).[8] This
could be then ascertained by single-crystal X-ray diffraction
analysis of 1d, 1 f, and 8a(SbF6) (Figure 3). The geometry
Table 1. Electrochemical redox potentials and Tolman cone angles of se-
lected phosphines.
Entry
Ligand[a]
Ep(Ox) [V][b]
q [8]
1
2
3
4
5
6
7
8
9
1a
1b
1 f
PPh3
P(tBu)3
9
JohnPhos
P(OMe)3
10
1.541
1.526
1.425
0.687
0.534
2.330
0.393
1.287
1.207
183
–
249
145
182
–
246
107
169
[a] JohnPhos=(2-biphenyl)di-tert-butylphosphine. [b] Oxidation peak po-
tentials are calibrated versus ferrocene/ferrocenium, (Bu)4NPF6 (0.1m) in
CH2Cl2.
the third substituent at phosphorus still plays a role. The phos-
phine 1b containing an aliphatic substituent is more readily
oxidized than its aromatic congener 1a and, consequently, it is
a slightly stronger donor ligand. The easier oxidation measured
for 1 f can be rationalized considering a repulsive interaction
between the p system of the additional phenyl group and the
phosphorus electron pair. A similar effect is observed when the
Ep(ox) of P(tBu)3 and (2-biphenyl)di-tert-butylphosphine (John-
Phos) are compared (Table 1, Entries 5 and 7). Additional in-
sights about the donor capacity of 1a–f could be obtained by
comparison of their Ep(ox) with those of the reference com-
pounds. For example the Ep(ox) of the dicationic phosphine 1a
lies much closer to that of P(OMe)3 (1.287 V) than to the value
determined for the also dicationic 9 (2.330 V), a bisimidazoli-
um-substituted phosphine with no reported ability to coordi-
nate metals. This nicely explains the exquisite behavior of 1a
and analogues; they still retain the ability to form coordination
complexes but act as very strong p-acceptor ligands.
Figure 3. Molecular structures of compounds 1d (up, left), 1 f (up, right),
and 8a(SbF6) in the solid state. Anisotropic displacement parameters are
shown at 50% probability level and hydrogen atoms are omitted for clari-
ty.[6]
around the phosphorus atom in these three compounds re-
mains pyramidal, although slightly flattened compared to 5
(sum of angles around P1: 314.48, 308.08, and 306.08 in 1d, 1 f,
and 8a(SbF6), respectively). This is probably a direct conse-
quence of the augmented steric congestion around the central
phosphorus atom due to the additional cyclopropenium sub-
stituent. All CꢀP bond lengths are in accord with typical single
bonds.[9]
Also interesting are the relationships obtained when steric
parameters are compared along the entries in Table 1. For ex-
ample, the phosphine 1 f and JohnPhos are characterized by
very similar Tolman cone angles (q=2498 and 2468, respective-
ly);[10] however, the oxidation potential of 1 f lies no less than
Chem. Eur. J. 2016, 22, 1 – 9
www.chemeurj.org
3
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
1 V higher, which indicates a significantly reduced donor ability
of the former. This aspect has deep implications in catalysis, as
will be described along the following sections.
oxidized with XeF2 to the corresponding difluorophosphorane
12a in good yield. The characteristic 31P NMR signal at d=
1
ꢀ75.3 ppm (triplet) with a JPꢀF coupling constant of 684 Hz, in-
dicated the presence of two equivalent fluorine substituents
directly attached to the central phosphorus atom. After many
attempts, a crystal suitable for X-ray analysis of 12a could be
obtained and, thus, its connectivity was confirmed.[12] This
compound is extremely sensitive to moisture and readily con-
verts to the oxide in air.
