Destiny of transient phosphenium ions generated from the addition of electrophiles to phosphaalkenes: Intramolecular C-H activation, donor-acceptor formation, and linear oligomerization

Article abstract of DOI:10.1021/om0495022

The reaction of electrophiles and phosphaalkenes and the cationic polymerization of phosphaalkenes were investigated. It was demonstrated that Lewis or protic acids can add across the P double bond C, generating highly reactive phosphenium ions which was

Full text of DOI:10.1021/om0495022

Organometallics 2004, 23, 5913-5923
5913
Articles
Destiny of Transient Phosphenium Ions Generated from
the Addition of Electrophiles to Phosphaalkenes:
Intramolecular C-H Activation, Donor-Acceptor
Formation, and Linear Oligomerization
Chi-Wing Tsang, Crystal A. Rohrick, Tejindra S. Saini, Brian O. Patrick, and
Derek P. Gates*
Department of Chemistry, University of British Columbia, 2036 Main Mall,
Vancouver, British Columbia, Canada V6T 1Z1
Received July 5, 2004
The reaction of the phosphaalkenes MesPdCPh2 (Mes ) 2,4,6-Me3C6H2) and Mes*PdCH2
(Mes* ) 2,4,6-tBu3C6H2) with Lewis (AlCl3, GaCl3, InCl3) and protic (HOTf) acids has been
examined to evaluate the feasibility of cationic polymerization for PdC bonds. Addition of
GaCl3 to Mes*PdCH2 generates the adduct Mes*(Cl3Ga)PdCH2, which can be detected
spectroscopically at 193 K. At higher temperatures, GaCl3 migrates from phosphorus to
carbon to afford the fleeting phosphenium zwitterion Mes*PCH2GaCl3. This undetected
transient species immediately oxidatively adds to a C-H bond of an o-tBu group in the P-Mes*
substituent, resulting in a GaCl3-coordinated ylide that has been characterized crystallo-
graphically. The analogous reaction of GaCl3 with MesPdCPh2 gives stable Mes(Cl3Ga)Pd
CPh2, for which a crystal structure determination has been conducted. Significantly, treating
a highly concentrated solution of Mes*PdCH2 with substoichiometric quantities of GaCl3
leads to linear dimerization following a cationic chain growth mechanism; however, the
oligomerization is terminated by intramolecular C-H activation. The novel coordinated linear
dimer (C-H activated Mes*)PCH2PH(Mes*)CH2GaCl3 has been characterized crystallo-
graphically. Interestingly, mechanistic studies reveal that the diphosphiranium ring
Mes*PCH2P(Mes*)CH2GaCl3 derived from the reaction of Mes*PCH2GaCl3 with Mes*Pd
CH2 is an intermediate in this transformation. The reaction of phosphaalkenes with
phosphenium species appears to be a general method to prepare diphosphiranium ions. In
one case, NMR spectroscopic data suggests that treating MesPdCPh2 with HOTf gives both
the diphosphiranium species [MesPCPh2P(Mes)CPh2H]OTf and the adduct Ph2Cd(Mes)PfP-
(Mes)(CHPh2)]OTf. Remarkably, treating concentrated Mes*PdCH2 solutions with HOTf
results in oligomers of up to six repeat units, as determined by ESI mass spectrometry.
These results suggest that it may be possible to initiate the polymerization of PdC bonds
using cationic initiators and that the propagating species will be a cationic phosphenium
moiety.
Introduction
an olefin with the isolobal and isovalent phosphinidene
(RP) group. Therefore, it might be expected that there
would be numerous parallels between the chemistry of
PdC bonds in phosphaalkenes and that of CdC bonds
The investigation of compounds containing a multiple
bond involving a heavier element of the p block often
leads to unexpected results that can challenge our ideas
of structure, bonding, and reactivity.1 One of the most
thoroughly studied classes of low-coordinate compounds
of the main-group elements are the phosphaalkenes:
compounds possessing a (3p-2p)π bond between phos-
phorus and carbon.2 Phosphaalkenes can formally be
derived by simply replacing a carbene (R2C) moiety in
(2) For reviews on phosphaalkenes, see: (a) Mathey, F. Angew.
Chem., Int. Ed. 2003, 42, 1578. (b) Dillon, K. B.; Mathey, F.; Nixon, J.
F. Phosphorus: The Carbon Copy; Wiley: New York, 1998. (c) Weber,
L. Eur. J. Inorg. Chem. 2000, 2425. (d) Yoshifuji, M. J. Chem. Soc.,
Dalton Trans. 1998, 3343. (e) Weber, L. Angew. Chem., Int. Ed. Engl.
1996, 35, 271. (f) Gaumont, A. C.; Denis, J. M. Chem. Rev. 1994, 94,
1413. (g) Mathey, F. Acc. Chem. Res. 1992, 25, 90. (h) Appel, R. In
Multiple Bonds and Low Coordination in Phosphorus Chemistry;
Regitz, M., Scherer, O. J., Eds.; Thieme: Stuttgart, Germany, 1990.
(i) Appel, R.; Knoll, F. Adv. Inorg. Chem. 1989, 33, 259. (j) Scherer, O.
J. Angew. Chem., Int. Ed. Engl. 1985, 24, 924. (k) Appel, R.; Knoll, F.;
Ruppert, I. Angew. Chem., Int. Ed. Engl. 1981, 20, 731.
* To whom correspondence should be addressed. Tel: 604-822-9117.
Fax: 604-822-2847. E-mail: dgates@chem.ubc.ca.
(1) Power, P. P. Chem. Rev. 1999, 99, 3463.
10.1021/om0495022 CCC: $27.50 © 2004 American Chemical Society
Publication on Web 11/03/2004

5914 Organometallics, Vol. 23, No. 25, 2004
Tsang et al.
Scheme 1. Phosphorus-Carbon Isolobal Analogy
Chart 1. Possible Coordination Modes for the
Reaction of a Phosphaalkene with an Electrophile
(E+)
in alkenes (Scheme 1). The quest for such parallels
between phosphaalkenes and olefins has attracted
considerable attention, and thus far, these developments
have been limited primarily to molecular chemistry,2a,b,g
although very recently π-conjugated PdC analogues of
poly(p-phenylenevinylene) have been developed.3,4
The addition polymerization of olefins is a textbook
reaction for CdC double bonds; however, it remains
limited primarily to CdC and in some instances CdO
bonds. We have been interested in expanding addition
polymerization to phosphaalkenes, thereby extending
the phosphorus-carbon analogy to polymer science. To
this end, we have recently reported the polymerization
of phosphaalkene (1a: Mes ) 2,4,6-trimethylphenyl) to
afford the unprecedented poly(methylenephosphine) 2,
Scheme 2. Proposed Mechanism for the
Hypothetical Cationic Polymerization of a
Phosphaalkene Initiated by an Electrophile (E+)
served coordination modes are η1 (Chart 1, mode I), and
η2 (Chart 1, mode II) which is reminiscent of the side-
on binding of olefins. Neither of these two modes would
be expected to initiate cationic chain growth, although
mode II is of potential interest in coordination poly-
merization. A possible mechanism for the putative
cationic polymerization of a phosphaalkene is shown in
Scheme 2. Although propagation could occur through
either a phosphenium ion (Chart 1, mode III) or a
carbocation (Chart 1, mode IV), we hypothesized that
mode III would be most likely on the basis of electrone-
gativity differences (øP < øC).10
an inorganic polymer composed of alternating phospho-
rus and carbon atoms.5 At present, we have successfully
polymerized 1a using either anionic or radical methods
of initiation. A thorough understanding of the nature
of the interaction between phosphaalkenes and potential
polymerization initiators at the molecular level will be
essential to fully develop and optimize this new chem-
istry. In the case of the anionic polymerization of 1a,
we have shown that MeLi adds to the PdC bond,
generating a carbanion, and preliminary investigations
of the products of chain growth using MALDI-TOF mass
spectroscopy suggest that backbiting may compete with
linear chain growth.6 The work described herein will
assess the potential for cationic polymerization of phos-
phaalkenes 1a (Mes ) 2,4,6-trimethylphenyl) and 1b
(Mes* ) 2,4,6-tri-tert-butylphenyl).
The cationic initiators used to polymerize alkenes
tend to be either strong protic acids with relatively
weakly coordinating anions (i.e. HX: X ) OTf-, BF4
)
-
or Lewis acids (i.e. AlCl3). Surprisingly, there are very
few studies in the literature of reactions of phosphaalk-
enes with such potential polymerization initiators,
although it has been shown that tBuPdCHtBu will form
an η1 complex with AlCl3.11 The generation of phosphen-
ium ions (i.e. Chart 1; mode III) by addition of an
electrophile to a PdC bond is rare; however, addition
of HOTf to Cp*PdCR2 (Cp* ) C5Me5; R ) SiMe3) gives
the phosphenium species [Cp*PCHR2]OTf, which was
characterized by NMR spectroscopy.12 It is also possible
(7) The electronic structure of phosphaalkenes has been studied
extensively using photoelectron spectroscopy and molecular modeling.
