Reactions of gaseous, halogenated propene radical cations with ammonia: A study of the mechanism by Fourier transform ion cyclotron resonance

Article abstract of DOI:10.1002/(SICI)1521-3765(19980904)4:9<1799::AID-CHEM1799>3.0.CO;2-7

The gas-phase ion-molecule reactions of the radical cations of 2-chloropropene (1^+.), 2-bromopropene (2^+.), and 2-iodopropene (3^+.), as well as of the corresponding three 3,3,3-trifluoropropenes (4^+.- 6^+.) with ammonia have been studied by FT-ICR mass spectrometry complemented by ab initio calculations of the reaction thermochemistry. In all cases a deprotonation of the 2-halopropene radical cations by ammonia is distinctly exothermic. In spite of this, the substitution of the halo substituent by NH₃ is the main reaction pathway for 1^+. and 2^+. and is still competing for the slowly reacting iodo derivative 3^+.. In the latter case deprotonation generates not only NH(4/⁺), but also the proton-bridged dimer [H₃N·· H⁺··NH₃]. These effects prove that the first addition step of the substitution by an addition-elimination mechanism of the haloalkene radical cations can compete effectively with exothermic deprotonation and occurs without noticeable activation energy. In fact it appears likely that the deprotonation of the unsaturated radical cations proceeds also by an addition-elimination process. The calculation of the reaction enthalpy shows that addition of NH₃ to the ionized 3,3,3-trifluoro-2-halopropenes 4^+.-6^+. is especially exothermic. Experimentally this effect is not only reflected in the increased reaction efficiency of the substitution product, even in the case of the iodo derivative 6^+., but also in competing fragmentations of the strongly excited distonic intermediates generated by the addition step. This corroborates the postulate that the variation of the rate constants with the substituents, which is observed for the reactions of ionized haloalkenes with ammonia, is caused by the excess energy released in the initial addition step.

Full text of DOI:10.1002/(SICI)1521-3765(19980904)4:9<1799::AID-CHEM1799>3.0.CO;2-7

FULL PAPER
Reactions of Gaseous, Halogenated Propene Radical Cations with Ammonia:
A Study of the Mechanism by Fourier Transform Ion Cyclotron Resonance
Michael Büchner, Andreas Nixdorf, and Hans-Friedrich Grützmacher*
Abstract: The gas-phase ion-molecule
also the proton-bridged dimer [H3N ´´
H ´´ NH3]. These effects prove that the
shows that addition of NH3 to the ion-
.
‡
‡
reactions of the radical cations of 2-
ized 3,3,3-trifluoro-2-halopropenes 4 ±
.
.
‡
‡
chloropropene (1 ), 2-bromopropene
first addition step of the substitution by
an addition-elimination mechanism of
the haloalkene radical cations can com-
pete effectively with exothermic depro-
tonation and occurs without noticeable
activation energy. In fact it appears
likely that the deprotonation of the
unsaturated radical cations proceeds
also by an addition-elimination process.
The calculation of the reaction enthalpy
6
is especially exothermic. Experimen-
.
.
‡
‡
(2 ), and 2-iodopropene (3 ), as well
tally this effect is not only reflected in
the increased reaction efficiency of the
substitution product, even in the case of
as of the corresponding three 3,3,3-
.
.
‡
‡
trifluoropropenes (4 ± 6 ) with ammo-
nia have been studied by FT-ICR mass
spectrometry complemented by ab initio
calculations of the reaction thermo-
chemistry. In all cases a deprotonation
of the 2-halopropene radical cations by
ammonia is distinctly exothermic. In
spite of this, the substitution of the halo
substituent by NH3 is the main reaction
.
‡
the iodo derivative 6 , but also in
competing fragmentations of the strong-
ly excited distonic intermediates gener-
ated by the addition step. This corrobo-
rates the postulate that the variation of
the rate constants with the substituents,
which is observed for the reactions of
ionized haloalkenes with ammonia, is
caused by the excess energy released in
the initial addition step.
Keywords: ion-molecule reactions ´
.
.
‡
‡
pathway for 1 and 2 and is still
mass spectrometry
substitution radical ions
reaction mechanisms
´ nucleophilic
competing for the slowly reacting iodo
´
´
.
‡
derivative 3 . In the latter case depro-
‡
tonation generates not only NH4 , but
are easily investigated by appropriate mass spectrometric
techniques, and a study of the gas phase ion-molecule
reactions of unsaturated organic radical cations can display
many details of the reaction mechanisms.
Introduction
The detection of radical cations as reactive intermediates has
initiated an increasing number of studies of their reactivities
in solution[1] and in the gas phase.[2] These studies reveal that,
in particular, organic radical cations exhibit a high reactivity
that is attributed to their electron deficiency and which has
been termed electron hole catalysis.[3] In line with this concept
unsaturated organic radical cations display fast reactions with
electron-rich donor molecules and generate a variety of
reaction products by addition, substitution, and/or fragmen-
tation reactions. However, owing to the short lifetime of most
organic radical cations in solution it is difficult to obtain
information about the details of the reaction mechanisms and
to determine which structural factors control the reaction
rates.[4] In contrast, even very reactive gaseous radical cations
Recently, we have performed a systematic study of the gas
phase reactions of mono- and dihalogenated arenes[5] with
ammonia and simple amines via radical cations using Fourier-
transform ion cyclotron resonance (FT-ICR) spectrometry.
These reactions result in the substitution of one halogen
substituent by the nucleophile through an addition-elimina-
tion mechanism. The kinetics and the rather low reaction
efficiency of less than 15% show that the first addition step of
the aromatic substitution process is rate-determining. This
was explained by the configuration mixing model of Shaik and
Pross[6] of polar organic reactions. However, further studies
revealed that the mechanisms of the aromatic substitution via
radical cations is more complex and includes rearrangement
steps between addition of the nucleophile and elimination of
the halo substituent.[5b,c]
[*] Prof. Dr. H.-F. Grützmacher, M. Büchner, Dr. A. Nixdorf
Lehrstuhl I für Organische Chemie, Fakultät für Chemie
der Universität Bielefeld
In the case of gas phase substitution reactions of olefinic
radical cations the addition of the nucleophile is not expected
to be the rate-determining step.[2a] This was corroborated by
FT-ICR studies of the ion-molecule reactions of the radical
Postfach 10 01 31, D-33501 Bielefeld (Germany)
Fax : (‡ 49)521-106-6417
E-mail: gruetzmacher@chema.uni-bielefeld.de
Chem. Eur. J. 1998, 4, No. 9
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1998
0947-6539/98/0409-1799 $ 17.50+.25/0
1799

H.-F. Grützmacher et al.
FULL PAPER
cations of vinyl halides with ammonia,[7] and by a
preliminary investigation of the reactions of 1,2-
dichloro- and dibromoethenes with ammonia and
amines through radical cations.[8] In these reac-
tions substitution of one halogen substituent
occurs with a high reaction efficiency although
the reaction rates are still distinctly below the
collision rate. Verification has been obtained that
the product of the reaction of vinyl halide radical
cations with ammonia corresponds to the
Scheme 1. Substitution reaction of haloethene radical cations with ammonia.
N-protonated vinyl amine as expected for an addition-
elimination mechanism.[7] A vinyl halide radical cation
produces two isomeric intermediate b-distonic ions by addi-
tion of ammonia that are analogous to a Markovnikov and
anti-Markovnikov orientation, but only the less stable anti-
Markovnikov adduct has the correct configuration for the
final loss of the halogen atom (Scheme 1). A calculation of the
minimum energy reaction pathway (MERP) by ab initio
methods suggested that the rearrangement of the intermedi-
ate b-distonic ions by 1,2-NH3 shifts is important for the
efficiency of the total reaction.[7] Therefore, the apparently
simple substitution of haloalkene radical cations by a
nucleophile is also very likely to be a complex multistep
reaction. The deprotonation of vinyl halide radical cations by
ammonia was calculated to be exothermic, but nevertheless
deprotonation was found to be only a minor process. There-
fore it had to be assumed that the addition of electron-rich
reactants like ammonia or amines to the ionized double bond
of the radical cations of haloalkenes is very fast in order to
compete successfully with proton transfer and may occur even
without any activation barrier, in line with theoretical
predictions.[2] Further, a highly energized b-distonic ion is
generated by the exothermic addition, and the excess
energy of this intermediate addition product should drive
the reaction further and may in fact be responsible for
the sometimes puzzling outcome of the reactions of
unsaturated organic radical cations with electron-rich reac-
tants.
