Phosphaalkenes with inverse electron density: Electrochemistry, electron paramagnetic resonance spectra, and density functional theory calculations of aminophosphaalkene derivatives

Article abstract of DOI:10.1021/jp030023a

Cyclic voltammetry of Mes*P=C(NMe₂)₂ (1) and Mes*P=C(CH₃)NMe₂ (2) shows that, in solution in DME, these compounds are reversibly oxidized at 395 and 553 mV, respectively. Electrochemical oxidation or reaction of 1 (or 2) with [Cp₂Fe]PF₆ leads to the formation of the corresponding radical cation, which was characterized by its electron paramagnetic resonance (EPR) spectra. Experimental ³¹P and ¹³C isotropic and anisotropic coupling constants agree with density functional theory (DFT) calculations showing that the unpaired electron is strongly localized on the phosphorus atom, in accord with the description Mes*P-(C(NMe₂)₂)⁺. Electrochemical reduction of 1 is essentially irreversible and leads to a radical species largely delocalized on the C(NMe₂)₂ moiety; this neutral radical results from the protonation of the phosphorus atom and corresponds to Mes*(H)P-C(NMe₂)₂. No paramagnetic species is obtained by reduction of 2. The presence of the amino groups, responsible for the inverted electron distribution at the P-C double bond (P^--C⁺), confers on 1 and 2 redox properties that are in very sharp contrast with those observed for phosphaalkenes with a normal π electron distribution (P⁺-C^-): no detection of the radical anion but easy formation of a rather persistent radical cation. For 1, this radical cation could even be isolated as a powder, 1⁺PF₆^-. As shown by DFT calculations, this behavior is consistent with the decrease of the double bond character of the phosphorus-carbon bond caused by the presence of the amino groups.

Full text of DOI:10.1021/jp030023a

J. Phys. Chem. A 2003, 107, 4883-4892
4883
Phosphaalkenes with Inverse Electron Density: Electrochemistry, Electron
Paramagnetic Resonance Spectra, and Density Functional Theory Calculations of
Aminophosphaalkene Derivatives
Patrick Rosa,*,† Cyril Gouverd,† Ge´rald Bernardinelli,‡ The´o Berclaz,† and Michel Geoffroy*,†
Department of Physical Chemistry, 30 Quai Ernest Ansermet, and Laboratory of Crystallography,
24 Quai Ernest Ansermet, UniVersity of GeneVa, 1211 GeneVa 4, Switzerland
ReceiVed: January 9, 2003
Cyclic voltammetry of Mes*PdC(NMe2)2 (1) and Mes*PdC(CH3)NMe2 (2) shows that, in solution in DME,
these compounds are reversibly oxidized at 395 and 553 mV, respectively. Electrochemical oxidation or
reaction of 1 (or 2) with [Cp2Fe]PF6 leads to the formation of the corresponding radical cation, which was
characterized by its electron paramagnetic resonance (EPR) spectra. Experimental 31P and 13C isotropic and
anisotropic coupling constants agree with density functional theory (DFT) calculations showing that the unpaired
electron is strongly localized on the phosphorus atom, in accord with the description Mes*P•-(C(NMe2)2)+.
Electrochemical reduction of 1 is essentially irreversible and leads to a radical species largely delocalized on
the C(NMe2)2 moiety; this neutral radical results from the protonation of the phosphorus atom and corresponds
to Mes*(H)P-•C(NMe2)2. No paramagnetic species is obtained by reduction of 2. The presence of the amino
groups, responsible for the inverted electron distribution at the P-C double bond (P--C+), confers on 1 and
2 redox properties that are in very sharp contrast with those observed for phosphaalkenes with a normal π
electron distribution (P+-C-): no detection of the radical anion but easy formation of a rather persistent
-
radical cation. For 1, this radical cation could even be isolated as a powder, 1•+PF6 . As shown by DFT
calculations, this behavior is consistent with the decrease of the double bond character of the phosphorus-
carbon bond caused by the presence of the amino groups.
1. Introduction
oxidation wave at a potential value (-0.08 V vs SCE) that is
appreciably inferior to those reported by Schoeller et al. for
Phosphaalkenes have raised a great deal of interest in the
last two decades for the similarities found between the PdC
and the CdC bonds.1,2 Moreover, the presence of a low lying
π* orbital confers to the PdC bond interesting electron acceptor
properties that are easily demonstrated by the easy formation
of radical anions and their detection by electron paramagnetic
resonance (EPR) spectroscopy.3 Within this kind of compound,
an especially interesting class was observed as early as 1986
by Regitz4 and by Grobe:5 whereas “common” phosphaalkenes
are polarized in the sense P+-C-, phosphatriafulvenes and
C-amino substituted phosphaalkenes6-8 exhibit an inverted
electron density.9 The corresponding mesomeric structures a and
b suggest that the electron transfer behavior of such compounds
typical phosphaalkenes (1.07-2.94 V).11 We decided, therefore,
to study the cyclic voltammetry of two phosphaalkenes present-
ing all the properties associated with an inverse electron density,
and to use EPR to identify their oxidation and reduction
products. Two C-amino substituted phosphaalkenes were chosen
for this purpose: the novel compound Mes*PdC(NMe2)2 (1)
and the phosphaalkene 2 whose synthesis has already been
reported.12
The experimental hyperfine couplings are interpreted in terms
of electronic configurations that are rationalized by density
functional theory (DFT) calculations. It will be shown that the
persistence of the radical species formed from these two
compounds drastically differs from that observed with “classi-
cal” phosphaalkenes and that it is even possible, with 1, to isolate
the corresponding radical cation.
differs from that of “classical” phosphaalkenes and that both
the formation and the electronic structure of the corresponding
radical ions should be appreciably affected by this particular
polarization.
As far as we know, no extensive study of the redox properties
of these polarity-reversed phosphaalkenes has been made so far;
the only indication to date is phenyl(1,3-dimesitylimidazolin-
2-ylidene)phosphane,10 which shows an irreversible one-electron
2. Results
2.1. Synthesis and Chemical Properties. 2.1.a. Phospha-
alkene 1. Phosphaalkene 1 was synthesized according to the
method reported by Oehme et al.13 starting from easily available
† Department of Physical Chemistry.
‡ Laboratory of Crystallography.
* Corresponding author.
10.1021/jp030023a CCC: $25.00 © 2003 American Chemical Society
Published on Web 05/23/2003

4884 J. Phys. Chem. A, Vol. 107, No. 24, 2003
Rosa et al.
SCHEME 1
SCHEME 2
Figure 1. Cyclic voltammetry of the oxidation of a solution of 1 (‚‚‚)
in DME and of a solution of 2 in THF (s).
and relatively innocuous Mes*PH2 and bis(dimethylamino)-
ethoxycarbenium tetrafluoroborate (Scheme 1).
Using the corresponding isotopically enriched ureas led to
compounds 1-d12 and 1-13C,d12.
Reaction of Protonation. A solution of 1 in THF treated with
gaseous HCl turned quickly colorless. 31P NMR showed the
unequivocal formation of a new compound at δ ) -60.1 ppm.
Removal of the solvent yielded, quantitatively, a white powder
that could be identified as the phosphinocarbenium chloride 1b
1
(Scheme 2). Protonation at phosphorus was visible in the H
NMR spectrum, with a resonance at δ ) 6.20 ppm (1JP-H
)
261.60 Hz). This proton exchanges quickly with deuterium in
the NMR time scale in acetone-d3. Monoprotonation was
evidenced by the detection of the corresponding molecular ion
by mass spectroscopy. 1H and 13C resonances of the two amino
groups coalesce and are indistinguishable at room temperature.
This shows a stronger delocalization of the nitrogen pairs over
the C(NMe2)2 moiety compared to the parent compound and
indicates that, in the protonated compound the rotation around
the phosphorus-carbon bond is free. This agrees with proto-
nation at phosphorus and disappearance of the double bond
character of the phosphorus-carbon bond. These results are
reminiscent of a previous report on RC(O)PdC(NMe2)2.17
1 is a yellow powder, easily handled in air, that is purified
without any special precaution by silica gel chromatography.