The two filled frontier orbitals (the lone pair on phosphorus and the
(2p-3p)π bond) are closely spaced in energy, and the order is highly
dependent on the nature of the substituents. (See, for example:
Lacombe, S.; Gonbeau, D.; Cabioch, J.-L.; Pellerin, B.; Denis, J.-M.;
Pfister-Guillouzo, G. J. Am. Chem. Soc. 1988, 110, 6964 and references
therein).
The cationic polymerization of olefins involves the
addition of an electrophile across the CdC bond to
generate a highly reactive carbocation which acts as the
propagating species. By comparison with olefins, the
reaction of a phosphaalkene with an electrophile (E+)
is more complex, due to the possible interaction with
the lone pair at phosphorus or the PdC π-bond. Thus,
several possible products may be envisaged (Chart 1).7
Most studies of reactions of PdC bonds with electro-
philes have focused on their use as ligands for transition
metals or on the simple addition of RX across the PdC
bond (R ) H, alkyl; X ) halide).2j,8,9 In phosphaalkene-
transition-metal complexes, the most commonly ob-
(8) Nixon, J. F. Chem. Rev. 1988, 88, 1327.
(9) Most studies of the coordination chemistry of phosphaalkenes
have involved transition metals. Selected recent examples: Daugulis,
O.; Brookhart, M.; White, P. S. Organometallics 2002, 21, 5935. Weber,
L.; Dembeck, G.; Lo¨nneke, P.; Stammler, H.; Neumann, B. Organo-
metallics 2001, 20, 2288. Bridgeman, A. J.; Mays, M. J.; Woods, A. D.
Organometallics 2001, 20, 2076. Ikeda, S.; Ohhata, F.; Miyoshi, M.;
Tanaka, R.; Minami, T.; Ozawa, F.; Yoshifuji, M. Angew. Chem., Int.
Ed. 2000, 39, 4512. van der Sluis, M.; Beverwijk, V.; Termaten, A.;
Bickelhaupt, F.; Kooijman, H.; Spek, A. L. Organometallics 1999, 18,
1402. Benvenutti, M. H. A.; Cenac, N.; Nixon, J. F. Chem. Commun.
1997, 1327. Hobbold, M.; Streubel, R.; Benvenutti, M. H. A.; Hitchcock,
P. B.; Nixon, J. F. Organometallics 1997, 16, 3726. Jouaiti, A.; Geoffroy,
M.; Terron, G.; Bernardinelli, G. J. Am. Chem. Soc. 1995, 117, 2251.
(10) The Pauling electronegativities (P ) 2.19; C ) 2.55) suggest
that the polarity should be P+dC-. This is true for the σ-system;
however, evidence from UV and magnetic circular dichroism of
phosphinines suggests that the effective π-electronegativity of P is
similar to or higher than that of C. For a useful discussion of this see:
Waluk, J.; Klein, H.-P.; Ashe, A. J., III; Michl, J. Organometallics 1989,
8, 2804.
(3) Wright, V. A.; Gates, D. P. Angew. Chem., Int. Ed. 2002, 41, 2389.
(4) Smith, R. C.; Chen, X.; Protasiewicz, J. D. Inorg. Chem. 2003,
42, 5468.
(5) Tsang, C.-W.; Yam, M.; Gates, D. P. J. Am. Chem. Soc. 2003,
125, 1480.
(6) Gillon, B. H.; Gates, D. P. Chem. Commun. 2004, 1868.
(11) Niecke, E.; Symalla, E. Chimia 1985, 39, 320.

Addition of Electrophiles to Phosphaalkenes
Organometallics, Vol. 23, No. 25, 2004 5915
to generate phosphenium cations from the reaction of
electrophiles with the diphosphene Mes*PdPMes*.13,14
The synthesis of stable phosphenium ions is of consider-
able fundamental interest, and a fascinating chemistry
for these species, which are isovalent with carbenes, has
emerged.15-17 To date, most isolable phosphenium ions
possess two adjacent π-donor substituents (i.e. NR, O,
S, P), although stable systems are known with one
π-donor and one bulky aryl substituent.18-20 A phos-
phenium ion generated by adding E+ across the PdC
bond of either 1a or 1b would have no adjacent π-donors
and thus will be highly reactive, which we hope will lead
to polymerization.21
In this paper we will provide evidence that transient
phosphenium ions can be generated by the addition of
Lewis or protic acids to the PdC bonds in 1a and 1b:
the first step in a putative cationic polymerization. A
portion of this work has appeared as a preliminary
communication.22 Three different fates for these highly
reactive phosphenium ions have been observed: (i)
intramolecular C-H activation (in the case of 1b), (ii)
donor-acceptor coordination of phosphaalkenes to phos-
phenium ions, affording novel diphosphiranium (P2C)
rings (in the case of both 1a and 1b) and/or phosphaalk-
ene-phosphenium adducts (in the case of 1a), and (iii)
linear oligomerization (in the case of 1b).
Figure 1. Molecular structure of 1b (50% probability
ellipsoids). Hydrogen atoms are omitted for clarity. Selected
bond lengths (Å) and angles (deg): P(1)-C(12) ) 1.643(3);
C(1)-P(1)-C(12) ) 104.1(1).
iron.24,25 This is in stark contrast to the widely studied
halogen-substituted systems Mes*PdCXH and Mes*Pd
CX2 (X ) halogen).26 Compound 1b was prepared by the
reaction of CH2Cl2 and Mes*PH2 (31P NMR: δ -127;
1JPH ) 205 Hz) in the presence of KOH.24 Interestingly,
31P NMR analysis of the initial reaction mixture re-
vealed a previously undetected intermediate (δ -66
1
ppm, JPH ) 226 Hz) that was consistent with Mes*P-
(H)CH2Cl. After 1 week, the 31P NMR spectrum of the
reaction mixture revealed that phosphaalkene 1b had
been formed quantitatively. Recrystallization of the
crude product afforded pale yellow crystals of 1b suit-
able for X-ray diffraction (yield 59%). The molecular
structure of 1b is shown in Figure 1. The PdC bond
length of 1.643(3) Å is in the shorter end of the range
found for the PdC double bonds in phosphaalkenes
(1.61-1.71 Å)2h and is significantly shorter than that
in 1a (1.692(3) Å).27
Results and Discussion
Phosphaalkenes 1a and 1b were chosen for cationic
PdC bond activation studies since they are isolable yet
possess minimal steric bulk and do not contain electron-
rich heteroatom substituents (i.e. O, N, P) that may
interfere in their reactions with electrophiles. Com-
pound 1a was prepared by following a modification to
a literature method23 which we have reported previ-
ously.5 The synthesis and characterization of phos-
phaalkene 1b, a rare example of a phosphaalkene with
H substituents, was first reported in 1984.24 Surpris-
ingly, the reactivity of compound 1b has received little
attention, since its initial synthesis and coordination to
Reaction with Lewis Acids: Intramolecular C-H
Activation. Initially, we chose to explore the possible
cationic activation of the PdC bond in 1b with the Lewis
acid GaCl3. Unexpectedly, the 31P NMR spectrum of the
reaction mixture revealed that 1b (δ 289.0) was com-
pletely consumed and that a new species with a doublet
(12) Gudat, D.; Niecke, E.; Krebs, B.; Dartmann, M. Chimia 1985,
39, 277.
(13) Cowley, A. H.; Kilduff, J. E.; Norman, N. C.; Pakulski, M. J.
Am. Chem. Soc. 1983, 105, 4845.
(25) (a) Appel, R.; Casser, C.; Knoch, F. J. Organomet. Chem. 1985,
293, 213. (b) Etemad-Moghadam, G.; Rifqui, M.; Layrolle, P.; Berlan,
J.; Koenig, M. Tetrahedron Lett. 1991, 32, 5965. (c) Breen, T. L.;
Stephan, D. W. J. Am. Chem. Soc. 1995, 117, 11914. (d) Ito, S.;
Yoshifuji, M. Chem. Lett. 1998, 651. (e) Termaten, A. T.; Nijbacker,
T.; Schakel, M.; Lutz, M.; Spek, A. L.; Lammertsma, K. Organome-
tallics 2002, 21, 3196.
(14) Loss, S.; Widauer, C.; Gru¨tzmacher, H. Angew. Chem., Int. Ed.
1999, 38, 3329.
(15) For reviews, see: (a) Cowley, A. H.; Kemp, R. A. Chem. Rev.
1985, 85, 367. (b) Sanchez, M.; Mazie`res, M.-R. In Multiple Bonds and
Low Coordination in Phosphorus Chemistry; Regitz, M., Scherer, O.
J., Eds.; Thieme: Stuttgart, Germany, 1990; p 129. (c) Driess, M.;
Gru¨tzmacher, H. Angew. Chem., Int. Ed. Engl. 1996, 35, 828.
(16) Burford, N.; Ragogna, P. J. Dalton 2002, 4307.
(17) Gudat, D. Coord. Chem. Rev. 1997, 163, 71.
(26) See, for example: Ito, S.; Toyota, K.; Yoshifuji, M. J. Organomet.
Chem. 1998, 553, 135. van der Sluis, M.; Klootwijk, A.; Wit, J. B. M.;
Bickelhaupt, F.; Veldman, N.; Spek, A. L.; Jolly, P. W. J. Organomet.
Chem. 1997, 529, 107. Ito, S.; Toyota, K.; Yoshifuji, M. Chem. Commun.