More experimental information is needed to understand
the extraordinary reactivity of olefinic radical cations and the
course of their reactions with nucleophiles. In particular, the
successful competition of addition to the ionized double bond
of the haloalkene radical cation against deprotonation by a
basic nucleophile should be verified unequivocally. Further,
the role of the excess energy imparted to the inter-
mediate distonic ion by an exothermic addition step
should be examined further. Thus, we complemented the
previous investigations by a study of the ion-molecule
reactions of ionized 2-halopropenes 1 ± 3 and 3,3,3-tri-
fluoro-2-halopropenes 4 ± 6 with ammonia using FT-ICR
spectrometry.
Abstract in German: Die Ion/Molekül-Reaktionen von Ra-
.
.
‡
The radical cations of the 2-halopropenes 1 ± 3 should be
distinctly more acidic than the previously studied ionized
vinyl halides because deprotonation of a propene radical
cation yields a stabilized allyl radical. Therefore, the radical
cations 1 ± 3 are well-suited for an investigation of the
competition between addition and deprotonation of unsatu-
rated organic radical cations by basic substrates. It will be
shown that, surprisingly, the substitution reaction of radical
cations 1 ± 3 is still observed in spite of the exothermic
deprotonation by ammonia, thus corroborating a fast addition
step of the nucleophile to the ionized double bond. In the case
of all ionized 2-halopropenes the addition step generating a b-
distonic ion by addition of ammonia is expected to be more
exothermic than for ionized haloethenes. This effect should be
especially increased for the radical cations of the 3,3,3-
trifluoro-2-halopropenes 4 ± 6. Further, deprotonation of 4 ± 6
is not expected to interfere with substitution because of the
missing methyl group. Hence, the radical cations 4 ± 6 are
expected to reveal more clearly effects of a chemical
activation during the addition step on the further fate of the
excited intermediate distonic ions. It will be shown that this is
indeed the case, and that the large excess energy present in the
distonic ion results in additional fragmentations. The results of
the kinetic experiments for 2-halopropene radical cations
‡
dikalkationen der 2-Halogenpropene 1 ± 3 und der 3,3,3-
.
.
‡
‡
Trifluor-2-halogenpropene 3 ± 6 mit Ammoniak in der
Gasphase werden mit Hilfe der FT-ICR-Massenspektrometrie
untersucht und durch Ab initio-Rechnungen zur Thermo-
chemie der verschiedenen Prozesse ergänzt. Die Deprotonie-
rung der 2-Halogenpropen-Radikalkationen durch Ammoniak
ist danach in allen Fällen deutlich exotherm. Trotzdem ist für
.
.
‡
‡
1
und 2 die Substitution des Cl- bzw. Br-Substituenten die
Hauptreaktion und auch im Fall des langsamer reagierenden
.
‡
I-Derivats 3 eine intensive Konkurrenzreaktion. Im letzteren
‡
Fall wird bei der Deprotonierung nicht nur NH4 ge-
bildet, sondern auch das protonenverbrückte Dimere
‡
[H3N ´´ H ´´ NH3]. Diese Befunde beweisen, daû der erste
Additionsschritt im Additions/Eliminierungsmechanismus der
Substitution wie die Deprotonierung ohne gröûere Aktivie-
rungsschwelle verläuft. Berechnungen der Reaktionsenthalpie
zeigen, daû die Addition von NH3 an die ionisierten 3,3,3-
.
.
‡
‡
Trifluor-2-halogenpropene 4 ± 6 besonders exotherm ist. Im
Experiment zeigt sich dies nicht nur durch erhöhte Reaktions-
geschwindigkeitskonstanten des gesamten Prozesses, sondern
auch durch zusätzliche Fragmentierungen des hoch angeregten
distonischen Zwischenprodukts. Dies stimmt mit der Vorstel-
lung überein, daû die Abstufung der Geschwindigkeitskon-
stanten der Reaktionen ionisierter Halogenalkene mit Ammo-
niak durch die unterschiedliche Überschuûenergie im Addi-
tionsprodukt bestimmt wird.
.
.
‡
‡
1
± 3 will be discussed first and will be followed by a
.
.
‡
‡
discussion of the reactions of the radical cations 4 ± 6 with
ammonia.
1800
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1998
0947-6539/98/0409-1800 $ 17.50+.50/0
Chem. Eur. J. 1998, 4, No. 9

Ion-Molecule Reactions
1799±1809
Appendix. In most cases more than one isodesmic reaction
Results and Discussion
has been calculated to arrive at a value of a specific DHf. The
1
results deviate by less than Æ 10 kJmol , which also corre-
Thermochemistry: In discussing ion-molecule reactions in the
gas phase at very low pressures studied by FT-ICR spectrom-
etry, it should be noted that these reactions occur in isolated
collision complexes of the reactants without energy exchange
with the surroundings. These collision complexes are electro-
statically activated by the energy gained from ion/dipole and
ion/induced dipole interactions, and this excess energy may be
used to induce chemical transformations within the
complex. A consequence of the lack of any additional
(thermal) activation is that only exothermic processes
are observed, and that the experimental rate constants
of the chemical reactions are not directly related the
internal barrier or activation energy of a chemical
reaction within the complex. Hence it is useful to
accompany the experimental rate measurements by a
theoretical calculation or estimation of the minima and
other stationary points along the MERP of the ion-
molecule reaction to get a better basis for the discussion
of the experimental results.
sponds to the estimated limits of error of most of the
experimental data used, and this is sufficient for a discussion
of the trends of the thermochemistry between the closely
.
‡
related reactions studied here . The DHf of 1 , of the reaction
intermediates, and of reaction products obtained are given in
Table 1. These data have been used for the construction of the
Table 1. Calculated heats of formation of reactants, intermediates, and products of
.
.
‡
‡
the reactions of 2-halogenopropene radical cations 1 ± 3 and 3,3,3-trifluoro-2-
.
‡
chloropropene radical cation 4 with ammonia.
DHf
[kJmol
Reaction
DHr
[kJmol ]
1
1
]
.
.
‡
‡
2-chloropropene (1
)
940
46[a]
691
NH3
distonic adduct (Markovnikov)
addition, Markovnikov
addition, anti-
203
172
distonic adduct (anti-Markovnikov) 722
Markovnikov
2-propenyl-ammonium cation
Cl atom
2-chloroallyl radical
641
121[a]
112
substitution
132
150
The parallel reactions shown in Scheme 2 for 2-
Ammonium cation
632[a]
deprotonation
.
.
‡
‡
halopropene radical cations 1 ± 3 and ammonia are
.
.
‡
‡
2-bromopropene (2
)
990
NH3
46[a]
743
distonic adduct (Markovnikov)
2-propenyl-ammonium cation
Br atom
2-bromoallyl radical
Ammonium cation
addition, Markovnikov
substitution
201
191
141
641
112[a]
171
632[a]
deprotonation
.
.
‡
‡
2-iodopropene (3
NH3
)
994
46[a]
786
distonic adduct (Markovnikov)
2-propenyl-ammonium cation
I atom
2-iodoallyl radical
Ammonium cation
addition, Markovnikov
substitution
162
200
89
641
107[a]
227
632[a]
deprotonation
.
‡
trifluoro-2-chloropropene (4
NH3
)
507
46[a]
85
distonic adduct (Markovnikov)
addition, Markovnikov
addition, anti-
376
330
distonic adduct (anti-Markovnikov) 131
Markovnikov
2
(3,3,3-trifluoro-propenyl-
61
Scheme 2. Substitution and deprotonation of halopropene radical
cations 1 ± 6 with ammonia.
ammonium cation
Cl atom
121[a]
substitution
279
[a] taken from ref. [10]
expected to occur because of the distinct gas-phase
acidity of the propene radical cation[9] and also from
consideration of the results obtained for ion-molecule reac-
tions of ionized vinyl halides with ammonia.[7, 8] The heat of
formation (DHf) of chloroethene, bromoethene, and their
radical cations are known with high accuracy,[10] and the
MERP of their reaction with ammonia has been calculated by
high-quality ab initio methods.[7] Further, DHf of many C2, C3,
and C4 alkenes, their radical cations, and of other radicals are
known reliably.[10] These data can be combined to calculate
the minima along the MERP of the reactions of 2-chloropro-
schematic MERP shown in Figure 1. The activation energy for
the 1,2-NH3 shift between the isomeric distonic adduct ions
was shown previously not to depend much on the substitution
type[7] and was taken from the published MERP of the
chloroethene radical-cation reaction.
