Synthesis of the first C2-asymmetric phosphinine and its pyrylium precursor

Article abstract of DOI:10.1016/j.tet.2009.08.085

The synthesis of the first C₂ chiral phosphinine (phosphabenzene) and its pyrylium salt precursor is reported. The chiral phosphinine is derived from (+)-camphor, is a crystalline air-stable solid and is shown to effectively form complexes with metals. A preliminary result using the phosphinine as a ligand for asymmetric catalysis is reported and rationalized structurally.

Full text of DOI:10.1016/j.tet.2009.08.085

Tetrahedron 65 (2009) 9368–9372
Contents lists available at ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
Synthesis of the first C2-asymmetric phosphinine and its pyrylium precursor
*
Jason R. Bell, Andreas Franken, Charles M. Garner
Department of Chemistry & Biochemistry, Baylor University, One Bear Place #97348, Waco, TX 76798, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 8 July 2009
Received in revised form 28 August 2009
Accepted 31 August 2009
Available online 4 September 2009
The synthesis of the first C2 chiral phosphinine (phosphabenzene) and its pyrylium salt precursor is
reported. The chiral phosphinine is derived from (þ)-camphor, is a crystalline air-stable solid and is
shown to effectively form complexes with metals. A preliminary result using the phosphinine as a ligand
for asymmetric catalysis is reported and rationalized structurally.
Ó 2009 Elsevier Ltd. All rights reserved.
1. Introduction
O
O
Catalytic asymmetric reactions are the most efficient methods
for the preparation of high enantiomeric purity organic com-
pounds. The majority of such processes rely on chiral ligands, many
of them phosphorus-based. Phosphinines (phosphabenzenes)
represent a little-studied class of phosphorus ligands, especially
with regard to asymmetric variants.
O
N
P
P
N
1
2
O
Achiral phosphinines have been shown to be effective ligands
for several catalysts,1 most notably Breit et al.’s use of triphenyl
phosphinine derivatives in hydroformylation.2 The first report of
a chiral phosphinine came in 1996 when Breit synthesized three
phosphinines (1–3, Fig. 1) with bound chiral auxiliaries.3 However,
no enantioselectivity was observed in the hydroformylation re-
action. The first successful asymmetric reaction using a phosphi-
nine ligand was reported by Mu¨ller et al., where a bidentate
phosphinine/phosphite ligand (4) achieved asymmetric hydroge-
nations of dimethylitaconate in as high as 79% ee.4 Recently this
same group reported5 the first atropisomeric phosphinine (5),
though its use in asymmetric reactions has yet to be described.
In order to further the development of phosphinines as ligands
for asymmetric catalysis, we undertook the development of C2
chiral phosphinine ligands. It has been recognized6 that C2 chiral
ligands are generally superior to other types of ligands in asym-
metric reactions, probably because the approach of a reactant to
either face of a C2 complex will be equivalent, limiting the number
of possible diastereomeric transition states. The best method of
preparing highly substituted phosphinines is via the ring trans-
formation reaction of pyrylium salts.7 Pyrylium salts can be syn-
thesized by a variety of methods that allow various substitution
patterns, though in practice aryl substituents are far more common
than alkyl groups.8 Polymerization can compete with pyrylium
PPh2
O
O
P
P
O
O
P
3
O
O
4
P
5
Figure 1. Existing chiral phosphinine ligands.
formation, and tars are a common result, especially from alkyl-
substituted substrates. Perhaps not unexpectedly, chiral pyrylium
salts are almost entirely unknown, with only one non-racemic
example9 and one racemic case5 having been reported to date. Yet,
by the proper choice of starting materials, we hoped to incorporate
C2 chirality into a pyrylium salt. We chose (þ)-camphor as our
chiral unit, since ketones are common starting points for building
* Corresponding author. Tel.: þ1 254 710 6862; fax: þ1 254 710 4272.
E-mail address: charles_garner@baylor.edu (C.M. Garner).
0040-4020/$ – see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2009.08.085

J.R. Bell et al. / Tetrahedron 65 (2009) 9368–9372
9369
pyrylium rings, and camphor is an inexpensive member of the
chiral pool.
