P=C bonds as building blocks for three- and four-membered heterocyclic cations: Synthesis, structures and mechanistic studies

Article abstract of DOI:10.1002/chem.201101535

The activation of the P=C bond of phosphaalkenes with electrophiles is investigated as a means to prepare and characterize unusual organophosphorus compounds. Treatment of RP=CHtBu (1 a: R=tBu; 1 b: R=1-adamantyl) with HOTf (0.5 equiv) affords diphosphiranium salts [RP-CHtBu-PR (CH₂tBu)]OTf ([2 a]OTf and [2 b]OTf), each containing a three-membered P₂C ring. In contrast, the addition of MeOTf (0.5 equiv) to either 1 a or 1 b affords diphosphetanium salts [RP-CHtBu-P(Me)R-CHtBu]OTf ([3 a]OTf and [3 b]OTf) containing four-membered P₂C₂ heterocycles. The phosphenium triflate [tBuP(CH₂tBu)]OTf ([5 a]OTf) and methylenephosphonium triflate [tBu(Me)P=CHtBu]OTf ([7 a]OTf) are identified spectroscopically as intermediates in the formation of [2 a]⁺ and [3 a]⁺, respectively. The phosphenium triflate intermediate can be trapped with 2-butyne to afford phosphirenium salt [MeC=CMe-tBuPCH ₂tBu]OTf ([6 a]OTf). Copyright

Full text of DOI:10.1002/chem.201101535

DOI: 10.1002/chem.201101535
P=C Bonds as Building Blocks for Three- and Four-Membered Heterocyclic
Cations: Synthesis, Structures and Mechanistic Studies
Joshua I. Bates and Derek P. Gates*[a]
ꢀ
ꢀ
ꢀ
Abstract: The activation of the P=C
bond of phosphaalkenes with electro-
philes is investigated as a means to pre-
pare and characterize unusual organo-
phosphorus compounds. Treatment of
RP=CHtBu (1a: R=tBu; 1b: R=1-
adamantyl) with HOTf (0.5 equiv) af-
um
salts
[RP CHtBu P(Me)R
the formation of [2a]+ and [3a]+, re-
spectively. The phosphenium triflate in-
termediate can be trapped with 2-
butyne to afford phosphirenium salt
CHtBu]OTf ([3a]OTf and [3b]OTf)
containing four-membered P2C2 hetero-
cycles. The phosphenium triflate [tBuP-
ꢀ
(CH2tBu)]OTf ([5a]OTf) and methyle-
[MeC=CMe tBuPCH2tBu]OTf
nephosphonium triflate [tBu(Me)P=
CHtBu]OTf ([7a]OTf) are identified
spectroscopically as intermediates in
([6a]OTf). Treatment of diphospheta-
nium [3a]OTf with an excess MeOTf
ꢀ
ꢀ
ꢀ
ꢀ
fords diphosphiranium salts [RP
affords
CHtBu]
[Me2P CHtBu PMetBu
(OTf)2 ([4a](OTf)2), com-
ꢀ
CHtBu PR (CH2tBu)]OTf ([2a]OTf
G
N
a
and [2b]OTf), each containing a three-
membered P2C ring. In contrast, the
addition of MeOTf (0.5 equiv) to
either 1a or 1b affords diphosphetani-
pound containing a diphosphetanium
dication. The molecular structures are
Keywords: multiple bonds · phos-
phaalkenes · phosphetanes · phos-
phiranes · phosphorus heterocycles
reported for [2a]OTf, [2b][H
ACHTUNGTRENNUNG(OTf)2],
[3a]I, [3b]I, [4a](OTf)2, and [6a]OTf.
AHCTUNGTRENNUNG
Introduction
Coordinatively unsaturated compounds, such as olefins, are
important building blocks in the construction of rings,
chains, and polymers. The development of a similar diverse
chemistry for inorganic multiple bonds is of interest, owing
to the prospect of accessing unique compounds with fasci-
nating structures and useful properties.[1] Perhaps most
widely investigated are phosphaalkenes, for which a strong
analogy has been established between the chemistry of C=C
and P=C bonds (Scheme 1).[2–4] This relationship is so com-
pelling that it has led to phosphorus being coined the
“carbon copy” and to the proposal of a new field of research
called “phospha-organic” chemistry. Our group has been in-
terested in the employment of P=C compounds as precur-
sors to new functional polymers through addition-polymeri-
zation processes and we have studied their reactions with
radical, anionic, and cationic initiators.[5] Although cationic
initiation has not yet afforded macromolecules, we have
shown that reactions of the P=C bond with electrophiles
afford unexpected strained heterocyclic cations.[6,7]
Scheme 1. Examples of reactions illustrating the P=C/C=C analogy: i) ad-
dition of HX (X=halide), ii) Diels–Alder reaction with 1,3-butadiene, iii)
epoxidation (provided the phosphorus lone pair is protected through co-
ordination), and iv) addition polymerization.
no isolable counterparts in nitrogen chemistry, they illustrate
the striking differences between the chemistry of the first
and second periods. A dramatic illustration of this difference
is provided by the fascinating isomeric possibilities for
[E2CR5]+ ions (E=N or P, Scheme 2). Whereas a symmetri-
cally substituted diphosphiranium cation A has been isolat-
ed,[9,10] analogous diaziridinium cations B are unknown, with
the acyclic amidinium cation D being 129.2 kcalmolꢀ1 lower
in energy for the parent system [N2CH5]+. In contrast, the
acyclic phosphinomethylenephosphonium form C is just
24.5 kcalmolꢀ1 lower in energy than cyclic A.[11] Thus,
through judicious choice of substituents, both cyclic A and
The study of strained phosphorus homo- and heterocycles
has attracted considerable attention, owing to their novel
structures, bonding, and reactivity.[8] Moreover, because
cyclic phosphorus species often behave differently or have
[a] J. I. Bates, Prof. Dr. D. P. Gates
Department of Chemistry
University of British Columbia
2036 Main Mall, Vancouver, BC V6T 1Z1 (Canada)
E-mail: dgates@chem.ubc.ca
1674
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 1674 – 1683

FULL PAPER
erocycle, thereby enabling its isolation. Thus, phosphaalkene
1a,[18,19] bearing bulky and electron-donating tBu substitu-
ents, was chosen for initial investigations. A solution of
phosphaalkene 1a in CH2Cl2 was treated with HOTf
(0.5 equiv) and subsequently, the reaction mixture was ana-
lyzed by 31P{1H} NMR spectroscopy. Interestingly, the spec-
trum revealed that 1a (d=274 ppm) was consumed and was
replaced by two sets of doublets with large coupling con-
stants (Jꢁ250 Hz), suggesting the presence of two species
ꢀ
(major: 60%; minor: 40%) each with P P bonds. Single
Scheme 2. Cyclic and acyclic isomers of [P2CR5]+ and [N2CR5]+.
crystals suitable for X-ray crystallography were obtained
from a concentrated solution of the sample in toluene/hex-
anes. The molecular structure is shown in Figure 1 and con-
acyclic C[12] isomers of [P2CR5]+ are synthetically accessible,
owing to the similar energies of the parent systems
[P2CH5]+.
Intriguingly, the asymmetrically substituted diphosphirani-
um cation E has been calculated to be 47.1 kcalmolꢀ1 lower
in energy than the known symmetrically substituted A.[11] It
is therefore surprising that, until the present studies, a di-
phosphiranium cation of type E had eluded isolation. There
have been two reports of the spectroscopic detection of spe-
cies of type E. The first involved methylating the neutral di-
phosphirane (Mes*P)2CH2 with MeOTf to afford diphos-
phiranium [Mes*(Me)PCH2PMes*]OTf (cf. E), which was
detected spectroscopically at low temperatures.[11] The
second involved treating the phosphaalkenes Mes*P=CH2 or
MesP=CPh2 with electrophiles (E=GaCl3, H+, or [P-
ACHTUNGTRENNUNG
(NEt2)2]+) to afford diphosphiranium species that were de-
tected spectroscopically in solution, but could not be isolat-
ed.[7] Presumably, the diphosphiranium cation is derived
from the cyclization of a phosphinophosphenium species G
through intramolecular donor/acceptor-adduct formation.
Perhaps not surprisingly, the analogous asymmetric diaziridi-
nium cations F and aminonitrenium cations H have not
been isolated.[13] Notably, there is considerable interest in
P,N-hybrid systems (e.g., [CNPR5]+) with isolable examples
of compounds of type C/D[14] and spectroscopic evidence for
species of type E/F.[15–17] The aforementioned examples fur-
ther demonstrate the striking differences between the
chemistry of the first and second periods.
Herein, we report our studies of the reactions of phos-
phaalkenes with electrophiles to afford isolable asymmetric
diphosphiranium salts of type E. Surprisingly, variation of
the electrophile affords four-membered 1,3-diphosphetani-
um [P2C2R7]+ ions rather than diphosphiranium [P2CR5]+
ions. These studies illustrate the unexpected reactivity of
phosphaalkenes and the utility of P=C bonds as simple
building blocks for fascinating organophosphorus heterocy-
cles.