Reactivity towards oxidants
Encouraged by the relatively low oxidation potentials mea-
sured for 1-type compounds, we first decided to explore
whether they could react with strong oxidants to afford the
corresponding dicationic PV species. Reaction of the salt 1 f
with a stoichiometric amount of m-chloroperbenzoic acid (m-
CPBA) in dichloromethane resulted in the formation of a new
species 11 f, that gave rise to a 31P NMR signal at d=
ꢀ1.4 ppm. (Scheme 2). The molecular structure of this new
Coordination chemistry
The ability of the newly prepared ligands to form coordination
complexes was confirmed through the preparation of several
AuI and PtII derivatives. Mixing of the ligands 1a, 7a, and 8a
with K2PtCl4 in acetonitrile led to the formation of the corre-
sponding platinum derivatives 13–15 in good to excellent iso-
lated yields. These complexes display downfield shifted
31P NMR signals (d=ꢀ21.2, ꢀ20.2, and ꢀ21.6 ppm for 13, 14,
and 15, respectively) when compared with the corresponding
free ligands. Additionally, the presence of 195Pt satellites in the
31P NMR spectra (1JPtꢀP =3989.6, 3980.0, and 3966.0 Hz for 13,
14, and 15, respectively) indicated the expected coordination
(Scheme 3).
In a similar manner, AuI complexes derived from dicationic
phosphines were prepared by reaction of 1a, 1c, 1 f, 7a, and
8a with one equivalent of (Me2S)AuCl in dichloromethane. As
expected upon coordination, these new species also depicted
downfield shifts of the phosphorus signals (d=ꢀ10.0, ꢀ21.8,
ꢀ18.4, ꢀ20.3, and ꢀ22.3 ppm for 16, 17, 18, 19, and 20, re-
spectively). We were able to obtain single crystals of 19 by
slow diffusion of diethyl ether into a saturated solution of the
complex in acetonitrile. Its structure exhibits a nearly linear ge-
ometry around the Au atom (P1ꢀAu1ꢀCl1 173.18) and a proxi-
mal location of Au to the arene ring with an Au1ꢀC13 distance
of 3.250 ꢁ, suggesting a weak attractive interaction between
the two atoms (Figure 5).
Scheme 2. Reactivity of bis[(dialkylamino)cyclopropenium]phosphines to-
wards oxidants. Reagents and conditions: a) S8, 1408C, 20 h, tetrachloro-
ethane; 11 b 47%. b) m-CPBA (m-chloroperbenzoic acid, 1 equiv), THF;
11 f 65%. c) XeF2 (1 equiv), CH2Cl2; 12a 68%.
compound, which corresponds to the desired dicationic phos-
phine oxide, was later confirmed by X-ray crystallography
(Figure 4). In 11 f, the measured P1ꢀO1 bond length
(1.457(3) ꢁ) is slightly shorter than the one reported for triphe-
nylphosphine oxide (1.479 ꢁ) and might suggest a stronger p
backdonation from the oxygen atom to phosphorus induced
by the two geminal positive charges.[11]
An oxidation with S8 is also possible. However, it requires
high temperatures and longer reaction times (20 h and 1408C
in the case of 11 b). The dicationic phosphine 1a could also be
Catalysis
The effect of the noncharged substituent on the cat-
alytic performance of complexes 16–19 was first eval-
uated in the hydroarylation of 1-furyl-2-alkynyl ben-
zene (22) into naphtho[1,2-b]furan (23). This process
is of interest, because it opens a straightforward ap-
proach to the synthesis of these tricyclic skeletons,
which are present in several natural products.[13,14]
Figure 6 shows the kinetic profiles compiled for
precatalysts 16 and 19 and other typical catalytic
mixtures under otherwise identical conditions
(2 mol% [Au], Cl(CH2)2Cl, 508C). As expected, the re-
activity of the corresponding catalyst is higher with
rising acceptor character of the ancillary ligand. This
indicates that the attack of the furan moiety to the
activated alkyne is probably the rate determining
Figure 4. Molecular structures of compounds 11 f (left) and 12a (right) in the solid state.
Anisotropic displacement parameters are shown at 50% probability level and hydrogen
atoms are omitted for clarity.[6]
step in this process. However, for the strongest p-ac-
ceptor ligands 16 and 19 this trend is inverted; the
&
&
Chem. Eur. J. 2016, 22, 1 – 9
www.chemeurj.org
4
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
Scheme 3. Coordination of bis[(dialkylamino)cyclopropenium]phosphines to
AuI and PtII. Reagents and conditions: a) K2PtCl4 (1 equiv), acetonitrile;
13 98%, 14 90%, 15 96%. b) (Me2S)AuCl (1 equiv), DCM; 16 95%, 17 84%,
18 79%, 19 91%, 20 88%, 21 86%. c) NaSbF6 (1 equiv), acetonitrile;
21(SbF6) 78%.