The substitution reaction of 1 by ammonia is distinctly
more exothermic than the reaction of the chloroethene
radical cation (132 kJmol 1 vs. 116 kJmol ), but the exother-
.
‡
1
micity of the addition step forming the Markovnikov adduct is
almost identical. The new feature of the MERP of the reaction
.
‡
pene 1 with ammonia at the MP2/6-31G* level of theory,
making extensive use of isodesmic reactions and anchoring
.
‡
of 1 with ammonia is the deprotonation reaction yielding the
.
‡
‡
2-chloroallyl radical and NH4 ; this is very exothermic, as
expected.
It is known that the result of isodesmic reactions often does
not depend on the level of the ab initio method used.[11] This
the relative heats of formation at DHf(chloroethene ) ˆ
1
1
999 kJmol
(238.8 kcalmol ).[10]
and
DHf(chloro-
.
ethene ) ˆ 1027 kJmol (245.4 kcalmol ).[10] The specific
‡
1
1
isodesmic reactions used and the relevant data are given in the
Chem. Eur. J. 1998, 4, No. 9
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1998
0947-6539/98/0409-1801 $ 17.50+.25/0
1801

H.-F. Grützmacher et al.
FULL PAPER
Compared with the 2-chloropropene reaction
.
‡
system of 1 the MERP of the reaction of the
.
‡
3,3,3-trifluoro-2-chloropropene radical cation 4
with ammonia exhibits a strong increase in the
exothermicity of the first addition step by
1
70 kJmol and also of the total substitution
reaction. These effects can be attributed to a
combination of a stabilization effect of the CF3
substituent on the radical site of the distonic
Markovnikov adduct ion and on a destabilization
of the reactant alkene radical ion. The first effect
is seen by the difference of 30 kJmol 1 of the DHf
of the 2,2,2-trifluoroethyl radical and the ethyl
radical[10] (relative to the corresponding ethanes),
and the second effect increases the ionization
energy of 3,3,3-trifluoropropene to 10.95 eV by
more than 1 eV, relative to propene.[10] This
corroborates the calculated CF3 effects on the
Figure 1. MERP (minimum reaction energy path) for deprotonation and substitution of 2-
chloropropene radical cations 1 by ammonia.
.
‡
.
‡
MERP of 4
Summarizing, the results of an estimation and
ab initio calculation of the MERPs of the
has been confirmed in the present case by a calculation the
.
‡
reaction of 2-halopropene radical cations with ammonia
assure the expected effects of the substituents on the reaction
thermochemistry. In particular the deprotonation of all three
halopropene radical cations by ammonia is distinctly exother-
mic and can be expected always to compete effectively with
the substitution reaction. Further, in the series of the ionized
2-chloro-, 2-bromo-, and 2-iodopropene the substitution
reaction becomes increasingly more exothermic, while the
effect of the halo substituents on the initial addition step is
opposite. Finally, the substitution reaction of the 3,3,3-
trifluoropropene radical cations is indeed distinctly more
exothermic than that of the ionized 2-halopropenes and, in
particular, exhibits a very exothermic initial addition step of
ammonia to the ionized double bond.
isodesmic reactions for 1 and its products also by the lower
level methods MP2/3-21G and MP2/STO-3G. Hence, the DHf
necessary for the construction of the MERPs of the reactions
.
.
‡
‡
of the bromo and iodo derivatives 2 and 3 , and of the 3,3,3-
.
‡
trifluoro-2-propene radical cation 4 with ammonia have
been determined only by these methods. While this may
introduce increased uncertainty into the DHf values obtained,
the main source of error is the limited number of DHf of
reference compounds for the iodo derivative 3, the trifluoro
derivative 4, and their radical cations. For example, the DHf of
iodoethene and iodopropene had to be estimated by Bensonꢁs
incremental method, and in the case of 4 only the DHf of the
3,3,3-trifluoropropene,[10] the 2,2,2-trifluoroethyl,[10] and the
3,3,3-trifluoropropyl radical[10] are known. Nonetheless, the
DHf obtained (see Table 1) can be used semiquantitatively to
discuss the variation of the reaction energies with changing
the substituents. For the reactions of the 2-bromopropene
Reactions of 2-halopropene radical cations 1 ± 3 with ammo-
nia: The rates of exothermic proton transfer between small
ions and basic molecules are expected to approach the ion-
molecule collision rate. Hence, deprotonation of the radical
cations of 1 ± 3 should override any other parallel ion-
molecule reaction that exhibitis a bottle neck on the MERP,
because of entropic effects or the presence of a substantial
activation energy. The experimental ion-intensity vs. reaction-
time curves (kinetic plot) obtained for the reaction of 2-
.
‡
radical cation 2 the MERP is essentially identical to that of
.
‡
the chloro derivative 1 with exception of the reaction
enthalpy (DHr) of the substitution, which is more exothermic
1
by approximately 60 kJmol . A similar result has been
obtained by high-quality ab initio calculations of the MERPs
of the corresponding haloethene radical cations.[7] The DHr of
.
‡
the substitution of the iodo substituent of 3 by ammonia
exhibits a further increase, as expected, owing to the small
bond energy of the C I bond cleaved, but the addition step
generating the Markovnikov adduct and the deprotonation
.
.
.
‡
‡
‡
halopropene radical cations 1 , 2 , and 3 with ammonia
are shown in Figure 2. In all three cases the decay of the 2-
halopropene radical cations follows exactly pseudo-first-order
kinetics, and the bimolecular rate constants (kbi), the reaction
efficiencies of the total reaction, and the branching ratio
between deprotonation and halogen substitution are given in
Table 2.
‡
yielding a 2-iodoallyl radical and NH4 are distinctly less
1
exothermic by about 40 and 60 kJmol , respectively. This
trend is expected because of the increased stabilization of the
reactant alkene radical cation by an iodo substituent at the
double bond, which is also reflected in the decreased
ionization energy of iodoethene[10] and the decreased stabili-
.
.
‡
‡
The efficiencies of the total reaction of 1 and 2 are
rather large and approach the collision rate, while that of the
.
.
‡
‡
zation of a radical by an a-iodo substituent. For 3 , a
iodo derivative 3 is distinctly smaller. It is clearly seen from
.
.
‡
‡
deprotonation yielding an allene and an iodine atom is also
Figure 2 that substitution of the halogen atom in 1 ± 3 by
the NH3 group to form the product ion with m/z ˆ 58
competes effectively with proton transfer to the ammonium
ion (m/z ˆ 18), in spite of the distinctly exothermic deproton-
1
slightly exothermic by 15 kJmol , whereas the analogous
process is endothermic and almost thermoneutral in the case
.
.
‡
‡
of 1 and 2 , respectively.
1802
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1998
0947-6539/98/0409-1802 $ 17.50+.50/0
Chem. Eur. J. 1998, 4, No. 9

Ion-Molecule Reactions
1799±1809
MERPs calculated for these processes (see previous section).
The efficiencies of 49% and 43% of the substitution process
.
.
‡
‡
of 1 and 2 , calculated from the branching ratio, are slightly
larger and smaller, respectively, than the values obtained for
this process of ionized vinyl chloride and vinyl bromide (57%
and 35%[7]). This suggests that the increased total reaction
.
.
‡
‡
efficiencies of the halopropenes 1 and 2 are due to the
additional deprotonation reaction.
In the case of the less reactive iodopropene radical cation
, deprotonation is the main process. Interestingly, depro-
.
‡
3
tonation yields not only ammonium ions (m/z ˆ 18), but also
‡
the proton-bridged dimer of ammonia [H3N ´´ H ´´ NH3]
(m/z ˆ 35). Under the experimental conditions used, the
.
‡
reaction of 3 proceeds to yield about equal amounts of the
substitution product (m/z ˆ 58), ammonium ions (m/z ˆ 18),
and the proton-bridged dimer (m/z ˆ 35). From this branching
ratio an efficiency of only about 13% is calculated for the
substitution process. A decrease of the efficiency of the gas-
phase substitution reaction of haloalkene radical cations with
decreasing strength of the carbon ± halogen bond has been
observed before[7] and will be discussed further in the next
section. The origin of this effect is very likely a smaller
amount of excess energy in the distonic addition product of
the first reaction step. Here it is of interest to note that even in
.