This stability is in marked contrast with usual phosphaalkenes
that are sensitive toward oxygen and humidity. Unfortunately,
crystals of X-ray diffraction quality could not be obtained.
The following NMR data confirm that b represents a
mesomeric form for 1 and are in accordance with values
previously reported for other phosphaalkenes with reverse
polarity. The 31P NMR of 1 shows a resonance at δ ) 28.3
ppm, well upfield of usual phosphaalkenes (100-400 ppm),1,2,14
consistent with an increased electron density at the phosphorus
2.1.b. Phosphaalkene 2. Phosphaalkene 2 was synthesized
according to the literature12,18,19 starting from Mes*PH2 and N,N-
dimethylacetamide dimethyl acetal and was purified by silica
gel chromatography.20 The 31P NMR shows a resonance at δ
) 94.3 ppm, closer to the values for “classical” phosphaalkenes,
thus showing a decreased electron density at phosphorus
compared with bis(amino) substituted 1. The 13C NMR reso-
atom. The 13C NMR resonance at δ ) 191.05 ppm (1JP-C
)
76.90 Hz) is typical of a sp2-hybridized carbon in a PdC bond,
while being quite downfield and in agreement with a strong
1
cationic character. Last, both H and 13C NMR resonances of
the dimethylamino groups distinguish cis and trans orientations
relative to the phosphorus lone pair, as expected for a PdC
bond, but do not distinguish each methyl group attached to the
nitrogen atoms, showing thus that for 1 free rotation around
the C-N bonds occurs at room temperature.
As shown below, the reactivity of 1 (Scheme 2) is consistent
with previous reports on inverse polarized phosphaalkenes and
agrees with the expected electronic properties of this compound:
Reaction with Oxygen. When a solution of 1 in ether was
left standing in contact with air, a very small quantity of white
crystals was obtained after several weeks. The crystal structure
of this compound was solved and showed that this oxidation
product was the zwitterionic amidinophosphinate 1a.15 This
reaction (see Scheme 2) is similar to the reaction reported by
Schmidpeter et al. for [bis(dimethylamino)methylene]phen-
ylphosphane16 with ozone.
nance of phosphaalkene carbon at δ ) 189.2 ppm (1JP-C
)
62.1 Hz) concurs with this conclusion, showing a stronger Pd
C bond.
2.2. Redox Properties. 2.2.a. Cyclic Voltammetry. The
voltammogram reported in Figure 1a shows that the oxidation
of a solution of phosphaalkene 1 in DME occurs at E°1/2
)
+395 mV (∆Ep ) 218 mV). A THF solution yields similar
results (E°1/2 ) +315 mV, ∆Ep ) 150 mV). Variations of the
current intensity with the scan rate indicate that this oxidation
wave is reversible.
The reductive behavior of 1 is much less clear-cut. A
nonreversible weak wave can be seen at Ep = -800 mV.
Increasing the scan rate upward of 1 V/s reveals a return anodic

Phosphaalkenes with Inverse Electron Density
J. Phys. Chem. A, Vol. 107, No. 24, 2003 4885
SCHEME 3
Figure 3. Frozen solution EPR spectra (120 K) obtained by electro-
chemical oxidation of 1.
TABLE 1: Experimental EPR Parameters for the Oxidation
Products of 1 and 2
liquid solution
frozen solution
31
precursor giso
A
( P) Aiso(13C)
g
31P-coupling (MHz)
iso
1
2.0077 296.2
23.8 gi ) 2.0081 Ti ) 1
gj ) 2.0136 Tj ) 31
τi ) -288
τj ) -258
Figure 2. EPR spectra obtained at room temperature by electrochemical
oxidation of (a) a solution of 1 in THF and (b) a solution of 1-13C,d12
in THF.
gk ) 2.0032 Tk ) 835 τk ) 546
gav ) 2.0083 Aav ) 289
2
2.0083 290.3
a
gi ) 2.0114 Ti ) 0
gj ) 2.0109 Tj ) 46
τi ) -286
τj ) -240
gk ) 2.0035 Tk ) 813 τk ) 527
gav ) 2.0086 Aav ) 286
wave with an estimated E°1/2 value equal to -650 mV. This
points to chemical irreversibility at low scan rates, very likely
due to a protonation reaction of the radical anion. In this context,
we have therefore carried out a voltammetric study of the
phosphinocarbenium chloride 1b. Two irreversible reduction
waves were detected at Ep) -610 mV and Ep) -970 mV vs
SCE.
Cyclic voltammetry (Figure 1b) indicates that a solution of
2 in THF undergoes an oxidation at E°1/2 ) +525 mV (∆Ep)
205 mV) (in DME: E°1/2 ) +553 mV, ∆Ep) 90 mV). As
shown by the variation of the current intensity with the scan
rate, these oxidation waves are reversible. No reduction wave
could be detected with 2.
a The 13C-enriched compound was not synthesized.
any EPR signal (vide infra). Oxidation with [Cp2Fe]PF6 is
accompanied by a drastic change in the color of the solution,
which progressively turns to deep red. The formation of the
cation is detected by EPR (vide infra).
2.3. EPR Spectroscopy. 2.3.a. Oxidation Products. At room
temperature, electrochemical oxidation of a solution of 1 in THF,
in situ in the EPR cavity, gives rise to the EPR spectrum shown
in Figure 2. The same spectrum is obtained by directly oxidizing
a THF solution of 1 with AgClO4 in the EPR tube or by
dissolving solid 1•+PF6 in carefully dried THF. In all cases,
-
2.2.b. Chemical Oxidation. Chemical Oxidation of 1. The easy
electrochemical oxidation of 1 prompted us to attempt the
synthesis of the corresponding radical cation by chemical
oxidation. Adding 1 equiv of AgClO4 in THF or [Cp2Fe]PF6 in
acetonitrile to 1 results in the immediate formation of a deep
violet solution, very sensitive to air. As shown by EPR (vide
infra) the radical cation 1•+ is indeed formed in this reaction in
accord with Scheme 3 and we could isolate the resulting salt.
The violet solid obtained after removal of the solvent can be
kept for a few weeks in a glovebox if synthesized starting from
[Cp2Fe]PF6. Unfortunately, we could not grow crystals suitable
for X-ray diffraction. Stored under an inert atmosphere, a
the color of the solution leading to the EPR spectrum is violet.
This spectrum consists of a doublet with a large splitting of
ca. 295 MHz, which is attributed to hyperfine coupling with
the 31P nucleus. The peak-to-peak line width is large, 10.5 G
(1.05 mT), and points to other nonresolved hyperfine interac-
tions. No significant shape alteration is seen upon cooling the
solution till freezing temperature is reached. The frozen solution
spectrum was recorded at 120 K and is shown in Figure 3a. As
shown in Figure 3b, it was satisfactorily simulated by using a
least-squares iterative procedure;21 the resulting g and 31P
hyperfine tensors are found to be oriented parallel to each other.
Assuming all the hyperfine components to be positive leads to
an isotropic coupling constant in agreement with the isotropic
constant measured in liquid solution.
solution of 1•+PF6 will lose its color after a few days. The
-
known oxidizing character of the perchlorate counteranion can
account for the lesser stability of the compound when AgClO4
is used in the synthesis.
The experimental EPR parameters are reported in Table 1.
Chemical Oxidation of 2. A solution of 2 in THF turns to
pale yellow by addition of AgClO4 but does not give rise to
The EPR spectrum obtained in liquid solution after oxidation
of the 2H enriched phosphaalkene 1-d12 is similar to the spectrum

4886 J. Phys. Chem. A, Vol. 107, No. 24, 2003
Rosa et al.
to investigate the effects of water on the spectra. Addition of 1
equiv of water resulted first in the disappearance of the signal;
however, when electrolysis was resumed, the same spectrum
was again observed with an increased intensity. Although no
coupling with an additional proton is observed, this result
suggests that the reduction products whose EPR spectra are
shown in Figure 4 probably result from the protonation of the
radical anion 1•- and correspond therefore to 1H•.