Figure 1. Molecular structure of the diphosphiranium cation [2a]OTf
(thermal ellipsoids set at 50% probability). The triflate anion and all hy-
drogen atoms, except H1, are omitted for clarity. Selected bond lengths
ꢀ
ꢀ
ꢀ
ꢀ
[ꢁ] and angles [8]: P1 P2 2.164(1), P1 C1 1.836(1), P1 C2 1.815(1), P1
ꢀ
ꢀ
C3 1.865(2), P2 C2 1.888(1), P2 C11 1.890(1); P2-P1-C2 55.8(1), P1-P2-
C2 52.7(1), P1-C2-P2 71.5(1).
firms the successful isolation of [2a]OTf, a diphosphiranium
cation of type E.[20] Interestingly, the molecular structure re-
veals that the P-tBu substituents are syn configured and
analysis of the crystals by 31P NMR spectroscopy revealed
that this is the major product (60%). Although we have not
been able to obtain crystals of the minor product, we specu-
late that it might be the isomer of [2a]OTf in which the
P-tBu substituents are anti configured.
Results and Discussion
In an effort to explore the generality of this new reaction,
a solution of P-Ad-substituted phosphaalkene 1b (Ad=1-
adamantyl)[19] in CH2Cl2 was treated with HOTf (0.5 equiv).
Analysis of the reaction mixture by 31P{1H} NMR spectros-
copy revealed that the signal assigned to phosphaalkene 1b
Asymmetric diphosphiranium cations: We hypothesized that
the incorporation of electron-donating P-alkyl substituents
may provide additional stabilization to the cationic P2C het-
Chem. Eur. J. 2012, 18, 1674 – 1683
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1675

D. P. Gates and J. I. Bates
(d=268 ppm) was consumed and was replaced by two sets
of doublets that, by analogy with [2a]OTf, were assigned to
the isomers of [2b]OTf. Attempts to grow crystals of
[2b]OTf from various combinations of CH2Cl2, toluene,
Et2O, and hexanes were unsuccessful. However, employing
a mixture of acetonitrile, CH2Cl2, and hexanes afforded
single crystals suitable for X-ray crystallographic analysis.
The molecular structure of the diphosphiranium cation in
an aliquot was removed from the reaction mixture for analy-
sis by 31P{1H} NMR spectroscopy. The spectrum featured a
single set of doublet resonances, suggesting that a single
ꢀ
product was formed. Strikingly, the P P coupling constant
ꢀ
of only J=15 Hz was not consistent with a direct P P bond
(1JPP >200 Hz), but suggested at least two-bond P P cou-
ꢀ
pling (nJPP with nꢂ2). Moreover, the product is stable to-
wards air and moisture in solution and in the solid state.
1
[2b][H
Ad substituents are syn configured. Interestingly, the [2b]+
ion crystallizes as a [H
(OTf)2]ꢀ salt. Unfortunately, our ef-
forts to characterize the crystals of [2b][H(OTf)2] by NMR
A
The H NMR spectrum shows a doublet resonance consis-
tent with the presence of a P-Me group (d=2.30 ppm,
2JPH =11 Hz). Clearly, this data was not consistent with the
product being a diphosphiranium species.
AHCTUNGTRENNUNG
AHCTUNGTRENNUNG
spectroscopy were precluded by their extreme sensitivity
both in the solid state and in solution.
Unfortunately, attempts to obtain crystals of the product
suitable for X-ray crystallographic analysis were unsuccess-
ful. Therefore, the triflate ion was exchanged with iodide by
repeated extraction of a solution of the product in CH2Cl2
with an aqueous solution of potassium iodide. Recrystalliza-
tion of the iodide salt afforded crystals that were analyzed
by X-ray crystallography as the 1,3-diphosphiranium salt
[3a]I (Figure 3). In an analogous fashion, treating phos-
phaalkene 1b with MeOTf (0.5 equiv), followed by anion
exchange with KI(aq) affords the diphosphetanium salt [3b]I
(Figure 4).
Although cationic P2C2 rings have been observed within
fused polycyclic structures,[23–25] [3a]+ and [3b]+ represent
the first structurally characterized examples of simple 1,3-di-
Figure 2. Molecular structure of the diphosphiranium cation in [2b][H-
A
(OTf)2]ꢀ ion
ACHTUNGTRENNUNG
ꢀ
ꢀ
ꢀ
lengths [ꢁ] and angles [8]: P1 P2 2.161(1), P1 C1 1.843(2), P1 C16
ꢀ
ꢀ
ꢀ
1.815(2), P1 C6 1.866(2), P2 C16 1.892(2), P2 C21 1.875(2); P2-P1-C2
56.0 (1), P1-P2-C2 52.7(1), P1-C2-P2 71.3(1).
To our knowledge, compounds [2a]OTf and [2b][H-
ACHTUNGTRENNUNG(OTf)2] are the first crystallographically characterized asym-
metric diphosphiranium cations (i.e., E) bearing only carbon
or hydrogen ring substituents.[20] The metrical parameters
within the P2C rings of [2a]+ and [2b]+ do not differ signifi-
ꢀ
cantly. The P P bond lengths (2.16 ꢁ) are shorter than typi-
[21]
+
ꢀ
ꢀ
cal P P single bonds (2.25 ꢁ),
but similar to the P P
bonds
observed
in
the
triphosphiranium
cation
[(tBuP)3Me]OTf (2.16 ꢁ).[22] The endocyclic P C bonds in-
ꢀ
+
ꢀ
volving the phosphonium (C P : av=1.82 ꢁ) are shorter
than those involving the phosphine (C P : av=1.89 ꢁ), but
0
ꢀ
[21]
ꢀ
are in the range for P C single bonds (1.85–1.95 ꢁ).
Within the P2C ring the bond angles at phosphorus in [2a]+
and [2b]+ (C-P+-P0: av=55.98; C-P0-P+: av=52.78) are
smaller than those at carbon (P+-C-P0: av=71.48).
Figure 3. Molecular structure of the diphosphetanium ion in [3a]I (ther-
mal ellipsoids set at 50% probability). Two virtually identical molecules
appear in the asymmetric unit. Metrical parameters are given for one of
the two molecules. The iodide atom and all hydrogen atoms except H1
and H2 are omitted for clarity. Selected bond lengths [ꢁ] and angles [8]:
1,3-Diphosphetanium cations: Having established that
HOTf can promote the cyclodimerization of P=C bonds,
treatment of phosphaalkenes with MeOTf was expected to
afford analogous diphosphiranium cations. Thus, phosphaal-
kene 1a was treated with MeOTf (0.5 equiv) in CH2Cl2 and
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
P1 C1 1.858(2), P1 C2 1.859(2), P1 C9 1.883(2), P1 C21 1.810(2), P2
ꢀ
ꢀ
C1 1.905(2), P2 C2 1.900(3), P2 C17 1.891(2); C1-P1-C2 91.3 (1), P1-C1-
P2 90.0(1), C1-P2-C2 88.6(1), P1-C2-P2 90.1(1), P1-C1-C13 130.2(2), P1-
C2-C5 127.6(2), P2-C1-C13 113.8(2), P2-C2-C5 113.4(2), P1-C1-P2-C2
2.4(1).
1676
www.chemeurj.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 1674 – 1683

P=C Bonds as Building Blocks for Heterocycles
FULL PAPER
two P-tBu resonances of methylated [3a]+. These spectro-
scopic results suggested that, although the product was
likely a 1,3-diphosphetanium dication, one of the P-tBu
groups in [3a]+ had been replaced by P-Me. Indeed, the
product was confirmed by X-ray crystallography as [4a]-
AHCTUNGRTEG(NUNN OTf)2 containing a fascinating 1,3-diphosphetanium dicat-
ion (Figure 5).