Figure 5. Molecular structure of compound 19 in the solid state. Anisotropic
displacement parameters are shown at 50% probability level and hydrogen
atoms are omitted for clarity.[6]
catalytic activity of the precatalyst 19 slightly overtakes that of
16, even though the Ep(ox) of ligand 1a is 116 mV higher than
that of 1 f. This behavior may indicate that, once the Lewis
acidity of Au exceeds a critical threshold, another step different
to the nucleophilic attack becomes rate determining. Naph-
tho[1,2-b]furans are relatively electron-rich heteroaromatics,
and for this reason it seems reasonable to assume that when
such Lewis acidic catalysts are used, the final release of the
product might be rate determining. This would explain why
the bulkiest ligand 19 is the best one to promote this process.
Having identified the optimal catalyst, further experiments
were conducted to explore the scope of this reaction. Hence,
a set of 1-furyl-2-alkynylbenzenes 24–28 were prepared and
a preliminary reactivity screen was performed. We were de-
lighted to see that all substrates, regardless of the substitution
pattern, were completely converted to the expected naphtho-
furanes 29–33 in very good to excellent yields. Positions 2-
and 3- of the furane moiety are able to conduct the desired
cyclisation. However, when both are available, the only prod-
uct observed is the one derived from the attack of the most
electron rich C2 position (Table 2, Entry 2). The molecular struc-
ture of 30, obtained by X-ray analysis, further demonstrates
the connectivity of the naphthofuran prepared (See the Sup-
porting Information).
Due to their saline nature, the utilization of chlorinated sol-
vents as reaction media is normally an indispensable require-
ment when cationic ancillary ligands are employed. For this
reason, we set up another series of experiments, now consider-
Figure 6. Ligand effect on the [Au]-catalyzed hydroarylation of 22 to naphtho[1,2-b]furane (23).
Chem. Eur. J. 2016, 22, 1 – 9 www.chemeurj.org
5
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
quires strong p-acid AuI catalysts was used as a model (34!
35; Figure 7); however, this time toluene was employed as the
solvent.
Table 2. Reaction scope of the naphthofurane synthesis.[a]
Entry
Substrate
Product
Yield [%][b]
As shown in Figure 7, precatalyst 16 was unable to catalyze
the hydroarylation under the new working conditions. We be-
lieve that the low solubility of 16 in toluene is responsible for
this low activity. Interestingly, precatalysts 20 and 21 exhibited
higher levels of efficiency. Since the stereoelectronic properties
of these three ligands are basically identical, we believe that
the enhanced reactivity of 21 can only be attributed to its
higher solubility in toluene, which is a consequence of the
hexyl substituents that it bears. Finally, the exchange of both
tetrafluoroborate counteranions in 21 by the softer hexafluor-
oantimonate also proved to be beneficial in terms of catalytic
activity, probably due to an additional enhancement of the cat-
alyst solubility. The use of this optimized catalyst, 21(SbF6),
guarantees a complete conversion of 34 into 35 under the de-
scribed conditions, even if the catalyst load is reduced to only
0.5 mol%.
1
91
2
3
94
95
4
5
84
95
Synthesis of Calanquinone C and Calanhydroquinone A
Once 21(SbF6) was identified as the optimal precatalyst for the
hydroarylation of ortho-alkynylated biaryls into twisted phe-
nanthrenes, its application to the synthesis of naturally occur-
ring phenanthrene architectures was explored. Specifically, we
focused our interest on the synthesis of 6-hydroxy,-5,7-dime-
thoxy-9,10-dihydro-1,4-phenanthrene-quinone (36), named as
Calanquinone C after isolation from the roots of the terrestrial
orchid Calanthe arisanensis.[15] Our synthesis of this product
started with the Suzuki coupling between 3-benzyloxy-2,4-di-
methoxyphenyl boronic acid (37) and the already known 3,6-
dibenzyloxy-2-bromobenzaldehyde 38 to afford the biphenyl
39 (Scheme 4).[16] Not unexpectedly, treatment of 39 with the
Ohira–Bestmann reagent transformed the aldehyde moiety
into the corresponding terminal alkyne 40, setting the stage
for the key Au-catalyzed intramolecular hydroarylation. Just
0.5 mol% of 21(SbF6) was enough to cleanly promote this step
in nearly quantitative yield. Treatment of 41 with H2 (20 bar)
using Pd/C, simultaneously promoted the deprotection of all
benzyl ethers and the reduction of positions 9 and 10 of the
phenanthrene skeleton in 41 affording Calanhydroquinone A
(42), the structure of which could be confirmed by X-ray analy-
sis (Figure 8). A final oxidation of the dihydroquinone fragment
with (NH4)2Ce(NO3)6 allowed us to elaborate 42 into Calanqui-
6
90
[a] Conditions: substrate (0.05m in CH2Cl2), 19 (2 mol%), AgSbF6
(2 mol%), rt. [b] Isolated yields.