Cl
O
2. Results and discussion
OH
PCl3
NaH, DME
PhCO2CH3
92%
reflux
42%
O
O
O
Our initial attempts to create the biscamphorpyrylium 6 (vide
infra) were based on the ‘one pot’ methods of creating triphenyl-
pyrylium10 (Scheme 1), which requires the formalloss of H2 to furnish
the third unit of unsaturation. We first attempted to react 2 equiv of
camphor with benzaldehyde under strongly acidic conditions (HBF4
or BF3) to form 3-benzylidenecamphor in situ, which theoretically
could continue to react with camphor, cyclize, and dehydrate to form
the pyrylium, to no avail. Alternative approaches starting with either
3-benzylidenecamphor or 3-benzoylcamphor produced only an in-
tractable tar, in which no downfield aromatic peaks indicative of the
pyrylium ring could be observed by 13C NMR spectroscopy. In the case
of BF3, the 3-benzoylcamphor route simply produced a crystalline
adduct.11
8
O
HBF4
1. KH, hexanes
2. 3, reflux, 18 h
toluene
O
Dean-Stark
- H2O, 55%
67%
9
P(TMS)3
Ar2
Ar2
CH3CN
82%
P
O
HBF4
-
BF4
6
10
Ar1
O
O
Ar3
Ar1
O
BF4
Ar3
Scheme 3. Synthesis of biscamphorpyrylium 6 and biscamphorphosphinine 10.
-
Ar2
P
P(TMS)3
acetonitrile
elimination by an excess of camphor enolate in refluxing hexanes to
give the corresponding 1,5-enedione (9) in 67% yield. When the
reaction was performed in polar solvents such as THF, and to some
extent in toluene, reduction of the vinyl halide 8 to 3-benzylidene-
camphor was observed to occur in high yield. By reducing solvent
polarity, this effect was mitigated. A mixture of two isomers of 9 was
formed, which was separated by fractional crystallization. The major
isomer (93% by GC/MS) was determined by X-ray crystallography to
be E-(9). 1H and 13C NMR spectroscopy of this compound displayed
temperature dependant behavior, at 25 ꢀC giving broadened peaks in
the 1H NMR spectrum and missing peaks in the 13C NMR spectrum.
We attributed this to hindered rotation in the molecule, probably
around the single bond between the vinyl carbon and the camphor
ring. In an attempt to obtain averaged spectra at higher temperature,
NMR spectra were obtained in C6D6 at 65 ꢀC, but neither the 1H nor
13C spectra were qualitatively different and there was evidence of E/Z
isomerization occurring. However, at ꢁ40 ꢀC, rotation was suffi-
ciently slowed that all expected 13C peaks were observed. Compound
9 then underwent a dehydration/cyclization using HBF4/diethyl ether
in refluxing toluene, using a Dean–Stark trap, to give the pyryliumsalt
6 in 55% yield. The 13C NMR spectrum displayed downfield aromatic
peaks indicative of the pyrylium cation (184.4,155.7, and 137.5 ppm).
The pyrylium 6 structure was also verified by X-ray crystallogra-
Ar1
Ar3
Scheme 1. A typical synthesis of a triarylpyrylium salt and triarylphosphinine.
Our next attempts were patterned after a procedure of Balaban
et al., for converting 1,5-diketones to pyryliums, using triphenyl-
carbenium (trityl) salts as hydride abstraction reagents.12 Following
this lead, the 1,5-diketone 7 (Scheme 2) was synthesized by literature
methods.13 All initial attempts at cyclization using trityl reagents in
acetic acid, acetic anhydride, and acetonitrile failed to produce any
pyrylium salts, furnishing again only the intractable tars previously
experienced. Simalty et al.9 had experienced similar difficulties in the
synthesis of a monocamphorpyrylium, all reactions ending in only
tar. However, they found that a mixture of 2:1 acetic acid to acetic
anhydride gave their product in w35% yield. Using this solvent
mixture finally produced the desired biscamphorpyrylium 6, albeit in
a miniscule 8% yield. The poor yield and difficult workup made this
route less than ideal, so a better method was sought.
O
Trityl BF4
phy, as shown in Figure 2. This represents the first C2 chiral pyrylium
salt to be reported. It displays a large optical rotation, [
20
a]
þ378
2:1 AcOH:Ac2O
D
O
(c 0.37, CH2Cl2), consistent with a highly polarized aromatic ring.