Figure 4. Molecular structure of the diphosphetanium ion in [3b]I (ther-
mal ellipsoids set at 50% probability). Two virtually identical molecules
appear in the asymmetric unit. Metrical parameters are given for one of
the two molecules. The iodide anion and all hydrogen atoms except
H(2a) and H(7a) are omitted for clarity. Selected bond lengths [ꢁ] and
ꢀ
ꢀ
ꢀ
ꢀ
angles [8]: P1a C1a 1.774(9), P1a C12a 1.86(2), P1a C2a 1.778(4), P1a
ꢀ
ꢀ
ꢀ
C7a 1.762(4), P2a C2a 1.958(4), P2a C7a 1.987(5), P2a C22a 1.87(2);
C2a-P1a-C7a 96.7(2), P1a-C2a-P2a 88.7(2), C2a-P2a-C7a 84.2(2), P1a-
C7a-P2a 88.2(2), P1a-C2a-C3a 152.9(6), P1a-C7a-C8a 151.0(8), P2a-C2a-
C3a 99.0(5), P2a-C7a-C8a 101.0(9), P1a-C2a-P2a-C7a 10.9(2).
phosphetanium [P2C2R7]+ ions. Owing to the disorder pres-
ent in the structure of [3b]I, this discussion will focus on the
pertinent metrical parameters of [3a]I, which crystallizes
with two unique molecules within the asymmetric unit. The
ꢀ
P C bond lengths within the P2C2 ring of [3a]I are in the
range normally found for single bonds (1.85–1.95 ꢁ)[21], with
+
Figure 5. Molecular structure of the diphosphetanium dication in [4a]-
AHCTUNRTGEG(NNNU OTf)2 (thermal ellipsoids set at 50% probability). Two virtually identical
molecules appear in the asymmetric unit. Metrical parameters are given
for one of the two molecules. The triflate counterions, the dichlorome-
thane solvate, and all hydrogen atoms except H3 and H13 are omitted
ꢀ
ꢀ
the endocyclic P C bond involving the phosphonium (C P
: av=1.86 ꢁ) being slightly shorter than those involving the
0
ꢀ
phosphine (C P : av=1.90 ꢁ). Interestingly, the P2C2 ring
in diphosphetanium [3a]+ is nearly planar (dihedral angle
ca. 38), with angles within the ring being <908 (sum of inter-
nal angles: 359.8(4)8). In contrast, neutral diphosphetanes,
ꢀ
ꢀ
for clarity. Selected bond lengths [ꢁ] and angles [8]: P1 C1 1.793(4), P1
ꢀ
ꢀ
ꢀ
ꢀ
C2 1.791(4), P1 C3 1.846(4), P1 C13 1.840(4), P2 C3 1.859(4), P2 C8
ꢀ
ꢀ
1.781(4), P2 C9 1.856(4), P2 C13 1.857(3); C1-P1-C2 109.4(2), C3-P1-
C13 88.5(2), C3-P2-C13 87.6(2), C8-P2-C9 110.6(2), P1-C3-P2 90.1(2), P1-
C13-P2 90.4(2), P1-C3-C4 123.1(3), P2-C3-C4 127.7(3), P1-C13-C14
122.0(3), P2-C13-C14 128.3(3), P1-C3-P2-C13 13.9(2).
such as [PACHTUNGTRENNUNG(CF3)CF2]2, show puckered structures (ring dihe-
dral angle: 35.48, C-P-C angle: 77.68).[26]
Given that the alkylation of phosphetanes offers a well-
known route to phosphetanium cations,[8b] it is perhaps sur-
prising that alkylation of simple 1,3-diphosphetanes has not
been reported as a means to generate 1,3-diphosphetanium
cations. By contrast, considerable success has been achieved
with the alkylation of polycyclic tetraphosphacubanes.[27–29]
Therefore, we attempted to methylate diphosphetanium
[3a]OTf by treating it with excess MeOTf in CH2Cl2. Sever-
al ratios of MeOTf to [3a]OTf were attempted and the ratio
of 5:1 afforded cleanly a single phosphorus-containing prod-
uct, as evidenced by the 31P NMR spectrum (d=74.9,
The molecular structure of the dication [4a]2+ reveals
that the two C-tBu moieties are anti configured with respect
to the remaining P-tBu substituent. In [3a]+ the phosphine
P-tBu is anti and the phosphonium P-tBu is syn configured
to the C-tBu moieties. Presuming that the initial step in the
conversion of [3a]OTf to [4a]ACHTNUTRGNENG(U OTf)2 involves methylation of
the anti-configured phosphino P-tBu, the fact that the P-tBu
substituent is anti configured in [4a]2+ suggests that it is the
syn P-tBu(Me) moiety in [3a]+ that becomes the P-Me2
moiety in [4a]2+. The metrical parameters within the P2C2
2
55.3 ppm, JPP =16 Hz). In comparison to those for [3a]OTf
(d=61.6, 18.1 ppm, 2JPP =15 Hz), which contains a phos-
phine and phosphonium moiety, the presence of two down-
field signals suggested a dication with two phosphonium
ring of [4a]2+ are unremarkable, with the P C bond lengths
ꢀ
being similar to those in [3a]+ and [3b]+. Interestingly, the
P2C2 ring is slightly puckered. as suggested by the non-zero
dihedral angle of the ring atoms (13.98) and the sum of the
internal angles (356.6(8)).
1
moieties. Surprisingly, the H NMR spectrum exhibited reso-
nances consistent with three P-Me resonances and only one
P-tBu resonance rather than the expected two P-Me and
Chem. Eur. J. 2012, 18, 1674 – 1683
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1677

D. P. Gates and J. I. Bates
Mechanistic studies: The remarkable umpolung behavior of
phosphaalkenes when treated with different electrophiles,
combined with the unprecedented nature of products [2a]+
and [3a]+, prompted us to probe these unexpected reactions
spectroscopically. Although inversely polarized phosphaal-
kenes are known,[30] it is not evident why the identical phos-
phaalkene reacts differently with Me+ than with H+. Based
upon the nature of the products formed from the reaction of
1a with HOTf compared to those with MeOTf, we proposed
the mechanism outlined in Scheme 3. Importantly, the reac-
signal (d=205.7 ppm) along with traces of [2a]OTf (Fig-
ure 6a). The H NMR spectrum of the reaction mixture was
consistent with [5a]OTf, exhibiting signals exclusively in the
1
Figure 6. a) 31P{1H} NMR spectrum (121.5 MHz, C6D6) of [5a]OTf, b)
13C{1H} NMR spectrum (75.5 MHz, C6D6) of [5a]OTf. * indicates signal
of the C6D6 solvent.
aliphatic region that were assigned to a CH2 moiety (d=
1.55 ppm, 2H) and two CACHTNUGTRNEUGN(CH3)3 moieties (d=1.04, 9H;
Scheme 3. Proposed mechanistic pathways from 1a to diphosphiranium
[2a]OTf, 1,3-diphosphetanium [3a]OTf, and phosphirenium [6a]OTf.
0.92 ppm, 9H). Further confirmation of the assignment of
the product as [5a]OTf was obtained from the 13C{1H} NMR
spectrum (Figure 6b). Importantly, the absence of a doublet
resonance assignable to a P=C moiety (d>180 ppm) and
tion of 1a with HOTf is believed to involve the phospheni-
um triflate [5a]OTf.[31] Based on previous studies of phos-
phenium triflates, a covalent rather than an ionic structure is
expected for [5]OTF, owing to a strong interaction between
the C-substituted phosphenium and the triflate.[32] In con-
trast to protonation, the reaction of 1a with MeOTf appa-
rently involves the methylenephosphonium intermediate
[7a]OTf. It is possible that both reactions initially proceed
ꢀ
the presence of a doublet attributed to a P CH2tBu func-
tionality (d=45.8 ppm, 1JPC =37 Hz) is consistent with
1
[5a]OTf. The H and 13C{1H} spectra each were complicated
ꢀ
slightly by the presence of P CDHtBu containing [5a]OTf,
which arises from H/D exchange of HOTf with the deuteri-
um atoms of C6D6.
Although the NMR spectra of solutions of [5a]OTf in
C6D6 remain unchanged upon standing for several days, at-
tempts to isolate the phosphenium triflate by removal of the
solvent in vacuo led to unidentified products. To confirm its
role as an intermediate, phosphenium triflate [5a]OTf was
trapped with 1a (1 equiv) to quantitatively afford [2a]OTf,
as evaluated by 31P NMR spectroscopy. Interestingly,
[5a]OTf could also be trapped with 2-butyne to afford phos-
phirenium [6a]OTf. The 31P NMR spectrum of [6a]OTf ex-
hibits a single high-field resonance (d=ꢀ86.3 ppm) charac-
teristic of a phosphirenium compound.[33–36] Crystals of
[6a]OTf obtained from the reaction mixture were analyzed
by X-ray crystallography and the molecular structure is de-
picted in Figure 7.
through
a
methylenephosphonium species; however,
+
ꢀ
[tBuP(H)=CHtBu] , which contains a labile P H bond, can
rearrange to phosphenium triflate [5a]OTf, whilst [7a]OTf,
ꢀ
which contains a relatively inert P CH3 bond, is unlikely to
undergo rearrangement. Our previous studies of the reac-
tion of GaCl3 with the Mes*P=CH2 provides some precedent
for this possibility.[7] In this reaction, the methylenephospho-
nium analogue Mes*
ACHTUGNRTNE(NUNG Cl3Ga)P=CH2 forms initially, but rear-
ranges to the transient zwitterionic phosphenium triflate
ꢀ
ꢀ
Mes*P CH2ACHTUNGTRENNUNG(GaCl3), before forming the intramolecular C
H-activated product. Keeping these possibilities in mind, we
set out to spectroscopically identify the products of the 1:1
reactions of 1a with HOTf and with MeOTf.
Detection of a phosphenium triflate intermediate: In an
effort to identify a phosphenium triflate, a solution of HOTf
in C6D6 was treated with phosphaalkene 1a (1 equiv) in
C6D6. The 31P{1H} NMR spectrum of an aliquot removed
from the reaction mixture shows that the signal for 1a (d=
275.7 ppm) had been consumed and replaced by a major
Detection of a methylenephosphonium intermediate: Having
established the intermediacy of phosphenium triflate
[5a]OTf in the protonation of 1a with HOTf, the nature of
the reaction of 1a with MeOTf was next considered. A solu-
tion of MeOTf in C6D6 was treated with a solution of 1a
(1 equiv) in C6D6. Analysis of the reaction mixture by
1678
www.chemeurj.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 1674 – 1683

P=C Bonds as Building Blocks for Heterocycles
FULL PAPER
Scheme 4. Synthesis of methylenephosphonium [7c]OTf from 1c.