‡
the case of the slowly reacting 3 , substitution is still
competitive with regard to deprotonation by ammonia.
Again, this observation excludes a direct substitution mech-
anism of the SN2 type in ionized haloalkenes, which is
entropically demanding and which would impose an activa-
tion barrier on the reaction path. Instead, the substitution
proceeds by at least two reaction steps of an addition-
elimination mechanism, with a fast first addition step that
evidently overrides even exothermic deprotonation. This is
only conceivable if the collision complex of a 2-halopropene
radical cation and ammonia collapses quickly to the b-distonic
addition product without any activation barrier. Thus, the
results obtained here confirm the ab initio calculations of the
MERP of ionized vinyl halides with ammonia[7] and of the
addition of ethene radical cations to neutral ethene,[12] which
show no activation barrier for the addition of the electron-rich
neutral reaction partner to the ionized alkene double bond.
This ease of addition of nucleophiles also explains the high
reactivity of alkene radical cations in solution generated
electrochemically or by single-electron-transfer reactions.
To complete the examination of the substitution reaction of
Figure 2. Plot of ion intensity vs. reaction time (kinetic plot) for the
.
‡
reaction of a) 2-chloropropene radical cations 1 , b) 2-bromopropene
.
.
‡
‡
radical cations 2 , and c) 2-iodopropene radical cations 3 with ammonia.
Table 2. Kinetic parameters for the reaction of 2-halogenopropene radical
cations with ammonia.
.
.
‡
‡
the 2-halopropene radical cations 1 ± 3 with ammonia, the
‡
structure of the C3H8N product ion was verified by a gas-
10
kbi  10
Efficiency Branching ratio
phase titration experiment[13] from which the proton affinity
(PA) of the neutral C3H7N precursor (conjugate base) of the
product ion can be determined. It has been shown previously
for the substitution-product ion of vinyl halide radical cation
with ammonia, that originally a vinyl ammonium ion is the
only reaction product, which isomerizes more or less quickly
by a shuttle mechanism[14, 7] into the more stable, protonated
aldiminium ion during deprotonation. Accordingly, the de-
[cm3 mol 1 sec
]
[%]
substitution/deprotonation
1
.
.
.
‡
‡
‡
1
2
3
17.8
15.7
7.47
84
77
37
58:42
56:44
35:31:34[a]
[a] m/z 35 (proton-bridged dimer of ammonia).
ation process. Substitution is still the main process for the
.
.
‡
‡
‡
chloro and the bromo derivatives 1 and 2 , which also show
protonation of the C2H6N product ions by a reference base
an almost identical branching ratio for the deprotonation and
with a PA between the PA of vinylamine and acetaldimine
exhibits bimodal kinetics.[7] If the 2-propenyl ammonium ion
.
.
‡
‡
substitution. This similar behavior of 1 and 2 in their
.
.
‡
‡
reactions with ammonia is in agreement with the similar
generated by the substitution from 1 ± 3 behaves analo-
Chem. Eur. J. 1998, 4, No. 9
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1998
0947-6539/98/0409-1803 $ 17.50+.25/0
1803

H.-F. Grützmacher et al.
FULL PAPER
gously, the proton shuttle during deprotonation converts the
2-propenyl ammonium ion P (initially formed) into an
acetone immonium ion A (Scheme 3), and deprotonation of
‡
the product ions C3H8N by a suitable base should again
exhibit bimodal kinetics; this can be used to estimate the PA
of the neutral C3H7N precursor of P and A.
Scheme 3. Shuttle mechanism for the base-catalyzed tautomerization of
protonated enamines and ketimines in the gas phase.
Figure 3. Plot of ion intensity vs. reaction time for the deprotonation (gas-
.
‡
‡
The PA of the appropriate conjugate bases of C3H7N were
calculated by AM1, which is known to reproduce the PA
phase titration) of the substitution product C3H8N (from 2 ) by a) ethyl-
amine
929.5 kJmol
(PA ˆ 911.8 kJmol 1[10]),
and
b) dimethylamine
(PA ˆ
1[10]
1 [15]
)
within Æ 20 kJmol ,
yielding PA(2-propenylamine) ˆ
1
1
881 kJmol and PA(acetone imine) ˆ 923 kJmol (experi-
mental value 934 kJmol 1[10a]). As expected, the PA of
acetone imine is considerably larger than that of 2-propenyl-
amine. From these values it is concluded that a reference base
with ammonia confirms that both substitution reactions take
identical routes[7] to homologous products, in spite of the
much more effective competition of the deprotonation of the
radical cations in the case of the 2-halopropenes. Therefore,
addition of ammonia results in isomeric b-distonic ammonium
ions, corresponding to Markovnikov and anti-Markovnikov
orientation, and interconversion by a 1,2-NH3 shift before the
halogen atom is lost from the less stable anti-Markovnikov
product as depicted in Scheme 4.
1
‡
with a PA < 870 kJmol does not deprotonate the C3H8N
1
product ions and a base with a PA > 940 kJmol quickly
‡
deprotonates all C3H8N product ions, while the deprotona-
1
tion by a base with a PA around 900 kJmol should exhibit
bimodal kinetics. Indeed, this is observed experimentally.
With ammonia (PA ˆ 854.0 kJmol 1[10a]) and methylamine
(PA ˆ 899.0 kJmol 1[10a]) no deprotonation is observed, while
complete deprotonation is achieved with diethylamine (PA ˆ
952.4 kJmol 1[10a]). The kinetic plots for the deprotonation
‡
of the C3H8N product ions by ethylamine (PA ˆ
1[10a]
911.8 kJmol
)
and
dimethylamine
(PA ˆ
929,5 kJmol 1[10a]) are shown in Figure 3. Deprotonation by
both amines exhibits clearly the bimodal kinetics of a fast
exothermic proton transfer and of a slow endothermic or
thermoneutral deprotonation. However, owing to the differ-
ent capabilities of the amines to act as a catalyst for the proton
shuttle, the fractions of quickly and slowly deprotonated ions
‡
C3H8N are different.
From these kinetic data the PA of the two isomers of
Scheme 4. Mechanism of nucleophilic substitution reaction of ionized
the
neutral
C3H7N
precursors
are
limited
and
to
.
.
‡
‡
halopropenes 1 ± 3
.
1
899 kJmol 1 < PA(C3H7N(I)) < 914 kJmol
930
kJmol 1 < PA(C3H7N(II)) < 952 kJmol or 906 Æ 7 kJmol
1
1
1
and 941 Æ 11 kJmol , respectively; this is in reasonable
In addition to confirming the addition-elimination mecha-
nism for the nucleophilic substitution of aromatic and olefinic
radical cations, the results presented here also suggest a
special mechanism of the deprotonation reaction of these
unsaturated radical cations. In particular the deprotonation of
agreement with the calculated PA of the conjugate bases, 2-
propenylamine and acetone imine, of the expected products.
This absolutely analogous behavior of the reaction products
of the substitution of ionized haloethenes and halopropenes
1804
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1998
0947-6539/98/0409-1804 $ 17.50+.50/0
Chem. Eur. J. 1998, 4, No. 9

Ion-Molecule Reactions
1799±1809
‡
the iodo derivative 3 yielding a proton-bridged dimer of
Reactions of 3,3,3-trifluoro-2-halopropene radical cations
4
± 6 with ammonia: The 3,3,3-trifluoro-2-halopropene
.
.
‡
‡
ammonia (m/z ˆ 35) is not explained easily by a simple
proton-transfer mechanism. Generally it is assumed that
deprotonation of radical cations by a base occurs through a
loose transition state structurally related to a proton-bridged
complex of the base and the radical formed by deprotonation.
This critical configuration is reached by a direct approach of
the base onto the acidic hydrogen atom (Scheme 5) within the
collision complex of the reactants, followed by a rapid proton
transfer if the reaction is exothermic.
radical cations lack the methyl group that could be easily
deprotonated with ammonia. Hence, substitution of the 2-
halogen substituent is expected to be the only efficient process
.
.