Figure 4. EPR spectra obtained by electrochemical reduction of a
solution of (a) 1 in THF, (b) 1-d12 in THF, and (c) 1-13C,d12 in THF.
Another mechanism can be invoked for the formation of this
reduction compound: the reduction of trace amounts of phos-
phinocarbenium 1b formed in a “wet” solution of phosphaalkene
1. To check this, we tried therefore to reduce phosphinocarben-
ium chloride 1b, directly in the EPR cavity. Because the cyclic
voltammogram of 1b exhibits two close irreversible reduction
waves, it was necessary to use a three-electrode EPR cell, which
enables control of the applied potential. When a potential close
to the first observed peak potential was applied, no EPR signal
was observed. Decreasing the potential further by 300 mV did
not give further results. It seems therefore that the EPR spectra
result from the protonation of 1•-. No EPR spectra could be
obtained by reduction of 2.
TABLE 2: Experimental EPR Parameters Obtained after
Reduction of a Solution of 1 in THF
isotropic hyperfine couplings (MHz)
precursor
g
31P
1H
2H
14N
13C
1
2.0033 53.8 (×1) 6.17 (×12)
12.1 (×2)
1-13C,d12 2.0033 53.8 (×1)
0.95 (×12) 12.1 (×2) 90.5 (×1)
obtained with the nondeuterated compound, the single difference
lies in a decrease of the line width to 8.7 G (0.87 mT), thus
supporting the existence of non resolved hyperfine couplings
with the dimethylamino protons. Nevertheless, the line width
remains quite large and implies some interaction with the two
14N nuclei and/or some protons of the Mes* group. To measure
the hyperfine coupling with the phosphaalkene 13C, spectra were
recorded after oxidation of 1-13C,d12. As shown in Figure 2b,
on the resulting spectrum the 13C interaction is revealed by
shoulders only, in line with a coupling of 24 MHz. The 13C
enrichment does not cause significant modifications of the frozen
solution.
2.4. DFT Calculations. With the aim of rationalizing the
redox behavior of phosphaalkenes containing NR2 groups bound
to the phosphaalkene carbon atom, we have carried out DFT
calculations on neutral RPdCRa′Rb′ model systems (M) as well
No EPR spectra could be obtained by oxidation of 2 with
AgClO4. However, oxidation of a solution of 2 in acetonitrile
with [Cp2Fe]PF6 or electrochemical oxidation of a solution of
2 in THF led to spectra that could be recorded in both liquid
and frozen solutions. The corresponding parameters, obtained
after simulation, are shown in Table 1.
2.3.b. Reduction Products. Whereas no EPR signal could be
obtained by reduction of 1 over a potassium mirror or by
reaction with a solution of sodium naphthalenide in THF,
interesting spectra could be recorded when electrochemical
reduction was carried out using the following conditions: (a)
separation of the anodic and cathodic compartments by a porous
frit to avoid the strong signal due to the diffusion of the radical
cation 1•+ and (b) use of THF that has not been thoroughly
dried. No coloration of the solution is observed at the cathode,
and the signals quickly disappear when the current is turned
off. As shown in Figure 4, the spectrum obtained with 1 (Figure
4a) exhibits a very rich hyperfine structure, which is consider-
ably simplified when the protons of the dimethylamino groups
have been replaced by deuteriums (1-d12, Figure 4b). An
additional drastic modification of the spectrum is observed after
13C enrichment of the phosphaalkene carbon (1-13C,d12, Figure
4c): obviously, this nucleus gives rise to a large coupling
constant of =90 MHz. Simulation22 of the three spectra reported
in Figure 4 could be performed by assuming hyperfine coupling
with one 31P, one13C, and the protons of the four dimethylamino
groups, as shown in Table 2.
as on both radical anion and radical cation derivatives. The
resulting structures will be compared with those recently
reported on neutral [bis(amino)methylidene]phosphanes.7,8
2.4.a. Neutral Phosphaalkenes. As shown in Table 3, the
geometric parameters obtained after optimization (B3LYP/6-
31+G) of the phosphaalkene models M1-M8 are in satisfactory
accord with the rare experimental data reported in the literature.
This table shows a progressive increase in both the experi-
mental and DFT CdP bond lengths when the methylene
hydrogens of phosphaalkenes are substituted by NMe2 groups
(e.g., M1, M2, M4 or M5, M6). This effect reflects the decrease
in the double bond character of the phosphaalkene bond and is
reminiscent of recent results on model molecules whose
phosphaalkene carbon atom was incorporated in an imidazoli-
nylidene structure.7,8 This decrease is illustrated by the conse-
quent diminution of the rotation barrier23 Vrot around the double
bond: CH2dCH2 (Vrot ) 97.2 kcal mol-1), HPdCH2 (M1, Vrot
The fact that the spectra shown in Figure 4 are more easily
detected when the solvent is not thoroughly dried prompted us

Phosphaalkenes with Inverse Electron Density
J. Phys. Chem. A, Vol. 107, No. 24, 2003 4887
TABLE 3: DFT Optimized Geometrical Parameters for Some Neutral Phosphaalkenes (ΣθN ) Sum of the Bond Angles for Na
and Nb; r: Angle between the Normals to the RPC and R′aCR′b Planes)
bond lengths (Å)
bond angles
compound
interplane angle,
sum of
RPdCR′aR′b
PdC
R-P
C-R′a
C-R′b
RPC
PCR′a
PCR′b
119.3
R
bond angles, ΣθN
M1
DFT
exp
1.677
1.673(2)
1.432
1.420(6)
1.089
1.088
97.6
97.4(4)
125.4
124.4(8)
0
0
118.4(12)
M2
DFT
M3
1.744
1.740
1.425
1.430
1.515
1.387
1.371
1.382
94.7
94.6
123.3
127.4
120.6
21.21
13.17
356.6 (Nb)
DFT
120.8
345.7 (Na)
344.1(Nb)
M4
DFT
1.750
1.424
1.392
1.392
96.0
127.1
119.8
13.63
25.12
353.2 (Na)
352.2 (Nb)
355.6(5) (Na)
354.3(6) (Nb)
exp
1.740(1)
1.42(3)
1.348(4)
1.391(4)
103(1)
124.2(3)
121.6(3)
M5
DFT
M6
1.742
1.752
1.862
1.860
1.512
1.384
1.379
1.397
103.6
105.5
124.7
128.4
119.5
118.4
29.6
355.5 (Nb)
DFT
21.12
359.5(Na)
352.4 (Nb)
M7
DFT
1.773
1.890
1.388
1.387
104.3
134.7
120.8
18.23
357.8 (Na)
360 (Nb)
TABLE 4: Adiabatic Vertical Ionization Potentials
Calculated by DFT (Experimental Values Given in
Parentheses) and Oxidation Potentials
atoms are practically coplanar. As shown by Tables 3 and 5,
oxidation increases the P-C bond length, moderately for HPd
CH2 (+0.04 Å) drastically for the amino substituted compounds
(ca. +0.1 Å). Moreover, the distance between the phosphaalkene
carbon and nitrogen atoms are shorter in the cation than in the
neutral molecule. These bond lengths (1.34 Å), shorter than for
a C(sp2)-N(sp2) single bond (1.355 Å), clearly reflect a strong
conjugation in the C(NR2)2 moiety. Nevertheless, as seen in
Table 5, the NR2 planes do not coincide with the NCN plane
and the angle æ formed by the normals to these two planes
increases when NH2 is replaced by NMe2. Replacing NH2 by
NMe2 also causes an increase in the dihedral angle R between
the RPC and NCN planes; this shows that, as for the neutral
molecule,31 nonnegligible steric interactions occur between the
phosphorus and nitrogen substituents. As expected, oxidation
of M2 or M5 hardly decreases the C-CH3 bond length.
