Gas-phase basicity and acidity trends in α,β-unsaturated amines, phosphines, and arsines

Article abstract of DOI:10.1021/ja982657e

The acidity and basicity trends in the series of α,β-unsaturated amines, phosphines, and arsines were analyzed through the use of G2 ab initio calculations and the examination of experimental data obtained by means of FT-ICR techniques. The α,β-unsaturated amines, phosphines, and arsines are less basic but significantly more acidic than the corresponding saturated analogues. However, while both vinyl- and ethynylamine protonate preferentially at the β-carbon atom, vinyl- and ethynylphosphine are phosphorus bases in the gas-phase. Arsines resemble closely the corresponding phosphines, although protonation at the Ca atom competes with protonation at the heteroatom. The enhanced acidity of unsaturated compounds can be attributed essentially to a stabilization of the anions because of a favorable interaction of the XH^- group with the C-C multiple bonds. This stabilizing effect is maximum for amines and minimum for arsines. The low relative basicity of these unsaturated compounds results from a destabilization of the protonated species due to unfavorable interactions of the XH₃⁺ group with the C-C π-system. Protonation at the β-carbon is strongly favored for amines but unfavorable for phosphines and arsines, because the carbocation formed is much less stabilized when the heteroatom of the XH₂ group belongs to the second or the third-row than when it is a first-row atom.

Full text of DOI:10.1021/ja982657e

J. Am. Chem. Soc. 1999, 121, 4653-4663
4653
Gas-Phase Basicity and Acidity Trends in R,â-Unsaturated Amines,
Phosphines, and Arsines
Otilia M o´ ,* Manuel Y a´ n˜ ez, Mich e` le Decouzon, Jean-Fran c¸ ois Gal,*
Pierre-Charles Maria, and Jean-Claude Guillemin*
Contribution from the Departamento de Qu ´ı mica, C-9, UniVersidad Aut o´ noma de Madrid, Cantoblanco.
2
8049 Madrid, Spain, Groupe FT-ICR, GREFCO-Chimie Physique Organique, UniVersit e´ de Nice-Sophia
Antipolis, Parc Valrose, 06108 Nice Cedex 2, France, and Laboratoire de Synth e` se et ActiVation de
Biomol e´ cules, ESA 6052, ENSCR, 35700 Rennes Cedex, France
ReceiVed July 27, 1998
Abstract: The acidity and basicity trends in the series of R,â-unsaturated amines, phosphines, and arsines
were analyzed through the use of G2 ab initio calculations and the examination of experimental data obtained
by means of FT-ICR techniques. The R,â-unsaturated amines, phosphines, and arsines are less basic but
significantly more acidic than the corresponding saturated analogues. However, while both vinyl- and
ethynylamine protonate preferentially at the â-carbon atom, vinyl- and ethynylphosphine are phosphorus bases
in the gas-phase. Arsines resemble closely the corresponding phosphines, although protonation at the CR atom
competes with protonation at the heteroatom. The enhanced acidity of unsaturated compounds can be attributed
-
essentially to a stabilization of the anions because of a favorable interaction of the XH group with the C-C
multiple bonds. This stabilizing effect is maximum for amines and minimum for arsines. The low relative
basicity of these unsaturated compounds results from a destabilization of the protonated species due to
+
unfavorable interactions of the XH3 group with the C-C π-system. Protonation at the â-carbon is strongly
favored for amines but unfavorable for phosphines and arsines, because the carbocation formed is much less
stabilized when the heteroatom of the XH2 group belongs to the second or the third-row than when it is a
first-row atom.
Introduction
almost complete, probably due to the low stability of the R,â-
unsaturated phosphines and arsines. This situation prompted us
to accomplish the synthesis and the study of these particular
low-stable systems. Unexpectedly, in our investigation of R,â-
The enamine and ynamine functional groups are good
examples of ambident reactants, since they exhibit nucleophilic
reactivity at both the nitrogen and the â-carbon atoms. Indeed,
it seems well established that protonation of enamines in the
gas phase occurs at carbon leading to the formation of the
1
2
5
unsaturated arsines, we found that the aforementioned interac-
tion between the heteroatom lone-pair and the unsaturated
4
moiety, which had been invoked to explain the anomalous
3
corresponding iminium ion. A similar conclusion was reached
by Smith and Radom on theoretical grounds with respect to
basicity and acidity of ethynylamine, seemed to depend on the
nature of the heteroatom. Actually, although vinyl- and ethy-
nylarsine, like the corresponding amines, are stronger acids and
weaker bases than their saturated counterpart ethylarsine,
protonation at the â-carbon is no longer the most favorable
processes, since there is a competition between protonation at
the heteroatom and protonation at the R-carbon. Hence, the main
question to be answered is whether this behavior is a peculiarity
of the arsines or, on the contrary, it is also found for other
unsaturated bases which have a Va group atom as basic center.
To answer this question it seems unavoidable to carry out a
systematic study of the basicity and acidity trends at least along
the series of R,â-unsaturated amines, phosphines, and arsines,
which will be afforded for the first time in this paper. To
complete the necessary set of data to carry out this analysis we
measure the gas-phase acidities of the R,â-unsaturated phos-
phines, as well as the gas-phase acidity of vinylamine, which
4
the protonation of ethynylamine as a prototype of ynamine
function. On the other hand, ethynylamine exhibits a high
relative acidity but a low relative basicity. Its high acidity has
been attributed to stabilization of the ethynylamide anion by
the acetylenic linkage, while its low relative basicity is mainly
due to a destabilization of the ethynylammonium cation. One
may assume that, in general, when an heteroatom bearing an
electron lone-pair is attached to an unsaturated moiety, the
interaction between both functions may have dramatic effects
on the reactivity of the system.
Quite surprisingly, the lack of information on analogous
systems containing phosphorus or arsenic as heteroatoms was
(1) For instance, see Dyke, S. F. The Chemistry of Enamines; Cambridge
University Press: London, 1973.
(
2) (a) Liler, M. AdV. Phys. Org. Chem. 1975, 11, 267. (b) Elguero, J.;
Jacquier, R.; Tarrago, G. Tetrahedron Lett. 1965, 471. (c) Elguero, J.;
Jacquier, R.; Tarrago, G. Tetrahedron Lett. 1966, 1112. (d) Alais, L.;
Michelot, R.; Tchovbar, B. Compt. Rend. Acad. Sci., Ser. C, 1971, 273,
6
were not reported in the literature so far.
(5) Guillemin, J.-C.; Decouzon, M.; Maria, P.-C.; Gal, J.-F.; M o´ , O.;
Y a´ n˜ ez, M. J. Phys. Chem. 1997, 101, 9525.
2
61.
(
3) Ellenberger, M. R.; Dixon, D. A.; Farneth, W. E. J. Am. Chem. Soc.
(6) National Institute of Standards and Technology. Chemistry WebBook;
NIST Standard Reference Database 69; National Institute of Standards and
Technology, Department of Commerce: Washington, DC, 1998.
1
981, 103, 5377.
(4) Smith, B. J.; Radom, L. J. Am. Chem. Soc. 1992, 114, 36.
1
0.1021/ja982657e CCC: $18.00 © 1999 American Chemical Society
Published on Web 05/01/1999

4654 J. Am. Chem. Soc., Vol. 121, No. 19, 1999
M o´ et al.
