Structural and conformational properties of 1,2-propadienylphosphine (allenylphosphine) studied by microwave spectroscopy and quantum chemical calculations

Article abstract of DOI:10.1021/jp047207p

1,2-Propadienylphosphine (allenylphosphine), H₂C=C=CHPH ₂, has been investigated by Stark and Fourier transform microwave spectroscopy. Two rotameric forms denoted syn and gauche have been assigned. The syn form has a symmetry plane (C_s symmetry) where the lone electron pair of phosphorus points toward the double bonds. The phosphino group is rotated roughly 120° from this position in the gauche retainer. The dipole moment of syn was determined to be μ_a = 1.613(23), μ_b = 2.347(24), μ_c = 0 (for symmetry reasons), and μ_tot = 2.848(28) × 10^-30 C m [0.854(8) D]. The energy difference between the two forms was found to be 2.1(4) kJ/mol from relative intensity measurements with syn as the more stable conformer. Extensive quantum chemical calculations have been carried out and accurate equilibrium structures have been determined for these two rotamers, as well as for the corresponding two conformers of vinylphosphine (H₂C=CHPH ₂).

Full text of DOI:10.1021/jp047207p

J. Phys. Chem. A 2005, 109, 115-121
115
Structural and Conformational Properties of 1,2-Propadienylphosphine (Allenylphosphine)
Studied by Microwave Spectroscopy and Quantum Chemical Calculations
Harald Møllendal,*,† Jean Demaison,‡ Denis Petitprez,‡ Georges Wlodarczak,‡ and
Jean-Claude Guillemin§
Department of Chemistry, UniVersity of Oslo, P.O. Box 1033 Blindern, NO-0315 Oslo, Norway, Laboratoire de
Physique des Lasers, Atomes et Mole´cules, UMR CNRS 8523, Baˆt. P5, UniVersite´ de Lille1,
F-59655 VilleneuVe d’Ascq, France, and Laboratoire de Synthe`se et ActiVation de Biomole´cules,
UMR CNRS 6052, Institut de Chimie de Rennes, ENSCR, F-35700 Rennes, France
ReceiVed: June 28, 2004; In Final Form: October 13, 2004
1,2-Propadienylphosphine (allenylphosphine), H2CdCdCHPH2, has been investigated by Stark and Fourier
transform microwave spectroscopy. Two rotameric forms denoted syn and gauche have been assigned. The
syn form has a symmetry plane (Cs symmetry) where the lone electron pair of phosphorus points toward the
double bonds. The phosphino group is rotated roughly 120° from this position in the gauche rotamer. The
dipole moment of syn was determined to be µa ) 1.613(23), µb ) 2.347(24), µc ) 0 (for symmetry reasons),
and µtot ) 2.848(28) × 10-30 C m [0.854(8) D]. The energy difference between the two forms was found to
be 2.1(4) kJ/mol from relative intensity measurements with syn as the more stable conformer. Extensive
quantum chemical calculations have been carried out and accurate equilibrium structures have been determined
for these two rotamers, as well as for the corresponding two conformers of vinylphosphine (H2CdCHPH2).
Introduction
The literature about the conformational and structural proper-
ties of aliphatic phosphines in the gas phase is not comprehen-
sive presumably because these compounds are often toxic,
pyrophoric, and unstable in air and they smell obnoxious.
Microwave (MW) spectroscopic studies of CH3CH2PH2,1 H2-
PCH2CH2CtN,2 H2PCH2CH2PH2,3 HCtCCH2PH2,4 H2Cd
7
CHCH2PH2,5 H2CdCHPH2,6 and HCtCPH2 have already
demonstrated that aliphatic phosphines have unique conforma-
tional and structural properties different from those of the
corresponding amines.
It is felt that the limited knowledge of this interesting class
of compounds warrants further investigation. In this work, the
studies of the aliphatic phosphines are extended to include 1,2-
propadienylphosphine (allenylphosphine), H2CdCdCHPH2.
The properties of this compound have previously been inves-
Figure 1. Possible rotameric forms of allenylphosphine. The syn and
tigated by nuclear magnetic resonance, and infrared and high-
gauche conformers were assigned in this work. Anti represents a
resolution mass spectrometry,8 as well as by photoelectron
transition state.
spectroscopy supported by theoretical MNDO calculations.9 The
It should be mentioned that an extensive delocalization of
the lep of phosphorus into the π-electron system of the allenyl
moiety might lead to yet a fourth planar (or nearly planar)
conformation for the -CdCHPH2 group of atoms. This
hypothetical conformation is not drawn in Figure 1.
In the vinylphosphine molecule (H2CdCHPH2), two con-
formers were identified in MW investigations.6,7 Interestingly,
the most stable one of these is similar to the syn form of Figure
1, whereas the less stable form is similar to gauche. No
experimental determination was made of the energy difference,
but the former conformer is the more stable one. Ab initio
calculations yielded an energy difference in the 1.7-5.4 kJ/
mol range between these two forms depending on the level of
calculation.7
photoelectron spectrum indicates that more than one conformer
may be present in the gas phase.9
Rotation around the C-P bond leads to rotational isomerism
in the title compound. Three typical conformations are drawn
in Figure 1. A distinction between them may be made with
reference to the lone electron pair (lep) of phosphorus. The Cd
C-P-lep has a syn-periplanar conformation in the syn con-
former. It is anti-clinal in the gauche form, and anti-periplanar
in the anti form. The syn and anti forms will thus each have a
symmetry plane (Cs symmetry), whereas the gauche form has
no symmetry (C1 point group).
* Corresponding author. E-mail: harald.mollendal@kjemi.uio.no. Tel:
+47 22 85 56 74. Fax: +47 22 85 54 41.
† University of Oslo.
The electronic environment in the allenyl group is of course
rather different from that of the vinyl group because two double
‡ Universite´ de Lille1.