1997, 1637. Niecke, E.; Fuchs, A.; Baumeister, F.; Nieger, M.; Schoeller,
W. W. Angew. Chem., Int. Ed. Engl. 1995, 34, 555. Yoshifuji, M.; Ito,
S.; Toyata, K.; Yasunami, M. Bull. Chem. Soc. Jpn. 1995, 68, 1206.
Ito, S.; Toyota, K.; Yoshifuji, M. Chem. Lett. 1995, 747. Jun, H.;
Angelici, R. J. Organometallics 1994, 13, 2454. Toyota, K.; Tashiro,
K.; Yoshifuji, M. Angew. Chem., Int. Ed. Engl. 1993, 32, 1163. Goede,
S. J.; van Schalk, H. P.; Bickelhaupt, F. Organometallics 1992, 11,
3844. Goede, S. J.; Bickelhaupt, F. Chem. Ber. 1991, 124, 2677. Jun,
H.; Young, V. G., Jr.; Angelici, R. J. J. Am. Chem. Soc. 1991, 113, 9379.
Yoshifuji, M.; Toyota, K.; Yoshimura, H. Chem. Lett. 1991, 491.
Yoshifuji, M.; Niitsu, T.; Inamoto, N. Chem. Lett. 1988, 1733. Markl,
G.; Kreitmeier, P. Angew. Chem., Int. Ed. Engl. 1988, 27, 1360. Appel,
R.; Hu¨nerbein, J.; Siabalis, N. Angew. Chem., Int. Ed. Engl. 1987, 26,
779.
(18) Reed, R. W.; Xie, Z.; Reed, C. A. Organometallics 1995, 14, 5002.
(19) Dumitrescu, A.; Gornitzka, H.; Schoeller, W. W.; Bourissou, D.;
Bertrand, G. Eur. J. Inorg. Chem. 2002, 1953.
(20) In contrast to the highly reactive phosphenium ions (R2P+),
phosphinium ions (i.e. C6H4P-Me+; cf. Chart 1, type I) are stable and
exhibit reactivity much different from that of phosphenium ions. See,
for example: Moores, A.; Ricard, L.; Le Floch, P. Angew. Chem., Int.
Ed. 2003, 42, 4940.
(21) A phosphenium salt without π-donors, [Cp*PtBu]AlCl4, has been
detected spectroscopically; however, its high reactivity results in C-H
activation of Cp*. See: Cowley, A. H.; Mehrota, S. K. J. Am. Chem.
Soc. 1983, 105, 2074. See also ref 12.
(22) Tsang, C.-W.; Rohrick, C. A.; Saini, T. S.; Patrick, B. O.; Gates,
D. P. Organometallics 2002, 21, 1008.
(23) Mundt, O.; Becker, G.; Uhl, W. Z. Anorg. Allg. Chem. 1986,
540/541, 319.
(27) van der Knaap, T. A.; Klebach, T. C.; Visser, F.; Bickelhaupt,
F.; Ros, P.; Baerends, E. J.; Stam, C. H.; Konijn, M. Tetrahedron 1984,
40, 765.
(24) Appel, R.; Casser, C.; Immenkeppel, M.; Knoch, F. Angew.
Chem., Int. Ed. Engl. 1984, 23, 895.

5916 Organometallics, Vol. 23, No. 25, 2004
Tsang et al.
Figure 3. 31P NMR spectra (CD2Cl2) showing reaction
progress: (a) 1b at 193 K; (b) 1b + GaCl3 f 5 (M ) Ga) at
193 K; (c) reaction in (b) with the temperature raised to
213 K; (d) reaction in (b) with the temperature raised to
233 K; (e) reaction in (b) with the temperature raised to
253 K; (f) reaction in (b) with the temperature raised to
273 K, showing quantitative formation of 3 (M ) Ga).
Figure 2. Molecular structure of 3 (M ) Ga) (50%
probability ellipsoids). All hydrogen atoms except H(20) are
omitted for clarity. Selected bond lengths (Å) and angles
(deg): P(1)-C(1) ) 1.773(3), P(1)-C(2) ) 1.786(3), P(1)-
C(9) ) 1.793(4), Ga(1)-C(1) ) 1.994(3); P(1)-C(1)-Ga(1)
) 112.70(14).
coordinate gallium compounds which exhibit slightly
longer Ga-C distances (i.e. Me3GaNMe3; Ga-C )
1.998(4) Å).31
We have carried out a similar reaction of 1b with
AlCl3 to afford the zwitterionic coordinated ylide 3 (M
) Al) (31P NMR: 16.9 ppm, 1JPH ) 489 Hz). In contrast
to 3 (M ) Ga), which is stable in air as a solid (g1 year
by NMR), compound 3 (M ) Al) is unstable in the solid
state even under inert conditions, and over several days
proton migration from P to C occurs quantitatively,
accompanied by AlCl3 loss. In solution, AlCl3 can be used
to promote catalytic intramolecular C-H activation in
1b, whereas GaCl3 does not show turnover under
ambient conditions.22 The reason for this difference in
stability is not obvious; however, we speculate that this
suggests a weaker Lewis acidity of AlCl3 than GaCl3
with respect to the ylide group in 3. We have detected
the analogous C-H activated product 3 (M ) In) by
resonance at 16.4 ppm and a very large P-H coupling
constant of 498 Hz had been formed. The large coupling
constant suggested the presence of a direct P-H bond;
however, its magnitude was much greater than the P-H
coupling in trivalent phosphines (i.e. Mes*PH2; 1JPH
)
1
205 Hz). In addition, the P-H signal in the H NMR
spectrum (7.4 ppm) was shifted downfield significantly
with respect to a phosphine (i.e. Mes*PH2; δ(1H) 4.30
ppm), suggesting a phosphonium moiety. For compari-
son, [(n-Bu)3PH]BF4 (δ(1H) 6.07 ppm, δ(31P) 14.1 ppm,
1JPH ) 484 Hz) and [(t-Bu)3PH]BF4 (δ(1H) 6.07 ppm, δ-
1
(31P) 51.7 ppm, JPH ) 465 Hz) show similar spectral
properties.28 In addition to signals in the P-H and aryl
1
NMR spectroscopy (31P NMR: 11.5 ppm, JPH ) 513
1
regions, the aliphatic region of the H NMR spectrum
Hz). Unfortunately, the reaction of 1b with InCl3 is very
slow and there is only partial conversion of 1b to 3 (M
) In) even with heating.
of the product showed three singlet resonances (δ 1.51,
6H; δ 1.48, 9H; δ 1.34, 9H) assigned to methyl groups
and a multiplet (δ 2.26, 2H) attributed to a methylene
group. This suggested that one of the tBu resonances of
Mes* had been activated, and we speculated that the
product was 3 (M ) Ga). Intramolecular C-H activation
Mechanism of C-H Activation: Is a Phosphen-
ium Ion Involved? On the basis of the fact that the
product of the reaction of 1b with GaCl3 contains a
Ga-C bond, we reasoned that the initial step in the
reaction must involve addition of GaCl3 to the PdC
bond, generating a phosphenium zwitterion (4; i.e.,
of the Mes* group is a reasonably common reaction in
low-coordinate phosphorus chemistry.29 Confirmation of
our assignment of structure 3 (M ) Ga) was obtained
through X-ray crystallography, and the molecular struc-
ture is shown in Figure 2. The Ga(1)-C(1) bond length
of 1.994(3) Å is significantly longer than for GaMe3
(Ga-C ) 1.967(2) Å)30 and is comparable to most four-
Chart 1; mode III) which subsequently inserts into the
C-H bond of Mes*. The potential involvement of 4 was
encouraging from the point of view of cationic poly-
merization, and therefore, variable-temperature NMR
studies were initiated to gain additional evidence for a
phosphenium intermediate. Thus, 1b was treated with
GaCl3 in d2-dichloromethane in a cooled NMR tube (163
K) and the product was immediately analyzed using 31P,
1H, and 13C NMR spectroscopy. Remarkably, only a
single signal is observed at 208.9 ppm in the 31P NMR
spectrum at 193 K (Figure 3 b). This species was stable
for several days if kept at 193 K; however, upon
(28) Netherton, M. R.; Fu, G. C. Org. Lett. 2001, 3, 4295.
(29) See for example: Shah, S.; Simpson, M. C.; Smith, R. C.;
Protasiewicz, J. D. J. Am. Chem. Soc. 2001, 123, 6925. Straddiotto,
M.; Fujdala, K. L.; Tilley, T. D. Helv. Chim. Acta 2001, 84, 2958.
Cowley, A. H.; Gabba¨ı, F.; Schluter, R.; Atwood, D. J. Am. Chem. Soc.
1992, 114, 3142. Gouygou, M.; Tachon, C.; Etemad-Moghadam, G.;
Koenig, M. Tetrahedron Lett. 1989, 30, 7411. Yoshifuji, M.; Sato, T.;
Inamoto, N. Chem. Lett. 1988, 1735. Karsch, H. H.; Reisacher, H.-U.;
Mu¨ller, G. Angew. Chem., Int. Ed. Engl. 1986, 25, 454. Cowley, A. H.;
Kilduff, J. E.; Norman, N. C.; Pakulski, M. J. Chem. Soc. Dalton Trans.
1986, 1801. Cowley, A. H.; Pakulski, M. K. J. Am. Chem. Soc. 1984,
106, 1491.