Comparison between experimental and calculated hyperfine
interactions is certainly an efficient method for asserting the
identification of a radical. However, in contrast with the dipolar
interaction, the 31P Fermi contact interaction is frequently poorly
calculated by DFT methods32-34 and can be appreciably affected
by the choice of the basis set, specially when spin polarization
effects are likely to occur. Taking some comments of Nguyen
et al.32 into account, we have used three different basis set for
the calculation of the 13C and 31P hyperfine constants for M2•+,
M4•+, M5•+, M6•+. The results are shown in Table 6. They
clearly indicate that, indeed, the IGLO-III35 and TZVP36 basis
sets lead to isotropic 31P constants appreciably larger than those
obtained with 6-31G(d).
compound
IPV (eV)
Ep°x(solvent) (V vs SCE)
C2H4
H2CdCHNH2
10.49 (10.51)27
8.35 (8.2)28
7.91
H2CdC(NH2)2
HPdCH2 (M1)
HPdC(NH2)2 (M3)
HPdC(NMe2)2 (M4)
MePdC(NH2)2
MePdC(NMe2)2
PhPdC(NH2)2
10.53 (10.3)29
7.61
7.09 (7.26)30
7.26
6.72
7.09
PhPdC(NMe2)2
(Me2C6H3)PdC(Me)NMe2
(M5)
6.53
6.85
+0.63 (THF) for 2
(Me2C6H3)PdC(NMe2)2 (M6) 6.52
M7 5.98
+0.39 (THF) for 1
-0.08 (MeCN) for M810
) 68.7 kcal mol-1), to HPdC(NH2)2 (M3, Vrot ) 28.4 kcal
mol-1) and HPdImidazolinylidene (Vrot ) 13.2 kcal mol-1).
This trend in the variation of Vrot agrees with the NMR data
reported above (synthesis of 1): the resonances of the cis- and
trans-dimethylamino groups do not coalesce, whereas this is
the case for the phosphaalkenes with a imidazolinylidene
ring.24,25
Because the formation of a radical cation of a phosphaalkene
constitutes an important aspect of the present study, we have
used DFT to calculate adiabatic vertical ionization potentials
(IPv) for several phosphaalkenes containing amino groups, DFT
calculations yielding correct IPvs for unsaturated molecules.26
These values, together with those that we have calculated for
some ethylene derivatives are given in Table 4 in the presence
of experimental values when available. The agreement between
calculated and experimental values is satisfactory. The expected
trends are well reproduced, with substitution by electron-
releasing amino groups sharply decreasing the IPv: -2.6 eV
for CdC and -2.9 eV for CdP. Alkylamino groups and more
electron-releasing substituents at phosphorus further decrease
the IPv.
2.4.c. Reduced Species. Because the reduction compound
detected by EPR (vide supra) could result either from the radical
anion [RPdC(NR′2)2]•- or from the protonation of this anion,
we have optimized the structure of four species: [HPdC(N-
Me2)2]•- (M4•-) and H2P-•C(NMe2)2 (M4H), [(Me2C6H3)Pd-
C(NMe2)2]•- (M6•-), and (Me2C6H3)HP-•C(NMe2)2 (M6H).
The various hyperfine couplings have been calculated for these
geometries and are given in Table 7.
2.4.b. Oxidized Species. The structures of M1•+, M2•+, M3•+,
M4•+, M5•+, and M6•+ have been optimized and some resulting
geometrical parameters are reported in Table 5. In these radical
cations the phosphaalkene P-C bond as well as the two nitrogen
3. Discussion
As shown in the previous section, two salient features
characterize the electron transfer behavior of phosphalkenes

4888 J. Phys. Chem. A, Vol. 107, No. 24, 2003
Rosa et al.
TABLE 5: DFT Optimized Geometrical Parameters for Some Phosphaalkene Radical Cations (ΣθN ) Sum of the Bond Angles
for Na and Nb; r ) Angle between the Normals to the Planes RPC and R′aCR′b; æ ) Angle between the Normals to the Planes
NaCNb and CNaC (or CNbC))
bond lengths (Å)
bond angles
interplane angle
compound
sum of
RPdCR′aR′b
PdC
R-P
C-R′a
C-R′b
RPC
PCR′a
PCR′b
R
æ
bond angles, ΣθN
M1•+
M2•+
M3•+
M4•+
1.721
1.837
1.864
1.875
1.428
1.420
1.423
1.423
1.095
1.506
1.328
1.343
1.093
1.320
1.330
1.343
99.3
93.2
93.1
94.0
121.3
121.6
123.2
121.4
121.9
119.5
116.4
117.2
36.33
1.9 1.2
0.0 0.0
32.03
1.2
0.0
26.8 (Na)
23.4 (Nb)
1.9
26.9 (Na)
23.9 (Nb)
362.2
359.3 (Na)
359.5 (Nb)
M5•+
M6•+
1.836
1.871
1.807
1.812
1.507
1.346
1.321
1.346
104.7
106.0
120.6
121.9
119.8
117.1
37.44
47.20
359.5 (Na)
352.4 (Nb)
TABLE 6: Calculated (DFT) Hyperfine Couplings (MHz)
for the Radical Cation M2•+, M4•+, M5•+, M6•+
and is consistent with the occupancy of each nitrogen lone pair,
which is substantially less than two electrons (average values:
1.813 for HPdC(NH2)2, 1.739 for HPdC(NMe2)2). These
properties correspond to those reported by Arduengo et al.25
for the structure of carbene-phosphinidene adducts, which point
out that in the adduct between dimesitylimidazolin-2-ylidene
and phenylphosphinidene, the pπ-pπ double bond between these
two moieties is not well developed.
13C coupling
31P coupling
dipole
dipole
Aiso
Aiso
M2•+
6-31+G(d) -21.9 -9.8 3.6 6.3 110.2 510.6 -264.5 -246.1
IGLO-III -25.4 -10.7 3.9 6.8 155.2 545.6 -282.0 -262.7
TZVP
M4•+
-28.4 -10.6 4.0 6.6 187.3 557.4 -288.4 -268.9
Reduction of “common” phosphaalkenes lead to radical
anions whose PdC π* orbital accommodates the extra electron.3
The fact that, in phosphaalkenes with inverse polarization, the
double bond character of the phosphorus-carbon bond dimin-
ishes does not favor, therefore, the formation of such radical
anions and, indeed, chemical or electrochemical reduction of 2
does not lead to any EPR signal. For the reduction product of
1, the spectrum is characterized by 31P and 13C isotropic
couplings (Aiso(31P) ) 58 MHz, Aiso(13C) ) 90.5 MHz), which
notably differ from those observed for the radical anion of
Mes*PdC(H)Ph (Aiso(31P) ) 152 MHz, Aiso(13C)) 16 MHz).3a
As shown in Table 7, the corresponding EPR parameters are
more in accordance with the formation of a carbon-centered
radical ArP(H)-•C(NH2)2 resulting from the protonation of the
radical anion. This interpretation agrees with the fact that traces
of water increase the intensity of the spectrum. However, in
contrast with the DFT predictions for M6H, no hyperfine
interaction with the additional proton is observed on the
spectrum. This can be due to the presence, in 1H•, of the tert-
butyl groups, which force the P-H bond to be oriented almost
perpendicular to the plane containing the phosphaalkene carbon
and the two nitrogen atoms. Because this proton coupling mainly
results from hyperconjugation, it is expected to vary as cos2
ê, where ê is the dihedral angle between the P-H bond
and the magnetic carbon 2pπ orbital. A rapid calculation on
H2P-•C(NMe2)2 shows that, indeed, for ê ) 75°, the coupling
is close to zero.
6-31+G(d) -17.5 -3.2 1.5 1.7 119.3 562.1 -290.7 -271.4
IGLO-III -20.5 -3.5 1.5 2.0 162.4 603.1 -312.5 -290.6
TZVP
M5•+
-22.6 -3.4 1.5 1.9 204.4 615.9 -318.3 -297.6
6-31+G(d) -10.9 -5.4 -0.3 5.7 129.8 421.4 -217.7 -203.7
IGLO-III -15.3 -5.6 -0.6 6.1 159.6 455.6 -236.6 -219.0
TZVP
M6•+
-17.2 -5.7 -0.5 6.1 188.6 464.6 -240.0
6-31+G(d) -9.6
3.4 -1.0 -2.4 137.7 461.7 -238.0 -223.7
3.8 -1.2 -2.6 170.6 500.2 -258.9 -241.2
3.7 -1.1 -2.6 203.1 510.0 -262.8 -247.2
IGLO-III -13.2
TZVP
-14.5
containing one or two amino groups bound to the phosphaalkene
carbon: the oxidation product is easily formed and readily
observed by EPR whereas the reduction product leads to EPR
signals that are either dramatically dependent upon the experi-
mental conditions (e.g., 1) or that remain undetectable (e.g.,
2). These results are in sharp contrast to those previously
reported for phosphaalkenes with a phenyl group bound to the
carbon atom:3 these “common” phosphaalkenes gave no re-
sponse by oxidation but led to intense and well-resolved spectra
by reduction. The origin of these differences in the redox
properties will be examined below.