The subsequent step will be the rationalization of the
under study and the reference compound, differing by at least
differences observed between the three families of compounds.
For this purpose we will take advantage of high-level ab initio
molecular orbital calculations which will allow us to explain in
a systematic manner the origin of these differences in terms of
dissimilarities in their bonding characteristics, charge distribu-
tions, and molecular orbital interactions.
a factor of 3, were used, with total pressures in the range (2-
-
5
8) × 10 Pa (as read on a Bayard Alpert ion gauge). Relative
(to N2) sensitivities Sr of the Bayard Alpert gauge have been
estimated using the Bartmess and Georgiadis equation:14
S
) 0.36R + 0.30
(1)
r
Experimental Section
The molecular polarizability R was taken as R(ahc), calculated
15
using the atomic hybrid component (τ) approach of Miller.
Safety Considerations. Phosphines are pyrophoric and
potentially highly toxic compounds. All reactions and handling
should be carried out in a well-ventilated hood.
Chemicals. The preparation of vinylamine, vinyl-, and eth-
ynylphosphines has been improved to obtain very pure com-
pounds. The ethenyl- and ethynylphosphines were prepared by
reduction of the corresponding phosphonates with AlHCl2 in
3/2
For phosphorus, Miller gave τP(ahc) ) 2.485 Å , derived from
experimental polarizabilities of a few phosphates. For the sake
5
,13
of consistency with studies on arsines we preferred to use a
3
/2
value τP(ahc) ) 3.637 Å , back-calculated from Sr(PH3) )
.975 ( 0.021, determined experimentally in our laboratory,
using a spinning rotor gauge (Leybold Vakuum GmbH, Cologne,
1
1
6
Germany).
7
,8
tetraethylene glycol dimethyl ether as previously described.
For acidity determinations of phosphines, negative ions were
generated by proton abstraction from the neutral reactant by
Ethylphosphine was prepared by reduction of the ethylphos-
phonic acid diethyl ester by LiAlH4 in tetraethylene glycol
dimethyl ether. Each compound was purified in a vacuum line
by trap-to-trap distillation. At a pressure of 10 Pa, the high
boiling impurities were selectively removed in a trap cooled at
-
t-BuO . This anion was obtained via electron ionization at 0.1
eV (nominal) of t-BuONO, introduced in the spectrometer at a
-
5
partial pressure of about 10 Pa. The proton-transfer reactions
were monitored for about 10 s without the manifestation of
significant secondary reactions. Vinylamine is not deprotonated
-100 °C and the substituted phosphine was condensed in a trap
cooled at -120 °C. The most volatile impurity (mainly PH3)
was not condensed at -120 °C and was thus removed. The cell
containing the expected phosphine was isolated from the vacuum
line by stopcocks and directly fitted to the inlet system of the
FT-ICR mass spectrometer.
-
-
by t-BuO ; consequently, we used instead MeO generated
1
7
according to Caldwell and Bartmess for the equilibrium
measurement against c-PrCN. No other equilibrium constant
could be determined. Vinylamine acidity was bracketed by
observing the reactions with t-BuO (from t-BuONO) and EtO
from ionization of EtOH by MeO ), isolated by broad rf
-
-
Vinylamine has been prepared, as previously reported, by
flash vacuum pyrolysis of cyclobutylamine at 800 °C at a
pressure of 10 Pa.9 Ethene was removed by trapping the
pyrolysis products in a trap cooled at -110 °C. The presence
of traces of ammonia and residual cyclobutylamine cannot be
completely avoided. Vinylamine can be vaporized and then
condensed at the liquid nitrogen temperature without isomer-
ization to ethylideneamine. Vinylamine decomposes slowly, in
-
(
resonant ejection. No equilibrium constants could be determined
for proton transfer between the anions of the two alcohols
(ethanol and tert-butyl alcohol) and vinylamine. A small quantity
of acetonitrile was present in the FT-ICR spectrometer, due to
either an impurity in the vinylamine, or to some pyrolysis on
the filament. The higher acidity of acetonitrile leads to disap-
pearance of all ions other than m/z ) 40 (NC-CH2-), after about
13
pure form at -80 °C, or diluted in CD2Cl2 at -10 °C. C NMR
2-4 s, and impeding the attainment of equilibrium. Equilibrium
1
(
CD2Cl2, 190 K): δ 84.2 (t, JCH ) 156.2 Hz, CH2), 138.4 (d,
constants were obtained at a ICR cell temperature of 338 K.
The trapping plate on the filament side is at about 20-30° above
the opposite trapping plate, as measured with thermocouples
1
JCH ) 165.8 Hz, CH). MS (Varian MAT 311) exact mass
+•
determination: [C2H5N] calcd, 43.04220; measured, 43.0425.
+
[C2H4N] : calcd, 42.03437; measured, 42.0344. Mass spec-
(
80-85 vs 55-60 °C, depending on the filament current). All
the other plates (excitation and detection electrodes) are at about
5 ( 5°. The weighted average taking into account the surfaces
trum: m/z(%), 44(3.0), 43(79), 42(100), 41(18), 40(16). The
bulb containing vinylamine at -90 °C, was isolated from the
vacuum line, fitted to the inlet system of the FT-ICR mass
spectrometer and kept at -90 °C during all the measurements.
The compounds used as references for the gas-phase acidity
determinations are of the highest purity commercially available.
They were used without further purification, except for several
freeze, pump, thaw cycles in the spectrometer inlet system.
6
of the six different plates, is actually 65 °C, which is taken as
the experimental temperature. Tabulated ∆acidG° of reference
compounds refer to the standard temperature of 298.15 K. As
5,13
explained previously,
temperature corrections are minor as
compared with other experimental uncertainties. Experimental
gas-phase acidities of ethyl-, vinyl-, ethynylphosphine, and
vinylamine are reported in Table 1, and the respective absolute
FT-ICR Measurements and Results
∆acidG° do not include such temperature corrections.
Proton-transfer equilibrium measurements were conducted on
an electromagnet Fourier transform ion cyclotron resonance
Computational Details
(
FT-ICR) mass spectrometer built at the University of Nice-
To obtain reliable estimates of the gas-phase basicity and
acidity of the amines and phosphines under investigation, we
have carried out ab initio calculations in the framework of the
Sophia Antipolis, using the methodology described pre-
viously.1
0-13
Variable pressure ratios between the unknown acid
(
(
7) Cabioch, J.-L.; Denis, J.-M. J. Organomet. Chem. 1989, 377, 227.
8) Guillemin, J.-C., Savignac, P.; Denis, J.-M. Inorg. Chem. 1991, 30,
(12) Decouzon, M.; Gal, J.-F.; Herreros, M.; Maria, P.-C.; Murrel, J.;
Todd, J. F. J. Rapid Commun., Mass Spectrom. 1996, 10, 242.
(13) Decouzon, M.; Gal, J.-F.; Guillemin, J.-C.; Maria, P.-C. Int. J. Mass
Spectrom. Ion Processes 1998, 179, 27.
(14) Bartmess, J. E.; Georgiadis, R. Vacuum 1983, 33, 149.