ing the catalyst solubility as the parameter to be optimized. In
this regard, our strategy consists of the use of the gold com-
plexes 20 and 21, both containing longer-chain dialkylamino
substituents (butyl- and hexyl-, respectively) attached to the
cyclopropenium moieties. A similar cyclization that also re-
1
none C (36). Comparison of the reported H and 13C NMR spec-
tra of the isolated product perfectly matches that of the syn-
thetic sample. Moreover, the connectivity of 36 could be addi-
tionally confirmed by X-ray diffraction experiments.
Conclusion
This study provides a straightforward route for the synthesis of
a series of strong p-acceptor dicationic phosphines of different
steric demands and solubilities. The utility of bulky ligands,
such as 1 f, is demonstrated on the 6-endo-dig cyclization of 1-
furyl-2-ethynylbenzenes into naphtho[1,2-b]furans and naph-
Figure 7. Ligand effect on the [Au]-catalyzed hydroarylation of the biphenyl
34 to the phenanthrene 35.
&
&
Chem. Eur. J. 2016, 22, 1 – 9
www.chemeurj.org
6
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
out in toluene are illustrative examples of this achievement.
Further studies towards the application of cationic ligands in
other contexts are currently underway in our research group.
Experimental Section
Synthesis of 5 f. A dried Schlenk flask was charged with 1,1’-(bi-
phenyl-2-yl)-phosphine (2.24 g, 12.0 mmol), chloro cyclopropenium
salt 2 (2.15 g, 6.0 mmol), and diglyme (20.0 mL), and the resulting
mixture was heated overnight at 1008C. After cooling of the reac-
tion mixture to room temperature, the solvent was removed in
vacuo, the residue dissolved in DCM, and the organic phase
washed with a saturated aqueous solution of NaBF4. Then, the or-
ganic phase was dried over sodium sulfate, and the solvent evap-
orated in vacuo to afford a residue that was finally purified by
column chromatography on silica gel (DCM/Acetone, 9:1). White
solid (1.86 g, 61%). 1H NMR (400 MHz, CDCl3): d=1.08 (d, J=
6.8 Hz, 6H), 1.16 (d, J=6.8 Hz, 6H), 1.34 (d, J=6.8 Hz, 6H), 1.37 (d,
J=6.8 Hz, 6H), 3.65 (septd, J=6.8, 1.8 Hz, 2H), 4.12 (sept, J=
6.9 Hz, 2H), 5.39 (d, J=233.3 Hz, 1H), 7.22–7.23 (m, 1H), 7.26 (m,
1H, overlaps with solvent signal) 7.31–7.51 (m, 6H), 7.69–7.75 ppm
(m, 1H); 13C{1H} NMR (101 MHz, CDCl3): d=21.0, 21.0, 21.5, 21.5,
Scheme 4. Synthesis of Calanhydroquinone A and Calanquinone C. Reagents
and conditions: a) 38 (1 equiv), 37 (1.5 equiv), Pd2(dba)3 (5 mol%), PCy3
(11 mol%), Cs2CO3, 1,4-dioxane/toluene, yield 84%. b) Ohira-Bestmann re-
agent (1.5 equiv), MeOH, K2CO3, yield 67%. c) 21(SbF6) (0.5 mol%), AgSbF6
(0.5 mol%), toluene, yield 89%. d) H2 (20 Bar), Pd/C (10 mol%), yield 89%.
e) (NH4)2Ce(NO3)6 (1 equiv), THF/H2O, yield 69%.