The first difference between the two types of phosphaalkenes
lies in the double bond character of the phosphorus-carbon
bond. The fact that this character considerably diminishes when
NR2 groups are bound to the carbon atom is clearly seen from
DFT results (Table 3) and, of course, agrees with previous
observations on other phosphalkenes with inverse electron
density:9 the P-C bond length increases with the presence of
amino groups, and this effect is accompanied by a decrease in
the rotation barrier around the phosphorus-carbon bond.
As seen from a population analysis (NBO37), the phosphorus
participation to the P-C π orbital increases from 47.0% in HPd
CH2 to 60.4% in HPdC(NH2)2. This variation reveals a
delocalization of the nitrogen electrons toward the carbon atom
The formation of a one-electron oxidation compound of
phosphaalkenes 1 and 2 is particularly interesting because, as
far as we know, 1•+ is the first dicoordinated phosphorus radical
cation to be isolated as a powder; moreover, radical cations
derived from systems involving PdC bonds could never be
detected, except with phosphaallenes,38 a highly delocalized
system. The hyperfine couplings given in Table 1 show that
the oxidation products of 1 or 2 are characterized by axial 31P
TABLE 7: Hyperfine Couplings (MHz) Calculated for the Anionic and Neutral Species Likely To Result from the Reduction of
Phosphaalkenes Containing Two Amino Substituents (DFT Calculations at the B3LYP/TZVP//B3LYP/6-31+G(d) Level)
14
A
( N)
iso
31
A
iso
( P)
Aiso(13C)
14Na
14Nb
A
iso(1H (CH3))
Aiso(1H (P))
[HPdC(NMe2)2]•- (M4•-
H2P-•C(NMe2)2 (M4H)
)
26.8
69.2
29.0
135.3
99.8
95.0
86.5
-2.2
15.3
0.8
12.4
15.9
10.5
18.7
-0.36 to +5.16
0.22-13.06
-0.11 to +5.77
1.0-9.0
23.7-139.6
[(Me2C6H3)PdC(NMe2)2]•- (M6•-
(Me2C6H3)HP-•C(NMe2)2 (M6H)
)
-28.0
13.7
80.8

Phosphaalkenes with Inverse Electron Density
J. Phys. Chem. A, Vol. 107, No. 24, 2003 4889
Figure 5. Representation of the SOMO for M3•+, M2•+, and M4•+
.
dipolar coupling tensors corresponding to a spin density of 0.75
in a phosphorus p orbital.39 This strong localization of the
unpaired electron induces an appreciable inner shell polarization,
giving rise to an isotropic coupling constant that is quite higher
than those observed with other radical species derived from
phosphaalkenes. The absence of resolved dipolar coupling with
13C together with the small Aiso(13C) constant confirm that the
unpaired electron is hardly delocalized on the phosphaalkene
carbon. These properties, together with the fact that the axial
g-tensor is characterized by a g| value close to 2.0023 aligned
along 31P-T|, are typical of a phosphinyl radical.40 This result
agrees with the DFT calculations (Tables 3 and 5), which
indicate that the formation of the radical cation derived from
RPdC(NH2)2 is accompanied by an elongation of the P-C
bond, a shortening of the CN bonds (Tables 3 and 5) and a
delocalization of the partial positive charge of the phosphaalkene
carbon over the amino substituents. Indeed, natural population
analysis37 shows that electronic density loss at this carbon
through oxidation is greatly diminished by amino substitution
(carbon partial charges of +0.28 e- for M1 to +0.1 e- for M4),
this diminution being concomitant with an enhanced donation
of the nitrogen lone pairs (occupancy of 1.587 e- in M4•+). As
shown by the SOMO of M3•+ represented in Figure 5, the
unpaired electron is localized in a phosphorus pz orbital oriented
perpendicular to the RPC plane, the corresponding phosphorus-
pz spin density is equal to F ) 0.87 whereas the total spin density
on the phosphorus atom is equal to 0.89 (for M2•+ F-
(phosphorus-pz) ) 0.79, total phosphorus spin density 0.83; for
M4•+ F(phosphorus-pz) ) 0.87, total phosphorus spin density
0.91; for M6•+ F(phosphorus-pz) ) 0.69, total phosphorus spin
density 0.73).
Figure 6. Correlation of experimental anodic potentials of phosphaal-
kenes with their vertical ionization potentials Ip. (The values of Ip are
obtained from DFT calculations for M5, M6, and M8 and from
photoelectron spectroscopy for other compounds (This plot is an
extension of the correlation reported by Schoeller et al.11).
between these two values for “classically polarized” phos-
phaalkenes.11
In Figure 6, we have reported the corresponding results
together with those we have obtained for three phosphaalkenes
with inverse electron density: 1, 2, and M10 whose models
are respectively M6, M5, and M8. The data corresponding to
these three last compounds beautifully fit those obtained by
Schoeller et al.,11 they extend the range of the correlation over
3 V and point out the relatively high facility of NR2-containing
phosphaalkenes to form radical cations.
4. Conclusion
Clearly, such a radical can be seen as a phosphinyl radical40
bearing a stabilized carbenium group in the R position to the
phosphorus atom (e.g. for (Et3C6H2)2P•, Aiso (31P) ) 275
MHz40a). This identification and structure are confirmed by the
good agreement between experimental (Table 1) and calculated
(Table 6) hyperfine couplings. The electronic structures of these
radical cations containing an amino group are quite different
from the structure of cations derived from “classical” phos-
phaalkenes; e.g., for (HPdCH2)•+ (M1•+) the calculated spin
density on the carbon atom (FC ) 0.43) is close to the spin
density on the phosphorus atom (FP ) 0.52).41
The electronic behavior of inversely polarized phosphaalkenes
1 and 2 has proved to be unexpectedly rich. Oxidation products
1•+ and 2•+ as well as a transient species 1H•, resulting from
protonation of 1•-, have been characterized. In contrast to radical
cations derived from “classical” phosphaalkenes, 1•+ is easily
available and stable. It may be said to behave as a stabilized
analogue of transient phosphinyl radical. Further studies are now
in progress concerning the coordination chemistry of these
species and the electronic behavior of corresponding polyphos-
phaalkenes.
Because inverse electron density seems to be involved in the
ability of these phosphaalkenes to easily form persistent radical
cations, it is worthwhile comparing the calculated vertical
ionization potentials of several “common” phosphalkenes with
those of phosphaalkenes with reverse polarity. As seen in Table
4, the expected trends are well reproduced, with substitution
by electron-releasing amino groups decreasing the ionization
potential values. Still more interesting in the context of electron
transfer in solution, is the correlation between these vertical
ionization potentials and irreversible electrochemical oxidation
potentials. Schoeller et al. showed that a good correlation exists
5. Experimental Section
Instrumentation. Cyclic voltammetry measurements were
performed on a BAS station (model CV-50W), A standard three-
electrode cell was used, with positive feedback IR compensation.
All measurements were carried out under dry argon, The
solution was made with 1,2-dimethoxyethane or THF, dried in
a glovebox over activated 4 Å molecular sieves, as a solvent,
NBu4PF6 (0.1 M), twice recrystallized and melt-dried under
vacuum, as supporting electrolyte. A 1 mm diameter platinum
wire as working and counter-electrode, and a SCE reference

4890 J. Phys. Chem. A, Vol. 107, No. 24, 2003
Rosa et al.
electrode were used. The standard potential of ferrocene in THF
is equal to 575 mV in our experimental conditions.