‡
‡
observed for the reactions of 4 ± 6 with ammonia. The
bimolecular rate constants (kbi) and reaction efficiencies of
these reactions are presented in Table 3, and Figure 4 displays,
as an example, the kinetic plot for the reaction of the bromo
.
.
‡
‡
derivative 5 and the iodo derivative 6 with ammonia.
Table 3. Kinetic parameters for the reaction of 3,3,3-trifluoro-2-halogen-
opropene radical cations with ammonia.
10
kbi  10
Efficiency
[%]
[cm3 mol 1 sec
]
1
.
.
.
‡
‡
‡
4
5
6
17.9
17.4
10.7
88
87
54
.
.
‡
‡
Scheme 5. Mechanisms of deprotonation of ionized halopropenes 1 ± 3
by ammonia.
In the case of the alkene radical cations studied here, the
nucleophilic base is also trapped competitively in the collision
complex by addition to the double bond. However, if the
further substitution process of the addition product is slow, the
addition product will dissociate back to the complex of
reactants, which once again are available for a deprotonation
reaction. A consequence of this picture of the dynamics of the
reaction of the 2-halopropene radical cations with ammonia is
that the total reaction efficiency should always approach the
collision rate, because deprotonation is exothermic and fast
.
.
‡
‡
for all three radical cations 1 ± 3 . However, the experiment
.
‡
shows that in the case of the iodo derivative 3 not only is the
substitution process slow, but also the exothermic deprotona-
tion is slow, as shown by the decreased total reaction
efficiency. An obvious explanation of this effect is that
deprotonation does not occur independently of the substitu-
tion, but proceeds via the distonic addition product
(Scheme 5). A syn-1,2-elimination in the anti-Markovnikov
adduct as well as a 1,3-elimination in the preferred Markov-
nikov adduct should exhibit a considerable activation barrier
that requires sufficient excess energy in the addition product
Figure 4. Plot of ion intensity vs. reaction time (kinetic plot) for the
.
‡
reaction of a) 3,3,3-trifluoro-2-bromopropene radical cations 5 , and
.
‡
b) 3,3,3-trifluoro-2-iodopropene radical cations 3 with ammonia.
‡
For all three 3,3,3-trifluoro-2-halopropene radical cations
± 6 excellent pseudo-first-order kinetics were observed,
for the elimination of NH4 . Since the adduct ion derived from
.
.
‡
.
‡
‡
4
3
is less excited than in the case of the chloro and bromo
.
.
‡
‡
and the kbi obtained from the decay of these radical cation
derivatives 1 and 2 , deprotonation by this mechanism is
also reduced and may in fact need a second ammonia
molecule to surmount the activation barrier. However, addi-
tional and more focussed experiments are needed to sub-
stantiate these ideas about a special deprotonation mecha-
nism of unsaturated radical cations.
show that the reaction is very efficient for the chloro and
.
.
‡
‡
bromo derivatives 4 and 5 . The reaction again proceeds
essentially with the ion-molecule collision rate, while the iodo
.
‡
substituted radical cation 6 exhibits a reduced reaction
efficiency. Contrary to the original expectation, the ammo-
Chem. Eur. J. 1998, 4, No. 9
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1998
0947-6539/98/0409-1805 $ 17.50+.25/0
1805

H.-F. Grützmacher et al.
FULL PAPER
nium ion is the only abundant reaction product at long
calculated by ab initio methods indeed indicates an inter-
action between a proton of the NH3 group and a fluorine
atom from the CF3 group. It is not well understood, however,
.
.
‡
‡
‡
reaction times for all three radical cations 4 ± 6 (Figure 4).
However, an inspection of the ion-intensity vs. reaction-time
curves of the ammonium ion and the expected substitution-
.
.
‡
‡
why the C3H4NF2I and C3H3NFI ions are formed only
during the reaction of the 3,3,3-trifluoro-2-iodopropene
‡
product ion C3H5NF3 reveals that the ammonium ion clearly
.
‡
arises from a deprotonation of the substitution product
radical cation 6 . The addition of ammonia to the other
radical cations 4 and 5 is even more exothermic leading to
an even more chemically activated b-distonic ion. It should be
noted that the reaction efficiency for 4 and 5 corresponds
almost to the collision limit because of this large excess energy
available in the reaction intermediate. It may be that the large
excess energy of these intermediates reduces the life-time of
the distonic ions so much that elimination of HF is bypassed
because of the tight transition state necessary. Similarly, the
.
.
‡
‡
‡
C3H5NF3 , which accordingly exhibits a maximum of its
intensity during the reaction. The deprotonation of the
.
.
‡
‡
‡
C3H5NF3 ion was corroborated in a separate experiment by
‡
isolation of the C3H5NF3 ions and ejection of all other ions
from the FT-ICR cell after an appropriate reaction time
‡
yielding the product ion C3H5NF3 with sufficient intensity.
Following the isolation, a complete deprotonation of the
‡
C3H5NF3 ions by ammonia was observed. Evidently, the CF3-
.
‡
substituent lowers the PA of the conjugate base 2-(3,3,3-
trifluoropropenyl)amine below the PA of ammonia
(854 kJmol 1[10]), so that the corresponding ammonium
formation of CF3I (m/z ˆ 196) as a product of the reaction of
.
‡
6
with ammonia is unprecedented and surprising. Without
further information the reaction pathway yielding this ion
from the b-distonic intermediate is difficult to visualize, but
clearly requires the rearrangement of an excited ion. Hence,
the formation of all these additional product ions during the
reaction of the radical cations of 3,3,3-trifluoro-2-halopro-
penes 4 ± 6 with ammonia demonstrate convincingly the
presence of an especially large amount of excess energy in
the intermediates of these reactions, as expected from the
thermochemical calculations. Further, these results show that
the reactions following the formation of the excited b-distonic
ion by addition of ammonia to the alkene radical cations are
controlled at least as well by dynamic effects as by thermo-
chemistry. Thus, the efficiency of the reactions of the
haloalkene radical cations with ammonia increases form the
haloethenes to the corresponding 3,3,3-trifluoro-2-halopro-
penes as expected from the calculated MERPs. This shows
that in this series there is an increasing release of excess
energy during the first addition step to drive the subsequent
reaction. By the same effect of a decrease of the chemical
activation of the distonic intermediates in each series of
ionized chloro-, bromo-, and iodoalkene, the decrease of the
substitution reaction efficiency is explained. However, al-
though the distonic intermediate derived from the iodo
‡
C3H5NF3 ion, which is the expected substitution product of
.
.
‡
‡
4
± 6 , is quickly deprotonated with ammonia. This effect of
the strongly electron-withdrawing CF3 group on the PA is well
documented for carbonyl compounds, ethers, and amines,[10]
1
which exhibit PA values 50 ± 100 kJmol lower than the
corresponding CH3 substituted derivatives. The PAs of 2-
1
(3,3,3-trifluoropropenyl)amine (817 kJmol ) and trifluoro-
1
acetone imine (833 kJmol ), which are the conjugate bases of
the possible reaction products, are both below the PA of
ammonia.
‡
Besides the C3H5NF3 ion and the ammonium ion, the
.
.
‡
‡
reaction of 4 ± 6 with ammonia forms product ions with
.
‡
m/z ˆ 92 in moderate yields, and the iodo derivative 6
generates additional ions with m/z ˆ 219, m/z ˆ 199, and
m/z ˆ 196. Accurate mass determination of these ions, from
the high-mass-resolution capability of the FT-ICR spectrom-
eter of m/Dm ꢀ 300000, proves unequivocally the following
‡
elemental composition for these ions: m/z ˆ 92 ꢁ C3H4NF2 ,
.
.
‡
‡
m/z ˆ 219 ꢁ C3H4NF2I , m/z ˆ 199 ꢁ C3H3NFI , and m/z ˆ
.
‡
‡
196 ꢁ CF3I . The formation of a C3H4NF2 ion corresponds to
‡
the loss of HF from the regular substitution product C3H5NF3 .
An inspection of the ion-intensity vs. reaction-time curves
.
‡
‡
shows that this second product ion (C3H4NF2 ) is formed in
derivative 6 is less chemically activated than its chloro and
.
.
‡
‡
parallel to the main product ion and not by a consecutive
bromo analogues derived from 4 and 5 , it nevertheless
exhibits more side reactions with greater abundance. This
mixture of effects still makes it difficult to predict the course
and the rate constants of the total reaction of ionized alkenes
with electron-rich reactants.