Figure 7. Molecular structure of the phosphirenium cation in [6a]OTf
(thermal ellipsoids set at 50% probability). The triflate atoms and all hy-
drogen atoms are omitted for clarity. Selected bond lengths [ꢁ] and
ꢀ
ꢀ
ꢀ
ꢀ
angles [8]: P1 C1 1.842(2), P1 C(5) 1.796(2), P1 C10 1.740(2), P1 C11
ꢀ
1.736(2), C10 C11 1.315(3); C10-P1-C11 44.5(1), P1-C10-C11 67.6(1), P1-
C11-C10 67.9(1).
31P NMR spectroscopy indicates the presence of unreacted
1a (d=275.7 ppm), diphosphetanium [3a]OTf, and a minor
previously unobserved product (d=177.1 ppm). In subse-
quent experiments, employing a significant excess of MeOTf
to 1a (5:1) afforded this previously unobserved species as
the major product. The 13C{1H} NMR spectrum showed
clearly a downfield doublet resonance that is characteristic
Figure 8. a) 31P{1H} NMR spectrum (121.5 MHz, CD2Cl2) of [7c]OTf, b)
13C{1H} NMR spectrum (75.5 MHz, CD2Cl2) of [7c]OTf.
* indicates
signal of the CD2Cl2 solvent, ¼ indicates signals for excess MeOTf.
1
of a P=C bond (d=153 ppm, JPC =98 Hz; cf. 1a (CDCl3):
ticular, doublet resonances assigned to the carbon atoms of
1
1
d=194.0 ppm, JPC =48.5 Hz).[18] Furthermore, a doublet res-
the P=C bond (d=176.2 ppm, JPC =97 Hz; cf. 1c (CDCl3):
onance was observed at d=7.5 ppm (JPC =35 Hz) that is
consistent with the presence of a P-Me moiety. Combined
with the 1H NMR spectral data, the aforementioned data
are indicative of a methylenephosphonium [7a]OTf. Impor-
tantly, if 1a was added to this solution, compound [3a]OTf
was detected in the 31P NMR spectrum.
d=193.4 pm, 1JPC =43.5 Hz)[41] and the H3C-P moiety (d=
2
11.9 ppm, JPC =48 Hz) in [7c]OTf were observed. Attempts
to isolate crystalline samples of [7c]OTf were unsuccessful,
with removal of the solvent resulting in an intensely colored
red oil.
Although there is considerable interest in methylenephos-
phonium cations,[37] owing to the polarity of the P=C bond,
the alkylation of phosphaalkenes was not believed to be a
viable route to them.[38,39] Therefore, it is remarkable that
[7a]OTf can be generated by the methylation of 1a. Un-
fortunately, it was not possible to isolate [7a]OTf and solu-
tions of this species were always contaminated with impuri-
ties. Given that a methylenephosphonium ion was postulat-
ed as an intermediate in the reaction of MesP=CPh2 (1c)
with MeI,[40] we attempted to generate [7c]OTf by treating
1c with MeOTf (2.5 equiv) in CD2Cl2 (Scheme 4). Interest-
ingly, the reaction proved to be quite sluggish at ambient
temperature and required heating to 808C. The
31P{1H} NMR spectrum of the reaction mixture is shown in
Figure 8a and shows that the signal for 1c (d=233 ppm)
was replaced by a singlet resonance at d=112.5 ppm. The
spectrum showed no additional signals, which suggests that
only a single product is formed. Conclusive evidence that
the product was methylenephosphonium [7c]OTf was ob-
tained from the 13C{1H} NMR spectrum (Figure 8b). In par-
Formation of 1,3-diphosphetanium: At first glance, the ar-
rangement of the tBu and/or Ad moieties in the molecular
structures of diphosphetanium compounds [3a]I and [3b]I is
quite perplexing. Remarkably, in both cases, albeit sterically
unfavorable, three of the four bulky substituents are found
on the same face of the P2C2 ring (i.e., syn configured).
Moreover, this is the only isomer observed by NMR spec-
troscopy. To account for these observations, the mechanism
of the formation of [3a]OTf from 1a and MeOTf must be
considered. In the previous section, we showed that methyl-
enephosphonium [7a]OTf is the key intermediate that
reacts with 1a. The cyclodimerization of phosphaalkenes to
diphosphetanes is well established.[42] Two plausible mecha-
nistic pathways are considered in Scheme 5. The
[p2s+p2a]cycloaddition involving a thermally allowed supra-
facial-antarafacial concerted cyclization could result in
AHCTUNGTRENNUNG
[3a]+.[47] A stepwise addition can also rationalize the ob-
served syn arrangement of tBu substituents. It is possible
that the first step involves binding the C atom of methylene-
phosphonium [7a]+ by the lone pair of phosphaalkene 1a.
Chem. Eur. J. 2012, 18, 1674 – 1683
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1679

D. P. Gates and J. I. Bates
Experimental Section
General considerations: Unless otherwise noted, all manipulations were
performed under an atmosphere of nitrogen. Hexanes, dichloromethane,
and toluene were deoxygenated with nitrogen and dried by passing
through a column containing activated alumina. C6D6 was stored over
molecular sieves (4 ꢁ) before use, and ampules of CD2Cl2 were used as
received from CIL. HOTf and MeOTf were used as received from Al-
drich. Phosphaalkenes tBuP=CHtBu (1a), AdP=CHtBu (1b), and MesP=
CPh2 (1c) were prepared following literature procedures.[19] 31P, 1H, and
13C{1H} NMR spectra were recorded at room temperature on a Bruker
300 MHz spectrometer with chemical shifts (d) in parts per million
(ppm). Chemical shifts are referenced and reported relative to 85%
H3PO4 as an external standard (d=0.0 ppm for 31P) or referenced to
TMS and measured relative to residual solvent peak (C6HD5 or CHDCl2:
d=7.16 or 5.32 ppm for 1H, respectively; C6D6 or CD2Cl2: d=128 or
53.8 ppm for 13C, respectively). Mass spectra were recorded on a Kratos
MS 50 instrument in EI mode (70 eV).
Synthesis of [2a]OTf: A solution of 1a (0.506 g, 3.20 mmol) in CH2Cl2
(3 mL) was added dropwise to a stirred solution of HOTf (0.238 g,
1.59 mmol) in CH2Cl2 (3 mL) at 258C. After the mixture was stirred for
20 min, the solvent was removed in vacuo, yielding a colorless powder.
The powder is dissolved in CH2Cl2 (1 mL) and precipitated with of hex-
anes (20 mL) twice to remove impurities. Residual solvent was removed
in vacuo. A sample of powder dissolved in CDCl3 and analyzed by
31P{1H} NMR spectroscopy was consistent with a mixture of two com-
pounds, both containing two phosphorus atoms (minor: d=ꢀ35.8,
ꢀ119.8 ppm, JPP =248 Hz; major: d=ꢀ44.2, ꢀ49.9 ppm, JPP =260 Hz) in
roughly a 2:3 ratio. Crude yield: 0.500 g (67%). Crystals of [2a]OTf suit-
able for X-ray crystallography were obtained. [2a]OTf (major isomer):
Scheme 5. Stereochemistry of stepwise and concerted mechanisms for the
reaction of phosphaalkene 1a with methylenephosphonium [7a]+ to
afford diphosphetanium [3a]+.
This key step sets the stereochemistry of the new phosphine
moiety. Importantly, the lone pair must end up on the oppo-
site face of the P=C bond in [7a]+ to which 1a adds. Subse-
quently, the phosphine moiety must rotate to permit intra-
molecular nucleophilic addition to the P=C bond of the
newly formed methylenephosphonium moiety. As a conse-
quence of this rotation, three of the four tBu substituents
end up with a syn arrangement in the product [3a]OTf.