§ ENSCR.
10.1021/jp047207p CCC: $30.25 © 2005 American Chemical Society
Published on Web 12/10/2004

116 J. Phys. Chem. A, Vol. 109, No. 1, 2005
Møllendal et al.
TABLE 1: Spectroscopic Constants and Dipole Moment of
Allenylphosphine as Calculated at the B3LYP/
6-311G(3df,2pd) Level of Theory
bonds adjacent to one another are present in the former
compounds, while only one such bond is found in the latter
molecules. This difference between allenic and vinylic phos-
phines might influence the conformational as well as other
molecular properties. Moreover, no allenic phosphine has
previously been studied by the very specific MW spectroscopic
method. These were the main reasons for choosing the prototype
of allenic phosphines, allenylphosphine, for study.
syn
gauche
A/MHz
B/MHz
C/MHz
25705.8
2740.4
2573.3
2.06
2.25
0.0
26021.4
2714.1
2544.8
2.37
0.10
1.51
µa/10-30 C m
µb/10-30 C m
µc/10-30 C m
µtot/10-30 C m
∆J/kHz
The methods of investigation were in addition to MW
spectroscopy, quantum chemical calculations. The extremely
high resolution and specificity of the former method are ideal
for analyzing conformational mixtures. Modern quantum chemi-
cal calculations are now rather reliable and may provide
information not readily available from experiments alone, thus
supplementing MW spectroscopy in an ideal manner. Compre-
hensive quantum chemical calculations have in the present case
been carried out to predict a number of molecular properties
including an accurate equilibrium structure not only for the title
compound, but for vinylphosphine as well.
3.29
2.82
1.180
-44.60
818.3
0.198
-2.347
1.066
-44.16
830.4
0.180
3.856
∆JK/kHz
∆K/kHz
δJ/kHz
δK/kHz
decomposition in the brass in the course of 6-8 h. The MW
spectra of impurities of 2-propenylphosphine, H2dCH-CH2-
PH2,5 as well as impurities of unidentified origin were noted.
Microwave Fourier Transform Spectrometer Experi-
ments. Rotational spectra in the 5-20 GHz spectral range were
recorded with the Lille microwave Fourier transform (MWFT)
spectrometer.15 A gas mixture of 100 Pa of allenylphosphine
completed with neon as carrier gas to a total pressure of 1 ×
105 Pa was prepared. Gas pulses were then expanded into the
vacuum tank through a pulsed nozzle to create a supersonic jet
parallel to the optical axis of the Fabry-Perot cavity. The pulsed
valve was opened for 600 µs at a repetition rate of 1.5 Hz. To
polarize the allenylphosphine molecule, 2 µs microwave pulse
were used with power fields of 0.8 mW for the syn form
transitions and 0.5 mW (1 mW) for the µa type (µc type,
respectively) gauche rotamer transitions. A total of 4096 data
points per free emission decay are collected at a 10 MHz sample
rate, which gives, after Fourier transformation, a spectral
resolution of 2.4 kHz per point. The accuracy of frequency
measurements is estimated to be better than 3 kHz.
Experimental Section
Synthesis of 1,2-Propadienylphosphine (Allenylphosphine).
Caution: The 1,2-propadienylphosphine is pyrophoric and
potentially highly toxic. All reactions and handling should be
carried out in a well-ventilated hood.
The synthesis of allenylphosphine has already been reported.8
The apparatus was similar to the one described for the prepara-
tion of 2-propynylphosphine.4 The reducing agent (AlHCl2) was
prepared as follows: in a 250-mL two-necked flask equipped
with a nitrogen inlet and a stirring bar were introduced LiAlH4
(1.14 g, 30 mmol) and tetraglyme (50 mL). The flask was
immersed in a cold bath (-30 °C) and aluminum chloride (12.0
g, 90 mmol) was added by portions.10 The flask was fitted on
the vacuum line, allowed to warm to -10 °C, and degassed.
The 5,5-dimethyl-2-oxo-2-(propa-1′,2′-diene)-1,3,2-dioxaphos-
phorinane8 (1.88 g, 10 mmol) dissolved in tetraglyme (50 mL)
was slowly added with a flex-needle through the septum in about
20 min. During and after addition, the formed allenylphosphine
was distilled off in vacuo from the reaction mixture. A cold
trap (-80 °C) selectively removed less volatile products and
the phosphine was condensed in a second cold trap (-120 °C)
to remove the most volatile products (mainly PH3). After the
sample was disconnected from the vacuum line by stopcocks,
the product was kept at low temperature (< -30 °C) before
analysis. The allenylphosphine was thus obtained in 92% purity
and in a 77% yield (0.55 g). The presence of several other
phosphines in a 0.1-5% yield cannot be avoided. The main
impurity was the 2-propenylphosphine (≈4%).
Results and Discussion
Computational Methods. The majority of the quantum
chemical calculations were made in Lille with the MolPro200016
and Gaussian03 programs.17 A few calculations were also
performed in Oslo using the latter program.
To obtain a first prediction of the rotational constants, density
functional theory with the hybrid functional B3LYP (Becke’s
three parameter18 functional employing the Lee, Yang and Parr
correlation functional19) was used with the 6-311G(d,p) basis
set as implemented in Gaussian 03. To improve the accuracy
of the predicted rotational spectra, the structure, the harmonic
force field, and the quartic centrifugal distortion constants were
next calculated at the B3LYP/6-311G(3df,2pd) level of theory.
The results are given in Table 1.