Haromatic); 13C (CDCl3) δ 32.15 (s, Cpara-CMe3), 34.10 (d, 4J(P-
C) ) 8.65, Cortho-CMe3), 35.45 (s, Cortho-CMe3), 39.30 (s,
3
Cpara-CMe3), 40.85 (s, NMe2), 43.25 (d, J(P-C) ) 18.60,
EPR spectra were recorded on a Bruker 200D spectrometer
(X-band, 100 kHz field modulation) with a Bruker ER-4111
VT variable temperature controller. Samples were usually
prepared by electrolysis in situ of a thoroughly argon degassed
solution of the compound. A two platinum electrodes cell was
used, the current intensity being adjusted till a signal appeared.
For reduction experiments, the counter electrode was isolated
from the working electrode by a small glass tube with a porous
frit. For potential-controlled experiments, a three-electrode cell
was used, with a 0.1 mm Ag wire treated with HCl/HNO3 as
pseudoreference. Full details for this cell will be published
elsewhere. NBu4PF6 0.1 M was again used as an electrolyte,
with THF or DME as a solvent for the two-electrode cell, and
CH2Cl2 for the three-electrode cell (THF does not allow a proper
tuning of the EPR cavity).
1
NMe2), 121.45 (s, Cmeta), 138.45 (d, J(P-C) ) 59.95, Cipso),
148.90 (s, Cpara), 157.00 (s, Cortho), 191.05 (d, 1J(P-C) ) 76.90,
12
PdC). HRMS: calculated for
C
1H4114N231P 376.30074,
23
measured 376.30200. Melting point: 93 °C. UV-vis (THF,
nm): 245.1 (ꢀ ) 9.3 × 105), 281.1 (7.7 × 105), 307.0 (8.1 ×
105), 355.
Tetramethylurea-d12. The synthesis was adapted from ref 57.
In a 100 mL flask, DMSO (17 mL) was added to KOH powder
(7.47 g, 133 mmol). The mixture was stirred for 10 min, and
then urea (0.5 g, 8.33 mmol) was added. A condenser with a
septum was then fit to the flask, the mixture was cooled with
an ice-water bath, and then CD3I (4.2 mL, 66.6 mmol) was
added dropwise through the septum. The ice was then removed
and replaced with cold water. After 1 h of stirring, the mixture
was dissolved in 160 mL of distilled water and then extracted
three times with 160 mL of dichloromethane portions. The
resulting colorless solution was washed five times with 80 mL
of distilled water portions and then dried over MgSO4. The
solvent was removed under water pump vacuum to give a
colorless liquid. Yield: 0.34 g, 32%. The 13C-labeled urea was
used in the same way to synthesize tetramethylurea-13C,d12.
NMR spectra were obtained on a Bruker 200AC and a Varian
Gemini 200 spectrometers. All shifts are given in ppm, and
coupling constants in Hz and referenced to internal solvent peaks
or external 85% H3PO4. UV-vis spectra were obtained on a
Varian Cary 50 Bio spectrometer.
Computation. Calculations were performed with the Gauss-
ian 98 set of programs (G98).42 All geometries were optimized
using the hybrid functional B3LYP,43,44 and the standard
6-31+G(d) basis set. Minima were characterized with harmonic
frequency calculations (no imaginary frequencies). Zero-point
energies thus calculated were corrected by a factor of 0.9804.45
Natural population analysis was performed as implemented in
Gaussian 98.46 Molecular orbital visualizations were performed
using the Molekel program.47,48 Ionization potentials were
calculated as the electronic energy difference between ionized
and neutral structure (with electronic relaxation but without
structural relaxation), without ZPE correction. Proton affinities
were calculated as the electronic energy difference between
optimized protonated and neutral compound, with ZPE correc-
tion, neglecting other terms.49,50 Hyperfine coupling constants
were calculated within the GIAO formalism implemented in
Gaussian 98.51-53
Corresponding labeled compounds 1-d12 and 1-13C,d12 were
then synthesized as previously.
Overall yield starting from urea for 1-d12 and 1-13C,d12: 0.25
g, 8%.
12
2
For 1-d12: HRMS calculated for
C
1H29 H1214N231P
23
388.37606, measured 388.37900.
For 1-13C,d12: NMR 31P (293 K, Et2O) δ 35.5 (d, 1J(P-13C)
12
2
) 79.4). HRMS: calculated for
389.37941, measured 388.38200.
C
13C1H29 H1214N231P
22
[(2,4,6-tri-tert-butylphenyl)phosphino]bis(dimethylamino)-
carbenium chloride (3). In a Schlenck tube, a solution of 75
mg of 1 in 5 mL of THF was exposed to gaseous HCl. The
solution turned quickly colorless. Remaining HCl was then
fluxed out with nitrogen, and the solvent was removed in vacuo
to give a white powder. 31P NMR in dichloromethane showed
the reaction to be quantitative.
Syntheses. All syntheses were made using standard Schlenck
line techniques under nitrogen atmosphere. Solvents were
NMR (293 K): 31P (CD2Cl2) δ -60.1; 1H NMR (CD2Cl2) δ
1.31 (s, 9H, Cpara-CMe3), 1.51 (s 18H (Cortho-CMe3), 3.04
54,55
distilled over sodium before use. Mes*PH2
and bis(di-
methylamino)ethoxycarbenium tetrafluoroborate56 were synthe-
sized according to literature procedures, CD3I and 13C-labeled
urea were purchased from Eurisotop. All other reagents were
commercial.
1
(broad s, 12H, NMe2, 6.20 (d, 1H, J(P-H) ) 261.60, P-H),
7.49 (d, 2H, 4P-H) ) 3.35, Haromatic); 13C (CD2Cl2) δ 31.20 (s,
4
Cpara-CMe3), 33.40 (d, J(P-C) ) 6.60, Cortho-CMe3), 35.50
(s, Cortho-CMe3), 39.60 (s, Cpara-CMe3), 44.70 (broad s, NMe2),
(2,4,6-Tri-tert-butylphenyl)[bis(dimethylamino)methylidene]-
phosphane (1). In a dry Schlenck tube, 1.38 g of Mes*PH2 (5
mmol) in 80 mL of ether was cooled to -80 °C. MeLi (1.6 M
in ether, 3.13 mL) was syringed in the solution. The yellow
solution was brought back to room temperature and then stirred
for an additiona1 h. The formation of the anion Mes*PHLi was
then checked by 31P NMR.54 The solution was then cooled to
-50 °C, and bis(dimethylamino)ethoxycarbenium tetrafluo-
roborate (1.16 g, 5 mmol) was added. The mixture was brought
slowly to room temperature over a couple of hours. 31P NMR
showed peaks corresponding to the desired product and Mes*PH2
(approximately 1/3 ratio). The product was purified by chro-
matography on silica gel, with hexane/ether 9:1 as an eluant,
to obtain a pure yellow powder. Yield: 560 mg, 30%.
118.90 (d, 1J(P-C) ) 26.60, Cipso), 124.20 (d, 3J(P-C) ) 3.15,
2
C
meta), 154.55 (s, Cpara), 157.45 (d, J(P-C) ) 9.95, Cortho),
1
182.90 (d, J(P-C) ) 58.60, P-C). HRMS: calculated for
12
C
23
1H4214N231P+ 377.30856, measured 377.30540.
[(2,4,6-Tri-tert-butyldeuteriophenyl)phosphino]bis(dimethyl-
amino)carbenium chloride. When product 3 was dissolved in
acetone-d6, 31P NMR did not show any spectrum. After removal
of the acetone and addition of dichloromethane, 31P NMR
showed the formation of the deuterated product. NMR (293
1
K): 31P (CH2Cl2) δ -61.1 (t, J(P-2H) ) 39.6).
Hexafluorophosphate (2,4,6-Tri-tert-butylphenyl)[bis(dime-
thylamino)methylidene]phosphanyl Radical Cation (4). In a
Schlenck tube, 75 mg (0.2 mmol) of 1 and 66 mg (0.2 mmol)
of hexafluorophosphate ferrocenium were dissolved in 1 mL
of acetonitrile. The solution became deep violet immediately.