31P{1H} NMR (121.5 MHz, CD2Cl2): d=ꢀ46.6 (d,
JPP =261 Hz),
ꢀ50.7 ppm (d, JPP =261 Hz). 1H NMR (300 MHz, CD2Cl2): d=2.83 (dd,
J
PH =4 Hz, JPH =7 Hz, 1H), 2.58 (ddd, JPH =7 Hz, JPH =11 Hz, JHH =
16 Hz, 1H), 2.28 (dd, JPH =14 Hz, JHH =16 Hz, 1H), 1.63 (dd, JPH =1 Hz,
PH =20 Hz, 9H), 1.43 (dd, JPH =3 Hz, JPH =16 Hz, 9H), 1.26 (s, 9H),
J
1.23 ppm (s, 9H); 13C{1H} NMR (75.5 MHz, CD2Cl2): d=40.3 (dd, JPC
5 Hz, PC =8 Hz), 36.9 (dd, PC =10 Hz, PC =42 Hz), 34.3 (dd, JPC
10 Hz, JPC =49 Hz), 34.0 (dd, JPC =4 Hz, JPC =12 Hz), 33.5 (dd, JPC
=
=
=
J
J
J
2 Hz, JPC =4 Hz), 32.2 (dd, JPC =10 Hz, JPC =18 Hz), 31.9 (dd, JPC =6 Hz,
PC =8 Hz), 31.2 (dd, JPC =8 Hz, JPC =17 Hz), 31.2 (dd, JPC =5 Hz, JPC
J
=
Conclusion
6 Hz), 30.0 ppm (dd, JPC =1 Hz, JPC =2 Hz); [2a]OTf (minor isomer):
31P{1H} NMR (121.5 MHz, CDCl3): d=ꢀ35.8 (d,
JPP =248 Hz),
ꢀ119.8 ppm (d, JPP =248 Hz); 1H NMR (300 MHz, CD2Cl2): d=3.19 (dd,
In summary, the reaction of alkyl-substituted phosphaal-
kenes with electrophiles gives surprisingly different products
depending on the nature of the electrophile. The protic elec-
trophile HOTf adds to the carbon atom of the P=C bond to
form a phosphenium triflate, which then reacts with a
second phosphaalkene to give a diphosphiranium salt. In
stark contrast, the analogous reaction with the electrophilic
carbenium of MeOTf affords a methylenephosphonium in-
termediate, which subsequently cyclizes with a second phos-
phaalkene to give a 1,3-diphosphetanium salt. This 1,3-di-
phosphetanium cation can be further alkylated with MeOTf
to afford the first structurally characterized diphosphetani-
um dication. The P2C and P2C2 ring systems reported herein
are unprecedented examples of highly strained phosphorus
heterocycles that have no isolable counterparts in nitrogen
chemistry. Owing to the unexpected umpolung reactivity of
the P=C bond with different electrophiles, studies were un-
dertaken to spectroscopically characterize the intermediates
in these novel transformations. The present work further
demonstrates the utility of phosphaalkenes, compounds with
a genuine (3p-2p)p bond, as building blocks for a plethora
of novel organophosphorus structures.
J
PH =2 Hz, JPH =10 Hz, 1H), 3.17 (dd, JPH =7 Hz, JHH =16 Hz, 1H), 2.23
(dd, JHH =16 Hz, JPH =17 Hz, 1H), 1.55 (dd, JPH =2 Hz, JPH =22 Hz, 9H),
1.31 (d, JPH =1 Hz, 9H), 1.30 (dd, JPH =1 Hz, JPH =16 Hz, 9H), 1.24 ppm
(s, 9H).
Synthesis of [2b]OTf: A solution of 1b (0.236 g, 1.0 mmol) in CH2Cl2
(2 mL) was added dropwise to a stirred solution of HOTf (0.075 g,
0.50 mmol) in CH2Cl2 (2 mL) at 258C. After the mixture was stirred for
1 h, an aliquot was removed and the 31P NMR spectroscopy was consis-
tent with [2b]OTf. Crude yield: 0.275 g (88%). Crystals of [2b][HACHTUNGTRENNUNG(OTf)2]
suitable for X-ray crystallography were obtained from a mixture of aceto-
nitrile, CH2Cl2 and hexanes. 31P{1H} NMR (121.5 MHz, CH2Cl2): minor
isomer (40%): d=ꢀ36.3, ꢀ127.1 ppm, JPP =250 Hz; major isomer (60%):
d=ꢀ48.2, ꢀ54.1 ppm, JPP =256 Hz.
Synthesis of [3a]OTf: A solution of 1a (0.200 g, 1.27 mmol) in CH2Cl2
(0.5 mL) was added dropwise to a stirred solution of MeOTf (0.104 g,
0.63 mmol) in CH2Cl2 (0.5 mL) at 258C. After the mixture was stirred for
1 h, the volatiles were removed in vacuo, yielding a colorless powder.
The solid was washed with hexanes (5 mL) and filtered dry. Yield:
2
0.210 g (69%). 31P{1H} NMR (121.5 MHz, CD2Cl2): d=61.6 (d, JPP
=
2
15 Hz), 18.1 ppm (d, JPP =15 Hz).
Synthesis of [3a]I: A solution of [3a]OTf (0.100 g, 0.21 mmol) in CH2Cl2
(10 mL) was treated with a saturated aqueous solution of KI (5ꢂ10 ml).
The organic fractions were collected and dried using MgSO4 and filtered
to remove insoluble material. The solvent was removed in vacuo to yield
a pale yellow powder. The product was recrystallized by slow evaporation
1680
www.chemeurj.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 1674 – 1683

P=C Bonds as Building Blocks for Heterocycles
FULL PAPER
of a concentrated solution in CH2Cl2 to yield crystals suitable for X-ray
diffraction; Yield: 0.035 g (38%); 31P{1H} NMR (121.5 MHz, CD2Cl2):
d=61.8 (d, 2JPP =15 Hz), 20.9 ppm (d, 2JPP =15 Hz); 1H NMR (CD2Cl2):
analysis calcd (%) C14H26PSO3F3: C 46.40, H 7.23; found: C 46.45, H
7.26.
Synthesis of [7a]OTf: A solution of 1a (0.100 g, 0.63 mmol) in C6D6
(1 mL) was added dropwise to a stirred solution of MeOTf (0.520 g,
3.15 mmol) in C6D6 (2 mL) at 258C. After the mixture was stirred for
30 min, an aliquot (0.5 mL) was removed and analyzed by NMR spec-
troscopy. 31P{1H} NMR (121.5 MHz, C6D6): d=177.1 ppm; 1H NMR
(300 MHz, C6D6): d=7.21 (d, JPH =11 Hz, 1H), 2.37 (d, JPH =23 Hz, 3H),
1.28 (d, JPH =20 Hz, 9H), 1.15 ppm (d, JPH =1 Hz, 9H); 13C{1H} NMR
(75.5 MHz, C6D6): d=152.7 (d, JPC =98 Hz), 122.1 (q, JFC =322 Hz), 40.0
d=3.51 (dd, JPH =1 Hz, JPH =19 Hz), 2.56 (d, JPH =11 Hz), 1.60 (d, JPH
=
17 Hz), 1.35 (d,
JPH =14 Hz), 1.26 ppm (dd, JPH =1 Hz, JPH =1 Hz);
13C{1H} NMR (75.5 MHz, CD2Cl2): d=49.6 (dd, JPC =23 Hz, JPC =25 Hz),
39.7 (dd, JPC =3 Hz, JPC =7 Hz), 35.5 (dd, JPC =4 Hz, JPC =21 Hz), 32.2
(dd, JPC =5 Hz, JPC =36 Hz), 31.2 (dd, JPC =7 Hz, JPC =9 Hz), 30.8 (dd,
J
J
PC =1 Hz, JPC =16 Hz), 28.6 (dd, JPC =1 Hz, JPC =4 Hz), 15.1 ppm (d,
PC =45 Hz); elemental analysis calcd (%) for C19H41P2I: C 49.78, H 9.02;
found: C 50.16, H 9.27.
(d, JPC =39 Hz), 39.3 (d, JPC =13 Hz), 30.5 (d, JPC =16 Hz), 28.0 (d, JPC
2 Hz), 7.5 ppm (d, JPC =35 Hz).
=
Synthesis of [3b]OTf: A solution of 1b (0.236 g, 1.00 mmol) in CH2Cl2
(0.5 mL) was added dropwise to a stirred solution of MeOTf (0.85 g,
0.52 mmol) in CH2Cl2 (0.5 mL) at 258C. After the mixture was stirred for
1 h, the volatiles were removed in vacuo, yielding a colorless powder.
Synthesis of [7c]OTf: A solution of 1c (0.100 g, 0.32 mmol) in CD2Cl2
(0.5 mL) was added to a solution of MeOTf (0.130 g, 0.79 mmol) in
CD2Cl2 (0.5 mL) in a bomb equipped with a magnetic stir bar. The mix-
ture was heated with stirring at 808C (1 h). Upon cooling, the mixture
was analyzed by NMR spectroscopy. 31P{1H} NMR (121.5 MHz, C6D6):
The solid was washed with hexanes (5 mL) and filtered dry. Yield:
2
0.280 g (88%); 31P{1H} NMR (121.5 MHz, CDCl3): d=58.4 (d, JPP
=
2
d=112.5 ppm; 1H NMR (C6D6): d=7.69–7.04 (m, 12H), 2.86 (d, JPH
26 Hz, 3H), 2.56 (d, JPH =3 Hz, 6H), 2.36 ppm (d, JPH =2 Hz, 3H);
13C{1H} NMR (75.5 MHz, C6D6): d=176.2 (d, JPC =97 Hz), 148.0 (d, JPC
=
15 Hz), 20.6 ppm (d, JPP =15 Hz).
Synthesis of [3b]I: A solution of [3b]OTf (0.200 g, 0.31 mmol) in CH2Cl2
(10 mL) was treated with a saturated aqueous solution of KI (5ꢂ10 ml).