22.1, 22.2, 49.4, 56.5, 103.4 (d, JPꢀC =58.4 Hz), 124.9 (d, JPꢀC =
8.1 Hz), 128.2, 128.7, 128.9 (d, JPꢀC =4.4 Hz), 129.1 (d, JPꢀC =3.0 Hz),
130.6 (d, JPꢀC =4.0 Hz), 131.0, 136.8 (d, JPꢀC =9.3 Hz), 138.9 (d,
J
PꢀC =2.0 Hz), 141.0 (d, JPꢀC =4.0 Hz), 147.9 ppm (d, JPꢀC =20.2 Hz);
31P{1H} NMR (161 MHz, CDCl3): d=ꢀ78.0 ppm; IR (solid): n˜ =702,
762, 861, 884, 1029, 1152, 1340, 1458, 1562, 1874, 2297, 2984 cmꢀ1
;
HRMS (ESI): m/z calcd for C27H38N2P+: 421.276711 [MꢀBF4]+; found:
421.276383.
Synthesis of 1 f. Dry THF (5 mL) was added at ꢀ408C to a solid
mixture of KHMDS (325 mg, 1.63 mmol) and phosphine 5 f
(830 mg, 1.63 mmol), and the resulting suspension was stirred at
this temperature for 2 h. Then, the desired chloro cyclopropenium
salt (580 mg, 1.63 mmol) was added, and the reaction warmed to
room temperature overnight. Removal of the solvent in vacuo af-
forded a residue that was dissolved in DCM and washed with a sa-
turated solution of NaBF4 (3x). For a final purification, the obtained
1
residue was washed with THF. White solid (660 mg, 49%). H NMR
(400 MHz, CD3CN): d=0.97 (d, J=6.8 Hz, 12H), 1.16 (d, J=6.8 Hz,
12H), 1.32 (d, J=6.8 Hz, 12H), 1.35 (d, J=6.8 Hz, 12H), 3.68–3.71
(mbr, 4H), 4.13 (sept, J=6.8 Hz, 4H), 7.17–7.19 (m, 2H), 7.44–7.50
(m, 3H), 7.52–7.56 (m, 1H), 7.60–7.64 (m, 1H), 7.67–7.70 (m, 1H),
7.73 ppm (td, J=7.5 Hz, 1.3 Hz, 1H); 13C{1H} NMR (101 MHz,
CD3CN): d=21.1, 21.2, 21.5, 21.6, 21.6, 21.7, 53.2, 56.0, 98.7 (d,
J
PꢀC =60.6 Hz), 125.7, 129.7, 129.8, 129.9 (d, JPꢀC =4.0 Hz), 130.2,
132.4 (d, JPꢀC =6.5 Hz), 134.0, 136.2 (d, JPꢀC =3.1 Hz), 140.6 (d,
PꢀC =2.0 Hz), 140.7, 150.0 ppm (d,
PꢀC =36.1 Hz); 31P{1H} NMR
J
J
Figure 8. Molecular structures of compounds 42 (left) and 36 (right) in the
solid state. Anisotropic displacement parameters are shown at 50% proba-
bility level and hydrogen atoms not directly bonded to an oxygen are omit-
ted for clarity.[6]
(162 MHz, CD3CN): d=ꢀ54.7 ppm; IR (solid): n˜ =440, 519, 549, 578,
616, 640, 666, 688, 705, 740, 761, 891, 1028, 1048, 1150, 1182,
1204, 1283, 1353, 1376, 1457, 1552, 1852, 2879, 2939, 2976 cmꢀ1
;
HRMS (ESI): m/z calcd for C42H65BF4N4P+: 743.497053 [MꢀBF4]+;
found: 743.498110.
tho[2,1-b]furans. In this reaction, the strong Lewis acidity gen-
erated at the AuI center probably makes the final release of
the product rate determining; hence, the bulky biaryl substitu- Acknowledgements
ent is essential for a faster transformation. On the other hand,
the introduction of long chain dialkylamino substituents to the
ligand structure increases the solubility of the resulting cata-
lysts and, consequently, chlorinated solvents are not necessary
as reaction medium. The syntheses of Calanquinone C and Cal-
anhydroquinone A through a key hydroarylation step carried
The X-ray structural determination of 36 was carried out at the
light source PETRA III at DESY, a member of the Helmholtz As-
sociation (HGF). We would like to thank A. Burkhardt and S.