(15) Miller, K. J. J. Am. Chem. Soc. 1990, 112, 8533.
(16) Decouzon, M.; Gal, J.-F.; G e´ ribaldi, S.; Maria, P.-C.; Rouillard,
M.; Vinciguerra, A. Analusis 1986, 14, 471.
(17) Caldwell, G.; Bartmess, J. E. Org. Mass Spectrom. 1982, 17, 456.
2
170.
9) Albrecht, B.; Allan, M.; Haselbach, E.; Neuhaus, L.; Carrupt, P.-A.
HelV. Chim. Acta 1984, 67, 220.
10) Berthelot, M.; Decouzon, M.; Gal, J.-F.; Laurence, C.; Le Questel,
J.-Y-; Maria, P.-C.; Tortajada, J. J. Org. Chem. 1991, 56, 4490.
11) Maria, P.-C.; Leito, I.; Gal, J.-F.; Exner, O.; Decouzon, M. Bull
Soc. Chim. Fr. 1995, 132, 394.
(
(
(

Unsaturated Amines, Phosphines, and Arsines
J. Am. Chem. Soc., Vol. 121, No. 19, 1999 4655
Table 1. Gas-Phase Acidities (kcal/mol) of Phosphines and
Vinylamine from Proton-Transfer Equilibrium Constant
Determinations and Bracketing Experiments
optimization was carried out at the B3LYP/6-31G* level. For
the anions, both MP2 and B3LYP geometry optimizations were
performed using a 6-31+G(d,p) basis set expansion. As
expected, for neutral and protonated species the inclusion of
diffuse components in the basis set leads to negligible changes
in the optimized geometries. The corresponding harmonic
vibrational frequencies were calculated at the same level of
theory used for the geometry optimization and the corresponding
ZPE corrections were scaled by the empirical factor 0.98.24 The
final energies were obtained using a 6-311+G(3df,2p) basis set
which was found to be adequate to reproduce the basicity and
∆
G°acid
∆∆G°acid
∆G°acid
a
b
(AH)c
AH
RefH
(RefH)
(338 K)
ethylphosphine
MeCO2Me
MeCN
i-PrCN
365.1
365.2
367.9
+0.92 ( 0.23
-0.28 ( 0.14
-1.39 ( 0.03
n-C5H11OH 368.2
CF3CH2OH 354.1
CF3CH2OH 354.1
HCONHMe 354.0
-2.17 ( 0.21 365.9 ( 0.8
-0.11 ( 0.06 354.0
-0.98 ( 0.04
HCONHMe
vinylphosphine
d
-1.22 ( 0.09 353.0 ( 0.3
+0.19 ( 0.07
ethynylphosphine n-PrSH
347.9
348.9
350.6
22
5
acidity of basis containing first- and third-row atoms. All the
ab initio and DFT calculations have been performed with the
EtSH
MeSH
-0.30 ( 0.04
-1.45 ( 0.08
2
5
CF3CH2OH 354.1 ,-3
348.6 ( 0.3
Gaussian-94 series of programs.
vinylamine
c-PrCN
t-BuOH
EtOH
367.8
367.7
371.7
+1.8 ( 0.2
g+2
g-2
Differences in reactivity usually reflect important dissimilari-
ties in charge distributions. We have used two different partition
techniques, namely the atoms in molecules (AIM) theory of
369.6
a
Absolute gas-phase acidities (Gibbs energies at 298.15 K for the
26
Bader and the natural bond orbital (NBO) analysis of Weinhold
-
+
b
reaction RefH f Ref + H ) from ref 6. Gibbs energies for the
27
-
-
et al. to analyze the charge distributions of amines, phosphines,
and arsines. The AIM theory is based in a topological analysis
reaction AH + Ref f A + RefH; quoted uncertainties correspond
to the standard deviation for three to four measurements. Upper and
c
2
lower limits were obtained from bracketing experiments. Absolute gas-
of the electron charge density F(r) and its Laplacian ∇ F(r).
phase acidities; no temperature correction applied, see text; quoted
uncertainties (standard deviations) correspond to the overlap quality,
and indicate the consistency with the existing absolute gas-phase acidity
scale. The uncertainty on this scale ((2 kcal/mol) should be added to
More specifically, we have located the so-called bond critical
points, i.e., points where F(r) is minimum along the bond path
and maximum in the other two directions. In general the values
d
2
evaluate the overall uncertainty on each value. Reevaluation, this work.
of F and ∇ F at these points provide useful information on the
bonding characteristics. In most cases negative values of the
Laplacian are associated with covalent linkages, while positive
values are usually associated with closed-shell interactions as
those found in ionic bonds, van der Waals complexes, and
hydrogen bonds. The NBO formalism permits a description of
the different bonds of the system in terms of the natural hybrid
orbitals centered on each atom and provides also useful
information on the charge distribution of the system.
G2-theory.18 G2 is a composite theory based on the 6-311G-
(d,p) basis set and several basis extensions, where electron
correlation effects are treated at the MP4 and QCISD(T) levels
of theory. The final energies are effectively at the QCISD(T)/
6
-311+G(3df,2p) level, assuming that basis set effects on the
correlation energies are additive. A small empirical correction
HLC) to accommodate remaining deficiencies is finally added
(
as well as the corresponding zero point energy (ZPE) correction,
estimated at the HF/6-31G* level. The reader is addressed to
ref 18 for a complete description of this method. Also, an
assessment of the G2 theory for the computation of enthalpies
of formation has recently been published.19
Since as mentioned above, electron correlation effects are
crucial for the systems under study, both population analyses
were performed at the MP2 level. For arsines the MP2-optimized
geometries reported in ref 5 were used. AIM calculations have
28
been carried out by using the AIMPAC series of programs.
For the particular case of phosphines we have considered it
of interest to investigate the performance of the B3LYP density
functional theory method,20 because it has been shown to yield
not only reliable geometries and vibrational frequencies for a
great variety of systems,21 but also to provide basicities and
acidities in fairly good agreement both with experimental
values,22 when available, and with high-level ab initio calcula-
tions, provided a flexible enough basis set is used. The B3LYP
approach combines the Becke’s three parameter non local hybrid
exchange potential with the non local correlation functional of
Lee, Yang, and Parr.23 In these DFT calculations, the geometry
Results and Discussion
Structures. Protonation and Deprotonation Effects. The
optimized geometries of ethyl-, vinyl-, and ethynylphosphine
and their protonated and deprotonated species are given as
Supporting Information. For the sake of comparison, the C-C
and C-X (X ) N, P, As) MP2(full)/6-31G* optimized bond
distances for the corresponding amines and arsines have been
summarized in Table 2.
It can be observed that the C-X bond length decreases in
the order ethyl > vinyl > ethynyl for the three series of
compounds. Consistently the charge density at the corresponding
bond critical points increases in the reverse order (See Table
(
18) Curtiss, L. A.; Raghavachari, K.; Trucks, G. W.; Pople, J. A. J.
Chem. Phys. 1991, 94, 7221.
19) Curtiss, L. A.; Raghavachari, K.; Redfern, P. C.; Pople, J. A. J.
Chem. Phys. 1997, 106, 1063.
20) (a) Becke, A. D. J. Chem. Phys. 1993, 98, 5648. (b) Becke, A. D.