The Gaussian-3 (G3) theory20 was employed to calculate the
energy differences between the conformers (Figure 1). The zero-
point energy (ZPE) is first calculated at the Hartree-Fock level
of theory (HF/6-31G(d)) followed by a scaling of the vibrational
frequencies (scaling factor: 0.8929). G3 theory uses, as refer-
ence geometry, the geometry from MP2/6-31G(d) level of
theory, and all electrons are correlated. The quadratic config-
uration interaction, QCISD(T), energy21 at a large basis set is
obtained assuming that a series of calculations at lower level
of theories are additive.22 The total energy at 0 K may be
expressed as
Stark Spectrometer Experiment. The MW spectrum was
studied using the Oslo Stark spectrometer, which is described
briefly in ref 11. A 3-m Stark cell made of brass was utilized.
Radio frequency microwave double resonance (RFMWDR)
experiments were carried out as described in ref 12 using the
equipment mentioned in ref 13. The 8-62 GHz spectral region
was investigated with the cell cooled to dry ice temperature
(-78 °C). The pressure was a few pascals when the spectrum
was recorded and stored electronically using the program written
by Waal.14 The accuracy of the spectral measurements was better
than (0.10 MHz and the maximum resolution was about 0.4
MHz.
The compound was kept for several months in a refrigerator
at about -30 °C when not in use. Polymeric products were seen
to form slowly. However, the compound showed no signs of

Structural Properties of 1,2-Propadienylphosphine
J. Phys. Chem. A, Vol. 109, No. 1, 2005 117
TABLE 2: Relative Energies (kJ/mol) of the Different
Conformers of Allenylphosphine and Vinylphosphine
E0(G3) ) MP4/d + [QCISD(T)/d - MP4/d] +
[MP4/plus - MP4/d] + [MP4/2df,p - MP4/d] +
[MP2(FU)/G3L - MP2/2df,p - MP2/plus + MP2/d] +
E(ZPE) + E(HLC)
allenylphosphinea
vinylphosphinea
vinylphosphineb
∆E
τc
∆E
τc
∆E
τc
syn
0.0
0.0
0.0
6.26
2.12
11.17
0.0
0.0
6.92
0.0
TSd
6.23
2.47
7.46
65.3
119.5
180.0
62.8
113.1
180.0
63.9
111.2
180.0
where MP4 ) fourth-order Møller Plesset,23 d ) 6-31G(d), plus
) 6-31+G(d), 2df,p ) 6-31G(2df,p), G3L ) G3Large basis
set, and E(HLC) is a higher level correction depending on the
numbers of R- and â-electrons. Thus, it cancels out when we
make the difference between the energies of two conformers.
Finally, FU ) full, meaning that all electrons are correlated.
The G3 calculations were repeated for the structurally similar
compound vinylphosphine, H2PCHdCH2, which was studied
previously.6,7 The results are given in Table 2. This table also
includes the results found in the MP2/6-311G(3df,p) calcula-
tions.24
gauche
2.15
antid
11.94e
a G3 calculations; see text. b MP2/6-311G(3df,p) calculations.24 c τ
is the dihedral angle (in degrees) between the lone pair and the
CdC(H)P skeleton. d TS and anti are transition states (imaginary
torsional frequency). e A MP4/6-311G(3df,p) calculation gives 11.44
kJ/mol.24
TABLE 3: Equilibrium Structuresa of the Syn and Gauche
Forms of Vinylphosphineb
sync
gauched,e
It is noted that the most stable form is obtained when the
phosphorus lep is syn-periplanar (syn) with respect to the
C-C-P link of atoms for both allenyl- and vinylphosphine.
The lep-anti-clinal form (gauche) is less stable and the lep-anti-
periplanar form (anti) is a transition state in both compounds.
The barrier between the two gauche forms is low. The anticlinal
form represents a maximum on the potential energy hyper
surface in both molecules. This conformation has an energy that
is 7.5 kJ/mol higher than the energy of syn in allenylphosphine
and 11-12 kJ/mol higher in vinylphosphine (Table 2). Another
transition state occurs about 65° from syn with an energy 6-7
kJ/mol higher than the energy of syn. For vinylphosphine, the
results are in extremely good agreement with the previous results
that were obtained at the MP2/6-311G(3df,p) level of theory.24
Structures. For the calculation of the geometrical structures,
the second-order Møller-Plesset perturbation theory, MP2,23
and the coupled-cluster method with single and double excita-
tion25 augmented by a perturbational estimate of the effects of
the connected triple excitations,26 CCSD(T), was used with the
Dunning’s correlation consistent polarized valence basis sets,
cc-pV(n+d)Z with n ) D, T, Q.27
The structure of both stable forms of allenylphosphine and
vinylphosphine was first calculated at the MP2/cc-pV(Q+d)Z
level of theory. This method is known to give reliable angles.28
On the other hand, the bond lengths are affected by a small
error, which is mainly systematic. This error (also called offset)
may be most easily estimated with the help of small molecules
whose structure is accurately known and can be used to correct
ab initio bond lengths. To determine the offsets of the r(CdC)
and r(C-H) bond lengths, we have used allene, H2CdCdCH2,
whose structure is well-known.29 Likewise, for the r(P-H) bond
length, we have used phosphine, PH3.30
r(C1dC2)
r(C1-P)
133.5
182.4
108.2
108.2
108.3
141.4
133.5
181.9
108.2
108.1
108.5
r(C2-Hcis)
r(C2-Htrans
r(C1-H)
r(P-H)
)
r(P-Hoop
)
141.3
141.3
121.6
121.2
119.0
125.4
r(P-Hip)
(C1C2Hcis)
(C1C2Htrans
(HC1C2)
(C2C1P)
121.1
121.6
119.6
120.3
97.4
)
(HPC1)
(HoopPC1)
(HipPC1)
(HC1PH)
(PC1C2Hcis)
(PC1C2Htrans
(HC1C2Htrans
(C2C1PHip)
(C2C1PHoop
97.7
97.0
46.8
5.6
)
)
-174.3
-0.6
-19.2
71.5
)
a See text. b Distances in pm and angles in degrees. c CCSD(T)/
V(T+d)Z + MP2[V(Q+d)Z - V(T+d)Z] + MP2(ae)/CVQZ -
MP2(fc)/CVQZ; see Table 7S in Supporting Information for details of
calculations. d Core correlation taken from the syn conformer; see Table
3. e CCSD(T)/V(T+d)Z + MP2[V(Q+d)Z - V(T+d)Z] + core cor-
relation; see also Table 8S in Supporting Information for details of
calculations.