The organic fractions were collected and dried using MgSO4 and filtered
to remove insoluble material. The solvent was removed in vacuo to yield
a pale yellow powder. The product was recrystallized by slow evaporation
of a concentrated CH2Cl2 solution to yield crystals suitable for X-ray dif-
fraction. Yield: 0.110 g (58%); 31P{1H} NMR (121.5 MHz, CDCl3): d=
58.7 (d, 2JPP =16 Hz), 21.0 ppm (d, 2JPP =16 Hz); 1H NMR (300 MHz,
CDCl3): d=3.49 (d, JPH =18 Hz, 2H), 2.64 (d, JPH =10 Hz), 2.24 (m, 6H),
2.21 (brs, 3H), 2.10 (brs, 3H), 1.90 (m, 6H), 1.79 (m, 6H), 1.75 (m, 6H),
=
4 Hz), 144.5 (d, JPC =6 Hz), 136.9 (d, JPC =3 Hz), 135.7 (d, JPC =2 Hz),
133.4 (d, JPC =6 Hz), 133.2 (d, JPC =5 Hz), 131.4, 131.2, 130.9, 130.8 (2C),
130.6, 130.4, 130.2 (2C), 129.8, 129.7, 121.5 (q, JFC =321 Hz), 117.4 (d,
J
PC =85 Hz), 23.3 (d, JPC =6 Hz), 21.8 (d, JPC =1 Hz), 11.9 ppm (d, JPC
=
48 Hz).
X-ray crystallography: All single crystals were immersed in oil and
mounted on a glass fiber. Data were collected at 173ꢃ0.1 K on a Bruker
X8 APEX 2 diffractometer with graphite-monochromated MoKa radia-
tion. Data was collected and integrated using the Bruker SAINT[43] soft-
ware package and corrected for absorption effects using SADABS.[44] All
structures were solved by direct methods[45] and subsequent Fourier dif-
ference techniques and refined anisotropically for all non-hydrogen
atoms using the SHELXTL[46] crystallographic software package from
Bruker AXS. All data sets were corrected for Lorentz and polarization
effects. The crystals of [2a]OTf, [3a]I, and [6a]OTf presented no crystal-
1.26 ppm (s, 18H); 13C{1H} NMR (75.5 MHz, CDCl3): d=47.9 (dd, JPC
=
23 Hz, JPC =23 Hz), 45.4 (dd, JPC =3 Hz, JPC =3 Hz), 41.6 (d, JPC =13 Hz),
36.8, 36.2, 35.7, 35.2 (dd, JPC =4 Hz, JPC =21 Hz), 34.9 (dd, JPC =4 Hz,
J
J
PC =37 Hz), 31.5 (dd, JPC =8 Hz, JPC =8 Hz), 28.9 (d, JPC =9 Hz), 28.2 (d,
PC =10 Hz), 13.0 ppm (d, JPC =46 Hz).
Synthesis of [4a]ACHTUNGTRENNUNG(OTf)2: A solution of [3a]OTf (0.350 g, 0.73 mmol) in
CH2Cl2 (2 mL) was added dropwise to a stirred solution of MeOTf
(0.600 g, 3.66 mmol) in CH2Cl2 (2 mL) at 258C. The mixture was stirred
overnight and then placed in a freezer (ꢀ308C) to afford colorless crys-
lographic complications. The structure of [3b]I and [4a]ACHTNUTRGNEG(UN OTf)2·CH2Cl2
exhibited some disorder. For [3b]I there are two overlapping cations,
each contributing 50% to the total electron density attributed to the
ring. The two cations relate through a pseudo-mirror plane that passes
through the cyclic carbon atoms at about 608 to the mean plane of the
ring, which results in the phosphino-Ad from one moiety to overlap with
tals suitable for X-ray diffraction. Yield: 0.260 g (61%); 31P{1H} NMR
2
(121.5 MHz, CD3CN): d=74.9 (d, 2JPP =16 Hz), 55.3 ppm (d, JPP
=
16 Hz); 1H NMR (300 MHz, CD3CN): d=4.49 (dd, JPH =19 Hz, JPH
=
22 Hz, 2H), 2.81 (d, JPH =14 Hz, 3H), 2.68 (d, JPH =13 Hz, 3H), 2.56 (d,
PH =14 Hz, 3H), 1.56 (d, PH =20 Hz, 9H), 1.31 ppm (s, 18H);
13C{1H} NMR (75.5 MHz, CD3CN): d=122.0 (q, JFC =319 Hz), 44.5 (dd,
PC =29 Hz, JPC =38 Hz), 37.9 (dd, JPC =3 Hz, JPC =4 Hz), 37.8 (dd, JPC
10 Hz, JPC =24 Hz), 31.3 (dd, JPC =5 Hz, JPC =6 Hz), 26.6 (d, JPC =2 Hz),
the phosphonium-Ad of the second and vice versa. In [4a]ACHTNUTRGNEG(UN OTf)2·CH2Cl2
J
J
one of the four triflate anions is disordered. There may also be disorder
in the methyl groups at C9; however, this was not modeled. The crystal
of [2b]OTf was a racemic twin with each component crystallizing as a
single enantiomer. The twin components were present in a 3:2 ratio,
where the minor component is related to the major component by a 1808
rotation about the (1 0 0) direct axis. That the ratio of the two compo-
nents is not 1:1 presumably reflects the composition of the selected crys-
tal and does not reflect the composition of the bulk sample, which is ex-
pected to be racemic.
J
=
16.7 (dd, JPC =9 Hz, JPC =39 Hz), 10.5 (d, JPC =37 Hz), 5.6 ppm (d, JPC
=
30 Hz).
Synthesis of [5a]OTf: A solution of 1a (0.100 g, 0.63 mmol) in C6D6
(1 mL) was added dropwise to stirred solution of HOTf (0.95 g,
a
0.63 mmol) in C6D6 (1 mL) at 258C. After the mixture was stirred for
15 min, an aliquot (0.5 mL) was removed and analyzed by NMR spec-
troscopy. 31P{1H} NMR (121.5 MHz, C6D6): d=205.7 ppm. 1H NMR
Crystal data for [2a]OTf: formula C19H39F3O3P2S, Mr =466.50 gmolꢀ1
,
monoclinic, space group P21/n, a=8.608(1), b=21.465(1), c=
(C6D6): d=1.55 (bs, 2H), 1.04 (d, JPH =1 Hz, 9H), 0.92 ppm (d, JPH
=
13.806(1) ꢁ, a=90, b=102.656(2), g=908, V=2489.1(2) ꢁ3, Z=4,
14 Hz, 9H); 13C{1H} NMR (75.5 MHz, C6D6): d=119.6 (q, JFC =318 Hz),
45.8 (d, JPC =37 Hz), 34.0 (d, JPC =24 Hz), 30.9 (d, JPC =8 Hz), 30.7 (d,
1calcd =1.245 gcmꢀ3
,
m=0.297 mmꢀ1
,
dimensions 0.5ꢂ0.35ꢂ0.1 mm3,
2qmax =55.68, total reflections collected=31780, independent reflec-
tions=5915 (Rint =0.0349), 277 parameters were refined and converged
J
PC =16 Hz), 24.5 ppm (d, JPC =17 Hz).
Synthesis of [6a]OTf: A solution of [5a]OTf was prepared by dropwise
addition of solution of phosphaalkene 1a (0.158 g, 1.00 mmol) in
at R1
ACHUTGTNRENNUG[I>2s(I)]=0.0338, wR2ACHTUNGTREN[NUNG all data]=0.0897.
a
Crystal data for [2b][HACTHNUGTENR(NUG OTf)2]: formula C32H52F6O6P2S2, Mr =
CH2Cl2 (2 mL) to a stirred solution of HOTf (0.150 g, 1.00 mmol) in
CH2Cl2 (2 mL) at 258C. After 1 h, a solution of 2-butyne (0.100 g,
1.85 mmol) in CH2Cl2 (2 mL) was added to the mixture. Slow evapora-
tion of the solvent afforded colorless crystals suitable for X-ray crystal-
lography. Yield: 0.250 g (69%); 31P{1H} NMR (121.5 MHz, CD2Cl2): d=
ꢀ86.3 ppm; 1H NMR (300 MHz, CD2Cl2): d=2.61 (d, JPH =10 Hz, 2H),
772.80 gmolꢀ1, monoclinic, space group P21, a=10.600(1), b=17.169(1),
c=11.212(1) ꢁ, a=90, b=118.210(2), g=908, V=1798.0(1) ꢁ3, Z=2,
1calcd =1.427 gcmꢀ3
, ,
m=0.310 mmꢀ1 dimensions 0.35ꢂ0.25ꢂ0.25 mm3,
2qmax =55.88, total reflections collected=29398, independent reflec-
tions=7760 (Rint =0.0194), 439 parameters were refined and converged
at R1ACTHNUGRTNENUG[I>2s(I)]=0.0234, wR2ACHTUGNTREN[NNGU all data]=0.0586.