J. Chem. Phys. 1992, 96, 2155.
21) Llamas-Saiz, A. L.; Foces-Foces, C.; M o´ , O.; Y a´ n˜ ez, M.; Elguero,
(
(
(24) Bauschlicher Jr., C. W. Chem. Phys. Lett. 1995, 246, 40.
(25) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Gill, P: M.W.;
Johnson, B. J.; Robb, M. A.; Cheeseman, J. R.; Keith, T. A.; Peterson, G.
A.; Montgomery, J. A.; Raghavachari, K.; Al-Laham, M. A.; Zakrzewski,
V. G.; Ortiz, J. V.; Foresman, J. B.; Cioslowski, J.; Stefanow, B. B.;
Nanayaklara, A.; Challacombe, M.; Peng, C. Y.; Ayala, P. Y.; Chen, W.;
Wong, M. W.; Andres, J. L.; Replogle, E. S.; Gomperts, R.; Martin, R. L.;
Fox, D. J.; Binkley, J. S.; Defrees, D. J.; Baker, J.; Stewart, J. P.; Head-
Gordon, M.; Gonzalez, C.; Pople, J. A.; Gaussian 94; Gaussian, Inc.:
Pittsburgh, PA, 1995.
(
J. J. Comput. Chem. 1995, 16, 263. Flori a´ n, J.; Johnson, B. G. J. Phys.
Chem. 1994, 98, 3681. Bauschlicher, C. W. Chem. Phys. Lett. 1995, 246,
4
1
0; Martell, J. M.; Goddard, J. D.; Eriksson, L. A. J. Phys. Chem. A 1997,
01, 1927. Cui, Q.; Musaev, D. G.; Svensson, M.; Seiber, S.; Morokuma,
K. J. Am. Chem. Soc. 1995, 117, 12366. Ricca, A.; Bauschlicher, C. W. J.
Phys. Chem. 1995, 99, 5922. Ziegler, T. Chem. ReV. 1991, 91, 651.
(22) (a) Smith, B. J.; Radom, L. J. Am. Chem. Soc. 1993, 115, 4885. (b)
Smith, B. J.; Radom, L. Chem. Phys. Lett. 1994, 231, 345. (c) Gonz a´ lez,
A. I.; M o´ , O.; Y a´ n˜ ez, M.; Le o´ n, E.; Tortajada, J.; Morizur, J.-P.; Leito, I.;
Maria, P.-C.; Gal, J.-F. J. Phys. Chem. 1996, 100, 10490. (d) Amekraz, B.;
Tortajada, J.; Morizur, J.-P.; Gonz a´ lez, A. I.; M o´ , O.; Y a´ n˜ ez, M.; Leito, I.;
Maria, P.-C.; Gal, J.-F. New J. Chem. 1996, 20, 1011.
(26) Bader, R. F. W., Atoms in Molecules. A Quantum Theory. Oxford
University Press: Oxford, 1990.
(27) Weinhold, F.; Carpenter, J. E.; The Structure of Small Molecules
and Ions; Plenum: New York, 1988.
(28) The AIMPAC programs package has been provided by J. Cheeseman
and R. F. W. Bader.
(23) Lee, C.; Yang, W.; Parr, R. G. Phys ReV. 1988, B37, 785.

4656 J. Am. Chem. Soc., Vol. 121, No. 19, 1999
M o´ et al.
Table 2. C-C and C-X(N,P,As) MP2(full)/6-31G* Optimized Bond Distances (in angstroms) for Ethyl-, Vinyl-, and Ethynylamine,
-Phosphine, and -Arsine and Their Protonated and Deprotonated Species. The Charge Density (in au) at the Corresponding bcp is Given within
Parentheses
+
NH-
+
PH-
+
AsH
3
AsH-
NH
2
NH
3
PH
2
PH
3
AsH
2
ethyl-C-C
C-X
1.519
0.242)
1.466
0.250)
1.340
0.336)
1.400
0.296)
1.218
0.395)
1.367
0.322)
1.513
(0.246)
1.519
(0.210)
1.326
(0.349)
1.483
(0.229)
1.210
1.556
(0.224)
1.417
(0.291)
1.386
(0.302)
1.340
(0.344)
1.271
1.527
(0.240)
1.862
(0.149)
1.339
(0.341)
1.824
(0.155)
1.224
1.533
(0.234)
1.809
(0.210)
1.341
(0.340)
1.776
(0.177)
1.222
1.528
(0.239)
1.900
(0.291)
1.369
(0.320)
1.788
(0.152)
1.242
1.525
(0.237)
1.968
(0.126)
1.338
(0.340)
1.932
(0.134)
1.222
1.529
(0.234)
1.909
(0.141)
1.339
(0.340)
1.875
(0.152)
1.221
1.524
(0.236)
2.007
(0.111)
1.364
(0.322)
1.907
(0.135)
1.224
(
(
(
(
(
(
vinyl-CdC
C-X
ethynylCtC
C-X
(0.401)
1.410
(0.278)
(0.353)
1.305
(0.373)
(0.400)
1.773
(0.156)
(0.401)
1.708
(0.181)
(0.385)
1.768
(0.148)
(0.403)
1.872
(0.139)
(0.405)
1.803
(0.167)
(0.389)
1.877
(0.133)
Scheme 1
Scheme 2
Scheme 3
2), reflecting the interaction between the lone pair of the
heteroatom and the C-C π-system, through the contribution
of b and d type valence structures (See Scheme 1). This is
consistent with the fact that for the three series of compounds
the lowest vibrational frequency, which is associated with the
XH2 torsion, increases in the order ethyl < vinyl < ethynyl.
For instance, while for ethylphosphine the C-PH2 torsion is
Protonation effects on amines are significantly different from
those observed in phosphines and arsines. Protonation at the
heteroatom in the latter leads to a significant shortening of the
C-X bond and to a lengthening of the C-C linkage, while in
amines the opposite effects are observed. These differences can
be explained in terms of the charge density redistribution which
takes place in the protonation process. As discussed by Alcam´ı
-
1
predicted to appear at 169 cm , for vinylphosphine it should
-
1
appear at 194 cm . Similarly, while the C-P stretching
-
1
frequency for ethylphosphine is 634 cm , for vinylphosphine
-
1
the frequency is 653 cm .
Interestingly, this shortening of the C-X bond distance in
R,â-unsaturated compounds is larger for phosphines and arsines
than for amines. Since phosphorus and arsenic are less elec-
tronegative than nitrogen they favor, to a greater extent, the
contribution of valence structure b (See Scheme 1) where the
heteroatom bears a formal positive charge. These differences
between amines, phosphines and arsines are also mirrored in
their NBO analysis (See Table 1 of the Supporting Information),
which shows that, whereas in amines the higher contribution
to the C-X bonding orbital comes from the N orbitals (55%),
in phosphines and arsines the highest contribution comes from
the CR orbitals (60%), which present also a much higher s
character. The C-C bonding characteristics of amines are on
the contrary rather similar to those found for phosphines and
arsines. Consistently, for the vinyl derivatives the differences
in the C-C bond lengths are negligible, and for the ethynyl
derivatives, the C-C bond is longer in phosphines and in ar-
sines as expected from a greater contribution of valence struc-
ture b. The NBO analysis also shows that the N-hybrids in-
volved in the bonding MOs have a much higher s character
2
9
et al., in general, protonation at the most electronegative
atom of a bond results in a weakening of the bond, while pro-
tonation at the less electronegative atom reinforces the linkage.