Panneerselvam for assistance in using the beamline P11. We
also acknowledge the NMR spectroscopy and X-ray crystallog-
Chem. Eur. J. 2016, 22, 1 – 9
www.chemeurj.org
7
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
for
a
second exception, see: b) M. P. Coles, P. B. Hitchcock, Chem.
raphy departments at the Max-Planck Institut fꢂr Kohlenforch-
ung for assistance, and C. Wille for the CV measurements. Fi-
nancial support from the Deutsche Forschungsgemeinschaft
(AL 1348/5-1) is gratefully acknowledged.
Commun. 2007, 5229–5231.
[6] CCDC 1472384, 1472385, 1472386, 1472387, 1472388,1472389,
1472390, 1472391, 1472392, 1472393, 1472394, 1472395 and 1473951
contain the supplementary crystallographic data for this paper. These
data are provided free of charge by The Cambridge Crystallographic
Data Centre.
Keywords: cationic ligands
hydroarylation · ligand design · p-acid catalysis
·
homogeneous catalysis
·
[7] a) R. Weiss, C. Priesner, H. Wolf, Angew. Chem. Int. Ed. Engl. 1978, 17,
446–447; Angew. Chem. 1978, 90, 486–487; b) R. Weiss, M. Hertel, H.
Wolf, Angew. Chem. Int. Ed. Engl. 1979, 18, 473; Angew. Chem. 1979, 91,
506; c) V. Lavallo, Y. Canac, B. Donnadieu, W. W. Schoeller, G. Bertrand,
Science 2006, 312, 722–724.
ˇ
[1] a) J. Petuskova, H. Bruns, M. Alcarazo, Angew. Chem. Int. Ed. 2011, 50,
3799–3802; Angew. Chem. 2011, 123, 3883–3886; b) J. Carreras, M.
Patil, W. Thiel, M. Alcarazo, J. Am. Chem. Soc. 2012, 134, 16753–16758;
[8] C. Maaliki, Y. Canac, C. Lepetit, C. Duhayon, R. Chauvin, RSC Adv. 2013,
3, 20391–20398.
ˇ
c) ꢃ. Kozma, T. Deden, J. Carreras, C. Wille, J. Petuskova, J. Rust, M. Alcar-
[9] A. F. Hollemann, N. Wiberg, Lehrbuch der Anorganischen Chemie, de
Gruyter, Berlin, 2007.
azo, Chem. Eur. J. 2014, 20, 2208–2214; d) H. Tinnermann, C. Wille, M.
Alcarazo, Angew. Chem. Int. Ed. 2014, 53, 8732–8736; Angew. Chem.
2014, 126, 8877–8881; e) E. Haldꢄn, ꢃ. Kozma, H. Tinnermann, L. Gu, R.
Goddard, M. Alcarazo, Dalton Trans. 2016, 45, 1872–1876. For cationic
arsines, see: f) J. W. Dube, Y. Zheng, W. Thiel, M. Alcarazo, J. Am. Chem.
Soc. 2016, 138, 6869–6877.
[10] a) S. Rousseaux, M. Davi, J. Sofack-Kreutzer, C. Pierre, C. E. Kefalidis, E.
Clot, K. Fagnou, O. Baudoin, J. Am. Chem. Soc. 2010, 132, 10706–10716;
b) J. M. Smith, N. J. Coville, Organometallics 2001, 20, 1210–1215.
[11] K. A. Al-Farhan, J. Chryst. Spect. Res. 1992, 22, 687–689.
[12] For the oxidation of monocationic phosphines with XeF2, see: M. H.
Holthausen, M. Mehta, D. W. Stephan, Angew. Chem. Int. Ed. 2014, 53,
6538–6541; Angew. Chem. 2014, 126, 6656–6659.
[2] a) Y. Canac, C. Maaliki, I. Abdellah, R. Chauvin, New J. Chem. 2012, 36,
17–27; b) S. Gaillard, J. L. Renaud, Dalton Trans. 2013, 42, 7255–7270;
c) M. Alcarazo, Chem. Eur. J. 2014, 20, 7868–7877.
[13] a) K. Ishiguro, Y. Ohira, H. Oku, J. Nat. Prod. 1998, 61, 1126–1129; b) J.