The biscamphorphosphinine 10 was prepared from the pyrylium
salt 6 by reaction with tris-(trimethylsilyl)phosphine, P(TMS)3. The
yield of this transformation, which is typically only from 15–40%, in
this case was remarkably and reproducibly high (82%).2 This may be
due to the meta substituents preventing attack on the para position,
which would lead to byproducts. This compound is easily purified by
silica gel chromatography followed by recrystallization, and is
obtained as an air-stable crystalline solid. Interestingly, the 13C NMR
spectrum of this compound shows carbon–phosphorus coupling as
far as five bonds. The X-raycrystal structure of 10 is shown in Figure 3,
and represents the first C2 chiral phosphinine reported to date.
31P NMR spectroscopy showed that this ligand binds strongly to
late transition metals, such as palladium(II) and rhodium(I). When
O
8% yield
-
BF4
7
6
Scheme 2. Initial synthesis of biscamphorpyrylium 6.
We theorized that the difficult step in the cyclization/aromatiza-
tion process was probably the formal loss of hydrogen, so we sought
to incorporate the unsaturation into the structure earlier in the
synthesis, which proved to be successful (Scheme 3). 3-Benzoyl-
camphor was refluxed in PCl3 for 2 h, and then neutralized to furnish
the known (E)-3-(
This compound was then subjected to conjugate addition/
a
-chlorobenzylidene)camphor (8) in 42% yield.14

9370
J.R. Bell et al. / Tetrahedron 65 (2009) 9368–9372
quantitatively at room temperature in 5 days, or at 40 ꢀC in 24 h to
give the trichlorosilane product; however, no reaction occurred
for 1-octene. Under either reaction conditions, after isolation of
the trichlorosilane product by Kugelrohr distillation, followed by
Tamao–Fleming oxidation16 to the alcohol, the 1-phenylethanol
product was found to have 27% ee by chiral GC analysis.17
The low level of asymmetric induction can be attributed to two
features that are evident from the X-ray structure of 10. The bis-
camphor ligand influences the metal environment primarily
through the bridgehead methyl groups. Unfortunately, these
methyl groups cannot extend appreciably into the metal’s co-
ordination sphere (Fig. 4), which limits the ligand’s ability to en-
force an asymmetric environment. In addition, the methyl groups
deviate only marginally (w18ꢀ) from the plane of the phosphinine
ring. If the methyl groups were coplanar with the phosphinine ring,
there would be no effective asymmetry at the metal. We believe
ligands, which extend further into the metal’s environment and
with a greater degree of twist (i.e., deviation from the phosphinine
plane) will be more effective in asymmetric reactions. Thus, ligands
with groups significantly larger than methyl (e.g., phenyl) posi-
tioned so as to extend further into the metal’s coordination sphere
should be superior. In addition, models suggest that bicyclic rings
([2.2.1] or [2.2.2]) fused to the phosphinine cannot provide these
extending groups more than the shallow twist angle observed for
10. We believe a significantly greater angle, perhaps two to three
times greater than found in ligand 10, is necessary for effective
asymmetric induction. We are pursuing non-bicyclic C2 ligands, as
well as the utilization of chlorobenzylidene 8 in the rapid prepa-
ration of a wide variety of C1 asymmetric phosphinines.
Figure 2. X-ray crystal structure of biscamphorpyrylium tetrafluoroborate 6. The tet-
rafluoroborate anion has been omitted for clarity.
Figure 3. X-ray crystal structure of biscamphorphosphinine 10.
10 was added to a solution of Pd(COD)Cl2, the 31P NMR signal shifts
upfield from 138.9 ppm for the free phosphinine to 119.1 ppm on
formation of the di(phosphinine 10)–dichloropalladium(II) com-
plex. When 10 was reacted with rhodium(dicarbonyl) chloride di-
mer, the 31P NMR spectrum shows a doublet at 122.5 (J¼152 Hz),
due to Rh–P coupling, consistent with literature reports for other
phosphinines.4
Figure 4. Face-on space-filling view of phosphinine 10. Phosphorus atom is at bottom
center.