Crystal data for [3a]I: formula C19H41IP2, Mr =458.36 gmolꢀ1, ortho-
rhombic, space group Pca21, a=19.631(1), b=9.746(1), c=23.632(1) ꢁ,
a=90, b=90, g=908, V=4521.5(3) ꢁ3, Z=8, 1calcd =1.347 gcmꢀ3, m=
2.45 (d, JPH =14 Hz, 6H), 1.31 (d, JPH =21 Hz, 9H), 1.01 ppm (d, JPH
=
1 Hz, 9H); 13C{1H} NMR (CD2Cl2): d=131.5 (d, 1JPC =3 Hz), 121.4 (q,
1JFC =321 Hz), 36.1 (d, 1JPC =38 Hz), 31.4 (d, 2JPC =7 Hz), 31.2 (d, JPC
=
1
25 Hz), 30.5 (d, 2JPC =8 Hz), 26.8, 11.5 ppm (d, 2JPC =3 Hz); elemental
Chem. Eur. J. 2012, 18, 1674 – 1683
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1681

D. P. Gates and J. I. Bates
1.556 mmꢀ1, dimensions 0.5ꢂ0.5ꢂ0.2 mm3, 2qmax =55.68, total reflections
collected=31854, independent reflections=10524 (Rint =0.0282), 439 pa-
2006, 118, 7429; Angew. Chem. Int. Ed. 2006, 45, 7271; h) C. W.
Tsang, B. Baharloo, D. Riendl, M. Yam, D. P. Gates, Angew. Chem.
2004, 116, 5800; Angew. Chem. Int. Ed. 2004, 43, 5682; i) C. W.
Tsang, M. Yam, D. P. Gates, J. Am. Chem. Soc. 2003, 125, 1480.
[6] J. I. Bates, D. P. Gates, J. Am. Chem. Soc. 2006, 128, 15998.
[7] C. W. Tsang, C. A. Rohrick, T. S. Saini, B. O. Patrick, D. P. Gates,
Organometallics 2004, 23, 5913.
[8] For reviews, see for example: a) Phosphorus–Carbon Heterocyclic
Chemistry: The Rise of a New Domain (Ed.: F. Mathey), Pergamon,
Amsterdam, 2001; b) C. A. Dyker, N. Burford, Chem. Asian J. 2008,
3, 28; c) D. Bourissou, G. Bertrand, Top. Curr. Chem. 2002, 220, 1;
d) R. Streubel, Coord. Chem. Rev. 2002, 227, 175; e) D. Bourissou,
G. Bertrand, Acc. Chem. Res. 1999, 32, 561; f) F. Mathey, Chem.
Rev. 1990, 90, 997.
rameters were refined and converged at R1
data]=0.0507.
ACHTUGTNREN[UNG I>2s(I)]=0.0235, wR2ACHTUGNTREN[NUGN all
Crystal data for [3b]I: formula C31H53IP2, Mr =466.50 gmolꢀ1, monoclin-
ic, space group P21/n, a=13.232(1), b=12.130(1), c=18.335(1) ꢁ, a=90,
b=90.057(2), g=908, V=2942.8(2) ꢁ3, Z=4, 1calcd =1.387 gcmꢀ3
, m=
1.215 mmꢀ1, dimensions 0.8ꢂ0.2ꢂ0.1 mm3, 2qmax =56.08, total reflections
collected=28106, independent reflections=6809 (Rint =0.0250), 618 pa-
rameters were refined and converged at R1
data]=0.1037.
ACHTUGTNREN[UNG I>2s(I)]=0.0471, wR2ACHTUGNTREN[NUGN all
Crystal data for [4a]ACHTUNGTRENNUNG(OTf)2·CH2Cl2: formula C39H78Cl2F12O12P4S4, Mr =
1290.09 gmolꢀ1, monoclinic, space group Cc, a=20.876(3), b=18.823(2),
c=16.796(2) ꢁ, a=90, b=109.847(2), g=908, V=6208(2) ꢁ3, Z=4,
1calcd =1.380 gcmꢀ3
,
m=0.427 mmꢀ1
,
dimensions 0.3ꢂ0.25ꢂ0.25 mm3,
[9] T. Kato, H. Gornitzka, A. Baceiredo, W. W. Schoeller, G. Bertrand,
Science 2000, 289, 754.
[10] T. Kato, H. Gornitzka, A. Baceiredo, W. W. Schoeller, G. Bertrand,
J. Am. Chem. Soc. 2002, 124, 2506.
2qmax =55.68, total reflections collected=29244, independent reflec-
tions=13644 (Rint =0.0366), 719 parameters were refined and converged
at R1
ACHTUNGTRENNUNG[I>2s(I)]=0.0529, wR2ACHTUNGTREN[NUNG all data]=0.1243.
Crystal data for [6a]OTf: formula C14H26F3O3PS, Mr =362.38 gmolꢀ1
,
[11] S. Loss, C. Widauer, H. Rꢃegger, U. Fleischer, C. M. Marchand, H.
Grꢃtzmacher, G. Frenking, Dalton Trans. 2003, 85.
[12] D. Martin, A. Baceiredo, G. Bertrand, Eur. J. Inorg. Chem. 2004,
3533.
monoclinic, space group P21/n, a=8.752(1), b=16.218(1), c=
12.982(1) ꢁ, a=90, b=92.657(3), g=908, V=1840.62(2) ꢁ3, Z=4,
1calcd =1.308 gcmꢀ3
,
m=0.298 mmꢀ1
,
dimensions 0.3ꢂ0.2ꢂ0.2 mm3,
[13] Diaziridium cations (F) have not been isolated, but have been
postulated as reaction intermediates; see: E. Schmitz, Angew. Chem.
1964, 76, 197; Angew. Chem. Int. Ed. Engl. 1964, 3, 333; I. Washing-
ton, K. N. Houk, A. Armstrong, J. Org. Chem. 2003, 68, 6497.
[14] See, for example: a) S. Goumri, Y. Leriche, H. Gornitzka, A. Baceir-
edo, G. Bertrand, Eur. J. Inorg. Chem. 1998, 1539; b) L. Weber,
M. H. Scheffer, H. G. Stammler, B. Neumann, W. W. Schoeller, A.
Sundermann, Organometallics 1999, 18, 4216; c) L. Weber, S. Uth-
mann, H. G. Stammler, B. Neumann, W. W. Schoeller, R. Boese, D.
Blaser, Eur. J. Inorg. Chem. 1999, 2369; d) N. Merceron, A. Baceire-
do, H. Gornitzka, G. Bertrand, Chem. Commun. 2002, 2250; e) V. P.
Morgalyuk, P. V. Petrovskii, K. A. Lysenko, E. E. Nifant’ev, Russ. J.
Gen. Chem. 2010, 80, 100.
2qmax =56.08, total reflections collected=32301, independent reflec-
tions=4449 (Rint =0.0595), 303 parameters were refined and converged
at R1
CCDC-637772 ([2a]OTf), CCDC-825738 ([2b][H
([3a]I), CCDC-825739 ([3b]I), CCDC-825740 ([4a]AHCTUNGTRENNUNG
ACHTUNGTRENNUNG[I>2s(I)]=0.0417, wR2ACHTUNGTREN[NUNG all data]=0.1052.
AHCTUNGTRENNUNG
CCDC-825741 ([6a]OTf) 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.
[15] H. Helten, M. Engeser, D. Gudat, R. Schilling, G. Schnakenburg, M.
Nieger, R. Streubel, Chem. Eur. J. 2009, 15, 2602.
Acknowledgements
[16] S. Fankel, H. Helten, G. von Frantzius, G. Schnakenburg, J. Daniels,
V. Chu, C. Muller, R. Streubel, Dalton Trans. 2010, 39, 3472.
[17] A. Espinosa, R. Streubel, Chem. Eur. J. 2011, 17, 3166.
[18] E. Niecke, E. Symalla, Chimia 1985, 39, 320.
[19] J. I. Bates, B. O. Patrick, D. P. Gates, New J. Chem. 2010, 34, 1660.
[20] An Al2P3C3 cage compound which formally contains the type E
moiety has been reported, however, the ring carbon possesses a
zwitterionic Al substituent; see: B. Breit, A. Hoffmann, U. Berg-
strꢄsser, L. Ricard, F. Mathey, M. Regitz, Angew. Chem. 1994, 106,
1541; Angew. Chem. Int. Ed. Engl. 1994, 33, 1491.
[21] CRC Handbook of Chemistry and Physics, 84th ed. (Ed.: D. R.
Lide), CRC, Boca Raton, 2003, Section 9.
[22] N. Burford, C. A. Dyker, M. Lumsden, A. Decken, Angew. Chem.
2005, 117, 6352; Angew. Chem. Int. Ed. 2005, 44, 6196.
[23] Although a simple 1,3-diphosphetanium is not known, the alkylation
of tetraphosphacubane affords cations with a P-methylated P2C2
moiety; see: K. K. Laali, B. Geissler, M. Regitz, J. J. Houser, J. Org.
Chem. 1995, 60, 47.