Consistently, protonation of amines at the nitrogen atom implies
-
1
significant red shifts (106, 230 and 88 cm ) of the C-N
stretching mode, while protonation at the heteroatom in phos-
-
1
phines and arsines implies a blue shift (80, 23, and 90 cm
-
1
and 77, 111, and 86 cm , respectively) of the C-P and
C-As stretching modes. This is also in agreement with the
fact, shown by Boyd and Boyd, that protonation of amines at
the nitrogen atom favors the heterolytic cleavage of the C-X
bond, while for phosphines the C-X heterolytic bond dissocia-
tion energy is greater for the protonated than for the neutral
species.
3
0
Deprotonation effects, on the contrary, are qualitatively
similar for R,â-unsaturated amines, phosphines, and arsines in
the sense that in all cases a shortening of the C-X bond takes
place, while the C-C linkage becomes slightly longer. These
effects are more important for the ethynyl than for the vinyl
function, and point to a significant participation of type d
valence structures (See Scheme 2). These effects are again
reflected in concomitant shifts of the C-C stretching frequen-
cies, which for the anion appear systematically at lower
frequencies than for the neutral.
(33%) than the P (17%) or As hybrids (13%) due to the higher
electronegativity of nitrogen. The necessary orthogonality
conditions force the s character of the P- and As-lone pairs to
be significantly higher than the s character of the N-lone pairs
and therefore, phosphines and arsines must be, in general,
weaker bases than amines, when protonation takes place at the
heteroatom, in agreement with the experimental evidence (vide
infra).
(29) Alcam ´ı , M.; M o´ , O., Y a´ n˜ ez, M.; Abboud, J.-L. M.; Elguero, J. Chem.
Phys. Lett. 1990, 172, 471.
(30) Boyd, S. L.; Boyd, R. J. J. Am. Chem. Soc. 1997, 119, 4214.

Unsaturated Amines, Phosphines, and Arsines
J. Am. Chem. Soc., Vol. 121, No. 19, 1999 4657
Table 3.-1 Ta otal Energies (E, hartrees) and Zero-Point Energies (ZPE, hartrees) of the Species Investigated. Entropies (S) are Given in cal
-
1
mol
K
B3LYP/6-311+G(3df,2p)
G2
G2(MP2)
HF/6-31G*
ZPE
system
E
∆E
ZPE
S
E
∆E
E
S
phosphines
EtPH
EtPH
EtPH
2
-421.835 23
-422.175 53
-421.161 17
-420.599 22
-420.933 11
-420.913 75
-420.921 65
-420.018 80
-419.968 24
-419.953 76
-419.349 41
-419.665 77
-419.659 19
-419.666 81
-418.777 83
-418.750 27
0.083 88
0.094 94
0.071 71
0.060 07
0.071 36
0.070 95
0.073 17
0.048 61
0.044 86
0.044 39
0.036 05
0.047 55
0.047 21
0.049 39
0.023 99
0.023 97
69.306
70.183
71.174
66.551
67.506
66.879
64.487
64.918
65.894
66.330
64.846
65.747
65.627
62.777
62.807
64.142
-421.129 60
-421.457 33
-420.538 01
-419.917 43
-420.238 59
-420.220 69
-420.229 37
-419.345 01
-419.299 38
-419.286 22
-418.695 12
-418.999 71
-418.988 50
-418.999 21
-418.133 28
-418.110 73
-421.123 47
-421.452 04
-420.530 80
-419.911 66
-420.233 57
-420.215 63
-420.224 40
-419.338 66
-419.292 45
-419.278 57
-418.689 51
-418.995 00
-418.983 73
-418.994 41
-418.126 95
-418.102 76
0.089 58
0.101 56
0.077 00
0.064 71
0.077 04
0.075 98
0.077 77
0.052 48
0.048 91
0.048 31
0.039 57
0.051 88
0.050 42
0.053 75
0.027 21
0.026 31
68.622
69.639
68.336
66.059
66.686
68.910
66.926
64.360
65.213
66.229
63.785
64.802
65.333
61.971
62.837
63.629
+
3
-
2
2
3
CdCHPH
CdCHPH
2
3
2
+
+
0.0
11.9
8.3
0.0
11.2
5.8
CdCHPH
+
-
H
H
2
CdCHPH
-
2
CdCPH
2
-
2
HCdCHPH
HCtCPH
HCtCPH
2
+
3
0.0
3.9
0.4
0.0
7.0
0.3
+
H
2
CdCPH
2
+
2 2
PH
(CH)
-
HCtCPH
CtCPH
-
2
amines
EtNH
EtNH
EtNH
2
-134.894 58
-135.240 44
-134.261 51
-133.691 29
-134.017 61
-134.041 07
-134.007 93
-133.096 19
-132.455 14
-132.755 19
-132.792 86
-132.745 96
-131.883 11
-134.891 37
-135.237 31
-134.258 24
-133.688 22
-134.014 60
-134.037 98
-134.005 19
-133.093 30
-132.452 08
-132.752 39
-132.789 87
-132.743 39
-131.880 41
0.099 67
0.115 71
0.081 42
0.074 24
0.090 09
0.088 49
0.091 15
0.057 91
0.048 28
0.064 46
0.061 63
0.063 32
0.031 46
64.233
65.101
63.229
61.564
63.070
63.187
59.794
60.735
59.471
60.601
60.345
59.053
57.647
+
3
-
H
2
H
2
H
3
CdCHNH
CdCHNH
CdCHNH
2
3
2
+
+
0.0
-14.7
6.1
+
2 2 2
(CH ) NH
-
2
H CdCHNH
HCtCNH
HCtCNH
2
+
3
0.0
-23.6
5.8
+
H
2
CdCNH
2
+
(CH) NH
2 2
-
HCtCNPH
a
The relative stability of the different protonated species (∆E, kcal mol-1) is referred always to the heteroatom protonated structure. These
values include the scaled ZPE correction.
For the corresponding saturated counterparts the effects
are the opposite and the C-X bond becomes longer, while the
C-C bond remains practically unperturbed, ratifying the
at the â-carbon, whereas the corresponding phosphines are phos-
1
3
phorus bases. For the arsines there is a less clear-cut between
both possibilities, and actually for both vinylarsine and ethy-
nylarsine, high-level ab initio and DFT calculations predicted5
both the CR- and the As-protonated forms to be practically de-
generate. It is also interesting to note that for phosphines and
arsines the protonation at the â-carbon is the less favorable
process.
-
significant interaction between the XH and C-C double and
triple bonds.
It is worth noting also that protonation of vinyl and ethynyl
derivatives at the R-carbon yields a three-membered ring struc-
ture in which the XH2 group interacts with a H2CCH2 or a
HCCH moiety (See Scheme 3). As shown in ref 5, for arsines
this interaction is weak and the corresponding protonated species
can be roughly viewed as tightly bound complexes between
In all deprotonation reactions, the deprotonation of the XH2
is the most favorable process.