3. Experimental section
3.1. General
Phosphinine 10 was briefly studied in one catalytic asymmetric
reaction,
a palladium-catalyzed asymmetric hydrosilylation
Reagents and solvents were generally purchased from the Aldrich
Chemical Company or from Alfa Aesar, and were used as received
unless otherwise noted. Tris-(trimethylsilyl)phosphine was obtained
from Strem Chemical. NMR spectra were obtained at 500 MHz for 1H,
125 MHz for 13C, and 202 MHz for 31P. Spectra obtained in CDCl3 were
referenced to TMS (0 ppm) for 1H and to solvent (77.0 ppm) for 13C.
31P spectra were referenced to an external standard of 85% H3PO4
(0 ppm). GC–MS was done on a Hewlett-Packard GCD using
a 30 mꢂ0.25 mm HP-5 capillary column with helium carrier and EI
ionization. Analysis for enantiomeric purity was by capillary gas
(Scheme 4). Styrene and 1-octene were treated with 0.5 mol %
allylpalladium(II) chloride dimer and 1.75 mol % of phosphinine
10, conditions patterned after those employing other mono-
dentate phosphine ligands.15 For styrene, the reaction proceeds
HSiCl3
10
OH
(1.75 mol%)
KF, KHCO3
(C3H5PdCl)2
(0.5 mol%)
H2O2. MeOH
chromatography using an HP5890 with
a Restek Rt-bDEXsa,
27% ee
30 mꢂ0.25 mm column. High resolution mass spectra were obtained
from the Baylor University mass spectroscopy facility. Elemental
Scheme 4. Hydrosilylation/oxidation of styrene using phosphinine ligand 10.

J.R. Bell et al. / Tetrahedron 65 (2009) 9368–9372
9371
analyses were performed by Atlantic Microlabs, Norcross, GA. Melt-
ing points were calibrated against accepted standards.
2H), 1.89 (tdd, J¼11.2, 3.6, 2.2, 2H), 1.43 (s, 6H), 1.16 (ddd, J¼11.5, 9.3,
3.2, 2H), 1.07 (ddd, J¼11.5, 9.1, 3.3, 2H), 0.90 (s, 6H), 0.55 (s, 6H); 13
C
NMR (125.7 MHz, CDCl3):
d
176.7 (2C, d, J¼49.3), 148.9 (1C, d,
3.2. Enedione 9
J¼12.1), 140.3 (d, J¼1.9), 132.6 (d, J¼15.0), 129.4 (2C, d, J¼1.9, CH),
128.0 (2C, CH), 126.6 (CH), 57.7 (2C, d, J¼16.7), 56.8 (2C, d, J¼2.3),
53.1 (2C, d, J¼1.4, CH), 34.5 (2C, d, J¼3.3, CH2), 25.9 (2C, d, J¼1.8,
CH2), 19.9 (2C, CH3), 19.3 (2C, d, J¼1.4, CH3), 13.6 (2C, d, J¼11.2, CH3);
To 1.1 g (27 mmol, 2.5 equiv) of KH in 30 mL of hexanes, 5.42 g
(35.7 mmol, 3 equiv) of (þ)-camphor was added and refluxed for
2 h. Then 3.00 g (10.9 mmol, 1 equiv) of (þ)-(E)-3-(
a
-chloro-
31P NMR (202.3 MHz, CDCl3):
d 138.9; IR (KBr): 3075, 3020, 2964,
benzylidene) camphor14 in 20 mL of hexanes was slowly added by
syringe, and the reaction mixture was refluxed for 18 h. The re-
action was quenched by addition of saturated ammonium chloride
solution, and brought to neutral pH with 6.0 M HCl. The mixture
was extracted with diethyl ether, dried with magnesium sulfate,
filtered, and evaporated to a solid. The mixture was dissolved in
1491, 1437, 1385, 1376 cmꢁ1; HRMS (EI) calcd for C27H33P [Mþ]
388.2320, found 388.2320. Elemental analysis calcd (%) for C27H33P:
C, 83.47; H, 8.68. Found: C, 83.65; H, 8.68.