The solvent was removed in vacuo to give a violet solid, which
was washed twice with 5 mL portions of dry ether to remove
1
NMR (293 K): 31P (CDCl3) δ 28.3; H NMR (CD2Cl2) δ
1.28 (s, 9H, Cpara-CMe3), 1.55 (d, 18H, 5J(P-H) ) 1.00,
4
Cortho-CMe3), 2.16 (d, 6H, J(P-H) ) 1.40, NMe2), 2.79 (d,
4
4
6H, J(P-H) ) 2.65, NMe2), 7.25 (d, 2H, J(P-H) ) 1.55,

Phosphaalkenes with Inverse Electron Density
J. Phys. Chem. A, Vol. 107, No. 24, 2003 4891
(15) Crystal data for 1a: (C23H43N2O2P) (H2O); Mr ) 426.6; µ ) 0.13
mm-1, dx ) 1.123 g‚cm-3, monoclinic, P21/c, Z ) 4, a ) 17.2208(12), b
) 11.3657(6), c ) 13.3749(10) Å, â ) 105.475(8)°, V ) 2522.9(3) Å3.
Data were collected at 200 K on a Stoe IPDS diffractometer. R ) 0.039,
ωR ) 0.038, S ) 1.26(2) for 2526 contributing reflections and 274 variables.
Additional details are given as Supporting Information.
(16) Schmidpeter, A.; Zirzow, K.-H.; Willhalm, A.; Holmes, J. M.; Day,
R. O.; Holmes, R. R. Angew. Chem., Ind. Ed. Engl. 1986, 25, 457; Angew.
Chem., Ind. Ed. Engl. 1986, 25, 457.
(17) Weber, L.; Uthmann, S.; Stammler, H.-G.; Neumann, B.; Schoeller,
W. W.; Boese, R.; Bla¨ser, D. Eur. J. Inorg. Chem. 1999, 1369.
(18) Navech, J.; Majoral, J. P.; Kraemer, R. Tetrahedron Lett. 1983,
52, 5885.
(19) Navech, J.; Revel, M.; Kraemer, R. Phosphorus Sulfur 1984, 21,
105.
(20) In contrast to refs 19 and 20, 2 was not found to be an oil, but a
crystalline compound (mp ) 80 °C).
(21) Soulie´, E.; Berclaz, T.; Geoffroy, M. In AIP Conference Proceed-
ings 330; Bemardi, F., Rivail, J.-L., Eds.; Computers and Chemistry; 1996;
p 627.
(22) WinEPR Simfonia, VJ.25, Bruker Analytische Messtecknik Gmbh,
1996.
(23) Rotation barriers V have been calculated (B3LYP/6-31+G(d)) as
the electronic energy difference betwenn the optimized structure and a
nonoptimized structure where P-H or CH2 has been rotated by 90°. See:
Schmidt, M. W.; Truong, P. N.; Gordon. M. S. J. Am. Chem. Soc. 1987,
109, 5217.
(24) Arduengo, A. J., III; Rasika Dias, H. V.; Calabrese, J. C. Chem.
Lett. 1997, 143-44.
(25) Arduengo, A. J., III; Calabrese, J. C.; Cowley, A. H.; Rasika Dias,
H. V.; Goerlich, J. R.; Marshall, W. J.; Riegel, R. Inorg. Chem. 1997, 36,
2151.
all traces of ferrocene. UV-vis (THF, nm): 255, 371.1 (ꢀ )
5.4 × 104), 555.9 (2.1 × 104), 760.
(2,4,6-Tri-tert-butylphenyl)[(dimethylamino)ethyl]phos-
phane20 (2). In a dry Schlenck tube, to 500 mg of Mes*PH2
(1.8 mmol) was added dropwise 1.5 mL (10 mmol) of N,N-
dimethylacetamide dimethyl acetal. After addition, the solution
was stirred under nitrogen (100 °C for 5 days). The mixture
was brought slowly to room temperature over a couple of hours.
31P NMR showed peaks corresponding to the desired product
and Mes*PH2 (approximately 1/3 ratio). The product was
purified by chromatography on silica gel, with hexane/ether 1:1
as an eluant, to obtain a pure white powder. Yield: 210 mg,
34%. Melting point: 80 °C.
NMR (293 K): 31P (CDCl3) δ 94.3; 1H NMR (CDCl3) δ 1.24
5
(s, 3H, Me), δ 1.32 (s, 9H, Cpara-CMe3), 1.55 (d, 18H, J(P-
H) ) 1.00, Cortho-CMe3), 3.05 (d, 6H, 4J(P-H) ) 5.00, NMe2),
4
7.39 (d, 2H, J(P-H) ) 1.55, Haromatic); 13C (CDCl3) δ 21.50
(d, 2J(P-C) ) 11.4, PdC-Me), 31.05 (s, Cpara-CMe3), 31.45
4
(s, Cortho-CMe3), 32.50 (d, J(P-C) ) 7.90, Cortho-CMe3),
3
33.45 (s, Cpara-CMe3), 38.40 (s, NMe2), 42.05 (d, J(P-C) )
1
18.50, NMe2), 121.30 (s, Cmeta), 138.95 (d, J(P-C) ) 62.50,
1
Cipso), 148.90 (s, Cpara), 155.50 (s, Cortho), 189.20 (d, J(P-C)
12
) 62.10, PdC). HRMS: calculated for
347.27414, measured 347.27098.
C
1H3814N31P
22
(26) Joante´guy, S.; Pfister-Guillouso, G.; Chermette, H. J. Phys. Chem.
A 1999, 103, 350.
(27) Ohno, K.; Okamura, K.; Yamakado, H.; Hoshino, S.; Takami, T.;
Yamauchi, M. J. Phys. Chem. 1995, 99, 14247.
(28) Lias, S. G.; Bartmess, J. E.; Liebman, J. F.; Holmes, J. L.; Levin,
R. D.; Mallard, W. G. J. Phys. Chem. Ref Data 1988, 17 (Suppl. No. 1).
(29) Lacombe, S.; Gonbeau, D.; Cabioch, J.-L.; Pellerin, B.; Denis, J.-
M.; Pfister-Guillouzo, G. J. Am. Chem. Soc. 1998, 110, 6964.
(30) Voyna, V. I.; Alekseiko, L. N.; Okhota, B. V.; Ruban, A. V.;
Romanenko, V. D.; Markovskii L. N. J. Gen. Chem. USSR (Engl. Transl.)
1991, 61, 463.
(31) Chernega, A. N.; Antipin, M. Y.; Struchkov, Y. T.; Sarina, T. V.;
Romanenko, V. D. J. Struct. Chem. 1986, 27, 741.
(32) Nguyen, M. T.; Creve, S.; Eriksson, L. A.; Vanquickenbome, L.
G. Mol. Phys. 1997, 91, 537.
(33) Sidorenkova, H. Ph.D. Thesis, University of Geneva, Switzerland,
Acknowledgment. We gratefully acknowledge support from
Swiss National Science Foundation. P.R. thanks the portuguese
Fundac¸ao da Cie`ncia e Tecnologia for financing part of his post-
doctoral fellowship.
Supporting Information Available: Crystal data, intensity
measurement and structure refinement, atomic coordinates,
displacement parameters, bond distances, and bond angles for
1a (PDF); X-ray crystallographic file (CIF). Representation of
the SOMO for M1•+. This material is available free of charge
via the Internet at http://pubs.acs.org.
2001.
(34) Cramer, C. J.; Lim, M. H. J. Phys. Chem. 1994, 98, 5024.
(35) Kutzelnigg, W.; Fleischer, U.; Schindler, M. NMR-Basic Princip/
es and Progress; Springer-Verlag: Heidelberg, 1991; Vol. 23, p 167.
(36) Godbout, N.; Salahub, D. R.; Andzelm, J.; Wimmer, E. Can. J.
Chem. 1992, 70, 560.