(T)/cc-pV(T+d)Z level, indicates that nondynamical electron
correlation is not important and that the CCSD(T) results should
be reliable.
To estimate the core and core-valence correlation effects on
the computed molecular geometry, the correlation-consistent
polarized core-valence quadruple-ú basis set, cc-pCVQZ,33 was
employed with the MP2 method which is known to be a rather
satisfactory approximation,31 at least for bonds involving only
first-row atoms. This correction leads to the expected bond
shortenings. The derived equilibrium structure for the syn and
gauche forms of vinylphosphine is given in Table 3. Further
data are found in Supporting Information, Tables 7S and 8S.
The equilibrium structure of the gauche form was calculated
from the ab initio CCSD(T) structure assuming that the core
correlation is the same as for the syn conformer, which is likely
to be a safe approximation.
Finally, the structures of the syn and gauche forms of
allenylphoshine are given in Table 5. They are calculated from
the ab initio MP2/cc-pV(Q+d)Z structures with the offset
corrections given in Table 4 and calculated as discussed above.
See also Supporting Information, Tables 9S and 10S.
A careful examination of the offsets shows that they are not
exactly constant for a given bond. For instance, for the P-H
For the r(C-P) bond length, we have calculated the ab initio
equilibrium structure of vinylphosphine. As this molecule is
smaller than allenylphosphine, it is more easily amenable to
high level ab initio calculations. For this reason, the structure
of the syn form of vinylphosphine was first calculated at the
CCSD(T) level of theory, but the expensive calculation at the
CCSD(T)/cc-pV(Q+d)Z level was replaced by simpler calcula-
tions because it is well established that the variation from CCSD-
(T)/cc-pV(T+d)Z to CCSD(T)/cc-pV(Q+d)Z may be accurately
predicted with MP2 calculations.29,31 The following approxima-
tion formula was used:
r[CCSD(T)/cc-pVQZ] ) r[CCSD(T)/cc-pVTZ] +
r[MP2/cc-pVQZ] - r[MP2/cc-pVTZ]
The coupled cluster T1 diagnostic,32 which is 0.012 at the CCSD-

118 J. Phys. Chem. A, Vol. 109, No. 1, 2005
Møllendal et al.
TABLE 6: Experimental Ground State and Calculated
TABLE 4: Determination of the Offsets for the MP2/
cc-pV(Q+d)Z Ab Initio Structures (Distances in pm and
Angles in Degree)
Equilibrium Rotational Constants (MHz)
Gsa
Eqb
Gs-Eqc
Gs-Eqd
molecule
bond
ab initioa
re
offsetb
refc
Vinylphosphine Syn
H2CdCHPH2
PH3
H2CdCdCH2 CdC
H2CdCdCH2 C-H
C-P
P-H
182.09
140.93
130.59
108.00
182.44
141.14
130.67
108.11
0.35
0.21d
0.08e
0.11
this work
A
B
C
39899.31
5521.90
5055.04
40246.92
5542.60
5077.83
-347.6
-595.25
-13.51
-33.42
29
28
28
-20.7
-22.8
Vinylphosphine Gauche
a MP2(fc)/cc-pV(Q+d)Z. b re - r (ab initio). c For the equilibrium
structure. d 0.29 pm from H2CdCHPH2; see Table 3. e 0.04 pm from
H2CdCHPH2; see Table 8S in Supporting Information for details of
calculations.
A
B
C
40484.15
5458.08
4974.19
40699.8
5489.31
5004.62
-215.6
-215.31
-42.36
-39.25
-31.2
-30.4
Allenylphoshine Syn
24836.1
A
B
C
25004.08
2765.72
2586.27
168.0
-15.6
-14.0
-175.18
-5.37
-8.46
TABLE 5: Equilibrium Structuresa of the Syn and Gauche
Forms of Allenylphosphineb
2781.34
2600.23
sync
gauchec
Allenylphosphine Gauche
A
B
C
25334
2735.98
2555.22
25051.1
2755.3
2571.39
282.9
-19.3
-16.2
-18.27
-14.01
-14.00
r(C1-P)
r(C1dC2)
r(C2dC3)
r(C1-H)
r(P-H)
r(P-Hoop
r(P-Hip)
r(C3-H)
r(C3-H(-)
r(C3-H(+)
(PC1C2)
182.9
130.9
130.7
108.4
141.3
182.9
130.7
130.9
108.5
a Experimental ground-state constants taken from refs 6 and 7 in
the case of vinylphosphine, or from Tables 7 and 9 in the case of the
title compound. b Equilibrium constants from the estimated equilibrium
)
141.3
141.0
structures; see Tables 3 and 5. c B0 (exp) - Beg(calc). d B0 - Beg
calculated at the MP2/6-31G* level of theory. For the syn form of
vinylphosphine, the corrections at the MP2/6-311G(3df,2p) are A0 -
Ae ) -565, B0 - Be ) -12.6, and Ce - C0 ) -32.3 MHz, respectively.
g
g
108.1
)
)
108.2
108.2
123.5
178.6
119.2
119.3
178.6
119.3
97.2
(C1C2C3) c
(C2C1H)
(C1PH)
(C1PHoop
(C1PHip)
(HPH)
lated from the equilibrium structure are generally larger than
the ground-state rotational constants.
)
97.9
96.5
94.3
The two kinds of rotational constants are compared in Table
6. It is seen that they agree well within the 1% limit. However,
it is seen that the experimental values for the A rotational
constants of the two conformers of allenylphosphine is larger
than the A rotational constants calculated from the equilibrium
structure. This is contrary to the usual situation, as already noted.