3.5. General method for asymmetric hydrosilylation
warm methanol and, after cooling to ꢁ20 ꢀC, filtered to give crys-
To 0.5 mg (1.3
with reflux condenser under nitrogen was added 3.7 mg
(9.5 mole) of biscamphorphosphinine 10, followed by 25 mmol of
mmol) of [(allyl)Pd(II)Cl]2 in a 10 mL flask fitted
20
talline enedione 9 (2.87 g, 68% yield). Mp 142–144 ꢀC; [
a]
þ82.6 (c
a
D
1
0.33, EtOAc); H NMR (500 MHz, CDCl3, ꢁ40 ꢀC):
d
7.38–7.32 (m,
m
4H, ArH), 7.13 (br d, J¼7.8 Hz, 1H, ArH), 5.32 (br d, J¼3.6 Hz, 1H),
2.38 (br t, J¼3.8 Hz, 1H), 2.17 (br d, J¼4.0 Hz, 1H), 1.99–1.91 (m, 1H),
1.76–1.68 (m, 1H), 1.56–1.39 (m, 4H), 1.23–1.15 (m, 1H), 1.04 (s, 3H),
neat olefin (styrene or 1-octene). After stirring at room temperature
for 10 min, 3 mL (30 mmol) of trichlorosilane was syringed into the
flask, and the mixture was brought to reflux overnight. The alkyl-
trichlorosilane product was isolated by Kugelrohr distillation
(100 ꢀC at w0.5 Torr). The isolated product was subjected to
Tamao–Fleming oxidation conditions. The resulting alcohol was
0.98 (s, 3H), 0.93 (s, 3H), 0.88 (s, 3H), 0.85 (s, 3H), 0.82 (s, 3H), 0.80–
13
0.74 (m, 1H); C NMR (125.7 MHz, CDCl3, ꢁ40 ꢀC):
d 217.9, 209.9,
143.6, 142.9, 140.2, 128.5, 128.0, 127.7, 127.6, 127.1, 59.2, 59.0, 52.6,
50.5, 49.4, 46.4, 46.0, 30.0, 29.0, 26.4, 21.7, 20.3, 19.3, 18.8, 18.1, 9.7,
9.5; IR (KBr) 2960, 1741, 1711, 1622 cmꢁ1; HRMS (EI) calcd for
C27H34O2 [Mþ] 390.2559, found 390.2560. Elemental analysis calcd
(%) for C27H34O2: C, 83.03; H, 8.77. Found: C, 82.86; H, 8.91.
analyzed by GC/FID for enantiomeric purity using
a Restek
Rt- DEXsa column using hydrogen carrier gas. The enantiomers
b
were well-resolved (12.18 min vs 12.48 min, resolution¼3.18,
temperature program 80–180 ꢀC at 4 ꢀC per min).
3.3. Biscamphorpyrylium tetrafluoroborate 6
3.6. X-ray crystallography
To 0.431 g (1.1 mmol,1 equiv) of 9 dissolved in 20 mL of toluene in
a flask equipped with a Dean–Stark trap was added 1.00 mL
(5.9 mmol, 5.4 equiv) of w52% HBF4 in diethyl ether (w5.92 M) and
the solution refluxed overnight. The toluene solution was diluted
with 100 mL of diethyl ether, cooled to 0 ꢀC, and filtered through
a Celite pad, trapping a fine precipitate. The pyrylium salt was then
washed through the Celite with dichloromethane into a separate
flask. The solvent was removed under vacuum, and the compound
was recrystallized from acetone/diethyl ether to yield biscam-
phorpyrylium tetrafluoroborate 6 (0.278 g, 55%). Mp 272–274 ꢀC;
Crystallographic data (excluding structure factors) for the struc-
tures 6, 8, 9, and 10 in this paper have been deposited with the
Cambridge Crystallographic Data Centre as supplementary publica-
tion nos. CCDC 745389, 745390, 745391, and 745392, respectively.
Copies of the data can be obtained, free of charge, on application to
CCDC, 12 Union Road, Cambridge CB2 1EZ, UK (fax: þ44 (0)1223
336033 or e-mail: deposit@ccdc.cam.ac.uk). Abbreviated crystallo-
graphic data are given below.