(30) Beagley, B.; Schmidling, D. G.; Steer, I. A. J. Mol. Struct. 1974,
21, 437.
(31) Golubinskaya, L. M.; Golubinskii, A. V.; Mastryukov, V. S.;
Vilkov, L. V.; Bregadze, V. I. J. Organomet. Chem. 1976, 117, C4.

Addition of Electrophiles to Phosphaalkenes
Organometallics, Vol. 23, No. 25, 2004 5917
length of 1.687(3) Å is comparable to the PdC bond
length in the starting material 1a, 1.693(3) Å,27 consis-
tent with significant multiple-bond character in 6. To
our knowledge, tBu(Cl3Al)PdCHtBu is the only related
η1 complex of a phosphaalkene to a main-group metal;
however, that compound was not characterized crystal-
lographically.11,32,33 In contrast, adducts with transition
metals are quite common. For comparison, the PdC
bond lengths found in (Ph3P)2PtP(Mes)dCPh2 (1.65(1)
Å)34 and related Fe, Ni, and Cr phosphaalkene com-
plexes34-39 are in the same range as that in 6. Interest-
ingly, 6 is stable under an inert atmosphere and there
is no evidence for C-H activation of the o-Me groups.
It is possible that the increased stability of 6 over 5 (M
) Al, Ga) is a consequence of the steric bulk of the larger
Mes* group and the absence of large C substituents in
5, which effectively favors migration of GaCl3 from P to
C to generate the fleeting phosphenium 4, which im-
mediately undergoes C-H activation.
We still had only circumstantial evidence that phos-
phenium ions were involved in these reactions and
thought that treating 1b with triflic acid, another
potential cationic initiator, may give us additional
insight into the reaction mechanism. In this case, an
analogous C-H activation reaction occurred, giving
phosphonium salt 7 quantitatively from 1b by 31P NMR.
Figure 4. Molecular structure of 6 (50% probability
ellipsoids). All hydrogen atoms are omitted for clarity.
Selected bond lengths (Å) and angles (deg): P(1)-C(1) )
1.687(3), P(1)-Ga(1) ) 2.3938(7), P(1)-C(14) ) 1.798(3);
C(14)-P(1)-C(1) ) 112.8(1), Ga(1)-P(1)-C(1) ) 129.33(9),
P(1)-C(1)-C(2) ) 119.3(2), P(1)-C(1)-C(8) ) 121.4(2).
warming to room temperature 3 (M ) Ga) was formed
quantitatively, confirming the intermediacy of 4 in the
C-H activation reaction (Figure 3). Initially, we specu-
lated that the intermediate might be phosphenium 4
(M ) Ga); however, 1H and 13C NMR spectroscopy
(including COSY) suggested that the intermediate was
adduct 5 (M ) Ga) (i.e. Chart 1; mode I). Convincing
evidence for this assignment came from the observation
of two inequivalent methylene hydrogens (δ 7.5 and 7.1;
The product 7 was characterized crystallographically,
and the molecular structure is shown in Figure 5. Of
particular interest is that there is hydrogen bonding
between the cation and the anion, as evidenced by a
close H1-O3A contact (2.38(2) Å), which is within the
sum of the van der Waals radii (rvdw ) 2.70 Å). We were
unsuccessful in detecting intermediates by using VT
NMR techniques, and even at -110 °C in CDCl2F
product 7 formed cleanly from 1b. Nevertheless, the
intermediacy of a fleeting phosphenium ion in the
formation of 7 was confirmed by a deuterium labeling
study. Treatment of 1b with DOTf yields 7 exclusively
1
2JHH ) 8 Hz) in the H NMR spectrum and a doublet
resonance at 148.7 ppm (1JCP ) 51 Hz) in the 13C NMR
spectrum. This was consistent with retention of the Pd
CH2 moiety in 5 (M ) Ga).
The low-temperature reaction of 1b with AlCl3 was
similarly observed to give the analogous adduct 5 (M )
Al), which underwent C-H activation upon warming
to room temperature. In particular, the 31P NMR
spectrum of 5 (M ) Al) in CD2Cl2 (193 K) showed only
a singlet resonance at 220.5 ppm. This species was also
characterized using 1H and 13C NMR, and analogous to
5 (M ) Ga), inequivalent methylene protons were
observed and a 13C signal at 152.0 assigned to the Pd
CH2 moiety confirmed its identity as 5 (M ) Al).
Neither species of 5 (M ) Al, Ga) could be isolated to
confirm their structures, since they were quantitatively
converted to 3 when warmed to 213 K or above.
However, when 1a was treated with GaCl3 (1 equiv),
the only product observed was the adduct 6 (31P NMR:
δ 132), which was stable at room temperature and could
be analyzed crystallographically. The molecular struc-
(32) Ga(III) adducts of heterocycles containing low-coordinate phos-
+
phorus atoms (i.e. 1,1′-diphosphaferrocene) are known, and a GaCl2
-
bridged [1]diphosphaferrocenophane has been structurally character-
ized. See, for example: Sava, X.; Melaimi, M.; Me´zailles, N.; Ricard,
L.; Mathey, F.; Le Floch, P. New J. Chem. 2002, 26, 1378.
(33) The adduct H3BP(Mes*)dCH2 has been detected spectroscopi-
cally as an intermediate in the hydroboration of 1b with excess BH3‚
THF. See: Yoshifuji, M.; Takahashi, H.; Toyota, K. Heteroat. Chem.
1999, 10, 187.
(34) van der Knaap, T. A.; Bickelhaupt, F.; Kraaykamp, J. G.; van
Koten, G.; Bernards, J. P. C.; Edzes, H. T.; Veeman, W. S.; de Boer,
E.; Baerends, E. J. Organometallics 1984, 3, 1804.
(35) Cowley, A. H.; Jones, R. A.; Lasch, J. G.; Norman, N. C.;
Stewart, C. A.; Stuart, A. L.; Atwood, J. L.; Hunter, W. E.; Zhang, H.-
M. J. Am. Chem. Soc. 1984, 106, 7015.
(36) Al-Resayes, S. I.; Klein, S. I.; Kroto, H. W.; Meidine, M. F.;
Nixon, J. F. J. Chem. Soc., Chem. Commun. 1983, 930.
(37) Gudat, D.; Nieger, M.; Schmitz, K.; Szarvas, L. Chem. Commun.
2002, 1820.
(38) van der Knaap, T. A.; Bickelhaupt, F.; van der Poel, H.; van
Koten, G.; Stam, C. H. J. Am. Chem. Soc. 1982, 104, 1756.
(39) van der Knaap, T. A.; Jenneskens, L. W.; Meeuwissen, H. J.;
Bickelhaupt, F.; Walther, D.; Dinjus, E.; Uhlig, E.; Spek, A. L. J.
Organomet. Chem. 1983, 254, C33.
ture of 6 is shown in Figure 4. The P(1)-C(1) bond

5918 Organometallics, Vol. 23, No. 25, 2004
Tsang et al.
Figure 6. Molecular structure of 8 (50% probability
ellipsoids). All hydrogen atoms except H(1) are omitted for
clarity. Selected bond lengths (Å) and angles (deg): P(1)-
C(19) ) 1.776(2), P(2)-C(20) ) 1.919(9), P(1)-C(20) )
1.815(3), P(1)-H(1) ) 1.31(2), Ga(1)-C(19) ) 1.990(2);
C(19)-P(1)-C(20) ) 114.07(11), Ga(1)-C(19)-P(1) )
115.89(14), C(20)-P(2)-C(21) ) 100.4(4), P(1)-C(20)-P(2)
) 110.1(2).
Figure 5. Molecular structure of 7 (50% probability
ellipsoids). All hydrogen atoms except H1 are omitted for
clarity. Selected bond lengths (Å) and angles (deg): P(1)-
C(19) ) 1.782(2), P(1)-C(1) ) 1.803(3), P(1)-C(2) )
1.803(2); C(2)-P(1)-C(1) ) 95.42(11), C(2)-P(1)-C(19) )
114.43(11), C(19)-P(1)-C(1) ) 111.49(12).
formation of a dimeric species; however, the coupling
constant was too large to be a P-C-P-C linkage and
suggested a direct P-P bond. The only other compounds
present in the crude reaction mixture were unreacted
1b (δ 289) and zwitterionic phosphonium 3 (M ) Ga)
(δ 16.4).
Over the few hours necessary to separate 1b and 3
(M ) Ga) from the dimeric intermediate mentioned
above, the intermediate was transformed quantitatively
to a new isolable compound. Crystals suitable for X-ray
diffraction were obtained by slow evaporation of an
acetonitrile solution. The molecular structure is shown
in Figure 6 and confirms its identity as a linear dimer
coordinated to GaCl3 (8). No analogous compounds with
with a -CH2D substituent and no evidence for a P-D
bond in the product, as determined using NMR spec-
troscopy. Therefore, even if deuterium initially reacts
at phosphorus to form an adduct, it must migrate to
carbon, forming a phosphenium ion prior to C-H
activation. Otherwise, the product 7 would possess both
P-H and P-D bonds.