Acidity and Basicity Trends. Our calculated gas-phase
basicities and acidities for amines, phosphines and arsines are
compared with the experimental values in Table 4. The
agreement between theory and experiment is fairly good. It is
important to note that the estimates obtained at the G2(MP2)
level differ from those obtained at the G2 level by less than 0.5
kcal/mol. This implies that the more economic G2(MP2)
procedure can be used with confidence to get absolute gas-phase
basicities and acidities for larger systems, for which the G2
approach can be unaffordable. A similarly good performance
is exhibited by B3LYP approach, although in this case the
differences with respect to the G2 values are slightly larger.
For the sake of consistency, in what follows we will refer
systematically to the G2 values.
It must be mentioned that for vinylamine (H2CdCHNH2) two
quite different experimental proton affinities are available. From
the experimental heats of formation of the neutral and its
protonated form reported in ref 31, a PA of 215.1 kcal/mol can
be obtained, which indicates that vinylamine should be only
11.1 kcal/mol more basic than ammonia. Also in the most recent
compilation of gas-phase basicities a similarly low value (214.8
kcal/mol) is reported. The value for the GB of vinylamine
+
AsH2 and ethylene or acetylene, respectively. For phosphines
and amines, where the heteroatom is more electronegative, the
interactions are stronger, and the charge densities at the C-X
bond critical points (bcps) are typical of normal covalent
linkages.
Relative Stability. The total energies of the species inves-
tigated have been summarized in Table 3. The first conspicuous
fact of this table is that the stability order of the protonated
species for R,â-unsaturated phosphines does not coincide with
that found for the corresponding amines. Actually, while in
+
amines the Câ-protonated species (CH3-CH2NH2 and H2Cd
+
CNH2 ) are the most stable ones, for phosphines protonation
is favored at the heteroatom. In this respect it should be noted
that for ethynylphosphine the energy gap between the phos-
+
+
phorus (HCtCPH3 ) and the CR ((CH)2PH2 ) protonated spe-
cies is quite small (0.3 kcal/mol). However, this gap is sizably
larger (1.1 kcal/mol) in terms of Gibbs free-energies since the
+
entropy of CR-protonated structure (CH)2PH2 , due to its cyclic
character, is smaller than that of the phosphorus-protonated
+
species HCtCPH3 . These results are in agreement with the
3
experimental evidence which showed that enamines protonate

4658 J. Am. Chem. Soc., Vol. 121, No. 19, 1999
M o´ et al.
Table 4. Proton Affinities (PA), Gas-Phase Basicities (GB), Deprotonation Enthalpies (∆H×bcacid), and Deprotonation Gibbs Free Energies
-
1
(
∆G×bcacid). All Values in kcal mol
B3LYP/6-311+G(3df,2p)
PA
G2(MP2)
G2
exptla
system
amines
GB ∆H°acid ∆G°acid
PA
GB ∆H°acid ∆G°acid
PA
GB ∆H°acid ∆G°acid
GB
∆G°acid
391.2 210.0 ( 2c
367.4 207.1 ( 2c
391.7 ( 0.7
c
ethyl
vinyl
218.6(N) 211.1 398.8
221.0(N) 213.7 374.8
391.3 218.5(N)
367.3 220.9(N)
206.2(Câ) 198.9
200.2(CR) 191.9
353.0 189.8(N) 182.4 360.4
211.0 398.7
213.6 374.9
b
d
369.6
2
2
06.3(Câ) 199.0
00.4(CR) 192.1
ethynylb
189.9(N) 182.5 360.2
353.2
2
1
13.4(Câ) 205.9
84.3(CR) 176.4
213.4(Câ) 205.9
184.0(CR) 176.1
phosphines
ethyl
vinyl
d
d
208.1(P) 200.6 373.0
203.9(P) 196.4 358.5
364.7 207.5(P) 199.8 373.3
351.2 203.0(P) 195.4 360.8
192.2(Câ) 185.3
197.6(CR) 190.1
345.3 193.1(P) 185.6 354.5
185.8(Câ) 178.5
364.9 207.0(P)
353.5 202.9(P)
191.8(Câ) 184.9
197.1(CR) 189.6
347.0 192.5(P) 185.0 354.0
199.3 372.6
195.3 360.5
364.2 198.9 ( 2.1e 365.6 ( 2.6
353.2 194.6 ( 2.2e 356.3 ( 2.8
b
1
1
92.2(Câ) 184.5
96.0(CR) 187.6
ethynylb
192.8(P) 185.3 352.5
346.5 181.9 ( 3.7e 348.9 ( 2.5d
1
1
88.9(Câ) 181.4
92.8(CR) 184.4
185.3(Câ) 178.0
192.2(CR) 183.8
193.2(CR) 184.9
arsinesf
ethyl
197.8(As) 190.1 363.7
194.0(As) 186.6 352.6
356.0
345.0
199.2(As)g 190.1 363.9
195.1(As) 187.7 350.9
195.0(CR) 187.5
184.0(As) 176.5 346.6
180.8(Câ) 173.3
356.7 187.2
343.5 183.6
358.7
346.1
b
vinyl
196.4(CR) 188.9
ethynylb
181.7(As) 174.2 346.6
338.8
338.8 175.2
342.7
1
1
84.9(Câ) 177.4
84.7(CR) 177.6
183.5(CR) 176.4
a
b
Overall uncertainties are given, including measurements errors and estimated absolute scale uncertainties (See Table 1). The position which
undergoes protonation is indicated within parentheses. Values taken from ref 6. d This work. e Values taken from ref 13. All calculated and
experimental values for arsines were taken from ref 5. The value reported in ref 5 (197.8 kcal/mol) does not include, by error, the thermal
c
f
g
corrections.
6
reported in the NIST WebBook does not correspond to the
value reported by Ellenberger and co-workers,13 but the cor-
rection method is not given. The Ellenberger data is based on
1). This acidity enhancement is a consequence of the strong
favorable interaction of the NH group and the multiple bonds,
-
reflecting the benefits of charge delocalization, as explained
+
4
the deprotonation of the (CH3-CH-NH2) ion and the isomer-
before in the literature. The same behavior is found for the
ization energy of CH3-CHdNH into CH2dCH-NH2. Using
corresponding phosphines, as well as for the corresponding
arsines, although the energy gaps are sizably smaller. Vinyl-
and ethynylphosphine are predicted to be 12.1 and 18.6 kcal/
mol, respectively, more acidic than ethylphosphine (See Table
4), whereas for arsines these differences become 13.0 and 17.3
kcal/mol, respectively.