(37) Glendening E. D.; Reed, A. E.; Carpenter J. E.; Weinhold, F.
Natural Bond Orbital Analysis, NBO 3.1; Theoretical Chemistry Institute,
University of Wisconsin, Madison, WI. Calculations performed at the
B3LYP/6-31+G(d) level.
(38) (a) Chentit, M.; Sidorenkova, H.; Jouaiti, A.; Terron, G.; Geoffroy,
M.; Ellinger, Y. J Chem. Soc., Perkin Trans. 2 1997, 921. (b) Alberti, A.;
Benaglia, M.; D’Angelantonio, M.; Emmi, S. S.; Guerra, M.; Hudson, A.;
Macciantelli, D.; Paolucci, F.; Roffia, S. J Chem. Soc., Perkin Trans. 2
1999, 309.
References and Notes
(1) Dillon, K. B.; Mathey, F.; Nixon, J. F. Phosphorus: the Carbon
Copy; Wiley & Sons: Chichester, U.K., 1998.
(2) Multip/e Bonds and Low Coordination in Phosphorus Chemistry;
Regitz, M., Scherer, O. J., Eds.; Georg Thieme Verlag: Stuttgart, 1990.
(3) (a) Geoffroy, M.; Jouaiti, A.; Terron, G.; Cattani-Lorente, M.;
Ellinger, Y. J. Phys. Chem. 1992, 96, 8241. (b) Jouaiti, A.; Geoffroy, M.;
Terron, G.; Bernardinelli, G. J. Am. Chem. Soc. 1995, 117, 2251. (c) Al
Badri, A.; Jouaiti, A.; Geoffroy. M. Magn. Reson. Chem. 1999, 37, 735.
(4) (a) Fuchs, E.; Breit, B.; Heydt, H.; Schoeller, W.; Busch, T.; Kruger,
C.; Betz, P.; Regitz, M. Chem. Ber. 1991, 124, 2843. (b) Fuchs, E. P. 0.;
Heydt, H.; Regitz, M.; Schoeller, W. W.; Busch, T. Tetrahedron Lett. 1989,
20, 5111.
(5) (a) Grobe, J.; Le Van, D.; Nientiedt, J.; Krebs, B.; Dartmann. M.
Chem. Ber. 1988, 121, 655. (b) Grobe, J.; Le Van, D.; Krebs, B.; Fro¨lich,
R.; Schiemann, A. J. Organomet. Chem. 1990, 389, C29-C33. (c) Grobe,
J.; Le Van, D.; Lange, G. Z. Naturforsch. B 1993, 48, 58.
(6) Weber, L.; Uthmann, S.; Stamrnler, H.-G.; Neumann, B.; Schoeller,
W. W.; Boese, R.; Blaser, D. Eur. J. Inorg. Chem. 1999, 2369.
(7) Frison, G.; Sevin, A. J Phys. Chem. A 1999, 103, 10998.
(8) Frison, G.; Sevin, A. J. Organomet. Chem. 2002, 643-644, 105-
11.
(9) For a review about phosphaalkenes with inverse electron density:
Weber, L. Eur. J. Inorg. Chem. 2000, 2425.
(10) Arduengo, A. J., III; Carmalt, C. J.; Clybume, J. A. C.; Cowley,
A. H.; Pyati, R. Chem. Commun. 1997, 981.
(11) Schoeller, W. W.; Niemann, J.; Thiele, R.; Hang, W. Chem. Ber.
1991, 124, 417.
(12) Oehme, H.; Leissring, E.; Meyer, H. Tetrahedron Lett. 1980, 21,
1141.
(13) Oehme, H.; Leissring, E.; Meyer, H. Z. Chem. 1981, 21, 407.
(14) Appel, R.; Knoll, F.; Ruppert, I. Angew. Chem., Int. Ed. Engl. 1981,
20, 731.
(39) From a comparison with atomic coupling constants reported in:
Morton, J. R.; Preston. K. F. J. Magn. Reson. 1978, 30, 577.
(40) (a) Cetinkaya, B.; Hudson, A.; Lappert, M. F.; Goldwhite, H. J.
Chem. Soc., Chem. Commun. 1982, 609. (b) Geoffroy, M.; Lucken, E. A.
C.; Mazeline, C. Mol. Phys. 1974, 28, 839. (c) Kajita, M.; Endo, Y.; Hirota,
E. J. Mol. Spectr. 1987, 124, 66. (d) Bonnazzola, L.; Michaut, J. P.; Roncin,
J. J. Chem. Phys. 1981, 75, 4829. (e) Nelson, W.; Jackel, G.; Gordy, W. J.
Chem. Phys. 1970, 52, 4572. (f) Fullam, B. W.; Mishra, S. P.; Symons, M.
C. R. J. Chem. Soc., Dalton Trans., 1974, 2145. (g) Berclaz, T.; Geoffroy,
M. HelV. Chim. Acta 1978, 61, 684.
(41) A representation of the SOMO for (HPdCH2)•+ (M1•+) is given
as Supporting Information.
(42) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
M. A.; Cheeseman, J. R.; Zakrzewski, V. G.; Montgomery, J. A.; Stratmann,
R. E.; Burant, J. C.; Dapprich, S.; Millam, J. M.; Daniels, A. D.; Kudin, K.
N.; Strain, M. C.; Farkas, O.; Tomasi, J.; Barone, V.; Cossi, M.; Cammi,
R.; Mennucci, B.; Pomelli, C.; Adamo, C.; Clifford, S.; Ochterski, J.;
Petersson, G. A.; Ayala, P. Y.; Cui, Q.; Morokuma, K.; Malick, D. K.;
Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Cioslowski, J.; Ortiz, J.
V.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I..;

4892 J. Phys. Chem. A, Vol. 107, No. 24, 2003
Rosa et al.
Gomperts, R.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng,
C. Y.; Nanayakkara, A.; Gonzalez, C.; Challacombe, M.; Gill, P. M. W.;
Johnson, B. G.; Chen, W.; Wong, M. W.; Andres, J. L.; Head-Gordon, M.;
Replogle, E. S.; Pople, J. A. Gaussian 98, Revision A.7; Gaussian, Inc.:
Pittsburgh, PA, 1998.
(49) Widauer, C.; Shiahuy Chen, G.; Gru¨tzmacher, H. Chem. Eur. J.
1998, 4, 1154.
(50) DeFrees, D. J.; McLean, A. D. J. Comput. Chem. 1986, 3, 321.
(51) Barone, V. In Recent AdVances in Density Functiona/ Methods,
Part I; Chong, D. P., Ed.; World Scientific Publishing Co.: Singapore,
1996.
(43) Becke, A. D. J Chem. Phys. 1993, 98, 5648.
(44) Lee, C.; Yang, W.; Parr; R. G. Phys. ReV. B, 1988, 37, 785.
(45) Wong, M. W. Chem. Phys. Lett. 1996, 256, 391.
(46) Reed, A. E.; Curtis, L. A.; Weinhold, F. Chem. ReV. 1988, 88,
899.
(47) P. F. Flukiger, Development of the molecular graphics package
MOLEKEL and its application to selected problems in organic and
organometallic chemistry, The`se No. 2561, De´partement de chimie physique,
Universite´ de Gene`ve, Gene`ve, 1992.
(52) Rega, N.; Cossi, M.; Barone, V. J. Chem. Phys. 1996, 105, 11060.
(53) Barone, V. Chem. Phys. Lett. 1996, 262, 201.
(54) Yoshifuji, M.; Toyota, K.; Shibayama, K.; Inamoto. N. Chem. Lett.
1983, 1653.
(55) Cowley, A. H.; Kilduff, J. E.; Newman, T. H.; Pakulski, M. J. Am.
Chem. Soc. 1982, 104, 5820.
(56) Meerwein, H.; Florian, W.; Scho¨n, N.; Stopp, G. J. Liebigs Ann.
Chem. 1961, 641, 1.
(48) Portmann, S.; Luthi, H. P. Chimia 2000, 54, 766.
(57) Johnstone, R. A. W.; Rose, M. E. Tetrahedron 1979, 35, 2169.