In this section, we have provided strong evidence that
the mechanism of reaction of 2 with Lewis acids involves
initial formation of adduct 5, followed by GaCl3 migra-
tion from P to C to form the fleeting phosphenium 4,
which rapidly undergoes C-H activation. In the case
of triflic acid, a phosphenium ion must also be involved
in the reaction; however, although it is unlikely, it is
not clear whether [Mes*HPdCH2]+ is also involved.
With this encouraging evidence for phosphenium ion
formation from 1b, we set out to find conditions where
intermolecular reaction (i.e. 4 + 1b) could be favored
over intramolecular C-H activation (i.e. 3).
GaCl3-Initiated Linear Dimerization of a Phos-
phaalkene. We postulated that, if very concentrated
solutions of monomer 1b and substoichiometric quanti-
ties of electrophile were used, intermolecular addition
of 4 to 1b (i.e. chain growth) might be favored over
intramolecular C-H activation. Thus, phosphaalkene
1b in CH2Cl2 (ca. 1 M) was treated with GaCl3 (0.5
equiv) at -40 °C and warmed to room temperature.
Interestingly, the 31P NMR spectrum of the reaction
mixture revealed a new compound that showed doublets
at -72.3 and -92.6 ppm with a large coupling constant
(1JPP ) 245 Hz). These data were consistent with the
such a P-C-P-C-Ga backbone have previously been
characterized. The P-C-P bond angle in 8 (P(1)-
C(20)-P(2) ) 110.1(2)°) is slightly widened with respect
to less sterically hindered bis(diphenylphosphino)-
methane (106.2(3)°).40 However, the P-C-P angle in 8
is significantly smaller than the very bulky Cy2PCH2-
PCy2 (120.5(1)°),41,42 suggesting that the steric nature
of 8 is intermediate between Ph2PCH2PPh2 and Cy2-
(40) Schmidbaur, H.; Reber, G.; Schier, A.; Wagner, F.; Mu¨ller, G.
Inorg. Chim. Acta 1988, 147, 143.
(41) Nickel, T.; Goddard, R.; Kru¨ger, C.; Po¨rschke, K.-R. Angew.
Chem., Int. Ed. Engl. 1994, 33, 879.
(42) Wolf, J.; Manger, M.; Schmidt, U.; Fries, G.; Barth, D.;
Weberndorfer, B.; Vicic, D. A.; Jones, W. D.; Werner, H. J. Chem. Soc.,
Dalton Trans. 1999, 1867.

Addition of Electrophiles to Phosphaalkenes
Organometallics, Vol. 23, No. 25, 2004 5919
Scheme 3. Proposed Mechanism for the
GaCl3-Initiated Linear Dimerization of 1b
phaalkene-phosphenium complex 11 is also a possibility.
We favored structure 10 on the basis of the observed
31P chemical shifts (-72.3 and -92.6 ppm), which were
shifted upfield from that predicted for the PdC moiety
in 11 (>100 ppm). Moreover, recently the analogous
diphosphiranium heterocycle [Mes*(Me)PCH2PMes*]-
OTf was prepared from the reaction of Mes*PCH2PMes*
with excess MeOTf and has been characterized in
solution at low temperature (31P NMR: δ -80.5 and
PCH2PCy2. The Ga(1)-C(19) bond length of 1.990(2) Å
is comparable to that of compound 3 (M ) Ga) (1.994(3)
Å). The P-C bond lengths in 8 (average 1.836(9) Å) are
typical of P-C single bonds and are close to those in
dppm (1.848(5) Å)40 and differ only slightly from that
in Cy2PCH2PCy2 (1.858 Å).41,42
1
-95.4; JPP ) 261 Hz).48
Diphosphiranium zwitterion 10 may be viewed as an
intramolecularly stabilized phosphenium ion that ring
opens to 9 and, subsequently, rapidly undergoes C-H
activation, forming 8 (Scheme 3). From the perspective
of a cationic polymerization, this type of reversible
stabilization at the phosphenium end of a growing chain
may slow propagation and, in the absence of C-H
activation, lead to a novel controlled polymerization.49
The preparation of diphosphiranium (P2CR5)+ rings is
an area of growing interest,50-52 and compound 10 is a
Presumably, this product results from the termina-
tion of growing chain 9 by C-H activation of the o-tBu
group (Scheme 3). Interestingly, the phosphonium
hydrogen atom in 8 was found bound to P(1) rather
than P(2). Compound 8 was fully characterized by
1
2
31P and H NMR spectroscopy (1JPH ) 490 Hz; JPP
)
90 Hz). Although [2 + 2] cycloaddition reactions of
phosphaalkenes forming diphosphetaines are quite
common,2a,b to our knowledge, the linear dimerization
of a phosphaalkene has not been observed previously.43
Furthermore, the formation of 8 from 1b and GaCl3
represents a proof of concept for the potential use of
cationic species to initiate the polymerization of phos-
phaalkenes.
Mechanism of Linear Dimerization: Evidence
for Diphosphiranium Ions and Phosphaalkene-
Phosphenium Adducts. On the basis of our studies
of the mechanism of formation for 3 described above,
and the mechanism of cationic polymerization for ole-
fins,44 we propose that this reaction follows the addition
mechanism outlined in Scheme 3. However, the 31P
NMR data for the intermediate species mentioned above
(δ -72.3, -92.6; 1JPP ) 245 Hz) are not consistent with
any of the species in Scheme 3 and suggest that the
mechanism of formation for 8 is slightly more compli-
cated. It is well known that phosphenium ions can
behave as Lewis acid acceptors toward phosphorus
donors, forming phosphanyl-phosphenium ions.16,45-47
Since intermediate 9 contains both a phosphenium
moiety and a phosphine, we postulate that the spectro-
scopically detected intermediate could be the intramo-
lecular donor-acceptor complex 10, although the phos-
rare example of a (R2PCR2PR)+ heterocycle.48 Further-
more, we are not aware of any previous report of the
formation of a diphosphiranium ring from the reaction
of phosphenium cations with PdC bonds. It has been
reported that one of the products of the reaction of a
phosphenium ion and an iminophosphine was a three-
membered (P2N)+ ring similar to 10.53,54 Moreover, it is
well-known that alkynes react with RPCl2-AlCl3 com-
plexes to yield phosphirenium salts.50 Significantly,
recent work on the reversible exchange of free alkynes
with coordinated alkyne in MePhPCRCR′ ions has
implicated a transient phosphenium ion (MePPh)+ as
an intermediate after dissociation of RCtCR′.55
We have explored the potential synthetic utility of this
novel method to prepare diphosphiranium rings. Treat-
ment of 1b with 1256 in CD2Cl2 at -80 °C successfully
formed diphosphiranium 13 (31P NMR: δ 17.7, -81.6;
1JPP ) 225 Hz). The results of 31P-1H HSQC NMR
experiments allowed for the estimation of the 1H chemi-
cal shift of the ring-CH2 moiety at 2.2-2.3 ppm and a
2JPH value of ca. 40-50 Hz. Although diphosphiranium
(43) Recently, the ambient-temperature oligomerization of imino-
phosphines (RNdPX; R ) Dipp, X ) Cl) using GaCl3 has been reported
to give cyclic dimeric and trimeric species. See: Burford, N.; Cameron,
T. S.; Conroy, K. D.; Ellis, B.; Lumsden, M.; Macdonald, C. L. B.;
McDonald, R.; Phillips, A. D.; Ragogna, P. J.; Schurko, R. W.; Walsh,
D.; Wasylishen, R. E. J. Am. Chem. Soc. 2002, 124, 14012.
(44) Odian, G. Principles of Polymerization; Wiley: New York, 1991.
(45) Burford, N.; Cameron, T. S.; Ragogna, P. J.; Ocando-Mavarez,
E.; Gee, M.; McDonald, R.; Wasylishen, R. E. J. Am. Chem. Soc. 2001,
123, 7947.
(48) Loss, S.; Widauer, C.; Ru¨egger, H.; Fleischer, U.; Marchand,
C. M.; Gru¨tzmacher, H.; Frenking, G. Dalton 2003, 85.
(49) Matyjaszewski, K., Ed. Controlled Radical Polymerization; ACS
Symposium Series 685; American Chemical Society: Washington, DC,
1998.
(50) Bourissou, D.; Bertrand, G. Top. Curr. Chem. 2002, 220, 1.
(51) Mathey, F. Chem. Rev. 1990, 90, 997.
(52) Kato, T.; Gornitzka, H.; Baceiredo, A.; Schoeller, W. W.;
Bertrand, G. Science 2000, 289, 754.
(46) Abrams, M. B.; Scott, B. L.; Baker, R. T. Organometallics 2000,
19, 4944.
(47) Schultz, C. W.; Parry, R. W. Inorg. Chem. 1976, 15, 3046.

5920 Organometallics, Vol. 23, No. 25, 2004
Tsang et al.
such a time, the proposal of compounds 14 and 15,
although consistent with NMR observations, remains
tentative.
Cationic Linear Oligomerization of 1b Initiated
by HOTf. On the basis of the results described thus
far, triflic acid appears to be a more effective activator
of the PdC bond than Lewis acids, since the latter tend
to form η1 complexes through phosphorus. Therefore,
we attempted to prepare higher oligomers by treating
a concentrated CH2Cl2 solution of 1b with triflic acid
(0.39 equiv). A complex 31P NMR spectrum was ob-
served, displaying many multiplets due to P-P coupling
(range -50 to 50 ppm) in addition to dominant signals
for 1b and 7. Although we were unsuccessful in sepa-
rating these products, we observed ions consistent with
the phosphonium cation (Mes*PCH2)nH+ in the oligo-
mers (Mes*PCH2)nH+OTf- (n ) 2-6) using electrospray
mass spectrometry. These ES-MS results, although
preliminary, are important, since they suggest that it
may be possible to polymerize phosphaalkenes using
cationic initiators if the right initiator and monomer are
chosen.