3.6.1. Pyrylium 6. C27H33BF4O, MW¼460.34, orthorhombic, P212121,
20
20
[
a
]
þ309 (c 0.44, CH3CN), [
a
]
þ378 (c 0.37, CH2Cl2); 1H NMR
a¼14.931(3) Å, b¼7.4327(19) Å, c¼10.861(3) Å,
a
¼90.00ꢀ,
b
¼90.00ꢀ,
D
D
(500 MHz, CDCl3):
d
7.66–7.63 (m, 3H, ArH), 7.56–7.54 (m, 2H, ArH),
g
¼90.00ꢀ, V¼1205.3(5), Z¼2, T¼110 K,
m
¼0.096, rcalcd¼
3.25 (d, J¼4.0 Hz, 2H), 2.39 (ddt, J¼13.0, 9.8, 4.0 Hz, 2H), 2.22 (ddd,
J¼13.2, 9.8, 4.0 Hz, 2H), 1.79 (ddd, J¼13.2, 9.3, 3.9 Hz, 2H), 1.60 (ddd,
J¼13.0. 9.3, 3.9 Hz, 2H),1.50 (s, 6H),1.05 (s, 6H), 0.77 (s, 6H); 13C NMR
1.268 Mg mꢁ3, GOF on F2¼1.042, R¼0.0495, Rw¼0.1089 [I>2
s(I)].
3.6.2. Chlorobenzylidene 8. C17H19ClO, MW¼274.77, monoclinic,
(125.7 MHz, CDCl3):
d
184.4 (2C),155.7 (1C),137.5 (2C),132.5 (1C, CH),
P21, a¼6.9230(9) Å, b¼12.4494(15) Å, c¼16.783(2) Å,
a
¼90.00ꢀ,
131.1 (1C), 129.65 (2C, CH), 129.64 (2C, CH), 60.6 (2C), 57.0 (2C), 50.4
(2C, CH), 31.3 (2C, CH2), 25.0 (2C, CH2), 19.7 (2C, CH3), 19.0 (2C, CH3),
8.6 (2C, CH3); IR (KBr) 2960, 1608, 1415, 1055 cmꢁ1. HRMS (EI) calcd
for C27H33O [Mþ] 373.2531, found 373.2530. Elemental analysis calcd
(%) for C27H33OBF4: C, 70.44; H, 7.23. Found: C, 70.58; H, 7.33.
b
¼93.181(5)ꢀ,
g
¼90.00ꢀ, V¼1444.2(3), Z¼4, T¼110 K,
m¼0.0.254,
rcalcd¼1.264 Mg mꢁ3
,
GOF on F2¼1.035, R¼0.0271, Rw¼0.0608
[I>2s(I)].
3.6.3. Enedione 9. C27H34O2, MW¼390.54, orthorhombic, P212121,
a¼6.6892(14) Å, b¼15.409(3) Å, c¼21.528(4) Å,
a
¼90.00ꢀ,
b
¼90.00ꢀ,
3.4. Biscamphorphosphinine 10
g
¼90.00ꢀ, V¼2219.0(7), Z¼4, T¼110 K,
m
¼0.072, rcalcd¼
1.169 Mg mꢁ3, GOF on F2¼1.033, R¼0.0407, Rw¼0.0828 [I>2
s(I)].
To 0.913 g (1.98 mmol, 1 equiv) of biscamphorpyrylium tetra-
fluoroborate (6) dissolved in 10 mL of anhydrous acetonitrile placed
carefully under nitrogen was added 1.0 g (4.0 mmol, 2 equiv) of
P(TMS)3 and the solution was refluxed for 24 h. The solution turned
from orange to dark red/black. After cooling to room temperature,
the solvent was removed by rotary evaporation and the phosphi-
nine was purified by silica gel column chromatography (5% ethyl
acetate in petroleum ether). The resulting yellow solid was
3.6.4. Phosphinine 10. C27H33P, MW¼388.50, triclinic, P1, a¼
6.866(2) Å, b¼12.056(4) Å, c¼14.084(6) Å,
a
¼78.525(12)ꢀ,
b¼
82.284(14)ꢀ,
g
¼74.282(12)ꢀ, V¼1095.7(7), Z¼2, T¼110 K,
m
¼0.135,
rcalcd¼1.178 Mg mꢁ3, GOF on F2¼1.024, R¼0.00405, Rw¼0.0908
[I>2s(I)].