To check this unusual behavior, we have calculated the
difference between the experimental and calculated rotational
92.7
120.7
(C2C3H)
(C2C3H(-)
(C2C3H(+)
(HC3H)
)
)
120.9
120.7
118.4
118.6
133.2
(C2C1PH)
(C2C1PHoop
(C2C1PHip)
(PC1C3H)
(PC1C3H(-)
(HC1C3H(-)
(PC1C3H(+)
)
108.0
12.8
g
constants (B0 (exp), g ) a, b, c) at the MP2/6-31G* level. The
89.7
)
)
)
84.2
-89.2
-95.4
results are given in the last column of Table 5. It is seen that
the expected negative difference is indeed calculated for the A
rotational constants, and there is reasonably good agreement
for the other rotational constants.
a See text. b Distances in pm and angles in degrees. c MP2/cc-
pV(Q+d)Z with offsets of Table 4; see also Tables 9S and 10S in
Supporting Information. d P inside the acute angle.
The A rotational constants of syn and gauche allenylphosphine
are very sensitive to the coordinates of the phosphorus atom. A
slight modification of the C-P bond length and C-C-P bond
angle of the equilibrium structures given in Table 5 would
produce better agreement. The uncertainty of these two param-
eters are thus estimated to be 0.5 pm and 0.4°, compared to 0.3
pm and 0.2°, respectively, for the remaining bond lengths and
angles.
Stark Spectrum and Assignment of the Syn Rotamer. The
first survey spectra taken using Stark spectroscopy revealed a
rather rich and comparatively intense MW spectrum. The syn
conformer (Figure 1) was predicted by the quantum chemical
calculations above to be the preferred form of the molecule. µb
was calculated to be the largest component of the dipole moment
for this form (Table 1). Using the B3LYP/6-311G(3df,2pd)
rotational constants available in the initial phase of this work,
the strong bQ-branch series J0,J r J1,J-1 was predicted to have
a number of transitions at frequencies higher than 23 GHz. These
transitions were soon identified. Their assignments were con-
firmed by their fit to Watson’s Hamiltonian (A reduction, Ir
representation)35 and the fact that the Stark effects of several
of them were well-resolved.
bond length, the offset is 0.21 pm in PH3 and 0.35 pm in
vinylphosphine (syn form, Table 4). The offset is indeed only
approximately constant for a given bond and is known to be
sensitive to the chemical environment.
Furthermore, the core correlation was calculated at the MP2
level for vinylphosphine. This approximation is known to give
a core correlation slightly too large for bond lengths where a
second-row atom (here P) is involved.34
Accuracy of the Estimated Equilibrium Structures. The
accuracy of the estimated equilibrium structures in Tables 3
and 5 warrants comments. The bond lengths and angles not
involving the phosphorus atom are estimated to be accurate to
within 0.3 pm and 0.2°, respectively. Bond lengths and angles
involving the phosphorus atom are likely to be slightly less
accurate, as the following will show.
The ground-state rotational constants determined for the two
forms in the experimental section below are obviously insuf-
ficient to determine the full structure of allenylphosphine. These
experimental constants determined for the ground vibrational
state usually deviate by less than about 1% from the equilibrium
rotational constants. Moreover, the rotational constants calcu-
Searches for low-J a- and b-type R-branch lines were made
next and soon met with success. The assignments of some of

Structural Properties of 1,2-Propadienylphosphine
J. Phys. Chem. A, Vol. 109, No. 1, 2005 119
TABLE 7: Rotational (MHz), Quartic Centrifugal Distortion
(kHz), and Sextic Centrifugal Distortion Constants (Hz)a of
the Syn Conformer of Allenylphosphine
TABLE 8: Second-Order Stark Coefficientsa and Dipole
Moment of the Syn Conformer of Allenylphosphine
stark coefficients
∆E-2/10-5 MHz
V-2 cm2
vib state
ground
C-P torsion lowest bending vib
A
B
25004.0803(23) 24801.0003(87) 25363.120(11)
2765.72333(23) 2765.76500(97) 2774.8108(12)
transition
1,0 r 10,1
1,1 r 20,2
1,2 r 30,3
|M|
obsd
calcd
C
∆
∆
∆
2586.26670(27) 2586.6823(10)
2590.3128(13)
1.3031(15)
-49.358(31)
922.72(21)
0.23351(23)
4.433(27)
1
2
3
1
2
3
2
4
5
4
6
5
8.73(10)
1.76(3)
8.90
1.60
1.30883(65)
-48.964(15)
818.33(13)
1.3114(12)
-49.018(26)
740.85(15)
0.23279(23)
-1.900(51)
-0.00220(36)
-0.363(11)
0.000500(79)
7.726725(28)
229
J
JK
K
0.720(8)
0.427(6)
0.542(6)
0.453(5)
0.218(3)
0.441(2)
0.253(3)
0.736
0.440
0.532
0.459
0.208
0.421
0.256
δJ
0.23074(13)
1.570(39)
4
5
1,3 r 40,4
1,4 r 50,5
δK
ΦJ
ΦJK
φJb
0.00276(18)
-0.1396(54)
0.001036(50)
0.00075(51)
-0.152(18)
0.001121(82)
6.953250(33)
176
61,5 r 60,6
(Ia + Ib - Ic)c 7.532580(10)
dipole moment
Nd
344
µa ) 1.613(23), µb ) 2.347(24), µc ) 0b and
µtot ) 2.848(28) × 10-30 C m
rmse
0.016f
0.099g
0.108g
a A reduction Ir representation.35 B3LYP/6-311G(3df,2pd) predictions
of rotational constants and quartic centrifugal distortion constants are
found in Table 1. Rotational constants calculated from the equilibrium
structure are listed in Table 6. b Further sextic constants preset at zero.