Figure 7. 31P NMR spectrum (CD2Cl2; 298 K) of the
reaction mixture after phosphaalkene 1a was treated with
HOTf (0.5 equiv). The peak marked with an asterisk is
tentatively assigned to MesP(OTf)CHPh2.
13 was formed reasonably cleanly from 1b and 12 at
-80 °C, it was not formed quantitatively and is unstable
at room temperature.
Due to the fascinating linear dimerization of 1b
initiated by a cationic initiator (i.e. GaCl3), it seemed
that the logical next step would be to investigate the
reaction of 1a and HOTf to see if a similar linear
oligomerization would occur (recall that 1a + GaCl3
gives stable 6). We rationalized that the o-CH3 groups
of the Mes substituent in 1a would not undergo C-H
activation, which effectively terminates chain growth
with 1b. The 31P NMR spectrum of the reaction mixture
when phosphaalkene 1b was treated with HOTf (0.5
equiv) is shown in Figure 7. Significantly, two sets of
doublets with coupling constants greater than 500 Hz
in an approximate 1:1 ratio are observed. Confirmation
that this was P-P coupling rather than P-H coupling
was obtained from 31P{1H} and 31P-31P COSY experi-
ments. We speculate that these signals result from the
unprecedented donor-acceptor adduct 14 (δ 196.4 and
Summary
We have conducted a thorough investigation of the
reactions of electrophiles and phosphaalkenes in order
to demonstrate the feasibility of cationic polymerization
for phosphaalkenes. In particular, we have demon-
strated that Lewis or protic acids can add across the
PdC bond, generating highly reactive phosphenium
ions: the first step in a putative cationic polymerization.
In the case of the reaction of Lewis acids (MCl3: M )
Al, Ga) with 1b, the transient phosphenium (4) under-
goes intramolecular C-H activation to give 3. Signifi-
cantly, we have shown that, at high concentrations of
monomer 1b in CH2Cl2, GaCl3 initiates a linear dimer-
ization to afford 8 following a cationic chain growth
mechanism. We have identified the diphosphiranium
zwitterion 10 spectroscopically as an intermediate in
this reaction. Interestingly, the addition of phosphenium
ions to phosphaalkenes appears to be a general route
to (P2C)+ cycles, and we have prepared several novel
rings using this method (i.e. 10, 13, and 15). Finally,
we provide MS evidence that oligomers containing up
to six repeat units can be generated from the addition
of HOTf to 1b. We hope that these studies will form
the basis for the future development of cationic poly-
merization for PdC bonds.
1
130.5; JPP ) 509 Hz) and diphosphiranium cation 15
(δ 120.3 and 0.2; 1JPP ) 516 Hz). We speculate that the
signals at 196.4 and 0.2 ppm are due to the PdC in 14
1
and the phosphine moiety in 15, respectively. H{31P}
NMR selective decoupling experiments on the proton
signal from the CHPh2 moieties in 14 and 15 are also
consistent with this assignment. An additional minor
singlet resonance at 155 ppm in the 31P NMR spectrum
was observed, which we tentatively assign to the phos-
phenium triflate [MesP(OTf)CHPh2]. These species are
stable in solution for several days at room temperature,
and after solvent removal a red oil is obtained. Thus
far, we have been unsuccessful in obtaining crystals
suitable for X-ray diffraction from this mixture. Until
Experimental Section
General Procedures. All manipulations of air- and/or
water-sensitive compounds were performed under prepurified
nitrogen (Praxair, 99.998%) using standard high-vacuum or
Schlenk techniques or in an Innovative Technology Inc.
glovebox. 1H, 31P, and 13C NMR spectra were recorded at room
temperature on a Bruker Advance 300 MHz spectrometer.
Chemical shifts are reported relative to the following: CHCl3
1
or HC6D5 (δ 7.24 and 7.15 for H); 85% H3PO4 as an external
standard (δ 0.0 for 31P); CDCl3 or C6D6 (δ 77.0 and 128.0 for
13C). NMR assignments were made with the aid of multidi-
mensional NMR techniques.
(53) Gudat, D.; Nieger, M.; Niecke, E. J. Chem. Soc., Dalton Trans.
1989, 693.
(54) Roques, C.; Mazieres, M.-R.; Majoral, J.-P.; Sanchez, M. Tet-
rahedron Lett. 1988, 29, 4547.
Materials. Hexanes, dichloromethane, and toluene were
dried by passing through activated alumina columns.57 Tet-
(55) Brasch, N. E.; Hamilton, I. G.; Krenske, E. H.; Wild, S. B.
Organometallics 2004, 23, 299.
(56) Cowley, A. H.; Kemp, R. A.; Wilburn, J. C. Inorg. Chem. 1981,
20, 4289.
(57) Pangborn, A. B.; Giardello, M. A.; Grubbs, R. H.; Rosen, R. K.;
Timmers, F. J. Organometallics 1996, 15, 1518.

Addition of Electrophiles to Phosphaalkenes
Organometallics, Vol. 23, No. 25, 2004 5921
1
rahydrofuran was distilled from sodium/benzophenone im-
mediately prior to use; CDCl3 (CIL) was distilled from P2O5
and degassed. CD2Cl2 (CIL) was purchased in ampules and
used as received. GaCl3 and AlCl3 were purchased from Aldrich
and sublimed prior to use. HOTf was freshly distilled prior to
31P NMR (CDCl3, 300 K): δ 16.9 ppm (d, JPH ) 488 Hz).
1H NMR (CDCl3, 300 K): δ 7.62 (d, 1H), 7.20 (dm, 1JHP ) 488
Hz, 1H, PH), 7.27 (s, 1H, m-H), 2.77 (dd, 2JHP ) 11 Hz, 3JHH
)
5 Hz, 2H, PCH2C(CH3)2), 2.37 (m, 2H, PCH2Al), 1.52 (s, 6H,
C(CH3)2), 1.49 (s, 9H, o-C(CH3)3), 1.35 (s, 9H, p-C(CH3)3).
NMR-Scale Preparation of Intermediate 5 (M ) Ga).
To an NMR tube charged with Mes*PdCH2 (1b; 25.0 mg, 0.09
mmol) in CD2Cl2 (0.5 mL) and cooled to 163 K was added a
solution of GaCl3 (15 mg, 0.09 mmol) in CD2Cl2 (0.5 mL) by
syringe. The contents of the NMR tube were carefully shaken
at 193 K until the solids dissolved. The NMR tube was
introduced into the NMR spectrometer that had been precooled
to 193 K.
use, while DOTf was used as received. Mes*PH2 and 1a59
58
were prepared by following the literature procedures. The
KOH used in the preparation of 1a was made anhydrous by
recrystallization from EtOH.60
Preparation of 1b. This compound was prepared by
following a modification to the original literature procedure.24,61
A solution of Mes*PH2 (10.0 g, 0.0359 mol) and finely ground
undried 85% KOH (15 g, g0.23 mol) was stirred at room
temperature in CH2Cl2 (200 mL). When the reaction was
complete as determined by 31P NMR spectroscopy (ca. 7 days),
the solvent was removed in vacuo, the solid was dissolved in
hexanes (300 mL), and this solution was filtered through a
Celite to remove KOH. The solvent was removed from the
filtrate in vacuo, affording a pale yellow powder. The crude
product was recrystallized from a minimal amount of toluene/
CH3CN (9:1) to give light yellow crystals. Yield: 6.20 g (59%).
31P NMR (CDCl3, 300 K): δ 289.5 (t, 2JPH ) 31 Hz). 1H NMR
31P NMR (CD2Cl2, 193 K): δ 208.9. H NMR (CD2Cl2, 193
1
2
K): δ 7.5 (s, 2H, m-H), 7.5 (d, JHH ) 8 Hz, 1H, PdCH), 7.1
2
2
(dd, JHH ) 8 Hz, JPH ) 33 Hz, 1H, PdCH), 1.5 (s, 18H,
o-C(CH3)3), 1.3 (s, 9H, p-C(CH3)3). 13C NMR (CD2Cl2, 193 K):
1
δ 156.4 (s, o-Ar), 155.0 (s, p-Ar), 148.7 (d, JCP ) 51 Hz, Pd
3
1
CH2), 123.3 (d, JCP ) 8 Hz, m-Ar), 117.5 (d, JCP ) 14 Hz,
i-Ar), 37.5 (s, o-CMe3), 34.9 (s, p-CMe3), 33.0 (s, o-C(CH3)3),
30.1 (s, p-C(CH3)3).
NMR-Scale Preparation of Intermediate 5 (M ) Al).
To an NMR tube charged with Mes*PdCH2 (1b; 25.0 mg, 0.09
mmol) in CD2Cl2 (0.5 mL) and cooled to 163 K, a solution of
AlCl3 (12 mg, 0.09 mmol) in CD2Cl2 (0.5 mL) was added by
syringe. The contents of the NMR tube were carefully shaken
to ensure rapid mixing and were kept at 193 K. Then the NMR
tube was introduced into the NMR instrument equipped with
a low-temperature unit, and the experiments were run at 193
K.