Padwal, C. J. Moody, W. Lewis, J. Org. Chem. 2011, 76, 8082–8087; c) H.
Itokawa, Y. Qiao, K. Takeya, Phytochemistry 1991, 30, 637–640.
[14] For related syntheses of furanes and benzo-/naphthofuranes, see:
a) A. S. K. Hashmi, T. M. Frost, J. W. Bat, Org. Lett. 2001, 3, 3769–3771;
b) A. S. K. Hashmi, L. Schwarz, J.-H. Choi, T. M. Frost, Angew. Chem. Int.
Ed. 2000, 39, 2285–2288; Angew. Chem. 2000, 112, 2382–2385; c) B.
Martꢆn-Matute, C. Nevado, D. J. Cꢇrdenas, A. M. Echavarren, J. Am.
Chem. Soc. 2003, 125, 5757–5766; d) Y. Yamamoto, K. Matsui, M. Shi-
buya, Chem. Eur. J. 2015, 21, 7245–7255.
[3] a) M. Azouri, J. Andrieu, M. Picquet, P. Richard, B. Hanquet, I. Tkatchen-
ko, Eur. J. Inorg. Chem. 2007, 4877–4883; b) J. Andrieu, M. Azouri, P. Ri-
chard, Inorg. Chem. Commun. 2008, 11, 1401–1404; c) N. Debono, Y.
Canac, C. Duhayon, R. Chauvin, Eur. J. Inorg. Chem. 2008, 2991–2999;
d) Y. Canac, N. Debono, L. Vendier, R. Chauvin, Inorg. Chem. 2009, 48,
5562–5568; e) S. Saleh, E. Fayed, M. Azouri, J. C. Hierso, J. Andrieu, M.
Picquet, Adv. Synth. Catal. 2009, 351, 1621–1628; f) I. Abdellah, C. Lepe-
tit, Y. Canac, C. Duhayon, R. Chauvin, Chem. Eur. J. 2010, 16, 13095–
13108; g) E. Digard, J. Andrieu, H. Cattey, Inorg. Chem. Commun. 2012,
25, 39–42; h) B. Vabre, Y. Canac, C. Duhayon, R. Chauvin, D. Zargarian,
Chem. Commun. 2012, 48, 10446–10448; i) Y. Canac, C. Bijani, C. Duhay-
on, R. Chauvin, Organometallics 2013, 32, 4054–4057.
[15] C. L. Lee, F. R. Chang, M. H. Yen, D. Yu, Y. H. Liu, K. F. Bastow, S. L. Morris-
Natschke, Y. C. Wu, K. H. Lee, J. Nat. Prod. 2009, 72, 210–213.
[16] K. Hayashi, A. Yamazoe, Y. Ishibashi, N. Kusaka, Y. Oono, H. Nozaki,
Bioorg. Med. Chem. 2008, 16, 5331–5344.
[4] a) R. Weiss, S. Engel, Synthesis 1991, 1077–1078; b) G. Bouhadir, R. W.
Reed, R. Rꢅau, G. Bertrand, Heteroat. Chem. 1995, 6, 371–375; c) M.
Azouri, J. Andrieu, M. Picquet, H. Cattey, Inorg. Chem. 2009, 48, 1236–
1242.
ˇ
Received: April 14, 2016
[5] a) J. Carreras, G. Gopakumar, L. Gu, A. M. Gimeno, P. Linowski, J. Petus-
kova, W. Thiel, M. Alcarazo, J. Am. Chem. Soc. 2013, 135, 18815–18823;
Published online on && &&, 0000
&
&
Chem. Eur. J. 2016, 22, 1 – 9
www.chemeurj.org
8
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
FULL PAPER
The bulkier the better! A series of
strong p-acceptor dicationic phosphines
were prepared through a modular syn-
thetic route. Preliminary studies on Au
catalysis indicate that, once the Lewis
acidity on the metal surpasses a critical
value, it is the bulkiness of the ligand
that controls the overall reaction speed.
&
Synthetic Methods
G. Mehler, P. Linowski, J. Carreras,
A. Zanardi, J. W. Dube, M. Alcarazo*
&& – &&
Bis(cyclopropenium)phosphines:
Synthesis, Reactivity, and Applications
Chem. Eur. J. 2016, 22, 1 – 9
www.chemeurj.org
9
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