c Moments of inertia with dimension 10-20 m2 u. Conversion factor:
5053.7905 nm2 u MHz. d Number of lines used in the fit. e Root-mean-
square deviation. f From weighted fit; see text. g From fit with equal
weights.
a Uncertainties represent one standard deviation. 1 D ) 3.33564 10-30
C m. B3LYP/6-311G(3df,2pd) values are listed in Table 1. b Preset at
this value in the least-squares fit.
vibrationally excited states. The assignments were made in the
same manner as described for the ground vibrational state. Only
a- and b-type lines were found as expected. The spectroscopic
constants are shown in Table 7. The transitions of the excited
states were assigned unit weights in the fitting procedure. The
full spectra are listed in Supporting Information, Tables 2S and
3S.
them were confirmed both by Stark effect studies and fit to the
said Hamiltonian. The assignments of some of the aR-transitions
were also confirmed by RFMWDR experiments.
The assignments were then gradually extended to include
increasingly higher-J transitions. Ultimately, about 350 lines
of the a- and b-varieties with a maximum value of J ) 78 were
assigned. Even higher-J transitions were searched for, but their
intensities were presumably too low to allow unique identifica-
tions to be made. c-type lines were also searched for, but not
found presumably because the corresponding dipole moment
component is identical to zero owing to the Cs symmetry of
the syn conformer.
Another 14 lines were found next using Fourier transform
spectroscopy. Ultimately 344 a- and b-type transitions were used
to determine the spectroscopic constants shown in Table 7. Three
sextic centrifugal distortion constants had to be used in addition
to the five quartic centrifugal distortion constants to get a fit
consistent with the measurements accuracy. The accuracy of
the transitions measured using the Fourier transform and Stark
spectroscopy are different, and weights have therefore been
applied in the fitting procedure. The inverse square of the
uncertainties have been used. The full spectrum including the
uncertainties is shown in Supporting Information, Table 1S. The
maximum centrifugal distortion is rather prominent amounting
to more than 5 GHz for the highest-J transitions.
Comparison of the B3LYP/6-311G(3df,2pd) rotational con-
stants (Table 1) predicted for syn and their experimental
counterparts (Table 7) shows expected agreement (2.8% for A,
0.8% for B, and 0.5% for C, respectively) for calculations of
this level of theory. The agreement between the theoretical and
experimental quartic centrifugal distortion constants given in
the same tables is about 1 order of magnitude poorer, as
expected.
Vibrationally Excited States of Syn. The vibrationally
excited states of this rotamer was assigned using the Stark-
modulated spectrometer. The ground-state transitions of syn
were accompanied by two series of lines showing Stark and
RFMWDR patterns similar to the corresponding transitions
belonging to the ground vibrational state but having considerably
less intensity. These transitions were assumed to belong to
It is seen in Table 7 that one of the excited states is assigned
as the first excited state of the lowest torsional mode, viz the
torsion around the P-C bond. This assignment has been made
because Ia + Ib - Ic increases upon excitation to 7.62 from the
ground-state value of 7.53 × 10-20 u m2 (Table 8). An increase
is expected for a torsional vibration.36 Relative intensity
measurements were made observing the precautions of ref 37.
A frequency of 137(25) cm-1 was obtained for this normal
mode. This should be compared to 122 cm-1 obtained in the
theoretical B3LYP calculations above but not tabulated.
The other vibrationally excited state is assigned as the first
excited state of the lowest bending vibration. This assignment
has been made because Ia + Ib - Ic now decreases to 6.95 ×
10-30 u m2 as expected for this kind of vibration.37 Relative
intensity measurements yielded 156(25) cm-1 for this mode,
relatively close to the theoretical value (169 cm-1).
Dipole Moment. The dipole moment of syn was determined
in a least-squares fit using the second-order Stark coefficient
of the transitions shown in Table 8. The weight of each Stark
coefficient was taken to be the inverse square of the standard
deviation of that coefficient, which is also shown in the same
table. Only µa and µb were allowed to vary in the fit, while µc
was kept constant at zero because of the symmetry plane. The
cell was calibrated using OCS, whose dipole moment was taken
to be 2.3857(68) × 10-30 C m.38
The experimental values of the dipole moment along the
principal inertial axes µa ) 1.613(23), µb ) 2.347(24) and µtot
) 2.848(28) × 10-30 C m agree reasonably well with the both
the B3LYP predictions (Table 1) and with advanced MP2/cc-
pV(Q+d)Z calculations, which yielded µa ) 1.744, µb ) 2.813,
and µtot ) 3.310 × 10-30 C m. These theoretical values are
additional evidence that syn has indeed been assigned and not
confused with gauche or anti, which would have had signifi-
cantly different principal-axis dipole moment components.
Fourier Transform Spectrum and Assignment of the
Gauche Rotamer. There are two identical gauche forms
separated by a mirror plane. Tunneling of the phosphino group

120 J. Phys. Chem. A, Vol. 109, No. 1, 2005
Møllendal et al.
TABLE 9: Rotational (MHz) and Quartic Centrifugal
Distortion (kHz)a of the Ground Vibrational State of the
Gauche Conformer of Allenylphosphine
The gauche rotamer is predicted to have a sizable dipole
moment component along the c-axis according to the calcula-
tions (Table 1). The strongest c-type lines are in general several
series of different Q-branch transitions. The c-type transitions
were believed to be split into two components owing to
tunneling of the phosphino group, just as observed for the related
molecule vinylphosphine,6 where a splitting of about 9 MHz
was observed. Extensive searches were now carried out in an
attempt to find these split transitions. These attempts met with
partial success. Several lines of the J1,J r J0,J c-type series with
J between 4 and 22 are assumed to have been assigned. They
are listed in Table 6S (Supporting Information). All these
transitions are either of the (+) r (-) species, or of the (-) r
(+) species, but it is not possible to know which one of these
two alternatives is correct. Attempts to find the other species,
which would have allowed a unique identification, failed,
possibly because the splittings are large or perturbed. Searches
were also made for further c-type Q- and R-branch lines of this
rotamer, however, with a negative result. The dense and
complicated nature of this perturbed spectrum in addition to
the relative weakness of these c-type lines rendered this search
unsuccessful. Searches were also made for b-type lines, but none
were found presumably because µb is close to zero (Table 1).