2
(CDCl3, 300 K): δ 7.60 (s, 2H, m-H), 7.02 (dd, JPH ) 33 Hz,
2
2
2JHH ) 5 Hz, 1H, PdCH), 6.73 (dd, JPH ) 28, JHH ) 5 Hz,
1H, PdCH), 1.60 (s, 18H, o-C(CH3)3), 1.30 (s, 9H, p-C(CH3)3).
MS (EI): m/z [%] 290 [9, M+], 275 [13, M+ - CH3]. Anal. Calcd
for C19H31P: C, 78.58; H, 10.75. Found: C, 78.68; H, 10.77.
Preparation of C-H Activation Product 3 (M ) Ga).
A solution of GaCl3 (0.303 g, 1.72 mmol) in CH2Cl2 (20 mL)
was added dropwise to a stirred solution of 1b (0.500 g, 1.72
mmol) in CH2Cl2 (40 mL) at 20 °C. 31P NMR analysis showed
quantitative conversion of 1b to a new product. The reaction
mixture was stirred for several hours, and the solvent was
removed in vacuo to give a colorless powder. The product was
recrystallized by slow evaporation of a concentrated CH3CN
solution to yield crystals suitable for X-ray diffraction. Yield:
0.650 g (81%).
31P NMR (CD2Cl2, 193 K): δ 220.5. H NMR (CD2Cl2, 193
1
K): δ 7.5 (d, 4JPH ) 3 Hz, 2H, m-H), 7.4 (dd, 2JHH ) 7 Hz, 2JPH
) 13 Hz, 1H, PdCH), 7.0 (dd, 2JHH ) 8 Hz, 2JPH ) 32 Hz, 1H,
PdCH), 1.5 (s, 18H, o-C(CH3)3), 1.3 (s, 9H, p-C(CH3)3). 13C
NMR (CD2Cl2, 193 K): δ 155.6 (s, o-Ar), 153.6 (s, p-Ar), 152.0
(br, PdCH2), 124.1 (s, m-Ar) (i-Ar not observed), 37.7 (s,
o-CMe3), 34.9 (s, p-CMe3), 33.4 (s, o-C(CH3)3), 30.5 (s, p-
C(CH3)3).
1
31P NMR (CDCl3, 300 K): δ 16.4 ppm (d, JPH ) 498 Hz).
4
1H NMR (CDCl3, 300 K): δ 7.60 (d, JHP ) 6 Hz, 1H, m-H),
Preparation of Adduct 6. To a stirred solution of 1a (0.490
g, 1.55 mmol) in 20 mL of CH2Cl2 at room temperature was
added dropwise GaCl3 (0.290 g, 1.7 mmol) in 10 mL of CH2-
Cl2. The mixture was checked by NMR (31P: 132 ppm). The
product is stable in CH2Cl2 at room temperature indefinitely.
The product was recrystallized from CH2Cl2/pentane solution
to obtain single crystals suitable for X-ray diffraction. Yield:
0.65 g (85%).
1
7.4 (dm, JHP ) 498 Hz, 1H, PH), 7.28 (s, 1H, m-H), 2.84 (dd,
3
2JHP ) 11 Hz, JHH ) 5 Hz, 2H, PCH2C(CH3)2), 1.93 (m, 2H,
PCH2Ga), 1.51 (s, 6H, C(CH3)2), 1.48 (s, 9H, o-C(CH3)3), 1.34
(s, 9H, p-C(CH3)3). Anal. Calcd for C19H31Cl3GaP: C, 48.92;
H, 6.70; Cl, 22.80. Found: C, 49.09; H, 6.90; Cl, 22.83.
Preparation of C-H Activation Product 3 (M ) Al).
Under an inert atmosphere, AlCl3 (0.046 g, 0.34 mmol) in CH2-
Cl2 (5 mL) was added dropwise to a stirred solution of 1b (0.100
g, 0.34 mmol) in CH2Cl2 (10 mL) at 25 °C. After the mixture
was stirred for 1/2 h, the solvent was removed in vacuo, yielding
a pale yellow powder. The product was recrystallized from CH3-
CN to obtain colorless crystals. The crystals were further
washed with hexanes. Yield: 0.110 g (76%). This compound
is not stable and decomposes after ca. 1 week in the solid state
under N2 (see text for a discussion of the decomposition
product).
31P NMR (CDCl3, 300 K): δ 132 (s). H NMR (CDCl3, 300
1
4
K): δ 7.03-7.64 (m, 10H, aryl H), 6.96 (d, JPH ) 3 Hz, 2H,
Mes H), 2.53 (d, 4JPH ) 2 Hz, 6H, o-CH3), 2.28 (s, 3H, p-CH3).
13C NMR (CDCl3, 300 K): δ 193.0 (d, JPC ) 36 Hz, PdC),
1
143.5-121.9 (m, aryl C), 23.4 (s, o-CH3), 21.5 (s, p-CH3).
Preparation of C-H Activation Product 7. Freshly
distilled HOTf (0.284 g, 1.89 mmol) was added to a stirred
solution of 1b (0.500 g, 1.72 mmol) in CH2Cl2 (20 mL). The
reaction mixture was stirred for ca. 30 min, and the solvent
was removed in vacuo. The white solid obtained was recrystal-
lized from a minimum amount of CH3CN/toluene (3:1) at -40
°C. Yield: 0.45 g (59%).
(58) Cowley, A. H.; Norman, N. C.; Pakulski, M. In Inorganic
Synthesis; Ginsberg, A. P., Ed.; Wiley: New York; Vol. 27, p 235.
(59) Becker, G.; Uhl, W.; Wessely, H.-J. Z. Anorg. Allg. Chem. 1981,
479, 41.
1
31P NMR (CDCl3, 300 K): δ 13.6 ppm (d, JPH ) 539 Hz).
1H NMR (CDCl3, 300 K): δ 7.84 (dqdd, 1JHP ) 538 Hz, 3JHH
)
(60) Armarego, W. L. F.; Perrin, D. D. Purification of Laboratory
Chemicals, 4th ed.; Butterworth-Heinemann: Oxford, U.K., 1996. To
dry KOH we follow the procedure for drying NaOH on p 429.
(61) Appel’s procedure for 1b uses dry, powdered KOH; however, a
drying procedure was not provided. We have found that anhydrous,
powdered KOH (recrystallization from EtOH and subsequent heating)
contains a trace of KOEt impurity. The presence of this impurity led
to the formation of Mes*P(OEt)CH3, which was difficult to separate
from 1b. We have found that finely ground “wet” KOH pellets yield
1b free from impurities, although longer reaction times are necessary
(several days rather than several hours). Characterization data for
Mes*P(OEt)CH3 are as follows. 31P{1H} NMR: δ 110 (s). 1H NMR: δ
5 Hz (to CH3), 3JHH ) 5 and 2 Hz (to CH2), 1H, P-H), 7.55 (d,
4JHP ) 6 Hz, 1H, m-H), 7.20 (s, 1H, m-H), 2.95 (dm, 2JHP ) 16
Hz, 1H, 1H in PCH2), 2.68 (dm, 2JHP ) 14 Hz, 1H, 1H in PCH2),
2
3
2.30 (dd, JHP ) 15 Hz, JHH ) 5 Hz, 3H, CH3), 1.49 (s, 6H,
C(CH3)2), 1.42 (s, 9H, o-tBu), 1.29 (s, 9H, p-tBu). 19F NMR
(CDCl3, 300 K): δ -78.9 ppm (s, OTf).
Preparation of Linear Dimer 8. To a stirred solution of
1b (1.50 g, 5.16 mmol) in a minimal amount of CH2Cl2 (5 mL)
at -40 °C was added a saturated solution of GaCl3 (0.45 g,
2.58 mmol) in CH2Cl2 (ca. 2 mL), and the mixture was warmed
to room temperature over ca. 1 h. 31P NMR analysis of the
4
7.58 (d, JPH ) 4 Hz, 2H, m-Ar), 3.70 (mult, 2H, POCH2CH3), 1.68 (s,
18H, o-tBu), 1.40 (d, 2JPH ) 9 Hz, 3H, PCH3), 1.35 (s, 9H, p-tBu), 1.10
(t, POCH2CH3).

5922 Organometallics, Vol. 23, No. 25, 2004
Table 1. Details of Crystal Structure Determinations of 1b, 3 (M ) Ga), 6, 7, and 8
Tsang et al.