Acknowledgements
recrystallized from methanol to give phosphinine 10 as colorless
20
needles (0.631 g, 82%). Mp 123–124 ꢀC; [
a
]
D
þ51.8 (c 0.55, EtOAc);
We would like to thank the Donors of the American Chemical
Society Petroleum Research Fund for support of this research (grant
#47942-AC1), the Robert A. Welch Foundation (grant #AA-1395)
1H NMR (CDCl3, 500 MHz):
1H, ArH), 7.24–7.20 (m, 2H, ArH), 2.69 (d, J¼3.9, 2H), 2.00–1.93 (m,
d
7.43 (t, J¼7.4, 2H, ArH), 7.38–7.33 (m,

9372
J.R. Bell et al. / Tetrahedron 65 (2009) 9368–9372
5. (a) Mu¨ ller, C.; Pidko, E.; Totev, D.; Lutz, M.; Spek, A.; van Santen, R. A.; Vogt, D.
Dalton Trans. 2007, 5372–5375; (b) Mu¨ ller, C.; Pidko, E.; Staring, A. J. P. M.;
Lutz, M.; Spek, A.; van Santen, R. A.; Vogt, D. Chem.dEur. J. 2008, 14, 4899–
4905.
for partial support of this work, the National Science Foundation
(Award #CHE-0420802) for funding the purchase of our 500 MHz
NMR, and Dr. Gary Stidsen of Restek Corporation for the gift of
6. Whitesell, J. Chem. Rev. 1989, 89, 1581–1590.
a chiral GC column (Rt-bDEXsa).
7. Mathey, F.; Le Floch, P. 1l
3-Phosphinines. In Science of Synthesis: Houben-Weyl
Methods of Molecular Transformations; Thomas, E. J., Ed.; Georg Thieme: Stutt-
gart, 2003; Vol. 15, pp 1097–1155.
Supplementary data
8. Pyrylium Salts: Synthesis, Reactions, and Physical Properties; Katritzky, A. R., Ed.
Advances in Heterocyclic Chemistry; Academic: New York, NY, 1982; Supple-
ment 2; 434 pp.
9. Barnaud, Y.; Maroni, P.; Simalty, M.; Madaule, Y. Bull. Soc. Chim. Fr. 1970, 4,
1398–1403.
10. Balaban, T. S.; Balaban, A. T. Pyrylium Salts In. Science of Synthesis: Houben-
Weyl Methods of Molecular Transformations; Thomas, E. J., Ed.; Georg Thieme:
Stuttgart, 2003; Vol. 14, pp 11–200.
11. Cullen, W. R.; Rettig, S. J.; Trotter, J.; Wickenheiser, E. B. Can. J. Chem. 1988, 66,
2007–2013.
Procedures for the synthesis of pyrylium 6 from dione 7 and for
the preparation of chlorobenzylidene 8, X-ray structures of chloro-
benzylidene 8 and enedione 9, a table of X-ray data for compounds
6, 8, 9, and 10 and NMR spectra for compounds 6, 8, 9, and 10 are
provided. Supplementary data associated with this article can be
found in online version, at doi:10.1016/j.tet.2009.08.085.
12. Balaban, A. T.; Barbalescu, N. S. Rev. Roum. Chim. 1966, 11, 109–112.
13. Sotiropoulos, J.; Batouti, N.; Lamazoue`re, A. J. Heterocycl. Chem. 1987, 24,
907–912.
References and notes
14. Sotiropoulos, J. C.R. Acad. Sci. 1970, 270, 1727–1730.
15. Hayashi, T. Catal. Today 2000, 62, 3–15.
16. Tamao, K.; Ishida, N.; Tanaka, T.; Kumada, M. Organometallics 1983, 2, 1694–
1696.
1. For a review see: Mu¨ ller, C.; Vogt, D. Dalton Trans. 2007, 5505–5523.
2. Breit, B.; Winde, R.; Mackewitz, T.; Paciello, R.; Harms, K. Chem.dEur. J. 2001, 7,
3106–3121.
3. Breit, B. Chem. Commun. 1996, 2071–2072.
17. Enantiomeric purity was determined by chiral GC analysis using a Restek
4. Mu¨ ller, C.; Lopez, L. G.; Kooijman, H.; Spek, A.; Vogt, D. Tetrahedron Lett. 2006,
47, 2017–2020.
Rt-
bDEXsa chiral column.