‡
elimination of HF caused by attack of ammonia on C3H5NF3 .
Obviously the excess energy present in the b-distonic ion
created by a very exothermic addition of ammonia to the
.
.
‡
‡
3,3,3-trifluoro-2-halopropenes 4 ± 6 not only induces the
loss of the halogen substituent, but is sufficient to provoke
additional elimination of HF in the still energetically excited
substitution product before the products separate from the
reaction complex. In line with this argument, the excited b-
distonic intermediate arising from the addition of ammonia to
Experimental Section
.
FT-ICR spectrometry: All FT-ICR experiments were performed in a
Spectrospin Bruker CMS 47X FT-ICR instrument,[17] equipped with an
InfinityTM[18] cell of 6 cm length, a 4.7 T superconducting magnet, a 24 bit/
128 kword Aspect 3000 computer and an external ion source.[19]
‡
the iodo derivative 6 eliminates one and two molecules HF
instead of losing iodine to yield the ions m/z ˆ 219,
.
.
‡
‡
C3H4NF2I and m/z ˆ 199, C3H3NFI , respectively. The
structures of these ions were not examined, but the elimi-
nation of HF from assorted ions driven by the thermodynamic
stability of HF is well known in mass spectrometry.[16] The
most obvious mechanism is a 1,4-HF elimination from the
Markovnikov-adduct ion yielding the radical cation of a 3,3-
difluoro-2-haloallylamine, and the structure of the adduct ion
Kinetic measurements: Ions were generated by 25 eV electron impact (EI)
from the halopropenes and an ion source pressure of 10 6 mbar. The
mixture of the ions produced was focused into the FT-ICR cell by means of
a transfer optic. The trapping voltages of the back and front plates were set
to 1 Æ 0.1 Vand the voltages of the excitation plates to 0.0 Æ 0.1 V. The ions
under study were selected by broad-band (frequency sweep) ejection of
88 Vp±p and rf pulses of 14 Vp±p fixed-frequency (single shots) ejections. The
1806
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1998
0947-6539/98/0409-1806 $ 17.50+.50/0
Chem. Eur. J. 1998, 4, No. 9

Ion-Molecule Reactions
1799±1809
ejection process was finished after 15 ± 20 ms. To remove any excess kinetic
energy prior to reaction, the ions were thermalized by collisions with argon
added by a pulsed valve (opened for 8 ± 15 ms). The argon was removed
after a delay time of 0.5 ± 1.0 s. Fragment ions and product ions formed
during this period were again ejected by single shots of 14 Vp±p. This
method has been described in detail previously.[5a] Ammonia as reactant gas
was introduced into the FT-ICR cell continuously by a leak valve resulting
in a constant pressure of 3 Â 10 8 to 8 Â 10 8 mbar. The pressure readings of
the ionization gauge were corrected for the sensitivity of the neutral gas
General procedure for synthesis of 2-halo-3,3,3-trifluoropropenes: 1,1,1-
trifluoropropene was condensed at 308 to 408C into a two neck
reaction flask and the halogen was introduced as a gas (Cl2) or as a liquid
(Br2, ICl) into the stirred 1,1,1-trifluoro-propene until the typical color of
the halogen persisted in the reaction mixture. Then the reaction mixture
was stirred at 308C for another hour. After warming up, the excess
halogen was removed with a diluted aqueous solution of sodium bisulfite.
The aqueous layer was separated, and the organic layer was washed three
times with distilled water, dried over phosphorus pentoxide, filtered, and
fractionated by distillation at normal pressure to yield the 1,2-dihalo-3,3,3-
trifluoropropenes. These were converted into the 2-halo-3,3,3-trifluoro-
propenes by treatment with KOH/EtOH at 108C. After neutralization of
the reaction mixture with hydrochloric acid the 2-halo-3,3,3-trifluoropro-
penes were extracted with hexane and purified by distillation.
.
‡
used[20] and were calibrated by rate measurements of the reaction NH3
‡
.
‡
NH3 !NH4 NH2 (kbi ˆ 21  10 10 cm3 molecule 1 s 1).[21] The sensitivity for
NH3 was taken from ref. [20]. The reaction-time delay varied from 1.5 ms to
10 s. After the reaction time, all ions were excited by a frequency sweep of
88 Vp±p with a step width of 7.8 kHz and an excitation pulse of 8 ms. FT-ICR
spectra were averaged by 8 data acquisition circles and recorded by 32 k
data points for 30 different reaction times. Peak intensities were obtained
by exponential multiplication and Fourier transformation of the time
domain signal. For kinetic evaluation, peak intensities of the magnitude
spectra were normalized to the sum of all ions detected at each reaction
time. By fitting these data to an exponential function by means of the
Microcal Origin 4.5 program,[22] the reaction rate constants kexp were
obtained. The bimolecular rate constants kbi were calculated by use of the
number density of the neutral reactant derived from the corrected pressure.
The efficiencies of the reaction is given by kbi/kcol. The collision rate
constant kcol was calculated by the method of Su and Chesnavich.[23]
1,2-Dichloro-3,3,3-trifluoropropane: B.p. 538C; purity (by GC): 98,0%; 1H
NMR (250 Hz, CDCl3, 258C; TMS): d ˆ 4.34 (m; 1H; CH3), 3.97 (dd, 1H;
CH2), 3.73 (dd, 1H; CH1).
2-chloro-3,3,3-trifluoropropene (4): This very volatile compound (b.p.
158C) was distilled directly from the reaction mixture after warming up
and neutralization with hydrochloric acid. The yield was not determined.
1H NMR (250 Hz, CDCl3, 258C; TMS): d ˆ 6.07 (ddt, 2H; CH2), 5.79 (dd,
.
‡
2H; CH1); MS (70 eV): m/z (%): 132 (39)/130 (100) [M ], 113 (12)/ 111
.
‡
‡
‡
(37) [M
F], 95 (100) [C3F3H2 ]; 87 (11)/85 (34) [CF2Cl ], 75 (40)
.
‡
‡
‡
‡
[C3F2H ], 69 (87) [CF3 ], 63 (29)/61 (81) [M
CF3], 26 (20) [C2H2 ].
1
1,2-dibromo-3,3,3-trifluoropropane: B.p. 1168C; purity (by GC): 98%; H
NMR (250 Hz, CDCl3, 258C; TMS): d ˆ 4.44 (m, 1H; CH3), 3.88 (dd, 1H;
CH2), 3.64 (dd, 1H; CH1).
Gas phase titration experiments: Gas phase titration experiments were
performed by generation of the substitution product by chemical ionization
(CI) in the external CI source of ammonia and 2-bromopropene. The ions
were focused into the FT-ICR cell and deprotonated by the titration base
present in the FT-ICR cell at a constant pressure. The determination of the
deprotonation efficiency followed the procedure described above for the
kinetic measurements. However, the thermalization of the cations had to
be omitted, because otherwise an erroneous ratio of the isomers may be
observed due to uncontrolled reaction during this period. Amines with
2-bromo-3,3,3-trifluoropropene (5): Yield: 91%; b.p. 348C; purity (by GC):
>95%; 1H NMR (250 Hz, CDCl3, 258C; TMS): d ˆ 6.50 (q, 1H; CH2), 6.03
.
(q, 1H; CH1); MS (70 eV, EI): m/z (%): 176 (65) /174 (65), [M ], 157 (12)/
‡
.
.
‡
.
‡
‡
‡
155 (19) [M
F], 131 (14)/129 (14) [CF2Br ], 107 (24)/105 (25) [M
‡
‡
‡
CF3], 95 (100) [C3F3H2 ], 75 (39) [C3F2H ], 69 (71) [CF3 ], 26 (26) [C2H2 ].
1-chloro-3,3,3-trifluoro-2-iodopropane: Purity (by GC): 99.0%; 1H NMR
(250 Hz, CDCl3, 258C; TMS): d ˆ 4.45 (m, 1H; CH3), 3.95 (q, 1H; CH2),
3.84 (dd, 1H; CH1). The compound decomposed during heating.
1
proton affinities[10] from 899 to 998 kJmol were used as titration bases.