Energy Difference. The energy difference between the two
rotamers was determined by relative intensity measurements.37
The dipole moment of the gauche rotamer was not determined
experimentally owing to insufficient intensities. The ratio of
the dipole moments is important, and the theoretical dipole
moments shown in Table 1 were used to derive this ratio.
Relatively strong isolated lines were used. The statistical weight
of gauche was assumed to be twice the weight of the syn
conformer. A value of 2.1(4) kJ/mol was derived with syn as
the most stable form. The theoretical values for this energy
difference is 2.5 (G3; Table 2), 2.1 (MP2/V(Q+d)Z), and 2.1
kJ/mol (B3LYP/6-311G(3df,2pd)). There is thus good agreement
between theory and experiment in this case.
A
B
C
∆
∆
∆
25334(19)
2735.9789(49)
2555.2237(50)
1.163(12)
-46.874(32)
830.4b
J
JK
K
δJ
δK
0.180b
3.856b
6.882(15)
74
(Ia + Ib - Ic)c
Nd
rmse/MHz
0.097
a Comments as for Table 7. b Preset at this value taken from Table
1. c Moments of inertia with dimension 10-20 m2 u. Conversion factor:
5053.7905 nm2 u MHz. d Number of lines used in the fit. e Root-mean-
square deviation.
will transform one form into its mirror image. A double-
minimum potential is associated with this transformation. Each
vibrational state is thus split into two components. The lowest
energy component of the ground vibrational state is denoted
the (+)-state, whereas the higher energy component is denoted
the (-)-state. a-type transitions will obey ordinary rotational
transitional rules plus (+) r (+), resulting in two closely spaced
transitions separated by a few megahertz at most. The c-type
lines obey ordinary rotational selection rules plus (-) r (+),
or (+) r (-). This leads to comparatively much larger splittings
for c-type lines than in the case of the a-type transitions.
The J ) 1 r 0, J ) 2 r 1 and J ) 3 r 2 a-type series of
transitions of the gauche rotamer was readily assigned using
the Fourier transform spectrometer. Splittings were resolved for
most, but not all, of these transitions, as can be seen in Table
4S in Supporting Information. The largest splitting is 1.646 MHz
found for the 21,2 r 11,1 transition (Table 4S). The splittings of
those transitions that were not resolved are less than 10 kHz.
Attempts to fit these aR-lines to Watson’s Hamiltonian35
within the measurement accuracy of 3 kHz failed. The reason
for this is presumed to be coupling between rotation and
tunneling, which is often seen in such cases. This coupling
makes it impossible to assign correctly the (+) or the (-) label
to each of the split lines in Table 4S (Supporting Information).
The first tentative spectroscopic constants obtained from
Conclusions
There are two stable conformers of allenylphosphine, the more
stable syn form and the less stable gauche rotamer. The energy
difference is 2.1(4) kJ/mol. This is the same conformational
preference as that found in vinylphosphine.6 Accurate equilib-
rium geometries have been derived for the two stable rotamers
of allenylphosphine, and for the two corresponding conformers
of vinyl phosphine in high-level ab initio calculations. The most
important geometrical parameters warrant comments: The C-P
bond lengths are slightly longer in allenylphosphine (182.9 pm;
Table 5) than in the vinylphosphine conformers (182.4 pm in
syn, 181.9 pm in gauche; Table 3). The CdC double bond length
in vinylphosphine is 133.5 pm in both forms (Table 3). The
two CdC bond lengths are definitely shorter than this in
allenylphosphine (130.7 or 130.9 pm; Table 5), which on the
other hand is practically the same as in the parent allene
molecule.29
a
a-type lines in Table 4S were next used to predict further R-
transitions involving higher-J transitions. These transitions
appear at frequencies not accessible for the Fourier transform
spectrometer. Their assignments were now readily achieved
using the Stark-modulated spectrometer. Transitions involving
J up to J ) 11 were found in this way. None of these appeared
to be significantly perturbed by tunneling. The assignments of
these transitions were confirmed not only by their fit to the
Watson Hamiltonian,35 but by RFMWDR experiments, Stark
modulation patterns, partially resolved Stark effects and intensi-
ties as well. The spectrum used to determine the spectroscopic
constants of the gauche rotamer is found in Table 5S in
Supporting Information. Some of the low-J lines measured by
Fourier transform spectroscopy have been included in the fit.
Their average frequencies have been used in these cases. Unit
weights were applied in the fitting of these transitions because
the amount of tunneling perturbation is unknown. The spectro-
Acknowledgment. We thank Anne Horn for her most
helpful assistance. A grant from the French-Norwegian Aurora
Program to J.-C.G. and H.M. is gratefully acknowledged. J.-
C.G. thanks the Centre National d’Etudes Spatiales (CNES) for
financial support. H.M. thanks the Research Council of Norway
(Programme for Supercomputing) for a grant of computer time.
a
scopic constants derived from the R-lines are listed in Table
9. Only two quartic centrifugal distortion constants ∆J and ∆JK
were fitted. The remaining quartic constants were fixed at the
theoretical values shown in Table 1. The agreement between
the calculated (Table 1) and experimental values for ∆J and
Supporting Information Available: Assigned MW spectra
(Tables 1S-6S); additional results from the quantum chemical
∆
JK
(Table 9) is satisfactory.