1b
3 (M ) Ga)
6
7
8
formula
fw
cryst syst
space group
color
a (Å)
b (Å)
C
19H31
P
C19H31PCl3Ga
466.51
triclinic
P1h
colorless
10.0454(6)
11.0240(7)
11.1410(9)
94.935(4)
112.222(3)
94.093(2)
1130.6(1)
2
C22H21PCl3Ga
492.46
triclinic
P1h
yellow
9.882(1)
10.263(2)
12.204(2)
108.054(5)
94.760(3)
103.022(4)
1130.8(2)
2
C20H32O3F3PS
440.50
monoclinic
P21/n
colorless
15.515(1)
9.5822(5)
15.143(1)
90
92.118(4)
90
C38H62P2GaCl3
756.93
triclinic
P1h
colorless
9.8363(4)
11.1839(5)
18.9750(8)
87.698(6)
81.603(5)
85.666(5)
2058.2(2)
2
290.43
orthorhombic
Pnma
colorless
18.865(3)
15.809(1)
6.1010(4)
90
90
90
1820(1)
4
c (Å)
R (deg)
â (deg)
γ (deg)
V (Å3)
2249.8(2)
4
Z
T (K)
173(1)
1.42
173(1)
16.41
173(1)
16.46
173(1)
2.57
173(1)
9.64
µ(Mo KR) (cm-1
cryst size (mm)
)
0.50 × 0.30 × 0.20
0.25 × 0.10 × 0.05
0.23 × 0.12 × 0.08
0.40 × 0.25 × 0.10
0.20 × 0.20 × 0.10
calcd density (Mg m-3
2θ(max) (deg)
no. of rflns
no. of unique data
R(int)
)
1.060
1.370
1.446
1.300
1.22
50.1
13 855
1894
55.8
10 364
4599
0.043
55.8
9381
4618
0.039
55.8
20 436
5130
0.062
55.8
19 332
8464
0.044
0.062
rfln/param ratio
16.06
0.043; I > 3σ(I)
0.110
19.32
0.037; I > 2σ(I)
0.092
18.92
0.037; I > 3σ(I)
0.093
17.20
0.050; I > 2σ(I)
0.136
16.50
0.039; I > 2σ(I)
0.098
R1a
wR2 (all data)b
GOF
1.20
1.07
1.25
0.95
0.95
2
2
a R1 ) ∑||Fo| - |Fc||/∑|Fo|. b wR2 ) [∑(Fo - Fc )2/∑w(Fo2)2]1/2
.
reaction mixture showed signals for 1b, 3 (M ) Ga), and a
new product with doublets at -72.3 (1JPP ) 245 Hz) and -92.6
(1JPP ) 245 Hz) ppm (a 31P-31P COSY experiment con-
firmed the coupling). GaCl3 (ca. 0.6 mmol total in several
steps) in CH2Cl2 was added, and the reaction was moni-
tored by 31P NMR, until 1b was entirely converted to 3
(M ) Ga). The solvent was removed in vacuo, and the colorless
solid residue was extracted with hexanes (ca. 2 × 30 mL) to
remove 3 (M ) Ga). During this process the signals at -72.3
and -92.6 ppm slowly converted to compound 8. Yield: 0.30
g (15%).
Preparation of the Adduct 14 and the Diphosphira-
nium Ring 15. To a solution of 1a (0.25 g, 0.79 mmol) in CH2-
Cl2 (5 mL) was added dropwise HOTf (0.059 g, 0.39 mmol) in
CH2Cl2 (5 mL) over ca. 10 min at room temperature. 31P NMR
(CDCl3, 300 K) showed the clean formation of two major
products from this reaction in an approximate 1:1 ratio. The
solvent was removed in vacuo, and attempts to obtain crystals
from the oily red residue using CH2Cl2/hexanes, CH3CN, and
1,2-Cl2C2H4 have thus far proven unsuccessful.
1
Assignments are tentative. 31P NMR: 14, δ 196.4 (d, JPP
) 509 Hz), 130.5 (d, 1JPP ) 509 Hz); 15, δ 120.3 (d, 1JPP ) 516
Hz), 0.2 (d, 1JPP ) 516 Hz). 1H NMR (CDCl3, 300 K): two types
of CH moiety corresponding to 14, δ 5.35 ppm (d, 2JPH ) 16.8
1
31P NMR (CDCl3, 300 K): isomer 1, δ 3.5 (dd, JHP ) 490,
2JPP ) 95 Hz, P(1)), -9.4 (d, 2JPP ) 95 Hz, P(2)); isomer 2, 1.6
(dd, 1JHP ) 490 Hz, 2JPP ) 90 Hz, P(1)), -8.9 (d, 2JPP ) 90 Hz,
3
2
Hz, JPH was not observed); 15, δ 5.50 ppm (dd, JPH ) 14.6
1
3
1
Hz, JPH ) 4.1 Hz). H{31P} selective decoupling experiments
suggest that in 14 the 16.8 Hz coupling is to the 31P signal at
130.5 ppm, while the same experiment shows that in 15 the
14.6 Hz coupling is to the 31P signal at 0.2 ppm.
P(2)). H NMR (CDCl3, 300 K): δ 7.59, 7.45 (isomer 1/2, dm,
1JPH ) 490 Hz, 1H, PH), 7.63-7.52 (br, 1H, m-H), 7.26 (d, 4JPH
4
) 5 Hz, 1H, m-H), 6.97 (d, JPH ) 5 Hz, 2H, m-H), 2.66 (dm,
2
2
2JPH ) 13 Hz, 2H, PCH2(CH3)2), 2.37 (dd, JPH ) 37 Hz, JPH
) 15 Hz, 2H, PCH2P), 1.98 (m, 2H, PCH2GaCl3), 1.65, 1.63
(isomer 1/2, s, 6H, C(CH3)2), 1.48 (s, 9H, o-C(CH3)3), 1.45 (s,
9H, o-C(CH3)3), 1.35 (s, 9H, p-C(CH3)3), 1.34 (s, 9H, p-C(CH3)3),
1.27 (s, 18H, p-C(CH3)3).
Oligomerization of Mes*PdCH2 Initiated by HOTf. To
a stirred solution of 1b (1.0 g, 3.44 mmol) in a minimal amount
of CH2Cl2 (3 mL) was added HOTf (0.20 g, 1.33 mmol) in CH2-
Cl2 (ca. 1 mL) at room temperature. The 31P NMR spectrum
of the reaction mixture showed signals for 1b, 7, and several
complex multiplets in the range of -50 to 50 ppm with P-P
coupling (from 31P-31P COSY). Analysis of this mixture using
electrospray mass spectrometry revealed that the product
mixture contains oligomers (Mes*PCH2)nH+OTf- (n ) 2-6).
ES MS (m/z): 581.7 [C38H63P2]+, 871.8 [C57H94P3]+, 1161.9
[C76H125P4]+, 1451.9 [C95H156P5]+, 1741.9 [C114H187P6]+. The ions
listed are the data for the all 12C oligomer. Ions that we have
not been able to account for were observed 46 m/z units below
each ion.
Synthesis of [(Et2N)2P]GaCl4 (12).56 To neat (Et2N)2PCl
(20 mg, 0.095 mmol) at -78 °C was slowly added GaCl3 (17
mg, 0.096 mmol) in CD2Cl2 (0.2 mL). The compound was not
isolated, but the 31P NMR spectrum showed that the (Et2N)2-
PCl was consumed and 12 was formed quantitatively.
31P NMR (CD2Cl2, 300 K): δ 245 ppm (s). 1H NMR (CD2Cl2,
300 K): δ 3.67 (m, 8H, (CH3CH2)2N)2P+), 1.36 (t, 3JHH ) 7 Hz,
12H, (CH3CH2)2N)2P+). 13C NMR (CD2Cl2, 300 K): δ 46.5 (d,
3
2JPC ) 58 Hz, (CH3CH2)2N)2P+), 14.8 (d, JPC ) 13 Hz, (CH3-
CH2)2N)2P+).
Preparation of the Diphosphiranium 13. To the freshly
prepared solution of 12 (see above) in CD2Cl2 at -78 °C in an
NMR tube was slowly added a solution of 1a (0.028 g, 0.096
mmol) in CD2Cl2 (0.3 mL). NMR analysis of the reaction
mixture showed partial, but clean, conversion (ca. 30%) to 13.
31P NMR (CD2Cl2, 293 K): δ 17.7 (d, 1JPP ) 225 Hz), -81.6
X-ray Crystallography. All single crystals were immersed
in Paratone-N oil and were mounted on a glass fiber. Data
were collected at 293 K on a Rigaku/ADSC CCD area detector
with graphite-monochromated Mo KR radiation. Data were
collected and processed using the d*TREK program. All
structures were solved by direct methods and subsequent
Fourier difference techniques and refined anisotropically for
all non-hydrogen atoms by full-matrix least-squares calcula-
tions on F2 using the Siemens SHELXTL97 version. Crystal
data and details of the data collection and structure
1
(d, JPP ) 225 Hz). 31P-1H HSQC experiments show strong
cross signals between both the 31P NMR signals and a broad
signal at 2.2-2.3 ppm in the 1H NMR spectrum. This was
assigned to the ring CH2 moiety.

Addition of Electrophiles to Phosphaalkenes
Organometallics, Vol. 23, No. 25, 2004 5923
refinement are given in Table 1. Further details are included
in the Supporting Information.
Burlinson for assistance with some of the NMR experi-
ments.
Acknowledgment. We thank The Natural Sciences
and Engineering Research Council (NSERC) of Canada,
The Canada Foundation for Innovation (CFI), BC
Knowledge and Development Fund, and the BC
Ministry for Advanced Education and Technology for
financial support. We gratefully acknowledge Dr. Nick
Supporting Information Available: Full details of the
crystal structure investigations are available as CIF files. This
material is available free of charge via the Internet at
http://pubs.acs.org.
OM0495022