To gain some insight into the origin of these acidity trends
and into the origin of the differences between amines, phos-
phines, and arsines, we have used the following isodesmic
reactions:
the most recent NIST tables, the bracketing experiments of
3
Ellenberger and co-workers leads to a reevaluated value GB-
CH3-CHdNH) ) 207.3 kcal/mol. The G2 isomerization
(
energy of CH3-CHdNH into CH2dCH-NH2 was found to
be 4.0 kcal/mol, leading to a reevaluated GB(CH2dCH-NH2)
)
1
211.3 kcal/mol.. Our calculations predict vinylamine to be
6.9 kcal/mol more basic than ammonia2 and therefore are in
2a
reasonably good agreement with the original paper of Ellen-
berger et al., who estimated that vinylamine was 14.8 kcal/
3
mol more basic than ammonia. For the gas-phase acidity of this
compound the agreement between our theoretical estimates and
the FT-ICR measurements is very good. Unfortunately, no
experimental values for the gas-phase basicity and acidity of
ethynylamine (HCtCNH2) have been reported in the literature
so far. However, the generally good agreement between G2
estimated values and experimental ones make us confident in
the reliability of our calculated values. The good agreement
between experimental measurements and theoretical estimates
for the basicity of R,â-unsaturated phosphines, clearly confirm
that these compounds protonate preferentially at the heteroatom
rather than at the â-carbon. For the saturated systems, where
protonation and deprotonation take place at the heteroatom, the
gas-phase basicity decreases in the order amine > phosphine
CH
-CH
XH
+ CH
dCH
f
3
2
2
2
2
CH dCHXH + CH -CH (2)
2
2
3
3
-
CH -CH XH + CH dCH f
3
2
2
2
-
CH dCHXH + CH -CH (3)
2
3
3
3
+
CH -CH XH + CH dCH f
3
2
3
2
2
+
CH
dCHXH
+ CH
-CH
(4)
2
3
3
CH -CH XH + HCtCH f
3
2
2
HCtCXH + CH -CH (5)
2
3
3
-
-
CH -CH XH + HCtCH f HCtCXH + CH -CH
>
arsine, while the gas-phase acidity decreases in the reverse
3
2
3
3
order, in agreement with our previous discussion.
(6)
4
As it has been pointed out before by Smith and Radom, both
the unsaturated amines, namely, vinyl- and ethynylamine, are
predicted to be more acidic than the saturated analogue
ethylamine by 23.8 and 38.3 kcal/mol, respectively (See Figure
+
3
+
CH -CH XH + HCtCH f HCtCXH3 + CH -CH3
3
2
3
(7)
which allow us to estimate the effects of the XH2 substituents
on the CdC and CtC bonds, for neutral, protonated and
deprotonated species, taking as a reference the saturated ethyl
(31) Lias, S. G.; Bartmess, J. E.; Liebman, J. F.; Holmes, J. L.; Levin,
R. D.; Mallard, W. G. J. Phys. Chem. Ref. Data 1998, 23, 413.

Unsaturated Amines, Phosphines, and Arsines
J. Am. Chem. Soc., Vol. 121, No. 19, 1999 4659
Figure 1. G2 calculated relative enthalpies of species involved in the deprotonation processes of ethyl-, vinyl- and ethynyl-amines, phosphines,
and arsines. All values in kcal/mol. ∆Hacid gives the enthalpies (including ZPE, thermal, and P∆V corrections), corresponding to each deprotonation
process.
analogue. The G2 energies of C2H2, C2H4, and C2H6 were taken
from ref 18 those of the different arsines and their protonated
and deprotonated structures were taken from ref 5.
ments multiple bonds are less stabilizing, in relative terms than
for first-row elements. Actually, for the particular case of arsines,
the unsaturated neutrals are destabilized rather than stabilized
with respect to the saturated analogue. This results in a certain
enhancement of the intrinsic acidity of arsines, since, although
the stabilization of the anions is smaller than for phosphines,
the destabilization of the neutrals reduces the energy difference
between both forms and increases the acidity of the system.
Differences between amines, phosphines and arsines are more
pronounced as far as their gas-phase basicities are concerned.
In general, with the only exception of vinylamine, the R,â-
unsaturated compounds are less basic than the ethyl-derivative,
but as mentioned above, while R,â-unsaturated amines protonate
at the â-carbon, phosphines protonate at the heteroatom, and
for arsines, protonation at CR may compete with arsenic
protonation. These differences are clearly illustrated in Figure
In Figure 1 the stabilization of the anions is compared with
the stabilization of the neutrals for amines, phosphines, and
arsines. It is quite evident that, in all cases, as it was found
previously for amines, the stabilization of the anions is
significantly greater than the stabilization of the neutrals. This
behavior has been explained for the particular case of the amines
in terms of the σ-electron-withdrawing nature of the NH2 group
-
4
and the strong π-electron donor nature of the NH group.
For phosphines and arsines there are, however, some impor-
tant differences. In the first place, the stabilization of the anion
by the interaction between the multiple bonds and the XH-
group is for both series of compounds much smaller than for
amines, reflecting the fact that for second- and third-row ele-

4660 J. Am. Chem. Soc., Vol. 121, No. 19, 1999
M o´ et al.
Figure 2. G2 calculated relative enthalpies of vinyl-amine, -phosphine, and -arsine and their protonated species as compared with the corresponding
saturated ethyl-analog. All values (in kcal/mol) refer always to the most stable conformer. PA corresponds to the proton affinities (including ZPE,
thermal, and P∆V corrections) for the basic center indicated within parentheses.
Scheme 4
cation, for phosphines and arsines there is an almost equally
large destabilization (11.1 and 11.0 kcal/mol, respectively).
This significant difference points out to a quite different
ability of the NH2 and the PH2 (or AsH2) groups to stabilize
+
the carbocation [CH3-CH-XH2] formed when the vinyl-
derivative is protonated at the â carbon atom. This can be
quantitatively measured by means of the following isodesmic
reactions:
2
where we have shown diagrammatically the energetics of the
appropriate isodesmic reactions for the vinyl-derivatives. The
first conspicuous fact of Figure 2 is that the stabilization of the
neutral is much smaller for vinylphosphine (1.7 kcal/mol) than
for vinylamine (7.3 kcal/mol), while for vinylarsine there is a
destabilization effect of 5.2 kcal/mol. In all cases there is
+
+
3
CH -CH-NH + CH f CH -CH -NH + CH
3
2
4
3
2
2
∆H° ) 107.8 kcal/mol (8)
destabilization of the species when protonated at the heteroatom,
+
due to an unfavorable interaction of the XH3+ groups with the
+
3
CH -CH-PH + CH f CH -CH -PH + CH
3
2
4
3
2
2
double bond. The most significant difference, however, is related
with the â-protonated species. As shown in Figure 2, while for
amines there is a strong stabilization (14.7 kcal/mol) of the
∆H° ) 37.0 kcal/mol (9)
+
which show that the [CH3-CH-NH2] cation is significantly

Unsaturated Amines, Phosphines, and Arsines
J. Am. Chem. Soc., Vol. 121, No. 19, 1999 4661
Scheme 5
heteroatom and both ethylidenephosphine and vinylphosphine
are phosphorus bases, vinylphosphine being the most basic of
the two. The same conclusions apply to the corresponding
arsenic containing compounds (See Scheme 5). The behavior
predicted for the corresponding ethynyl-derivatives (See Figure
3
) is rather similar to that discussed above for the vinyl-
derivatives, and analogous arguments can be invoked to explain
it. Protonation at the â carbon atom yields a carbocation [H2Cd
+
C-XH2] which is more stabilized when the substituent is an
amino group than when the substituent is a PH2 group as
illustrated by the isodesmic reaction 10.
+
CH dCH-NH + CH dCH-PH f
2
2
2
2
+
CH dCH-NH + CH dCH-PH
∆H° ) 19.1 kcal/mol
10)
2
2
2
2
(
These protonated species can be also viewed as the result of
the protonation of the H2CdCdXH isomers at the heteroatom.