Structural Properties of 1,2-Propadienylphosphine
J. Phys. Chem. A, Vol. 109, No. 1, 2005 121
Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G. A.;
Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels,
A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.;
Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.;
Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz,
P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.;
Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson,
B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian 03,
Revision B.03; Gaussian, Inc.: Pittsburgh, PA, 2003.
(18) Becke, A. D. J. Chem. Phys. 1993, 98, 5648.
(19) Lee, C.; Yang, W.; Parr, R. G. Phys. ReV. B 1988, 37, 785.
(20) Curtiss, L. A.; Raghavachari, K.; Redfern, P. C.; Rassolov, V.;
Pople, J. A. J. Chem. Phys. 1998, 109, 7764.
(21) Pople, J. A.; Head-Gordon, M.; Raghavachari, K. J. Chem. Phys.
1987, 87, 5968.
(22) Curtiss, L. A.; Carpenter, J. E.; Raghavachari, K.; Pople, J. A. J.
Chem. Phys. 1992, 96, 9030.
(23) Møller, C.; Plesset, M. S. Phys. ReV. 1934, 46, 618.
(24) Mele´ndez, F. J.; Gallego-Luxan, B.; Demaison, J.; Smayars, Y. G.
J. Comput. Chem. 2000, 21, 1167.
(25) Purvis, G. D., III; Bartlett, R. J. J. Chem. Phys. 1982, 76, 1910.
(26) Raghavachari, K.; Trucks, G. W.; Pople, J. A.; Head-Gordon, M.
Chem. Phys. Lett. 1989, 157, 479.
(27) (a) Dunning, T. H., Jr. J. Chem. Phys. 1989, 90, 1007. (b) Dunning,
T. H., Jr.; Peterson, K. A.; Wilson, A. K. J. Chem. Phys. 2001, 114, 9244.
(28) Margule`s, L.; Demaison, J.; Boggs, J. E. J. Mol. Struct.
(THEOCHEM) 2000, 500, 245.
(29) Margule`s, L.; Demaison, J.; Boggs, J. E. Struct. Chem. 2000, 11,
145.
(30) McRae, G. A.; Gerry, M. C. L.; Cohen, E. A. J. Mol. Spectrosc.
1986, 116, 58.
(31) Margule`s, L.; Demaison, J.; Rudolph, H. D. J. Mol. Struct. 2001,
calculations (Tables 7S-10S). This material is available free
of charge via the Internet at http://pubs.acs.org.
References and Notes
(1) Groner, P.; Johnson, R. D.; Durig, J. R. J. Chem. Phys. 1988, 88,
3456.
(2) Marstokk, K.-M.; Møllendal, H. Acta Chem. Scand., Ser. A 1983,
37, 755.
(3) Marstokk, K.-M.; Møllendal, H. Acta Chem. Scand. 1996, 50, 875.
(4) Demaison, J.; Guillemin, J.-C.; Møllendal, H. Inorg. Chem. 2001,
40, 3719.
(5) Møllendal, H.; Demaison, J.; Guillemin, J.-C. J. Phys. Chem. A
2002, 106, 11481.
(6) (a) Dre´an, P.; Le Guennec, M.; Lo´pez, J. C.; Alonso, J. L.; Denis,
J. M.; Kreglewski, M.; Demaison, J. J. Mol. Spectrosc. 1994, 166, 210. (b)
Dre´an, P.; Colmont, J.-M.; Lesarri, A.; Lo´pez, J. C. J. Mol. Spectrosc. 1996,
176, 180.
(7) Cohen, E. A.; McRae, G. A.; Goldwhite, H.; Di Stefano, S.;
Beaudet, R. A. Inorg. Chem. 1987, 26, 4000.
(8) Guillemin, J.-C.; Savignac, P.; Denis, J. M. Inorg. Chem. 1991,
30, 2170.
(9) Lacombe, S.; Dong, W.; Pfister-Guillouzo, G.; Guillemin, J.-C.;
Denis, J. M. Inorg. Chem. 1992, 31, 4425.
(10) Ashby, E. C.; Prather, J. J. Am. Chem. Soc. 1966, 88, 729.
(11) Guirgis, G. A.; Marstokk, K.-M.; Møllendal, H. Acta Chem. Scand.
1991, 45, 482.
(12) Wodarczyk, F. J.; Wilson, E. B., Jr. J. Mol. Spectrosc. 1971, 37,
445.
(13) Marstokk, K.-M.; Møllendal, H. Acta Chem. Scand., Ser. A 1988,
42, 374.
(14) Waal, Ø., personal communication, 1994.
(15) Kassi, S.; Petitprez, D.; Wlodarzcak, G. J. Mol. Struct. 2000, 517-
518, 375.
(16) Hampel, C.; Peterson, K. A.; Werner, H. J. Chem. Phys. Lett. 1992,
190, 1.
599, 23.
(32) Lee, T. J.; Taylor, P. R. Int. J. Quantum Chem., Quantum Chem.
Symp. 1989, 23, 199.
(33) Peterson, K. A.; Dunning, T. H., Jr. J. Chem. Phys. 2002, 117,
10548.
(17) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
M. A.; Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.; Kudin, K.
N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.;
Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.;
Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.;
Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.; Li,
X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Adamo, C.; Jaramillo, J.;
Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.;
(34) Demaison, J.; Boggs, J. E.; Rudolph, H. D. J. Mol. Struct. 2004,
695-696, 145.
(35) Watson, J. K. G. Vibrational Spectra and Structure; Elsevier:
Amsterdam, 1977; Vol. 6.
(36) Laurie, V. W.; Herschbach, D. R. J. Chem. Phys. 1962, 37, 1687.
(37) Esbitt, A. S.; Wilson, E. B. ReV. Sci. Instrum. 1963, 34, 901.
(38) Muenter, J. S. J. Chem. Phys. 1968, 48, 4544.