Materials and reagents: Ammonia (99.8%; Merck), methanol (>99.9%;
3,3,3-trifluoro-2-iodopropene (6): The thermally labile iodide
6 was
Merck), 2-chloropropene
1 (>97%; Fluka) and 2-bromopropene 2
sampled into a liquid-nitrogen-cooled flask during distillation. Yield:
50%; b.p. 568C; purity (by GC): 98%; 1H NMR (250 Hz, CDCl3, 258C;
TMS): d ˆ 6.95 (q; 2H; CH2), 6.30 (dd, 2H; CH1); MS (70 eV): m/z (%):
(>99%; Fluka) were obtained commercially.
Synthesis of 2-iodopropene 3 and 2-halo-3,3,3-trifluoropropenes 4 ± 6: The
purity of 2-iodopropene and of the 2-halo-3,3,3-trifluoropropenes synthe-
sized in this work were analyzed by gas chromatography (Hewlett Packard
HP 5890 Series II equipped with a 30 m SE54 capillar column, a VG
Autospec mass spectrometer and a Bruker AC250P 1H NMR spectrom-
eter). For 1H NMR characteriza-
.
.
.
‡
‡
‡
‡
222 (100) [M ]; 203 (23) [M
F]; 177 (8) [CF2I ]; 153 (5) [M
CF3];
.
‡
‡
‡
‡
128 (25) [HI ]; 127 (71) [I ], 95 (24) [C3F3H2 ]; 75 (92) [C3F2H ], 69 (85)
‡
[CF3 ].
tion the protons of 4 ± 6 are la-
beled as indicated in the struc-
Appendix
tures given below. The 2-halo-
3,3,3-trifluoropropenes 4 ± 6 were
prepared by a two step synthesis,
starting with the addition of the
The experimental heats of formation (DHf) used for the calculation of the
.
‡
isodesmic reactions are given in Table 4. The DHf(chloroethene ) and
.
‡
DHf(bromoethene ) are well known. However, in the case of iodoethene
only the ionization energy of 9.30 Æ 0.10 eV is tabulated. Combining this
with the DHf of 2-iodoethene derived by the increment system of S. W.
halogen to 1,1,1-trifluoropropene, which generates the corresponding 1,2-
dihalo-3,3,3-trifluoropropanes. These were transformed by elimination of
HX to the corresponding 2-halo-3,3,3-trifluoropropenes according to a
modified method of R. N. Hazeldine and K. Leedham with KOH/EtOH as
the base.[24]
.
‡
Benson[26] yields DHf(iodoethene ) given in Table 4. The NIST Webbook
contains two values for DHf(2-chloropropene), and the value of
1
24.7 kJmol has been selected for the further calculations, since this
value is more consistent with Bensonꢁs increments.[26] No experimental
values for DHf(2-bromopropene) and DHf(iodopropene) are available, and
these DHf have been estimated with Bensonꢁs increments.[26]
Synthesis of 2-iodopropene 3: 2-chloro-2-iodopropane was synthesized by
the reaction of 2-chloropropene with concentrated hydroiodic acid.[25] 2,2-
diiodopropane was synthesized by stirring 2-chloro-2-iodopropane (4.5 g,
22.0 mmol), sodium iodide (15.0 g, 100.0 mmol) and acetone (50 mL) for
three days at ambient temperature. Distilled water (100 mL) was added to
the reaction mixture, and the aqueous layer was washed three times with
dichloromethane (100 mL). The combined dichloromethane layers were
washed three times with distilled water and dried over anhydrous sodium
sulfate. The solvent was removed by vacuum (15 mbar). The crude product
was distilled at 0.03 mbar into a liquid-nitrogen-cooled Schlenk tube
yielding 2,2-diiodopropane (0.23 g, 0.8 mmol). 1H NMR (250 Hz, CDCl3,
258C; TMS): d ˆ 3.00 (s, 6H; CH3); MS (70 eV, EI): m/z (%): 296 (23)
Isodesmic reactions of the following type [Eqs. (1) and (2)], in which the
halogen substituent and either an H atom or a CH3 substituent change their
.
‡
places, have been used for the calculation of DHf (2-halopropene ).
.
‡
.
.
.
‡
‡
‡
ˆ
ˆ
ˆ
ˆ
CH3 CX CH2 ‡ H CH CH2 ! H CX CH2 ‡ CH3 CH CH2 (1)
.
‡
.
‡
ˆ
ˆ
CH3 CX CH2 ‡ H C(CH3) CH2
!
‡
.
.
‡
ˆ
ˆ
H CX CH2 ‡ CH3 C(CH3) CH2
(2)
.
.
‡
.
.
.
‡
‡
‡
‡
‡
‡
[M ], 254 (7) [I2 ], 169 (100) [C3H6I ], 128 (8) [HI ], 127 (19) [I ]; 42 (15)
In the case of 2-bromopropene and 2-iodopropene the values obtained
in this way were confirmed by isodesmic reactions of a halogen interchange
.
‡
.
‡
‡
‡
[C3H6 ], 41 (67) [C3H5 ] 40 (6) [C3H4 ], 39 (30) [C3H3 ]. 2-iodopropene was
obtained from the 2,2-diiodopropane by elimination of hydroiodic acid.
.
.
‡
‡
with 2-chloropropene and chloroethene , which had been calculated at
Chem. Eur. J. 1998, 4, No. 9
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1998
0947-6539/98/0409-1807 $ 17.50+.25/0
1807

H.-F. Grützmacher et al.
FULL PAPER
Table 4. Experimental heats of formation[a] used for isodesmic reactions.
The DHf of the relevant species derived from 3,3,3-trifluoro-2-chloropro-
pene were calculated by analogous sets of isodesmic reactions with the
corresponding species derived from 2-chloropropene and 3,3,3-trifluoro-
propene, with its radical cation, the 2,2,2-trifluoroethyl radical and the
3,3,3-trifluoropropyl radical as references.
DHf
[kJmol
neutral
Ionization
energy
[eV][a]
DHf
1
[a]
1
]
[kJmol
radical
cation[b]
]
chloroethene
bromoethene
iodoethene
chloropropene
bromopropene
iodopropene
35.3 Æ 1.4
79.2 Æ 1.9
129[c]
9.99 Æ 0.02
9.82 Æ 0.03
9.30 Æ 0.1
999
1027
1026
Acknowledgments: The financial assistance of this work by the VW-
Stiftung (Schwerpunkt intra- und intermolekularer Elektronentransfer) is
gratefully acknowledged. The FT-ICR spectrometer used for this work was
a grant of the Deutsche Forschungsgemeinschaft. We thank the Fonds der
Chemischen Industrie for additional financial assistance.
24.7
27[c]
82[c]
ethene
propene
isobutene
3,3,3-trifluoropropene
ethyl radical
n-propyl radical
isopropyl radical
isobutyl radical
tert.butyl radical
2,2,2-trifluoroethyl radical
3,3,3-trifluoropropyl radical
allyl radical
52.5 Æ 0.5
20.4 Æ 0.5
18.0 Æ 1.0
614 Æ 7
10.5138
1014
939
509
441
Received: February 9, 1998 [F994]
9.73 Æ 0.01
9.22 Æ 0.02
10.95 Æ 0.1
118.8 Æ 4
100.0 Æ 2.1
92.0 Æ 2.1
66.9 Æ 2.1
46.o Æ 2.5
517 Æ 8
[1] a) L. P. Johnstone, N. P. Schepp, J. Am. Chem. Soc. 1993, 115, 6564 ±
6576; b) J. P. Dinnocenzo, D. R. Lieberman, T. R. Simpson, J. Am.
Chem. Soc. 1993, 115, 366 ± 367; c) G. Trampe, J. Mattay, S. S.
Steenken, J. Phys. Chem. 1989, 93, 7157 ± 7165; d) S. J. Steenken, C. J.
Warren, B. C. Gilbert, J. Chem. Soc. Perkin Trans. 2 1990, 335 ± 342;
e) M. Yasuda, Y. Matsuzaki, K. Shima, C. Pac, J. Chem. Soc. Perkin
Trans. 2 1988, 745 ± 751; f) O. Hammerich, V. D. Parker, Adv. Phys.
Org. Chem. 1984, 20, 56 ± 189; g) B. Reitstöen, V. D. Parker, J. Am.
Chem. Soc. 1991, 113, 6954 ± 6958; h) T. Drewello, N. Heinrich,
W. P. M. Maas, N. M. M. Nibbering, T. Weiske, H. Schwarz, J. Am.
Chem. Soc. 1987, 109, 4810 ± 4818.