In this case, as shown in Scheme 6, the isomerizations in-
volving the neutrals are predicted to be, at the G2 level (See
Table 5), significantly exothermic for nitrogen containing
species, but only slightly exothermic for P or As containing
systems. Therefore, in this case the H2CdCdNH is a nitrogen
base sizably weaker than ethynylamine which behaves as a
carbon base. As before the phosphorus (and arsenic) containing
compounds behave as phosphorus (or arsenic) bases in the gas
phase, the ethynyl derivative being much more basic than the
H2CdCdXH isomer.
Scheme 6
In agreement with our previous discussion, in both vinyl- and
ethynyl- derivatives, when the XH2 group of the â-protonated
species is replaced by hydrogen, the cation undergoes a dra-
matic destabilization, which is much larger for the amines (64.5
kcal/mol, and 61.9 kcal/mol, respectively) than for the phos-
phines (29.9 and 37.0 kcal/mol, respectively) or the arsines (15.3
and 24.8 kcal/mol, respectively). This explains the enhanced
gas-phase basicity of vinylamine and ethynylamine with respect
to ethylene and acetylene, respectively, even though all systems
are carbon bases. As illustrated in Figures 2 and 3 the
corresponding phosphorus and arsenic containing compounds
are also significantly more basic than ethylene and acetylene.
The second important dissimilarity of vinyl- and ethynyl-
amines with respect to phosphines and arsines is that, whereas
+
for amines, on going from the â-protonated (CH3-CHNH2
+
+
and H2CdCNH2 ) to the R-protonated structures ((CH2)2NH2
+
more stabilized, in relative terms, than [CH3-CH-PH2] . This
indicates that the NH2 group interacts in a more efficient way
with the electron deficient central carbon atom than does the
PH2 (or the AsH2) group, likely because of a much less efficient
overlap between the corresponding atomic orbitals when the
substituent contains a second (or a third) row atom.
+
and (CH)2NH2 ), there is a destabilization of the system by 20.7
and 29.4 kcal/mol, respectively (See Figures 2 and 3), and for
phosphines and arsines there is a stabilization of 5.3 and 6.9
kcal/mol for the former and 10.9 and 2.7 kcal/mol for the latter,
respectively. The origin of these differences is not obvious, since
they are related with the relative stability of the cyclic species
formed. To investigate this point we have envisioned the
following isodesmic reactions:
It is also important to note that Câ-protonated species (CH3-
+
CHXH2 ) can be alternatively viewed as the result of proto-
nating the corresponding imine isomer at the heteroatom (See
Scheme 4). This fact permits to predict also the basicity trends
within the CH3-CHdXH series.
As illustrated in Scheme 5, the isomerization from vinylamine
(
CH2dCH-NH2) into ethylidenimine (CH3-CHdNH) is en-
ergetically favorable. On the other hand, because of the enhanced
stability of the [CH3-CH-NH2] cation, the protonation of
+
ethylidenimine at the nitrogen atom is more exothermic that
the nitrogen protonation of vinylamine which, accordingly,
behaves as a slightly stronger carbon base (See Scheme 5). Since
+
for phosphorus-containing compounds the [CH3-CH-PH2]
cation is much less stabilized, the protonation is favored at the
which measure the relative stability of the protonated cyclic

4662 J. Am. Chem. Soc., Vol. 121, No. 19, 1999
M o´ et al.
Figure 3. G2 calculated relative enthalpies of ethynyl-amine, -phosphine and -arsine and their protonated species as compared with the corresponding
saturated ethyl-analog. All values (in kcal/mol) refer always to the most stable conformer. PA corresponds to the proton affinities (including ZPE,
thermal, and P∆V corrections) for the basic center indicated within parentheses.
Table 5. G2 Total Energies (hartrees) for the Isomers of Vinyl-
and Ethynyl-Derivatives
explained before, the Câ-protonated structure for the amine is
particularly stabilized while for the phosphine or the arsine it
is destabilized. Hence, although the cyclization step leads to a
stabilization of the system greater for amines than for phosphines
and arsines, this is not enough to bring the R protonated amine
lower in energy than the â-protonated amine.
CH
CH
CH
CH
CH
CH
3
3
3
2
2
2
-CHdNH
-CHdPH
-CHdAsH
dCdNH
dCdPH
dCdAsH
-133.697 685
-419.921 280
-2313.318 924
-132.479 518
-418.700 878
-2312.091 915
Conclusions
species with respect to the corresponding open-chain neutral.
It can be seen that both isodesmic processes are clearly endo-
thermic, so one must conclude that the nitrogen-containing cy-
cles are more stable, as expected, than their phosphorus or
arsenic analogues. However, the CR-protonated form of vinyl-
amine ((CH2)2NH2+) is still less stable than the â proto-
Our systematic study of the basicity and acidity trends along
the series of R,â-unsaturated amines, phosphines, and arsines
leads to the main conclusion that amines exhibit an exceptional
behavior with regards to the analogous bases of the Va group.
Actually, although R,â-unsaturated amines, phosphines, and
arsines are less basic but significantly more acidic than the
corresponding saturated analogues, vinyl- and ethynylamine are
â-carbon bases while the corresponding phosphorus containing
+
nated species (CH3-CHNH2 ), while the opposite is found for
the corresponding phosphine and arsine because, as we have

Unsaturated Amines, Phosphines, and Arsines
J. Am. Chem. Soc., Vol. 121, No. 19, 1999 4663
+
analogues protonate preferentially at the heteroatom. Arsines
resemble closely the corresponding phosphines, although for
arsines protonation at the CR atom competes with protonation
at the heteroatom. The enhanced acidity of unsaturated com-
pounds can be attributed essentially to a stabilization of the
anions due to a favorable interaction of the XH with the
multiple C-C bonds. This stabilizing effect is maximum for
amines and minimum for arsines.
important because they may easily dissociate in AsH2 and
1
5
ethylene or acetylene, respectively, explaining their rapid
decomposition.
Acknowledgment. This work has been partially supported
by the Spanish DGES Project PB96-0067. Pierre Guenot
CRMPO, Rennes) is gratefully acknowledged for recording the
high-resolution mass spectrum of vinylamine. J.-C.G. thanks
the Program National de Plan e´ tologie (INSU-CNRS) for finan-
cial support.
-
(
The low relative basicity of these unsaturated compounds is
mainly due to a destabilization of the protonated species owing
+
to unfavorable interactions of the XH3 group with the C-C
Supporting Information Available: Optimized geometries
for ethylphosphine, vinylphosphine, and ethynylphosphine and
their protonated species. Table of NBO population analysis
of R,â-unsaturated amines, -phosphines, and -arsines (FTP).
This material is available free of charge via the Internet at http://
www.pubs.acs.org.
π-system. However, as mentioned above, while protonation at
the â-carbon is strongly favored for amines, it is, on the contrary,
unfavorable for phosphines and arsines, because the carbocation
formed is less stabilized when the heteroatom of the XH2
belongs to the second or the third row than when it is a first-
row atom. For arsines the R-protonated structures and species
protonated at the heteroatom are estimated to be equally stable.
The high stability of the R-protonated species is particularly
JA982657E