[24] Reaction of tetraphosphacubane with “GaI” in presence of trace
water affords polycyclic compounds P-protonated P2C2 moieties;
see: R. J. Baker, H. Bettentrup, C. Jones, Acta. Crystallogr. Sect. C
2003, 59, m339.
We thank the Natural Sciences and Engineering Research Council
(NSERC) of Canada, the Canada Foundation for Innovation (CFI) and
the British Columbia Knowledge Development Fund (BCKDF) for sup-
port of this work. J.I.B. thanks NSERC for PGS M and D scholarships.
We thank Dr. Brian O. Patrick (UBC) for assistance in solving some of
the X-ray crystal structures and Prof. Jennifer Love for useful discus-
sions.
[1] For general reviews highlighting the field of inorganic multiple
bonds, see: a) R. C. Fischer, P. P. Power, Chem. Rev. 2010, 110, 3877;
b) P. P. Power, Chem. Rev. 1999, 99, 3463.
[2] See, for example: a) F. Mathey, Angew. Chem. 2003, 115, 1616;
Angew. Chem. Int. Ed. 2003, 42, 1578; b) K. B. Dillon, F. Mathey,
J. F. Nixon, Phosphorus: The Carbon Copy: From Organophospho-
rus to Phospha-Organic Chemistry, Wiley, New York, 1998; c) F.
Mathey, Acc. Chem. Res. 1992, 25, 90; d) R. Appel in Multiple
Bonds and Low Coordination in Phosphorus Chemistry (Eds.: M.
Regitz, O. J. Scherer), Thieme, Stuttgart, 1990.
[3] J. I. Bates, J. Dugal-Tessier, D. P. Gates, Dalton Trans. 2010, 39,
3151.
[4] D. P. Gates, Top. Curr. Chem. 2005, 250, 107.
[25] Interesting unsaturated ylide-substituted cationic P2C2 rings have
been reported; see: H.-P. Schrçdel, A. Schmidpeter, Chem. Ber.
1997, 130, 1519.
[26] B. Steger, H. Oberhammer, J. Grobe, D. Levan, Inorg. Chem. 1986,
25, 3177.
[27] L. Heuer, L. Ernst, R. Schmutzler, D. Schomburg, Angew. Chem.
1989, 101, 1549; Angew. Chem. Int. Ed. Engl. 1989, 28, 1507.
[28] S. D. Riegel, N. Burford, M. D. Lumsden, A. Decken, Chem.
Commun. 2007, 4668.
[5] a) K. J. T. Noonan, B. H. Gillon, V. Cappello, D. P. Gates, J. Am.
Chem. Soc. 2008, 130, 12876; b) B. H. Gillon, B. O. Patrick, D. P.
Gates, Chem. Commun. 2008, 2161; c) K. J. T. Noonan, D. P. Gates,
Macromolecules 2008, 41, 1961; d) K. J. T. Noonan, B. Feldscher, J. I.
Bates, J. J. Kingsley, M. Yam, D. P. Gates, Dalton Trans. 2008, 4451;
e) K. J. T. Noonan, B. O. Patrick, D. P. Gates, Chem. Commun. 2007,
3658; f) B. H. Gillon, K. J. T. Noonan, B. Feldscher, J. M. Wissenz,
Z. M. Kam, T. Hsieh, J. J. Kingsley, J. I. Bates, D. P. Gates, Can. J.
Chem. 2007, 85, 1045; g) K. J. T. Noonan, D. P. Gates, Angew. Chem.
1682
www.chemeurj.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 1674 – 1683

P=C Bonds as Building Blocks for Heterocycles
FULL PAPER
[29] J. J. Weigand, N. Burford, R. J. Davidson, S. Cameron, P. Seelheim,
J. Am. Chem. Soc. 2009, 131, 17943.
Moores, L. Ricard, P. Le Floch, Angew. Chem. 2003, 115, 5090;
Angew. Chem. Int. Ed. 2003, 42, 4940.
[30] L. Weber, Eur. J. Inorg. Chem. 2000, 2425.
[31] The formation of a phosphenium ion from a phosphaalkene has pre-
[40] T. A. Van der knap, F. Bickelhaupt, Tetrahedron Lett. 1982, 23, 2037.
[41] T. C. Klebach, R. Lourens, F. Bickelhaupt, J. Am. Chem. Soc. 1978,
100, 4886.
viously been observed in the reaction of Cp*P=CACTHNUGTRNEUNG(SiMe3)2 with
HOTf; see: D. Gudat, E. Niecke, B. Krebs, M. Dartmann, Chimia
1985, 39, 277.
[42] a) M. Yam, J. H. Chong, C. W. Tsang, B. O. Patrick, A. E. Lam, D. P.
Gates, Inorg. Chem. 2006, 45, 5225; b) X.-G. Zhou, L.-B. Zhang, R.-
F. Cai, Q.-J. Wu, L.-H. Weng, Z.-E. Huang, J. Organomet. Chem.
2000, 604, 260; c) M. Fild, P. G. Jones, K. Ruhnau, C. Thone, Z. Na-
turforsch. B: J. Chem. Sci. 1994, 49, 1361; d) L. N. Markovski, V. D.
Romanenko, Tetrahedron 1989, 45, 6019; e) J. Grobe, D. Le Van,
Angew. Chem. 1984, 96, 716; Angew. Chem. Int. Ed. Engl. 1984, 23,
710; f) G. Becker, W. Becker, O. Mundt, Phosphorus Sulfur Relat.
Elem. 1983, 14, 267; g) G. Becker, W. Becker, G. Uhl, W. Uhl, H.-J.
Wessely, Phosphorus Sulfur Relat. Elem. 1983, 18, 7; h) R. Appel, F.
Knoll, I. Ruppert, Angew. Chem. 1981, 93, 771; Angew. Chem. Int.
Ed. Engl. 1981, 20, 731; i) G. Becker, W. Uhl, H. J. Wessely, Z.
Anorg. Allg. Chem. 1981, 479, 41.
[32] For examples describing the covalency in phosphenium trilfates, see:
a) K. K. Laali, B. Geissler, O. Wagner, J. Hoffmann, R. Armbrust,
W. Eisfeld, M. Regitz, J. Am. Chem. Soc. 1994, 116, 9407; b) K. K.
Laali, B. Geissler, M. Regitz, J. J. Houser, J. Org. Chem. 1995, 60,
6362; c) V. A. Jones, S. Sriprang, M. Thornton-Pett, T. P. Kee, J. Or-
ganomet. Chem. 1998, 567, 199; d) D. Gudat, A. Haghverdi, H.
Hupfer, M. Nieger, Chem. Eur. J. 2000, 6, 3414; e) M. B. Abrams,
B. L. Scott, R. T. Baker, Organometallics 2000, 19, 4944; f) S. L.
Renard, T. P. Kee, J. Organomet. Chem. 2002, 643, 516.
[33] D. C. R. Hockless, M. A. Mcdonald, M. Pabel, S. B. Wild, J. Chem.
Soc. Chem. Commun. 1995, 257.
[34] D. C. R. Hockless, M. A. McDonald, M. Pabel, S. B. Wild, J. Orga-
nomet. Chem. 1997, 529, 189.
[43] G. M. Sheldrick, SAINT, Version 7.03A, Bruker AXS Inc., Madison,
1997.
[35] N. E. Brasch, I. G. Hamilton, E. H. Krenske, S. B. Wild, Organome-
tallics 2004, 23, 299.
[44] G. M. Sheldrick, SADABS, Version 2.03, Bruker AXS Inc., Madi-
son, 2001.
[36] K. Lammertsma, Top. Curr. Chem. 2003, 229, 95.
[37] The preparation of isolable methylenephosphonium cations is well
established; see: a) A. Igau, A. Baceiredo, H. Grꢃtzmacher, H.
Pritzkow, G. Bertrand, J. Am. Chem. Soc. 1989, 111, 6853; b) H.
Grꢃtzmacher, H. Pritzkow, Angew. Chem. 1992, 104, 92; Angew.
Chem. Int. Ed. Engl. 1992, 31, 99; c) H. Grꢃtzmacher, H. Pritzkow,
Angew. Chem. 1991, 103, 721; Angew. Chem. Int. Ed. Engl. 1991, 30,
709.
[45] A. Altomare, M. C. Burla, M. Camalli, G. L. Cascarano, C. Giaco-
vazzo, A. Guagliardi, A. G. G. Moliterni, G. Polidori, R. Spagna, J.
Appl. Crystallogr. 1999, 32, 115.
[46] G. M. Sheldrick, SHELXTL, Version 5.1, Bruker AXS Inc., Madi-
sion, 1997.
[47] R. B. Woodward, R. Hoffmann, Angew. Chem. 1969, 81, 797;
Angew. Chem. Int. Ed. 1969, 8, 781.
[38] O. Guerret, G. Bertrand, Acc. Chem. Res. 1997, 30, 486.
Received: May 19, 2011
Revised: September 16, 2011
Published online: January 2, 2012
+
ꢀ
[39] The preparation of 1-methylphosphinium ion (C5R5P Me) cannot
be performed by simple methylation of a phosphinine; see: A.
Chem. Eur. J. 2012, 18, 1674 – 1683
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1683