536[d]
163.2 Æ 2
methallyl radical
121.3 Æ 2
[a] Taken from ref. [10] if not otherwise stated. [b] Calculated from DHf
(neutral) and the ionization energy. [c] Calcaluated by the incremental
method of S. W. Benson.[26] [d] Calculated from the electron affinity data
given in ref. [10].
[2] a) N. Heinrich, W. Koch, J. C. Morrow, H. Schwarz, J. Am. Chem. Soc.
1988, 110, 6332 ± 6336; b) S. S. Shaik, A. Pross, J. Am. Chem. Soc. 1989,
111, 4306 ± 4312; c) J. K. Cho, S. S. Shaik, J. Am. Chem. Soc. 1991, 113,
9890 ± 9891.
[3] P. A. Pabon, N. L. Bauld, J. Am. Chem. Soc. 1984, 106, 1145 ± 1146.
[4] For a recent example see P. Maslak, W. H. Chapman, Jr., T. M.
Vallombroso, Jr., B. A. Watson, J. Am. Chem. Soc. 1995, 117, 12380 ±
12389.
[5] a) D. Thölmann, H.-F. Grützmacher, J. Am. Chem. Soc. 1991, 113,
3281 ± 3287; b) D. Thölmann, H.-F. Grützmacher, Int. J. Mass
Spectrom. Ion Processes 1992 117, 415 ± 440; c) D. Thölmann, S.
Hamann, H.-F. Grützmacher, Int. J. Mass Spectrom. Ion Processes
1994, 137, 43 ± 54.
[6] a) S. S. Shaik, Acta Chem. Scan. 1990, 44, 205 ± 221; b) S. S. Shaik, E. J.
Canadell, J. Am. Chem. Soc. 1990, 112, 1446 ± 1452; c) A. Pross, J. Am.
Chem. Soc. 1986, 108, 3537 ± 3538; d) S. S. Shaik, J. Am. Chem. Soc.
1981, 103, 3692 ± 3701; e) A. Pross, S. S. Shaik, J. Am. Chem. Soc. 1981,
103, 3702 ± 3709.
the MP2/6-31G* level of theory. Similarly, the distonic ions generated by
the addition of NH3 to the haloalkene radical cations in the Markovnikov
and anti-Markovnikov mode were treated in the isodesmic reactions as
substituted alkyl radicals [Eqs. (3), (4), (5), and (6)].
.
.
‡
CH3 C (X) CH2(NH3 ) ‡ H C H CH3
!
.
.
‡
H C (X) CH2(NH3 ) ‡ CH3 C H CH3
(3)
(4)
(5)
(6)
.
.
‡
CH3 C (X) CH2(NH3 ) ‡ H C (CH3) CH3
!
.
.
‡
H C (X) CH2(NH3 ) ‡ CH3 C (CH3) CH3
.
.
‡
CH3 C(NH3 )(X) C H2 ‡ H CH2 C H2
!
.
.
‡
H C(NH3 )(X) C H2 ‡ CH3 CH2 C H2
[7] A. Nixdorf, H.-F. Grützmacher, J. Am. Chem. Soc. 1997, 119, 6544 ±
6551.
[8] D. Thölmann, D. Flottmann, H.-F. Grützmacher, Chem. Ber. 1991 124,
2349 ± 2356.
.
.
‡
CH3 C(NH3 )(X) C H2 ‡ H CH(CH3) C H2
!
.
.
‡
H C (X) CH2(NH3 ) ‡ CH3 CH(CH3) C H2
[9] a) C. H. DePuy, S. Gronert, S. E.Barlow, V. M. Bierbaum, R. Dam-
rauer, J. Am. Chem. Soc. 1989, 111, 1968 ± 1973; b) F. G. Bordwell, J. P.
Cheng, J. Am. Chem. Soc. 1989, 111, 1792 ± 1795.
[10] a) NIST Chemistry WebBook (http//webbooknist.gov/chemistry),
NIST Standard Reference Database No. 69, August 1997. b) NIST
Standard Reference Database 19A, NIST Positive Ion Energetics
Database, Vers. 1.1, National Institute of Standards and Technology,
Gaithersburg (USA), 1989.
Since the DHf of the distonic ions formed by addition of NH3 to the radical
cation of iodoethene are not available, a second set of isodesmic reactions
were used, in which the distonic ions were equilibrated with the radical
cations of the appropriate 2-halopropenes. The DHf of substitution product
2-propenyl ammonium cation was obtained from the DHf of its lower
homologue vinyl ammonium cation by comparing with alkenes in the
isodesmic reactions given in Equations (7) and (8).
[11] W. J. Hehre, L. Radom, P. von R. Schleyer, J. A. Pople, Ab Initio
Molecular Orbital Theory, Wiley, New York, 1985.
[12] J. Alvarez-Idaboy, L. A. Erikson, T. Fangström, S. Lunell, J. Phys.
Chem. 1993, 97, 12737 ± 12741.
[13] H. H. Büker, H.-F. Grützmacher, M. E. Crestoni, A. Ricci, Int. J. Mass
Spectrom. Ion Processes 1997, 160, 167 ± 181.
[14] a) D. K. Bohme, Int. J. Mass Spectrom. Ion Processes 1992, 115, 95 ±
110; b) A. J. Chalk, L. Radom, J. Am. Chem. Soc. 1997, 119, 7573 ±
7578.
[15] a) M. J. S. Dewar, K. M. Dieter, J. Am. Chem. Soc. 1986, 108, 8075 ±
8086; b) D. M. Myton, R. J. OꢁBrien, Anal. Chem. 1991, 63, 1201 ±
1205.
[16] D. A. Stams, T. D. Thomas, D. C. MacLaren, D. Ji, T. H. Morton, J.
Am. Chem. Soc. 1990, 112, 1427 ± 1434.
‡
ˆ
ˆ
CH3 C(NH3 ) CH2 ‡ H CH CH2
!
‡
ˆ
ˆ
H C(NH3 ) CH2 ‡ CH3 CH CH2
(7)
(8)
‡
ˆ
ˆ
CH3 C(NH3 ) CH2 ‡ H C(CH3) CH2
!
‡
ˆ
ˆ
H C(NH3 ) CH2 ‡ CH3 C(CH3) CH2
Finally, the DHf of the 2-haloallyl radical cations were calculated from the
isodesmic reactions given in Equations (9) and (10).
CH2 CX CH2 ‡ CH3 CH CH2 ! CH2 CH CH2 ‡ CH3 CX CH2 (9)
.
.
ˆ
ˆ
ˆ
ˆ
.
.
ˆ
ˆ
ˆ
CH2 CX CH2 ‡ CH3 C(CH3) CH2CH2 C(CH3) CH2
ˆ
‡ CH3 C(X) CH2
(10)
1808
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1998
0947-6539/98/0409-1808 $ 17.50+.50/0
Chem. Eur. J. 1998, 4, No. 9

Ion-Molecule Reactions
1799±1809
[17] M. Allemann, H. Kellerhals, K. P. Wanczek, Int. J. Mass Spectrom. Ion
Phys. 1983, 91, 139 ± 142.
[22] Microcal Origin 41, Microcal Software Inc., Northampton MA 01060
(USA), 1996.
[18] P. Caravatti, M. Allemann, Org. Mass Spectrom. 1991, 26, 514 ± 518
[19] P. Kofel, M. Allemann, H. Kellerhals, K. P. Wanczek, Int. J. Mass
Spectrom. Ion Physics 1985, 65, 97 ± 103.
[20] J. E. Bartmess, R. M. Georgiadis, Vaccuum 1983, 33, 149 ± 153.
[21] V. G. Anicich, J. Phys. Chem. Ref. Data 1993, 22, 1506.
[23] T. Su, W. J. Chesnavich, J. Chem. Phys. 1982, 76, 5183 ± 5185.
[24] R. N. Hazeldine, K. Leedham, J. Chem. Soc. 1952, 3483 ± 3490.
[25] A. Oppenheim, Justus Liebigs Ann. Chem. 1868, Suppl. VI, 353 ± 375.
[26] S. W. Benson, Thermochemical Kinetics: Methods for the Estimation
of Thermochemical Data and Rate Parameters, Wiley, New York, 1976.
Chem. Eur. J. 1998, 4, No. 9
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1998
0947-6539/98/0409-1809 $ 17.50+.25/0
1809